MISPost 600

Management
Information
SyStemS
Solving BuSineSS ProBlemS with
information technology
Gerald V. Post
Sixth Edition
Management
Information
Systems
Solving Business Problems
With
Information Technology
Version 6.0.1
Gerald V. Post
University of the Pacifc
Management Information Systems
Solving Business Problems with Information Technology
Copyright 2012 by Gerald V. Post
All rights reserved. No part of this publication may be reproduced or distributed
in any form or stored in any database or retrieval system without the prior written
consent of Gerald V. Post.
The cover photo shows a bicycle hill in Northern California but is a metaphor for
the road to understanding information technology. It can be a fun road with scen-
ery, climbs, and turns.
Students:
Your honesty is critical to your reputation. No company wants to hire a thief
particularly for jobs as critical as application development and database adminis-
tration. If someone is willing to steal something as inexpensive as an e-book, how
can that person be trusted with billions of dollars in corporate accounts?
You are not allowed to share this book in any form with anyone else. You can-
not give or sell any information from this publication in any form to anyone else.
To purchase this book or other books: http://JerryPost.com/Books
To my parents.
v Preface
v
Preface
Managers and Information Technology
Today, everyone uses computers. But does everyone use them effciently? How
many business people turn to a spreadsheet when a database would be a better
tool? Do managers understand how information technology can solve business
problems? As a future manager, do you know about all of the technologies that
businesses are using? Can you spot a business problem and identify possible in-
formation technology solutions? Can you tell when a company is simply wasting
money on technology and not solving problems?
Over the past decade, technology has changed many jobs. Competition and an
economic downturn caused companies to cut costs wherever possible. In many
cases, that meant reducing the number of employees. In the past few years,
companies have used technology to alter and even eliminate management jobs.
Twenty years ago, new college graduates could get entry-level jobs as manag-
ers, often performing relatively simple analytical tasks or summarizing data. To-
day, software does those jobs and generates detailed information on demand for
executives.
If you want a job as a manager, you need to know how to use information tech-
nology. But it is not simply a matter of knowing how to use a word processor or
spreadsheet. You need to use the technology to collaborate with other workers, to
analyze data, and to fnd ways to improve your organization.
Continual changes in IT present two challenges: learning to use new technolo-
gies and fnding new opportunities to improve management. Most students have
taken a hands-on course that teaches them how to use a computer. Many expect
the introductory MIS course to be more of the samehands-on computer usage
tied to specifc needs. However, there are more complex and interesting problems
to be solved. Managers need to apply their knowledge of IT tools to solve man-
agement problems and fnd new opportunities to improve their organizations. The
focus of this book is to investigate the more complex question: How can informa-
tion technology be used to improve management and make companies more ef-
fcient or better than their competitors?
Learning Assessment
After fnishing the book, students should be able to evaluate common business
problems and identify information technology solutions that could help an orga-
nization. Students should be able to explain their choices and point out potential
problems or issues.
Anyone who teaches this material knows that this learning objective is diff-
cult to meetbecause of the huge number of possible issues and the fexibility
required in analyzing problems. To meet the objective, smaller, and more concrete
goals are presented in each chapter. Each of these is spelled out as a series of ques-
tions at the start of the chapter. By the end of each chapter, students should be able
to provide intelligent answers to the various questions.
The book provides several tools to achieve its objectives. The industry cases
and Reality Bytes show business problems and how organizations have attacked
them. The body of the text explains the problems in more detail and focuses on the
strengths and weaknesses of the various tools used to solve similar problems. The
Technology Toolboxes and chapter exercises show students how to apply specifc
tools and concepts to solve problems.
vi Preface
Organization
The organization of the text builds from basic computer concepts to teamwork and
strategy. Each section and chapter is focused on how managers can use informa-
tion technology to solve business problems. In particular, the frst section focuses
on the information technology infrastructurethe specifc tools and technologies
used to collect, transfer, and store data. These hardware and software tools form
the building blocks of any information system.
The second section explores how IT is used to support business integration and
teamwork. It begins by analyzing the importance of computer security when shar-
ing data, and includes chapters on enterprise resource planning (ERP) systems
and using the Internet for electronic business. The teamwork chapter examines the
role of recent tools for collaboration and teamworkcritical concepts in todays
business organizations. These tools constitute some of the most recent additions to
any companys toolbox. Any manager needs to be able to use these tools to work
with colleagues across the room and around the world.
The third section focuses on tools to analyze data and make decisions. As the
power of computers and software increases, more tasks get transferred to the ma-
chines. To adapt and grow, managers need to move to higher-level tasks including
evaluating the data, identifying strategies, and organizing new businesses.
The fourth section looks at issues directly affecting the management of infor-
mation technology resources. The chapters deal with how systems are developed
and the management issues of the development process. Organizing the resources,
particularly centralization and decentralization are also critical topics for manag-
ers who need to resolve conficts and determine the best ways to ft IT into an
organization. The last chapter looks at broader issues of how information technol-
ogy decisions affect society and how the political and social environments infu-
ence business technology decisions. Privacy, crime, jobs, and education are top-
ics that managers need to factor
into their business decisions.
The organization of the text
is based on two features. First,
each chapter emphasizes the
goal of the text: applying infor-
mation technology to improve
management and organizations.
Second, the text is organized
so that it begins with concepts
familiar to the students and
builds on them.
Each chapter is organized in
a common format: (1) the in-
troduction ties to the goal and
raises questions specifc to that
chapter, (2) the main discussion
emphasizes the application of
technology and the strengths
and weaknesses of various ap-
proaches, and (3) the applica-
Chapter 1: Introduction
Part 1: Information Technology Infrastructure
Chapter 2: Information Technology Foundations
Chapter 3: Networks and Telecommunications
Chapter 4: Database Management
Part 2: Business Integration
Chapter 5: Computer Security
Chapter 6: Transactions and ERP
Chapter 7: Electronic Business
Chapter 8: Teamwork
Part 3: Decisions and Strategies
Chapter 9: Business Decisions
Chapter 10: Strategic Analysis
Chapter 11: Entrepreneurship
Part 4: Organizing Businesses and Systems
Chapter 12: Systems Development
Chapter 13: Organizing MIS Resources
Chapter 14: Information Mgt. and Society
vii Preface
tion of technology in various real-world organizations is presented in the end-of-
chapter cases.
Chapter 1 (Introduction) remains an introduction to MIS and provides an ex-
planation of the goalsemphasizing the texts focus on how technology can help
managers perform their jobs and improve the companies they manage.
Chapter 2 (Information Technology Foundations) reviews the basic issues
in personal productivity and hardware and software. It emphasizes recent issues
such as the growing importance of wireless devices. It also discusses issues in
choosing computers used in e-business such as the importance of scalability in
servers.
Chapter 3 (Networks and Telecommunications) explains the role of net-
works in managing businesses. It also explains the foundations of the Internet, so
students can understand how problems can arise on the Internet and why certain
technologies have evolved to solve them.
Chapter 4 (Database Management) explains the importance of database sys-
tems in business. The focus remains on managers roles and uses of databases,
such as querying databases and building reports.
Chapter 5 (Computer Se-
curity) focuses on the business
problems and threats. It explains
the various tools and the fact
that business managers are criti-
cal in helping maintain computer
security.
Chapter 6 (Transactions and
Enterprise Resource Planning)
raises the main issues in opera-
tions and transaction processing.
It includes the role of account-
ing and HRM systems as well
as supply chain management
and customer relationship man-
agement, particularly payment
mechanisms. The goal is to show
managers how the systems are
used to solve common business
problems.
Chapter 7 (Electronic Busi-
ness) describes the various ways
that companies can use the Inter-
net to extend their businesses. It
includes traditional e-commerce
applications, as well as Web ser-
vices and mobile commerce.
Chapter 8 (Teamwork) ex-
plains the increasing importance
of teamwork and collaboration
tools. Managers need these tools
in almost any business, and al-
most no other textbooks even
Technology Toolbox
1 Choosing a Search Engine
Finding Information on the Internet
2 Voice Input
Creating Efective Charts
3 Creating Web Pages
Transferring Files on the Internet
4 Building Forms in Access
Creating Database Reports
5 Assigning Security Permissions
Encryption
6 Selecting an ERP System
Designing an EIS/Dashboard
7 Choosing Web Server Technologies
Paying for Transactions
8 Collaborating with SharePoint
Meeting Space and IM
9 Forecasting a Trend
Browsing Data with a PivotTable
10 Locating Customers with a GIS
Analyzing Businesses
11 Sending E-Mail Legally (CAN-
SPAM)
Creating a Business Plan
12 Creating Forms with InfoPath
Programming a New Function in Excel
13 Defning E-Mail Rules
Managing Projects
14 Privacy
Working in a Global Environment
viii Preface
talk about them. Here, an entire chapter ex-
plains the problems that need to be solved
and shows how the various tools support
communication, collaboration, and knowl-
edge management.
Chapter 9 (Business Decisions) begins
with the basic importance of using analyti-
cal tools to make decisions. It then explains
how basic tools help managers evaluate
data, from decision support systems to data
mining. Expert systems are introduced to
show how computers can make more ad-
vanced decisions. Specialized tools and
topics in artifcial intelligence research are
used to show how more intelligent soft-
ware can solve even relatively complex
problems.
Chapter 10 (Strategic Analysis) ex-
plores in detail how information systems
are used to support the functional business
areas to help gain a competitive advantage.
The chapter also explores how IT helps entrepreneurs create new businesses.
Chapter 11 (Entrepreneurship) examines how information technology is used
to start new companies and manage small businesses. By defnition, small busi-
nesses have few employees so it becomes critical for entrepreneurs to leverage
the technology to handle the thousands of details of managing a growing business.
Chapter 12 (Systems Development) looks at systems design and various de-
velopment alternatives. It examines the challenges faced in developing software
and the continuing movement to commercial off-the-shelf software.
Chapter 13 (Organizing MIS Resources) looks at the issues involved in the
organization and management of MIS resources. In examining MIS roles, it also
presents job opportunities. The chapter discusses how wireless, intranets, and
Web services are having important effects on the structure and management of IT
resources.
Chapter 14 (Information Management and Society) is an expanded discus-
sion of the impact of IT on societyparticularly the effects of the Internet. It
investigates the issue of privacy versus business, social, and governmental needs.
It examines the potential changes in a global society that is increasingly linked
online.
Features Tat Focus on Solving Problems
Why do you need a textbook? You have access to the Internet and you can fnd
millions of articles discussing technology, even online support for various soft-
ware tools. You can read several online magazines about information technolo-
gyfor free. Regardless of where you live, you can subscribe to manufacturer
press releases and watch the technology change. You can subscribe to the Wall
Street Journal and read about business, technology, governance, and host of other
enlightening subjects. And yes, if you are serious about business and informa-
tion technology, you should do all of these things. But you can still use a text-
Case focus: Industry
1 Fast Food
2 Computer Industry
3 Wholesale Suppliers
4 Pharmaceuticals
5 Professional Sports
6 Automobile Industry
7 Retail Sales
8 Package Delivery
9 Financial Services
10 Airline Industry
11 Entrepreneurship
12 Government Agencies
13 Energy Industry
14 Health Care
ix Preface
book. This textbook provides a structure to the information. It teaches you how
the many pieces ft together to solve business problems. It provides the context
and ideas on how to analyze the millions of tidbits of data foating on the Web.
Use the textbook to provide the foundation knowledge that describes the informa-
tion system discipline. Then when you read new facts and see new technologies
evolve, you can understand how to use them to improve your career.
Each chapter contains several unique features to assist in understanding the ma-
terial and in applying it to analyze and solve business problems. Each chapter
focuses on a specifc type of problem. These problems are highlighted by the in-
troduction and demonstrated in the business cases. The text dissects the problems
and explores how technology can be used to solve the problems.
What You Will Learn in This Chapter. A series of questions highlight the
important issues.
Lead Case. An introductory, real-world case illustrates the problems ex-
plored in the chapter.
Trends. Sidebar boxes present the major changes, brief history, or trends
that affect the topics in the chapter.
Reality Bytes. Brief applications, mini-cases, and discussions emphasize a
specifc point and highlight international issues, business trends, or ethics.
They also illustrate problems and solutions in the real world.
Technology Toolboxes. A short example and description of a software tool
that can be used to help managers solve specifc problems. They provide a
hands-on example of specifc projects. Students are encouraged to follow the
exercises and use the software tools to build the examples.
Chapter summary. A brief synopsis of the chapter highlightsuseful when
reviewing for exams.
A Managers View. A short summary of how the material in the chapter ap-
plies to building applications.
Key Words. A list of words introduced in the chapter. A full glossary is pro-
vided at the end of the text.
Website References. Web sites that provide discussions or links to useful
topics.
Review Questions. Designed as a study guide of the main topics in the
chapter..
Exercises. Problems that apply the knowledge learned in the chapter. Many
utilize common application software to illustrate the topics. Questions and
Exercises marked with a checkmark icon have at basic solutions available
online. Students can check their knowledge and progress by checking their
work on these exercises.
Additional Reading. References for more detailed investigation of the
topics.
Industry-Specifc Cases. In-depth discussion of the lead case and several
other companies. Each chapter highlights a specifc industry and compares
x Preface
different approaches
to the problems faced
by the frms..
Rolling Thunder Bi-
cycles Database Ap-
plication. A sample
company database
with integrated data
for several years. Use-
ful for demonstrating
transactions, creating
business queries, and
analyzing data.
Goals and Philosophy
All of the chapters emphasize the goal of understanding how information
technology can be used to improve management. The focus is on under-
standing the benefts and costs of technology and its application.
Emphasis is on the importance of database management systems. Increas-
ingly, managers need to retrieve data, and utilize a DBMS to investigate,
analyze and communicate.
Emphasis is also placed on the importance of communication, teamwork,
collaboration, and integration of data. Understanding information technol-
ogy requires more than knowledge of basic application packages. Students
need to use and understand the applications of groupware technologies.
In-depth cases illustrate the use of technology. By focusing each chapter
on a specifc industry, students can understand and evaluate a variety of ap-
proaches. Many cases illustrate companies varying over time, so students
can see the changes occurring in business and understand the evolving role
and importance of information technology.
The Rolling Thunder Database, a medium-size, detailed database application
of a small business, is available on disk. Specifc exercises are highlighted in
each chapter. The database contains data and applications suitable for oper-
ating a small (fctional) frm. It also contains data generation routines so that
instructors can create their own scenarios.
Tis Book Is Diferent from Other Texts
First, this book is a business text designed for an upper division or M.B.A. course,
so it does not have the hundreds of step-by-step instructions to show students how
to use a particular piece of software. It addresses the more diffcult question of
how to use information technology to solve business problems. Consequently, this
text focuses on business issues frst.
Second, this book is not simply a dictionary or encyclopedia that defnes tech-
nology terms in one or two sentences. Students need to understand the various
technology tools and see how they actually solve business problems. This book
uses detailed cases to show the business problems and the technological solutions
xi Preface
to illustrate each chapter. It also contains in-depth explanations of various technol-
ogy issues, showing their strengths and weaknesses. Where possible, these tools
are also demonstrated with hands-on applications, through the accompanying da-
tabases and through the Technology Toolboxes.
E-Book Edition
When Congress holds hearings about them you know that textbook prices have
become absurd. Simply applying the concepts in this subject and textbook it is
clear that e-books are the only viable answer. With this approach, prices are rea-
sonable and students can download the book anywhere in the world, 24-hours a
day. More important, students can keep the textbook for future reference. Yes,
some people still like paper. If you truly believe you prefer paper, go ahead and
print your own personal copy. But, before you kill a tree, check out the new tablet
PCs and the Sony e-book Reader or Kindle DX using the e-Ink technology.
Changes to the Sixth Edition
Updates
The sixth edition keeps the same overall structure as the ffth edition. All chapters
were revised and updated. The early chapters have more accurate and up-to-date
descriptions, along with more fgures describing the new technologies. All of the
data charts have been updated to the most current data. All examples are illus-
trated with Offce 2010 tools. There is a greater discussion of online tools. Almost
all of the Reality Bytes are new, and all cases have been updated with a few new
ones added. The data for Rolling Thunder has been improved and updated and
extended through 2011 to provide more realistic analyses. About one-third to one-
half of the exercises have been replaced or rewritten.
Newer topics have been added, including extensible business report language,
computer forensics, virtual machines, near-feld communications, telepresence,
augmented reality, e-discovery, bandwidth hogs, and search engine optimization.
A new section on cloud computing was added to each chapter. Cloud comput-
ing has been around for a few years, but more options and tools are available and
it is becoming popular, so managers need to understand the strengths and weak-
nesses of cloud computing. With the continued emphasis on mobile computing
using smartphones and tablets, cloud computing becomes more important.
Focus on Business Problems
MIS is a business class, so students need to begin with the business perspective
and then see how technology can solve the problems. The chapters emphasize this
focus by beginning with a series of questions. Specifc, real-world cases highlight
aspects of these problems so that students see the problems in context. Almost all
of the main cases have been replaced or updated. The chapter then addresses the
questions and demonstrates how technology can be used to solve the problems.
Each chapter begins with a set of questions that will be addressedproviding stu-
dents with a direct learning objective. The questions also serve to increase student
curiosity, by raising issues that they are not likely to have considered before. Sev-
eral chapters have been restructured and the overall book has been reorganized to
emphasize the business focus.
xii Preface
Technology Toolboxes
More details of technologies have been added. The goal is to build a stronger
bridge between the underlying technologies and hands-on tools to business appli-
cations and strategies. Students should be encouraged to understand the technical
as well as the business issues, and they should at least experiment with the tech-
nologies in the book. Chapters 2 and 3 include more technical details in particular.
Collaboration and Teamwork
Some of the biggest changes in software and technology in the last few years have
been designed to improve collaboration and teamwork. It is critical that manag-
ers be able to use these tools, and they are not taught in typical introductory tools
courses. This text covers these tools in Chapter 8 and uses Technology Toolboxes
to demonstrate the basic features. In addition, each chapter contains teamwork ex-
ercises designed for collaboration by a group of students. Ideally, students would
be able to use the collaboration tools while participating in these exercises.
New Cases and Reality Bytes
Most of the chapter cases and Reality Bytes examples have been replaced or re-
written. The book contains over 100 new Reality Bytes cases. All of the end-of-
chapter industry cases have been updated and are tied to the specifc problems in
each chapter.
Instructional Support
Instructors have access to the following resources for course presentation and
management. All the instructor supplements were created by the author, except
the test bank:
Instructors Manual includes answers to all end-of-chapter review questions,
exercises, and teaching notes for the industry-specifc cases. Teaching tips
and ties to the PowerPoint slides are included for each chapter.
A test bank contains true/false, multiple choice, and short answer questions.
Lecture notes are available as slide shows in Microsoft PowerPoint for-
mat. The slides contain all of the fgures along with additional notes. The
slides are organized into lectures and can be rearranged to suit individual
preferences.
Several databases and exercises are available online. The instructor can add
new data, modify the exercises, or use them to expand on the discussion in
the text.
The Rolling Thunder database application is available in Microsoft Access
format. It is a self-contained application that illustrates many of the concepts
and enables students to examine any facet of operating a small company.
The Instructors Manual includes further guidance on how to incorporate
this innovative tool into your course.
The books Web site at JerryPost.com provides resources for instructors and
students using the text.
xiii Preface
Te Online System
To students, the most important change in the ffth and sixth edition is placing
everything online. The main reason for this move is cost. The main textbook was
rewritten and expanded. The costs and student prices for the print book was out of
line. Electronically, it is possible to make the books available for an almost trivial
fee.
E-books provide additional benefts, including advanced search capabilities.
Students can also set bookmarks and highlight sections. More importantly, they
get to keep the books, instead of being forced to sell them back at the end of the
term. Database application development is an important topic, and the examples,
comments, and tips in the books will be valuable to students throughout their
careers.
xiv
Brief Contents
1 Introduction
Part One: Information Technology
Infrastructure
2 Information Technology Foundations
3 Networks and Telecommunications
4 Database Management
Part Two: Business Integration
5 Computer Security
6 Transactions and Entreprise
Relationship Planning
7 Electronic Business
8 Teamwork
Part Three: Decisions and
Strategies
9 Business Decisions
10 Strategic Analysis
11 Entrepreneurship
Part Four: Organizing Businesses
and Systems
12 Systems Development
13 Organizing MIS Resources
14 Information Management and Society
xv Contents
Contents
Managers and Information Technology, v
Learning Assessment, v
Organization, vi
Features That Focus on Solving Problems, viii
Goals and Philosophy, x
This Book Is Different from Other Texts, x
E-Book Edition, xi
Changes to the Sixth Edition, xi
Updates, xi
Focus on Business Problems, xi
Technology Toolboxes , xii
Collaboration and Teamwork, xii
New Cases and Reality Bytes, xii
Instructional Support, xii
The Online System, xiii
Introduction, 1
Introduction, 2
The Scope of MIS, 4
The Importance of Information Technology, 5
Productivity, 6
Teamwork and Communication, 7
Business Operations and Strategy, 7
The Role of the Internet in Business, 8
Technology Excesses, 9
So You Want to Be a Manager, 11
Traditional Management and Observations,
11
Making Decisions, 13
Business and Technology Trends, 14
Changes in Organizational Structure, 15
Business Trends, 16
Reengineering: Altering the Rules, 25
Management and Decision Levels, 27
Operations, 28
Tactics, 29
Strategy, 31
An Introduction to Strategy, 32
Searching for Ideas, 32
Strategy Example: Baxter Healthcare, 33
Cloud Computing, 35
Summary, 35
Key Words, 36
Web Site References, 37
Review Questions, 38
Exercises, 38
Additional Reading, 41
Cases: The Fast-food Industry, 41
Information Technology Infrastructure,
64
Information Technology Foundations,
65
Introduction, 66
Types of Data, 69
Object Orientation, 70
Numbers and Text, 72
Pictures, 74
Color, 78
Style Sheets, 79
Sound, 81
Video, 82
Size Complications, 82
Data Compression, 83
Hardware Components, 84
Processors, 86
Input, 92
Output, 95
Secondary Storage, 95
Operating Systems, 97
Computers in e-Business, 99
What Is a Browser?, 100
What Is a Server?, 102
Virtual Machines, 104
Application Software, 104
Research: Databases, 105
Analysis: Calculations, 105
Communication: Writing, 107
Communication: Presentation and Graphics,
107
Communication: Voice and Mail, 108
Organizing Resources: Calendars and
Schedules, 108
Augmented Reality, 109
The Paperless Offce?, 110
Open Software, 110
Cloud Computing, 111
Summary, 113
Key Words, 114
xvi Contents
Exercises, 116
Cases: The Computer Industry, 121
Networks and Telecommunications, 132
Introduction, 133
Network Functions, 134
Sharing Data, 135
Sharing Hardware, 142
Sharing Software, 144
Components of a Network, 145
Computers, 145
Transmission Media, 147
Connection Devices, 157
Network Structure, 159
Network Packets, 159
Shared Connections, 160
Managing Shared Networks, 163
Standards, 166
The Need for Standards, 166
A Changing Environment, 167
Internet TCP/IP Reference Model, 167
The Internet, 171
How the Internet Works, 172
Video and Heavy Use Complications, 174
Telephones and The Internet, 175
Internet Addresses, 177
Domain Names, 179
Internet Committees, 181
Internet 2, 181
Wireless Networks and Mobile Commerce,
183
Global Telecommunications, 186
Technical Problems, 186
Legal and Political Complications, 187
Cultural Issues, 188
Summary, 188
Key Words , 190
Exercises, 191
Cases: Wholesale Suppliers, 196
Database Management, 210
Introduction , 211
Relational Databases, 213
Tables, Rows, Columns, Data Types, 214
Data Quality, 215
Data Integrity, 216
Multitasking, Concurrency, and Integrity, 217
Data Volume, 218
The Database Management Approach, 218
Focus on Data, 219
Data Independence, 219
Data Integrity, 219
Speed of Development, 220
Control over Output, 221
Queries, 221
Single-Table Queries, 222
Different Types of Conditions, 230
Computations, 233
Computing Subtotals, 238
Joining Multiple Tables, 241
Examples, 241
Multiple Tables, GROUP BY, and WHERE,
243
Views, 247
Converting Business Questions to Queries,
248
Designing a Database (optional), 250
Notation, 252
First Normal Form, 252
Second Normal Form, 254
Third Normal Form, 254
Database Applications, 257
Data Input Forms , 257
Reports, 259
Putting It Together with Menus, 261
Database Administration, 262
Standards and Documentation, 263
Testing, Backup, and Recovery, 264
Access Controls, 264
Databases and e-Business , 264
Summary, 268
Key Words, 269
xvii Contents
Exercises, 270
Cases: Pharmaceuticals, 276
Business Integration, 292
Computer Security, 293
Introduction, 294
Threats to Information, 295
Disasters, 297
Employees and Consultants, 299
Business Partnerships, 301
Outsiders, 302
USB Drives, 303
Threats to Users, 303
Virus/Trojan Horse , 304
Phishing: Fake Web Sites, 306
Updates and Patches, 307
Intercepted Data, 308
Computer Security Controls, 310
User Identifcation, 311
Access Control, 313
Administrator Access Rights, 314
Single Sign-On and Lack of Standards, 314
Encryption, 315
Single Key, 316
Public Key Infrastructure, 318
Additional Security Measures, 322
Audits, 322
Physical Access, 323
Monitoring, 323
Hiring and Employee Evaluation, 325
Computer Forensics, 325
E-Commerce and Cloud Computing Security
Issues, 326
Theft of Data from Servers, 327
Firewalls, 328
Intrusion Detection and Prevention, 329
Denial of Service, 330
Privacy, 332
Summary, 333
Key Words, 334
Exercises, 335
Cases: Professional Sports, 339
Transactions and Enterprise Resource
Planning, 350
Introduction, 352
Data Capture, 353
Point of Sale, 354
POS Advantages, 358
Process Control, 359
Electronic Data Interchange (EDI), 360
Extensible Business Reporting Language
(XBRL), 364
Elements of a Transaction, 366
Vendor Perspective, 366
Customer Perspective, 366
Transaction Fees, 367
Government Perspective, 368
International Issues, 369
The Role of Accounting, 370
Input and Output: Financial Data and Reports,
371
Purchases, Sales, Loans, and Investments,
372
Inventory, 372
The Accounting Cycle, 373
Process: Controls, Checks, and Balances, 375
Human Resources and Transaction Processing,
375
Input: Data Collection, 376
Output: Reports, 377
Process: Automation, 377
Production Management, 378
Production Issues, 379
Distribution and Inventory Control, 380
Integration in Business, 382
Enterprise Resource Planning, 385
International Environment, 387
Financial Accounting, 387
Logistics, 388
Human Resource Management, 388
Integration, 389
Supply Chain Management, 390
SCM Changes the Focus, 391
SCM Challenges, 392
Integration across Systems, 394
Customer Relationship Management, 395
xviii Contents
Multiple Contact Points, 395
Feedback, Individual Needs, and Cross
Selling, 396
CRM Packages, 397
Summarizing ERP Data, 398
Digital Dashboard and EIS, 399
How Does an EIS Work?, 400
Advantages of an EIS, 401
Transaction Accuracy: Sarbanes-Oxley, 402
Summary, 404
Key Words, 406
Exercises, 407
Cases: Automobile Industry, 411
Electronic Business, 428
Introduction, 429
Selling Products Online, 432
Business to Consumer, 433
Business to Business, 438
The Production Chain and Disintermediation,
440
Three Stages of a Purchase, 443
Price Competition, 444
Services and Web 2.0, 446
Social Networking and Consumer Services,
446
Digital Products, 448
Business Services, 449
Network Services, 451
Search Engines, 454
Payment Mechanisms and Risk Mitigation in
e-Commerce, 454
Payment Risks, 455
Payment Mechanisms, 456
Advertising, 460
Traditional Media and Name Recognition, 461
Web Advertisements, 461
Privacy, 466
Developing and Hosting Web Sites, 467
Simple Static HTML, 468
Web Auctions, 470
Amazon WebStore, 471
Auction versus WebStore, 472
Web Commerce Servers, 472
Content Management Systems, 473
Application Service Providers, 474
Web Hosting Summary, 474
Mobile Commerce, 476
Taxes, 477
Global Economy, 478
Summary, 480
Key Words, 481
Exercises, 482
Cases: Retail Sales, 485
Teamwork, 505
Introduction, 506
Communication, 507
e-Mail, Instant Messaging, and Voice Mail, 509
Web Pages and Blogs, 511
Scheduling and Project Management, 512
Conferences and Meetings, 513
Peer-to-Peer Network Complications, 515
Mobile Systems, 516
Collaboration, 516
Documents and Changes, 517
Version Control, 518
Wikis and Online Applications, 520
Information Rights Management, 521
Workfow, 522
Group Decision Support Systems, 523
Knowledge Management, 525
Organizational Memory, 526
Service Processes, 527
Microsoft SharePoint, 528
Communication and Scheduling, 528
Collaboration, 532
Workfow, 535
SharePoint Confguration, 537
Peer-to-Peer Collaboration, 540
Sharing Documents, 540
Immediate Interaction: Microsoft SharedView,
542
xix Contents
Sharing with Multi-Touch Panels, 544
File Formats and Standards, 545
Cloud-Based Collaboration, 546
Summary, 549
Key Words, 550
Exercises, 551
Cases: Package Delivery, 555
Decisions and Strategies, 575
Business Decisions, 576
Introduction, 577
It is Hard to Make Good Decisions, 578
Human Biases, 581
Models, 584
Data Warehouse, 589
Online Analytical Processing (OLAP), 590
Decision Support System, 594
Marketing Forecasts, 594
Human Resources Management, 596
Geographical Information Systems, 597
Maps and Location Data, 599
Example, 601
Data Mining, 602
Expert Systems, 606
Specialized Problems, 608
Diagnostic Problems, 608
Speedy Decisions, 609
Consistency, 610
Training, 611
Building Expert Systems, 612
Knowledge Base, 613
Rules, 613
Creating an ES, 614
Limitations of Expert Systems, 615
Management Issues of Expert Systems, 616
Specialized Tools, 616
Pattern Recognition and Neural Networks, 617
Machine Vision, 619
Language Comprehension and Translation,
620
Robotics and Motion, 621
Machine Intelligence, 623
DSS, ES, and AI, 623
The Importance of Intelligent Systems in
e-Business, 625
Agents, 625
Support and Problem-Solving Applications,
627
Summary, 628
Key Words, 630
Exercises, 631
Cases: Financial Services Industry, 636
Strategic Analysis, 653
Introduction, 654
The Competitive Environment, 656
External Agents, 658
Buyers, 658
Suppliers, 660
Rivals, New Entrants, and Substitutes, 660
Government Regulations, 662
IS Techniques to Gain Competitive Advantage,
662
Barriers to Entry, 662
Distribution Channels, 664
Switching Costs, 666
Lower Production Costs, 667
Product Differentiation and New Products, 667
Quality Management, 669
The Value Chain, 670
The Search for Innovation, 671
Research, 672
Engineering and Design, 673
Manufacturing, 673
Logistics and Supply, 674
Marketing, 674
Sales and Order Management, 676
Service, 676
Management, 676
Costs and Dangers of Strategies, 678
High Capital Costs, 678
When the Competition Follows, 679
Changing Industry, 679
Sharing Data, 680
xx Contents
Government Intervention, 681
The Role of Economics, 682
Challenges of Strategy, 683
Sustainability, 685
Leader or Follower, 686
Change Agents, 688
Summary, 690
Key Words, 691
Exercises, 692
Cases: The Airline Industry, 696
Entrepreneurship, 715
Introduction, 716
Small Business Constraints, 717
Money, 718
Workers and Specialists, 720
Information Technology Expertise, 721
Strategic Power, 723
Information Tasks, 724
Operations and Transactions, 724
Tactics and Decisions, 727
Strategies, 728
Selecting Technology Levels, 729
Managing Consultants, 730
Internationalization, 731
Idea, 732
Strategy, 732
Research, 734
Plan, 736
Strategy, Competition, and Market Analysis,
738
Forecasts, Cash Flow, and Investment
Budget, 738
Marketing, 740
Online Advertising, 741
Organization and Timetable, 742
Implementation, 742
Ownership Structure, 743
Financing, 744
Accounting and Benchmarks, 747
Starting an e-Commerce Firm, 748
Analysis of Dot-Com Failures, 749
Pure Internet Plays, 749
Proft Margins, 750
Advertising Revenue, 751
Summary, 752
Key Words, 754
Exercises, 755
Cases: Entrepreneurship, 758
Organizing Businesses and Systems, 771
Systems Development, 772
Introduction, 773
Building Information Systems, 776
Custom Programming, 777
Outsourcing and Contract Programmers, 778
Assemble Applications from Components, 779
Purchase an External Solution, 780
Computer Programming, 781
Programming Logic, 781
Events, 783
Object-Oriented Programming, 785
From Programming to Development, 787
Systems Development Life Cycle, 788
Introduction to SDLC, 789
Feasibility and Planning, 789
Systems Analysis, 791
Systems Design, 792
Systems Implementation, 793
Maintenance, 796
Evaluation, 797
Strengths and Weaknesses of SDLC, 798
Alternatives to SDLC, 799
Prototyping or Iterative Development, 800
Extreme Programming and Agile
Development, 800
Developing Systems Requires Teamwork:
JAD and RAD, 802
Communication, 806
Object-Oriented Design, 806
Open Source Development, 807
End-User Development, 808
Development Summary, 809
xxi Contents
Process Analysis, 810
Input, Process, Output, 810
Divide and Conquer, 812
Goals and Objectives, 813
Diagramming Systems, 814
Summary: How Do You Create a DFD?, 818
Distributed Services, 819
Summary, 821
Key Words, 823
Exercises, 824
Cases: Government Agencies, 829
Organizing MIS Resources, 860
Introduction, 861
Managing the Information Systems Function, 862
Managing Workers, 863
Planning and Integration, 864
Evaluation, Oversight, and Control, 865
MIS Roles, 867
Hardware Administration, 867
Software Support, 868
Network Support, 869
Software Development, 869
Support for End-User Development, 869
Corporate Computing Standards, 870
Data and Database Administration, 871
Security, 872
Advocacy Role, 873
MIS Jobs, 873
Outsourcing, 875
MIS Organization: Centralization and
Decentralization, 880
Hardware, 884
Software and Data, 886
Personnel, 888
Recentralization with Cloud Computing, 889
Networks, 890
Hardware, 893
Data, 893
Confict Management, 894
Summary, 896
Key Words, 897
Exercises, 898
Cases: The Energy Industry, 903
Information Management and Society,
916
Introduction, 917
Individuals, 918
Privacy, 919
Privacy Laws and Rules, 930
Anonymity, 931
Jobs, 932
Loss of Jobs, 932
Physical Disabilities, 934
Telecommuting, 934
Business: Vendors and Consumers, 936
Intellectual Property, 936
Digital Rights Management, 940
Balance of Power, 942
Education and Training, 944
Social Interactions, 945
Social Group Legitimacy, 945
Access to Technology, 948
e-Mail Freedom, 949
Liability and Control of Data, 949
Government, 950
Government Representatives and Agencies,
950
Democracy and Participation, 950
Voting, 951
Information Warfare, 953
Rise of the World-State?, 955
World Government Cases, 956
Crime, 959
Police Powers, 959
Freedom of Speech, 960
Gambling, 961
Responsibility and Ethics, 962
Users, 962
Programmers and Developers, 963
Companies, 964
Governments, 964
xxii Contents
Some Computer-Related Laws, 965
Property Rights, 965
Privacy, 968
Information Era Crimes, 971
Data in Multiple Countries, 973
Threats to Shared Servers, 974
Subcontractors, 975
Summary, 976
Key Words, 977
Exercises, 979
Cases: Health Care, 983
1
What You Will Learn in This Chapter
How can MIS help you in your job?
What is MIS?
Why is information technology important? Why do all business majors need to study it?
How important is the Internet in retail sales?
Do you have too much technology?
Do you know what a manager does?
Do you know what a successful manager will do in the future?
How is business changing? What will managers need to know in the future?
Does technology alone improve a business?
How do you break businesses into smaller pieces to analyze them?
Why are strategic decisions so diffcult?
How do you begin searching for competitive advantage?
Chapter Outline
Introduction
1
Chapter
Introduction, 2
The Scope of MIS, 4
The Importance of Information Technology,
5
Productivity, 6
Teamwork and Communication, 7
Business Operations and Strategy, 7
The Role of the Internet in Business, 8
Technology Excesses, 9
So You Want to Be a Manager, 11
Traditional Management and
Observations, 11
Making Decisions, 13
Business and Technology Trends, 14
Changes in Organizational Structure, 15
Business Trends, 16
Reengineering: Altering the Rules, 25
Management and Decision Levels, 27
Operations, 28
Tactics, 29
Strategy, 31
An Introduction to Strategy, 32
Searching for Ideas, 32
Strategy Example: Baxter Healthcare,
33
Cloud Computing, 35
Summary, 35
Key Words, 36
Exercises, 38
Cases: The Fast-food Industry, 41
2 Chapter 1: Introduction
McDonalds
Introduction
How can MIS help you in your job? This is the ultimate question that
you must continually ask and answer as you pursue a management career. This
book explores many variations of this question and some useful answers. Infor-
mation technology offers two main features to managers: (1) productivity and (2)
innovation.
Productivity is the ability to accomplish more with fewer resources; and tech-
nology is often used to reduce costs. On a personal level, you are probably famil-
iar with the basic personal tools such as a word processor, e-mail, spreadsheet, and
Internet searches. All of these technologies improve your individual productivity.
What do customers want? McDonalds Corporation has sold billions of hamburgers.
Beginning in 1955 with a single drive-in in Des Plaines, Illinois, McDonalds has
grown to todays system of more than 25,000 restaurants across 115 countries. As
a brand, McDonalds is synonymous with a quality product at a reasonable price.
Equally important, McDonalds markets itself as more than a place to get a ham-
burger. Ronald McDonald, Happy Meals, the clean restaurants, and each new prod-
uct or promotional theme add to the fun that brings more than 40 million customers
of all ages to its restaurants around the world each day.Eighty percent of worldwide
McDonalds are franchiseyd. Each restaurant must meet strict requirements to make
it the same as all others. This ensures that each time you drive or walk into a McDon-
alds, no matter where you are, the Big Mac that you order will always be the same
taste, size, weight, and quality. It will also be competitively priced.
Eighty percent of worldwide McDonalds are franchised. Each restaurant must
meet strict requirements to make it the same as all others. This ensures that each time
you drive or walk into a McDonalds, no matter where you are, the Big Mac that you
order will always be the same taste, size, weight, and quality. It will also be competi-
tively priced.
Legal contracts, quality standards, and performance specifcations direct the in-
dividual restaurants in the effort to keep all the food orders the same. What most
individuals do not think about when they walk or drive into McDonalds is that Mc-
Donalds management information system (MIS) plays a critical role in ensuring
the quality and consistency of each sandwich. McDonalds Corporation maintains
a strict requirement that food be fresh and not stored more than a limited amount of
time. MIS applications direct managers in the management of employees and the
ordering and tracking of hamburgers, buns, potatoes, and soft drinks.
Competition is ferce in the fast food industry. The top chains have been strug-
gling over the last several years to fnd a way to attract new customers. McDonalds
experimented with changing menus. It altered its in-store system to focus on the
Made for You campaign. Concerns about healthy food have led the company to try
new foods and new ways of cooking. Following in the footsteps of Starbucks, Mc-
Donalds is offering wireless Internet access to customers for a fee. But is that what
McDonalds customers really want? Will you go to McDonalds, pay money for a
connection to the Internet, and then hang around to buy more food?
In the business world, these tools have largely eliminated the personal assistant,
reducing costs and improving the response time. But, MIS goes way beyond these
basic tools and the individual. As you will see, several technologies exist to help
teams and the entire organization improve productivity.
Innovation is the creation of new thingsnew services and products, new mar-
kets, new business methods, and even new industries. Instead of cutting costs,
information technology is used to open up new sources of revenue for business-
es. Think about companies like Google (and the frms it bought), Amazon, and
Match.com. These companies used information technology to create new ways to
make money. Although less radical, many existing frms have found new ways to
expand markets and make money with technology.
Innovation is sexier and more fun, but fnding innovative ideas is considerably
harder than using technology to increase productivity and reduce costs. These is-
sues are part of the challenge of business strategy. Leaders and managers need to
choose how their organization will view technology. Innovative uses require cre-
ativity and cost money, but can lead to gains against the competition.
As a manager, how can you use information technology to improve your busi-
ness, increase sales, and gain an edge on your competitors? Keep in mind that
technology changes rapidly. How do you know what tools to buy? Is it worth the
money and risk to buy the latest technology? Can you rebuild your company to
use the technologies? What do the customers want and how will they respond to
technology? These are questions that managers face every day. The questions are
challenging and the answers are hard to fnd. This book lays out a framework for
analyzing your business problems and evaluating technology solutions.
Trends
Economies, businesses, people, and societies all change over time. A cen-
tury ago, people were farmers and laborers. Most businesses were small.
Technologies changed and people moved to cities and manufacturing jobs.
Technology changed again, and the importance of the service sector grew.
Although digital computers were invented in the early 1940s, it was not until
the 1960s and 1970s that they became affordable for most businesses. At
frst, most companies used information technology to solve easy problems:
transaction processing. Before computers, companies needed hundreds of
bookkeepers in the back offce tracking sales, updating ledgers, and sum-
marizing data by hand. As hardware prices dropped in the 1980s and 1990s,
businesses integrated personal computers into management. Networksand
today, the Internetmade it easier to share data and communicate. Informa-
tion technology not only changes jobs, but also changes the way companies
are organized. If you want a job in business, you need to know how to use
these technologies to become a better manager and to improve the business.
Remember that technology can perform amazing tasks. Businesses no lon-
ger need people to perform many of the menial middle-management tasks.
Instead, businesses need managers who can think, adapt, and creatively fnd
new solutions.
Te Scope of MIS
What is MIS? You probably have some experience with using computers and
various software packages. Yet computers are only one component of a manage-
ment information system (MIS). As shown in Figure 1.1, an MIS consists of fve
related components: hardware, software (applications), people, procedures, and
collections of data. More importantly, an MIS is designed to solve problems for
the entire business. The term information technology (IT) represents the various
types of hardware and software used in an information system, including comput-
ers and networking equipment.
The physical equipment used in computing is called hardware. The set of in-
structions that controls the hardware is known as software. In the early days of
computers, the people directly involved in MIS tended to be programmers, design
analysts, and a few external users. Today, almost everyone in the frm is involved
with the information system, but some specialized MIS employees are needed to
create and manage the information systems. Procedures are instructions that help
people use the systems. They include items such as user manuals, documentation,
and procedures to ensure that backups are made regularly. Databases are collec-
tions of related data that can be retrieved easily and processed by the computers.
As you will see in the cases throughout the book, all of these components are vital
to creating an effective information system.
So what is information? One way to answer that question is to examine the use
of information technology on three levels: (1) data management, (2) information
Figure 1.1
MIS Components. Information technology consists of the hardware and software,
which includes networks. Information systems exist to process data. It takes people
and procedures to create and run an information system. Information systems need to
support users to help them make decisions.
Backup data
Restart job
Virus scan
Hardware
Software
People
Procedures
Data
systems, and (3) knowledge bases. Data consists of factual elements (or opinions
or comments) that describe some object or event. Following the comments by
Max Hopper (an early IT visionary at American Airlines) Data can be thought of
as raw numbers, text, images, or even video. Data management systems focus on
data collection and providing basic reports. Information represents data that has
been processed, organized, and integrated to provide insight. Information systems
are designed to help managers analyze data and make decisions. From a decision
makers standpoint, the challenge is that you might not know ahead of time which
information you need, so it is hard to determine what data you need to collect.
Knowledge represents a higher level of understanding, including rules, patterns,
and decisions. Knowledge-based systems are built to automatically analyze data,
identify patterns, and recommend decisions. Humans are also capable of wisdom,
where they put knowledge, experience, and analytical skills to work to create new
knowledge and adapt to changing situations. To date no computer system has at-
tained the properties of wisdom.
Information systems continue to change over time. Computer hardware has ad-
vanced rapidly for several decades. As you will see in Chapter 2, these changes
have enabled computers to handle increasingly complex data at faster speeds and
lower prices. Similar improvements in data storage and transmission (networks)
have opened up new options and methods of doing business. This increasing pen-
etration of technology in business continues to cause changes in how companies
operate. But, someone has to be in charge of the technology, evaluate the options,
and keep everything running. The role of the people managing the information
systems cannot be overemphasized. Many good careers exist in MIS and they are
explored in Chapters 12 and 13. Most of them require business knowledge and
skillsyou do not need to be a techie or geekand the pay has been relative-
ly good, so as you read the chapters, try picturing yourself in one of these roles.
Te Importance of Information Technology
Why is information technology important? Why do all business
majors need to study it? Productivity is the bottom-line issue in informa-
Reality Bytes: Data, Information, Knowledge, and Wisdom
Consider the case of a retail store that is trying to increase sales. Some of the data
available includes sales levels for the last 36 months, advertising expenses, and cus-
tomer comments from surveys. By itself, this data may be interesting, but it must be
organized and analyzed to be useful in making a decision. For example, a manager
might use economic and marketing models to forecast patterns and determine rela-
tionships among various advertising expenses and sales. The resulting information
(presented in equations, charts, and tables) would clarify relationships among the
data and would be used to decide how to proceed.
It requires knowledge to determine how to analyze data and make decisions. Edu-
cation and experience create knowledge in humans. A manager learns which data to
collect, the proper models to apply, and ways to analyze results for making better
decisions. In some cases, this knowledge can be transferred to specialized computer
programs (expert systems).
tion technology. Workers are expensive, diffcult to manage, and hard to remove.
Information technology continues to decline in price with increased power and
capabilities. Programmers are more creative, and new tools are being developed
to solve more complex problems. Organizations that use information technology
to reduce costs and provide better service can thrive. If you want a job in manage-
ment, you must learn to use the technologies. But it is not as simple as knowing
how to use a spreadsheet and a word processor. You have to understand how busi-
nesses work and determine how technology can be used to improve the business.
Productivity
Figure 1.2 shows the productivity index for U.S. non-farm workers. In the seven
years from 2000-2006, productivity grew at an average annual rate of 2.9 percent,
but it was closer to 5 percent at the start of the century. Over about 15 years, pro-
ductivity increased by about 47 percent. The other way to look at productivity is
that each year, a frm needs fewer workers to produce the same level of service
which means it can operate at lower costs. Basically, the only way to achieve this
continued growth in productivity is for workers to become more effcienteither
through learning or improved use of technology.
As a manager, it is critical that you understand two implications of this trend.
First, you have to use information technology to stay competitive. Today, comput-
ers are almost always cheaper than people. Second, at an average productivity rate
of about 3 percent per year, productivity will increase by 34 percent in a decade.
Firms would be able to produce the same output with 34 percent fewer workers.
What happens to those workers? Chapter 14 addresses the economic issue in more
Figure 1.2
Productivity is an index of the amount of work per employee. There is a 33
percent increase in output over the decade from 1996-2006, with an average
annual growth rate of 2.7 percent. Source: http://data.bls.gov/PDQ/servlet/
SurveyOutputServlet?request_action=wh&graph_name=PR_lprbrief.
0
20
40
60
80
100
120
1
9
9
4
-
Q
1
1
9
9
4
-
Q
4
1
9
9
5
-
Q
3
1
9
9
6
-
Q
2
1
9
9
7
-
Q
1
1
9
9
7
-
Q
4
1
9
9
8
-
Q
3
1
9
9
9
-
Q
2
2
0
0
0
-
Q
1
2
0
0
0
-
Q
4
2
0
0
1
-
Q
3
2
0
0
2
-
Q
2
2
0
0
3
-
Q
1
2
0
0
3
-
Q
4
2
0
0
4
-
Q
3
2
0
0
5
-
Q
2
2
0
0
6
-
Q
1
2
0
0
6
-
Q
4
2
0
0
7
-
Q
3
2
0
0
8
-
Q
2
2
0
0
9
-
Q
1
2
0
0
9
-
Q
4
2
0
1
0
-
Q
3
Output per Person
detail, but you are going to be better off if you are one of the workers who under-
stands and uses the technology to keep your job.
Teamwork and Communication
Teamwork is a key element in business. Teamwork means that tasks are divided
among team members. You are responsible for completing specifc items. You are
also responsible for helping the other members of the team to fnd the best solu-
tion to the problems. Teamwork requires cooperation, but with technology, it no
longer means that the teams have to work in the same room. Your team members
may participate from around the world. Even if everyone is in the same room, you
need technology to organize and share the contributions.
At a minimum, you already know how to use a word processor and a spreadsheet
to write reports and analyze data. Do you know what changes have been made
to these tools in the last few versions? The most important changes have been
support for collaboration tools that make it easier to share documents and work
as teams. Working on a team is diffcult. Not only do you have to produce your
own work, but you also have to communicate closely with the group to see what
problems and answers they have encountered and determine how to integrate all
of the pieces. Chapter 9 explores these problems and examines some of the tools
available to help you.
Business Operations and Strategy
Information technology is increasingly critical to the daily operations of a busi-
ness. Obviously, online businesses cannot live without technology, but neither can
the local grocery store, bank, or many other businesses. Computers process sales,
handle payments, and place new orders. They also analyze the sales data and help
set prices and predict trends. Information technology is also used to create new
Reality Bytes: Jobs
CareerCast.com a career Web site annually rates jobs on the basis of several factors
including income, stress, and job outlook. In 2011, based on a study of 200 jobs, the
top career was: software engineer. Jesse Severe is a software engineer in San Diego
working for ProFlight, LLC designing fight training software. He says My jobs
fexible, pays well and gives [me] a lot of job satisfaction. Typical salaries for soft-
ware engineers average about $87,000 up to $132,000. Many jobs exist for program-
mers across the U.S. and opportunities exist for entrepreneurs to develop their own
software and companies. One of the challenges with programming is that technol-
ogy, both hardware and software, continue to change. Consequently, the job requires
constant education. Mr. Severe observes that You cant think that youll learn a skill
and that it will be relevant in six months. Most people in the feld look on technol-
ogy change as a positive aspect. Tomorrow or next year there will be new tools and
new challenges. The top fve jobs on the list in 2011 were: (1) Software Engineer, (2)
Mathematician, (3) Actuary, ($) Statistician, (5) Computer Systems Analyst. Note
that all of them involve data and require analytical skills. The worst job on the list
was roustabout--typically handling entry-level work on oil rigs and pipelines.
Adapted from Adapted from Joe Light, The Best and Worst Jobs, The Wall Street
Journal, January 5, 2011.
products and services or to provide unique features to existing products. These
new features can give your company a strategic advantage and help the company
grow.
Te Role of the Internet in Business
How important is the Internet in retail sales? Today, almost every-
one is aware of the Internet and many of its opportunities. But, how important is
the Internet for sales? Electronic commerce, or e-commerce (EC), denotes the
selling of products over the Internet. These sales can be from a business to con-
sumers (B2C) or from one business to another (B2B). For a while in 1999, some
people believed that e-commerce would become the dominant form of business
where everyone bought all items over the Internet. Thousands of frms and Web
sites were created, trying to become the dominant frm in some niche. The group
was called dot-coms because almost all had an Internet address of something.
com. Many of the frms received huge amounts of funding from venture capital-
ists and experienced surprisingly high prices for their stock. Some foolish people
predicted a new economic world. But beginning in mid-2000, thousands of these
frms failed. Most had enormous expenses and huge losses. Many had been tak-
ing losses on every item they sold. And foolish people predicted the end of e-
commerce. Figure 1.3 shows the U.S. statistics collected by the census bureau.
The chart is somewhat misleading because EC sales are shown on the right-hand
y-axis, versus total retail sales on the left. These scales make it possible to see the
details in the EC data, but in reality, EC represents only about 5 percent of the
Figure 1.3
Retail and B2C e-commerce data in the United States. Although e-commerce sales
are only 5 percent of the total, the percentage has been steadily increasingdespite
the crash of the dot-coms in 2000 and 2001. Notice the seasonal peak in the fourth
quarteras more people go online to purchase holiday gifts. Source: http://www.
census.gov/retail/index.html#ecommerce.
0
10
20
30
40
50
60
0
200
400
600
800
1000
1200
1
9
9
9
-
4
2
0
0
0
-
2
2
0
0
0
-
4
2
0
0
1
-
2
2
0
0
1
-
4
2
0
0
2
-
2
2
0
0
2
-
4
2
0
0
3
-
2
2
0
0
3
-
4
2
0
0
4
-
2
2
0
0
4
-
4
2
0
0
5
-
2
2
0
0
5
-
4
2
0
0
6
-
2
2
0
0
6
-
4
2
0
0
7
-
2
2
0
0
7
-
4
2
0
0
8
-
2
2
0
0
8
-
4
2
0
0
9
-
2
2
0
0
9
-
4
2
0
1
0
-
2
2
0
1
0
-
4
R
e
t
a
i
l

B
i
l
l
t
o
n

$
U.S. Retail and E-Commerce Sales
Retail
Ecommerce
total retail sales. The main point of the chart is that although small, EC sales are
increasing at a faster rate than total sales. The second point to notice is the strong
seasonal effect in the salesparticularly in EC. Year-end holiday sales provide a
35 percent boost to the EC numbers. This large number indicates that EC is still
heavily used for gift and specialty items, instead of routine day-to-day purchases.
E-business is a more general term that encompasses e-commerce but also in-
cludes using the Internet for other business tasks, such as teamwork, communica-
tion, and new business services. As Internet technologies improve, more frms are
offering e-business servicessuch as digital maps, remote data backup, and sup-
ply chain management. Chapter 8 discusses e-commerce and e-business in more
detail.
Technology Excesses
Do you have too much technology? To some people (and organizations),
technology becomes a goal by itself. They strive to be the leaders in acquiring
technology and revel in the latest gadgets. The cell phone market is a leading
example. Check out the offerings of the major carriers, the constant changes, and
multitude of features. Now look around and notice that some people insist on hav-
ing the most recent, fanciest, and often largest cell phones. Perhaps it is the bling
factor, but do you really need to spend money to have the latest releases of tech-
nology? Bear in mind that most IT devices are most expensive when they are
released, with prices often dropping rapidly over time. Figure 1.4 illustrates the
common pattern in the release of new information technology. Prices decline over
time and a new model is released with additional features. The rate of decline
and the time frame (weeks, months, or years) depends on the specifc technology.
In all cases, you need to decide when you should purchase new technology. The
answer should depend on a careful assessment of your needs, the features offered,
Reality Bytes: You are Already Too Old for This Job
Digital advertising is a new frontier, for consumers, companies, and advertising
agencies. The big advertising agencies (Madison Avenue) have struggled to fnd ex-
perienced workers with knowledge of people who live and buy online. Jeff Tritt, an
HRM executive at Leo Burnett, a large Chicago-based marketing frm, noted that
The demand is greater than the supply so there is a big war for digital talent right
now. Some, such as WPP PLCs direct marketing unit Wunderman, are signing ap-
prenticeship agreements with schoolsgetting students to work for pay or credit.
Leo Burnett created a group of 35 young adults to function as a SWAT team to assist
on accounts that need advice concerning the digital world. In a different twist, JWT,
an ad frm of WPP created a reverse-mentor program where children ages 9 to 14 of
JWT employees were brought in to work on client projects. In a project for Nestle,
the kids feedback led to the development of a mobile game to be used to promote
items for the food giant. Beyond the children, the new workers not only need to
understand the technology, they have to be able to use it. Wunderman brought in
students to create a video to promote graphic design software from Microsoft. The
students were able to complete the task in three daysinstead of three weeks.
Adapted from Suzanne Vranica, Kids Lend a Digital Hand, The Wall Street Jour-
nal, January 10, 2011.
and your budget. This analysis is the same for computers, enterprise software,
and cell phones. But, with lower-priced items such as cell phones, the bling factor
(and marketing) can override the judgment of consumers.
Consider a simple cell-phone example. New cell phones are constantly being
introduced, and there might be a six-month (or year) lag in the release of new
technologies. If you purchase a new phone every six months, you will spend at
least $600 over the course of two years ($200 a phone but the frst one is often
free). If you keep the original phone and replace it every two years, you can gener-
ally get the phones free, or close to free. How many new features were introduced
over two years? Is it worth $300 a year (or more) to have the most recent phone?
Maybe it isparticularly when IT becomes a fashion accessory. But what about a
$2,000 laptop for business? Should every person get a new laptop every year? The
point is that these decisions depend on the individual. Until you need to evaluate
the purchases for an entire organization. If you need to make the same decision for
your company, you need to look at the true benefts to justify the additional costs.
To make those decisions, you need to understand the technology and how it adds
value to your frm. A key point in this book is that you need to begin thinking as a
managerresponsible for analyzing decisions for the entire frmnot just your-
self as an individual.
Another potential risk with information technology is that individuals might
isolate themselves and rely on e-mail and text messages, avoiding interperson-
al contact. Sometimes you need to meet face-to-face to get a true picture. Some
Figure 1.4
Information technology is often characterized by declining prices and new models.
You can lead everyone and spend large amounts of money to buy new models
whenever they are released, or you can keep technology until you truly need a new
model.
0
100
200
300
400
500
600
700
800
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
price
time
Typical Price Declines
communications are probably too important to use technology. Even in business,
you need to be aware of the different preferences of people and know when to use
technology and when to use a more personal approach.
So You Want to Be a Manager
Do you know what a manager does? Do you know what a suc-
cessful manager will do in the future? Answer: The good ones work
their tails off. But what exactly does a manager do? Managers actually have many
tasks, and the roles depend on the culture of the organization, the style of top
managers, and the skills of the individual manager. At one level managers are in
charge of other employees, so they must organize tasks, set schedules, direct the
processes, and communicate with other managers. On the other hand, managers
often have to help with the ongoing operations. For example, an accounting man-
ager also performs accounting tasks, and fnance managers continue to analyze
data. Finally, managers spend time communicating with others in the organiza-
tionboth formally and informally. You always need to know what is happening
in the company and how it will affect you.
Traditional Management and Observations
If you think of a traditional manager, you would consider tasks such as organiz-
ing work, planning jobs, and controlling workers. (Joe, go peel those potatoes.)
However, when observed at their jobs, managers appear to spend most of their
time in meetings, talking on the phone, reading or preparing reports, discussing
projects with their colleagues, explaining procedures, and participating in other
activities that are diffcult to ft into the traditional framework.
Henry Mintzberg, a psychologist who studies management, classifes manage-
rial tasks into three categories: (1) interpersonal, (2) informational, and (3) deci-
sional. Interpersonal roles refer to teaching and leading employees. Informational
Reality Bytes: Hidden Data
Managers need information to make decisionsand sometimes they spend consider-
able time searching for information. A study of 1,000 middle managers at large com-
panies by Accenture revealed that managers of American and English companies
spent an average of two hours a day in their searches. But then, over half the data
they fnd is worthless. Almost 60 percent said they had to use multiple sources to
fnd data, and they typically needed three different sources. Almost half (42 percent)
said they end up using the wrong data at least once a week. Much of the data, and
the problems, are stored within the company. Most of the problems arise between
managers. Almost half (40 percent) said that other managers are unwilling to share
information. But, the respondents did help matters. Most of them said they store their
most valuable information on their local computers and in their e-mail folders. It is
hard to blame the problem on searching skills. The IT workers surveyed were the
ones who most reported the data they found was worthless and they spent the most
time looking for it. As much as 30 percent of their time was spent searching for data.
Adapted from Justin Lahart, For Small Businesses, Big World Beckons, The Wall
Street Journal, January 26, 2011.
Technology Toolbox: Choosing a Search Engine
Problem: How do you fnd infor-
mation on the Internet? The Inter-
net is huge and getting bigger. It
contains an enormous number of
pages. Content is provided by mil-
lions of organizationseach using
different formats and terms.
Tools: Internet search engines
were created specifcally to crawl
the Web and capture key words
from the billions of pages they
fnd.
Choosing the right starting
point can be critical to a success-
ful search. Obtaining the most
number of hits is not as important
as a site that returns accurate re-
sults. The main search companies
constantly refne their methods to
improve the accuracy. Remember
the following hints when you are
searching:
1. Google is currently the most popular search site and often returns the most hits.
Bing and Yahoo are also good and sometimes the results are easier to read.
2. Some sites, such as Yippy, organize the results into categories.
3. Sometimes it is best to search an encyclopedia (wikipedia.org), dictionary (Wik-
tionary.com), a phone book (switchboard.com), the CIA World Factbook (www.
cia.gov/cia/publications/factbook), or government statistics (fedstats.gov).
4. Check specialist magazine sites for targeted information. For example, use for-
tune.com to fnd information on the largest companies.
5. The most accurate sources are not free. University libraries provide access to
huge commercial databases.
Quick Quiz: Where would you begin your search to answer the following ques-
tions?
1. Under the proposed IAU defnition, which planetoid falls between Mars and Ju-
piter?
2. By revenue, which was the largest company in the world in 2010?
3. Find the best price on a 512 GB SSD.
4. Which U.S. professional basketball team had the fewest wins in the 2010-2011
season?
5. Which celebrities are still alive? Sophia Loren, Harman Killebrew, Phyllis Diller,
Nancy Reagan, I.M. Pei.
General purpose search engines
Google, Bing, Yahoo
Meta-searches across multiple engines
Dogpile, Yippy
Encyclopedia
Wikipedia.org
Dictionary
Wiktionary
Phone book
Switchboard, Superpages
Products
Mysimon, Cnet
Government data
CIA.gov (World Factbook)
Fedstats.gov (main data source)
SEC.gov (EDGAR corporate flings)
Other
Your library databases
www.wolframalpha (math and data)
tasks are based on the transfer of information throughout the organization, such as
relaying information to subordinates or summarizing information for executives.
Decisions involve collecting information, evaluating alternatives, and choosing
directions that beneft the frm.
Other researchers have studied managers and developed alternative classifca-
tions. Fred Luthans uses three classifcations of management activities. He indi-
cates that approximately 50 percent of a managers time is spent on traditional
management activities (planning, organizing, and controlling), 30 percent in for-
mal communications, and 20 percent in informal networking. Formal communica-
tions include attending meetings and creating reports and memos. Informal net-
working consists of contacts with colleagues and workers that tend to be social in
nature but often involve discussions regarding business and jobs.
Making Decisions
In many ways managers expend a lot of their effort in making decisions or con-
tributing information so others can make decisions. When you look at courses
offered for future managers, you will fnd a focus on administration, human be-
havior, quantitative modeling and problem solving, decision theory, and elements
of business ethics and globalization. Typically, these courses are designed to help
managers solve problems and make decisions. However, if you ask managers how
much time they spend making decisions, they are likely to say that they seldom
make decisions. That seems like a contradiction. If managers and executives do
not make decisions, who does?
In many organizations, day-to-day decisions are embodied in the methodol-
ogy, rules, or philosophy of the company. Managers are encouraged to collect data
and follow the decisions that have resulted from experience. The managers are
Reality Bytes: Bicycles are Getting Expensive
Avid bicyclists have noticed huge increases in prices of high-end bicycles starting
around 2005. Lightweight bicycles with carbon fber or titanium frames could be
purchased for $3,000-$5,000. By 2007, similar bikes from top-line manufacturers
were commonly listed for $9,000-$10,000 or more. The pattern exists in road and
mountain bikes. News in 2006 indicated that prices would continue increasing
this time because of shortages of raw materials. In particular, carbon fber prices
increased by 25 percent. Demand for the material along with titanium skyrocketed
as airplane manufacturers increased production. Zsolt Rumy, CEO of carbon fber
manufacturer Zoltek Companies, Inc. in St. Louis said he was favoring his large cus-
tomers (airplane manufacturers) to keep them happy. He passed on most of the 60-
100 percent price increases to the sporting-goods customers, noting we really jack
up the price, for smaller customers. Titanium is relatively rare and in high demand,
driving prices to more than $32,000 a toncompared to $1,000 a ton for carbon
steel. Still, 19.8 million bicycles were sold in North America in 2005an 8.2 per-
cent increase from 2004. Manufacturers estimate that 30,000 cyclists a year spend
$3,000 or more on a bike.
Adapted from Paul Glader, Why Bike Prices Are Shifting Higher, The Wall Street
Journal, August 1, 2006.
directly involved in the decision process, even though they may not think they are
making the fnal choice.
The broader decision process involves collecting data, identifying problems,
and making choices. Making a decision also requires persuading others to accept
the decision and implement a solution. With this broader defnition, many of the
tasks performed by managers are actually steps in the decision process. Meet-
ings, phone calls, and discussions with colleagues are used to collect data, iden-
tify problems, and persuade others to choose a course of action. Each of these
steps may be so gradual that the participants do not think they are actually making
decisions.
Because of the subtlety of the process and the complexity of the decisions, it
is often diffcult to determine what information will be needed. Decisions often
require creativity. Because data generally need to be collected before problems
arise, it is challenging to design information systems to support managers and
beneft the organization. One important job of management is to examine the need
for information and how it can be used to solve future problems.
Business and Technology Trends
How is business changing? What will managers need to know
in the future? A key issue in management is that you must always work in the
future. If you spend all of your time running around trying to solve todays prob-
lems by putting out fres, you will never succeed. You must plan for tomorrow and
build the structure and processes to handle the day-to-day tasks.
Figure 1.5
In a traditional organizational structure, lower-level managers deal with customers
and collect basic data. Middle-level managers analyze the data, create reports,
and make suggestions to upper-level managers. The higher-level managers make
decisions and set rules to guide the other managers.
CEO
VP
Finance
VP
Marketing
VP
Accounting
VP
HRM
VP
MIS
Layers of middle managers
Customers
Commands
Analyze data
Condensed reports
Collect
data
Changes in Organizational Structure
Even without the Internet, management and companies are changing. The most
important change is the move away from the traditional hierarchical structure to a
team-based approach. Most of todays large companies developed years ago when
communications were limited and there were no computers. Most adopted a mil-
itary-inspired hierarchical command structure shown in Figure 1.5. The top-level
managers set policy and directed the vice presidents to carry out the mission of the
company. Sales staff dealt directly with customers, collected data, and passed it to
middle managers. The middle managers organized and summarized the data and
passed it up the chain. Little data was shared among the middle and lower levels.
In contrast, because it is easy to share data, information technology offers the
ability to alter the way companies are organized and managed. Figure 1.6 shows
the new approach. This method focuses on teamwork and a shared knowledge
of all relevant data. Some teams, like sales and accounting, will have ongoing
tasks. Other task forces will be formed to solve new problemsoften created
from managers across the company. Managers can expect to participate in many
teams, essentially at the same time. Data can be obtained and shared through the
information system, meetings can be held online, documents and comments can
be circulated electronically.
This structure enables companies to be run with a smaller number of managers.
Each manager is more productive because of the tools and the ability to perform
many jobs. Another strength of this approach is that it is easy to use consultants
Figure 1.6
In the last few years, many companies have moved toward a more decentralized
form of management. They have removed the middle layers of management and
replaced them with smaller teams. Franchises and smaller teams have become the
primary service contact with customers. Information sharing becomes crucial in this
environment. Teams communicate directly and share data across the company.
Management Team
CEO
Finance
Team
Marketing
Team
Accounting
Team
HRM
Team
Sales
Team
Franchise
Strategy
Methodology/Rules
Customers
Corporate
Database
&
Network
VP
Fin
VP
Mrkt
VP
Acct
VP
HRM
VP
MIS
and temporary workers for short-term projects. In todays legal climate, it is ex-
ceedingly diffcult to fre workers, so frms often use temporary workers for in-
dividual projects. Permanent workers, supplemented with specialized temporary
talent, can organize a team. The team disbands when the project is fnished.
Business Trends
As described in Figure 1.7, seven fundamental trends have been driving the econ-
omy and changing businesses: (1) specialization, (2) management by methodolo-
gy, (3) mergers, (4) decentralization and small business, (5) reliance on temporary
workers, (6) internationalization, and (7) the increasing importance of service-ori-
ented businesses. These trends will be discussed throughout the text to illustrate
how they affect the use of information systems and how managers can use infor-
mation systems to take advantage of these trends. Tightening job markets also
means that managers must continually work on self-improvement. To survive, you
must provide value to the organization.
Business Trend Implications for Technology
Specialization Increaseddemandfortechnicalskills
SpecializedMIStools
Increasedcommunication
Methodology and
franchises
Reductionofmiddlemanagement
Increaseddatasharing
Increasedanalysisbytopmanagement
Computersupportforrules
Reengineering
Mergers Fourorfvebigfrmsdominatemostindustries
Needforcommunication
Strategictiestocustomersandsuppliers
Decentralization
and small business
Communicationneeds
Lowercostofmanagementtasks
Lowmaintenancetechnology
Temporary workers Managingthroughrules
Findingandevaluatingworkers
Coordinationandcontrol
Personaladvancementthroughtechnology
Security
Internationalization Communication
Productdesign
Systemdevelopmentandprogramming
Salesandmarketing
Service orientation Managementjobsareinformationjobs
Customerservicerequiresbetterinformation
Speed
Figure 1.7
Changes occurring in the business world affect the use of information technology.
These trends and the implications are discussed throughout the book. Managers
who understand these trends and their relationship with technology will make better
decisions.
Specialization
Adam Smith described the advantages of specialization and division of labor in
manufacturing more than 230 years ago. The concepts are now being applied to
managers. As functional areas (such as marketing or fnance) become more com-
plex, they also become more specialized. Area managers are expected to under-
stand and use increasingly sophisticated models and tools to analyze events and
make decisions. As a result, the demand for managers with specifc technical skills
is increasing, while the demand for general business managers is declining. First
you get a job as an accountant (or whatever our specialty is), then you become a
manager. This trend is refected in MIS by the large number of specialized tools
being created and the increased communication demands for sharing information
among the specialists.
Management by Methodology and Franchises
Specializations advantage is that it reduces management tasks to smaller prob-
lems. Using specialization coupled with technology, frms have reduced many
management problems to a set of rules or standard operating procedures. Day-
to-day problems can be addressed with a standard methodology. For example,
the managers guidebook at Wal-Mart or McDonalds explains how to solve or
prevent many common problems. These rules were created by analyzing the busi-
ness setting, building models of the business, and then creating rules by anticipat-
ing decisions and problems. This approach gives less fexibility to the lower-level
managers but encourages a standardized product, consistent quality, and adher-
ence to the corporate philosophy.
Management by methodology also allows frms to reduce their number of mid-
dle managers. By anticipating common problems and decisions, there is no need
to call on trained managers to solve the daily problems. Franchises like McDon-
alds carry this technique one level further by making the franchisee responsible
for the fnancial performance of individual units. The common management tasks,
however, are defned by the central corporation.
Reality Bytes: American Workers Must be Crazy
The Center for Work-Life Policy surveyed 1,600 workers who earned more than
$75,000 a year. In 2007, almost half of those workers put in 60 hours a week and 10
percent were working more than 80 hours a week. Many of the workers (28 percent)
in banking and fnance had extreme jobs working more than 60 hours a week with
unpredictable demands, travel, and tight deadlines. Most of the workers putting in
over 60 hours a week said they loved their jobs. But half of those with extreme
jobs said they wanted to quit within a year. Still, in a different online survey, 29 per-
cent of 510 respondents said they would work more than 100 hours a week for their
dream job. Remember that a week has only 168 hours. In other studies, American
workers rarely use their full vacation allotment.
Adapted from Kyle Stock, Would You Work 100 Hours a Week for Your Dream
Job? The Wall Street Journal, March 29, 2011.
Technology Toolbox: Finding Government Data
Problem: Many business problems require data, particularly regarding demograph-
ics and the economy.
Tools: The federal government collects a huge amount of data; much of it is now
available online. The data is maintained by individual agencies and it helps if you
know what types of data are provided by each agency. However, the main site www.
fedstats.gov contains links to all of the agencies.
Agency Main Types of Data Site
Labor (BLS) Employment and Prices www.bls.gov/data
Census Demographic and maps www.census.gov
Economic Analysis (BEA) Economic summaries www.bea.gov
Transportation Statistics Airline, rail, and road www.bts.gov
Justice Statistics Crime and courts bjs.ojp.usdoj.gov
Economic Research (Ag) Food and farm economics www.ers.usda.gov
Health (CDC) Health and Healthcare www.cdc.gov.nchs
Securities and Exchange (SEC) Business flings www.sec.gov (EDGAR)
Most of the federal Web sites have the ability to download historical data, typi-
cally into spreadsheets. Some, as the BLS have search forms where you can select
exactly which data you want to download. At all of the sites it is important that you
read the descriptions carefully because there can be subtle but important differences
in the series.
In the past couple of years, the government agencies have opened their databases
to some of the search engines. In particular, Google and WolframAlpha can automat-
ically search the databases and retrieve the data. In most cases, the data is returned
as a chart, so if you truly want the underlying data, you might still have to go to the
government Web site. But sometimes it is helpful to use the search engines to fnd
the specifc agency.
Quick Quiz:
1. What was the U.S. monthly unemployment rate for the last year?
2. What is the current population of the U.S.?
3. What was the value of the U.S. trade defcit for the last year?
Merger Mania
Up to the late 1800s and early 1900s, most businesses were small, having markets
limited to small geographic regions. A brief history of industrial organization re-
veals four waves of mergers in the United States: (1) the horizontal mergers of the
late 1800s epitomized by the oil and banking industries; (2) the vertical integration
of the early half of the 20th century, illustrated by the oil, steel, and automobile
companies; (3) conglomerate mergers of the 1950s and 1960s, in which frms like
IT&T (an international telecommunications giant) acquired subsidiaries in many
different industries (including a bakery!); and (4) giant horizontal mergers at the
turn of the 21
st
century. All of these mergers arose to take advantage of economic
power, but technology made them possible. Without communication (telegraph
and telephones earlier, computer networks later), frms could not grow beyond a
certain size because managers could not monitor and control lower-level workers.
The most recent mergers have been impressive in terms of the size of the frms
and the sectors involved. The banking industry was one of the frst to begin con-
solidation. Relaxation of federal restrictions quickly led to large regional and na-
tional banks. The telecommunications industry also experienced several changes,
such as the ABC-Disney and AOL-Time/Warner merger between telecommunica-
tions and entertainment industries. Telephone, Internet, and cable companies also
were fertile ground for mergers, such as MCI and WorldCom or AT&T, TCI, and
the reconsolidation of the telephone companies. The horizontal mergers in the pe-
troleum, food production, automobile, and grocery industries represented major
consolidations of operations as well. Some of these combinations crossed interna-
tional boundaries (e.g., Daimler and Chrysler or Chrysler and Fiat). Some of these
trends were fueled by the high stock market valuations, which provided capital to
the successful frms and punished the weaker ones.
One of the important keys to these mergers was the improved capability of
information and communication technology. Without the IT structure, it would be
exceedingly diffcult to manage these combined frms. Most of the combinations
also resulted in a loss of middle-management jobs. The remaining workers relied
on technology to improve their productivity. The newly centralized frms also re-
lied on communication technology to provide customer service across the country
and around the world.
Decentralization and Small Business
Strangely, businesses today are becoming both larger and more decentralized. The
goal of decentralization is to push the decision-making authority down to the level
that deals with the customer. The top managers set strategy and establish projects
across the organization. The lower-level managers and salespeople solve prob-
lems and make decisions to improve sales and negotiate with customers. The gap
between these groups is bridged by information technology. Companies no longer
need hundreds of middle-level managers to organize data and interpret top man-
agement commands. In the past, with limited information technology, small divi-
sions were expensive to maintain because of the cost of collecting and processing
the basic accounting and operating data.
Within a frm, operations can be decentralized into teams of workers. In this
situation, departments operate relatively independently, selling services to other
departments by competing with other teams. They often perform work for outside
frms as wellessentially operating as an independent business unit within the
corporation. The main goal of decentralization is to push the decisions and the
work down to the level of the customer, to provide better customer service and
faster decisions. Information systems enable executives to gather and manipulate
information themselves or with automated systems. As a result, there is less need
for middle managers to prepare and analyze data.
At the same time that large business have consolidated and shed managers,
many of these people have started their own small companies. Sometimes work-
ers have become consultantsperforming tasks similar to the jobs they left, but
working part time for a variety of frms.
Temporary Workers
So what happens to the people who are no longer needed as middle-level manag-
ers? At various times in the past, some companies provided a form of lifetime em-
ployment for their workers. As long as workers continued to do their job and re-
mained loyal to the company, their jobs were secure. Even in more diffcult times,
when employees were laid off, they were often encouraged (through extensions
of unemployment benefts) to wait until the economy improved and they could be
rehired. Companies in other nations, especially Japan, had stronger commitments
to workers and kept them on the payroll even in diffcult times.
Today, in almost every industry and in many nations (including Japan), all jobs
are at risk. To compensate, companies increasingly rely on a temporary work-
force. Individuals are hired for specifc skills and tasks. When these basic tasks
are completed, the employees move on to other jobs. Increasingly, even execu-
tives are hired because of their specifc expertise. Consultants and other profes-
sionals are hired on a contract basis to solve specifc problems or complete special
assignments.
In many ways, it is more diffcult to manage a company that relies on temporary
workers. Special efforts must be made to control quality, keep employees work-
ing together, and ensure that contract provisions are met. Technology can play an
Reality Bytes: Visual Search
http://www.bing.com/visualsearch
http://image-swirl.googlelabs.com/
http://www.google.com/mobile/goggles/#text
http://images.google.com (Use Google Chrome)
Humans do not often remember items by name. People often visualize items and
remember things by association. So it would be useful to have search engines that
can fnd things based on images. The challenge is that computers are relatively weak
at image recognition. The code is getting better, but search results based on actual
photos are still relatively generic. Microsoft and Google both have tools to search
for collections of images based on word tags. The Microsoft Bing engine has some
useful collections, such as dog and cat breeds. Google tends to rely on images that
are tagged by other users. Google Goggles is an early version of a search engine that
uses an uploaded photo or a photo taken with a Smartphone to search for information
related to the object.
Adapted from Katherine Boehret, In Search Of...Images Worth 1,000 Results, The
Wall Street Journal, January 12, 2010.
important role in these situations. It can improve communications, maintain easy
(but controlled) access to data and contracts, and help to institute corporate stan-
dards. The Internet is beginning to play this management rolefnding contract
workers, negotiating the work, and distributing the fnished products.
To you as a worker, the loss of middle-management jobs and reliance on tem-
porary workers should be scary. It means more competition for jobsparticularly
higher-level careers. To obtain higher-level jobs, you will need to possess more
analytic skills than other potential employees. Even as a manager, you will need
your own competitive (professional) advantage. Along with additional education,
your use and knowledge of technology can give you an advantage.
The issue of contract workers is particularly critical with computer program-
mers and developers. Many frms want to hire contract programmers because they
will no longer be needed after an initial application is developed. This approach
works until the economy booms, then it becomes exceedingly diffcult to fnd
workers. In the end some companies fnd they have to hire permanent workers
simply to avoid the problems of trying to fnd new contractors every couple of
years.
Internationalization
Several events of the early 1990s demonstrated the importance of international
trade: closer ties forged with the European Union, creation of the North Ameri-
can Free Trade Area (NAFTA), and the continued relaxation of trade restrictions
through the General Agreement on Tariffs and Trade (GATT) and the World Trade
Organization (WTO). Although barriers to trade remain, there is no doubt that the
Reality Bytes: IT Spending
Forrester Research is a consulting frm that tracks trends and details in information
technology. The company regularly reports on IT spending. Note that software and
telecommunication services (phones and Internet) are the largest component of costs.
Although, combining hardware and network equipment into a single category makes
it about equal to the software costs.
Adapted from The Wall Street Journal, Where the Money Goes, April 25, 2011
international fow of trade and services plays an increasingly important role in
many companies. Even small frms are buying supplies from overseas and selling
products in foreign markets. Trade also brings more competition, which encour-
ages frms to be more careful in making decisions.
As Figure 1.8 shows, the role of exports and imports has expanded rapidly
in the United States since 1970. In European nations, international trade is even
more important. Today, internationalization is a daily fact of life for workers and
managers in almost every company. Even small businesses have links to frms in
other nations. Many have set up their own production facilities in other nations.
Much of this global expansion is supported by technology, from communication
to transportation, from management to quality control.
Communication facilities are one of the most prominent uses of information
technology to support the move to international operations. Communication tech-
nology is especially important for service industries such as consulting, program-
ming, design, marketing, and banking. Several years ago, services were often con-
sidered to be nontradable goods because they tended to have high transportation
costs, making them diffcult to export. Today, improved communication facilities
through the Internet have made certain types of services easy to export. For exam-
ple, fnancial institutions now operate globally. Today, software development has
a growing international presence. Many U.S. frms are turning to programmers in
Ireland, India, and Taiwan. Through the use of programmers in India, for example,
a U.S.-based frm can develop specifcations during the day and transmit them to
India. Because of the time difference, the Indian programmers work during the
U.S. night and the U.S. workers receive updates and fxes the next morning.
Figure 1.8
By almost any statistic, in almost every nation, the level of international trade has
increased dramatically during the last 20 years. International trade brings more
choices, more competition, more data, more complexity, and more management
challenges. Source: http://www.bea.gov/national/nipaweb/Index.asp.
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
P
e
r
c
e
n
t
Year
U.S. Trade Imports/GDP
Exports/GDP
Internationalization also plays a role in selling products. Groups of countries
have different standards, regulations, and consumer preferences. Products and
sales techniques that work well in one nation may not transfer to another culture.
Information technology can track these differences, enabling more fexible manu-
facturing systems that can customize products for each market.
Figure 1.9 shows one way to look at internationalization on the Web. It shows
the percentage of Web browsers confgured for the top ten languages in 2010.
English is still at the top of the list, but it is followed closely by Chinese. These
numbers are relatively accurate because servers can be confgured to recognize
the language of the client browser and the statistics are collected automatically.
Figure 1.10 shows a broader picture by counting the number of Internet users
in each geographical region. These numbers are somewhat more subjective be-
cause someone (typically the Nielsen agency) has to estimate the number of peo-
ple in each nation and region who use the Internet on a regular basis. This survey
approach can miscount peopleparticularly those who get access through work
or school. Still, it shows the international character of the Internet. It does not take
a high percentage of users for regions with large populations to dominate. Asias
825 million users is only about 30 percent of the people versus about 75 percent
for North America Do the statistics mean that you can create a Web site and begin
selling items to the hundreds of millions of Chinese and Indian Web users? Based
Figure 1.9
Internationalization of the Web. Percent of user languages estimated from Web
browser settings. Source: http://www.internetworldstats.com/stats7.htm (2010).
English, 27.3
Chinese, 22.6
Spanish,
7.8
Japanese, 5
Portuguese,
4.2
German, 3.8
Arabic, 3.3
French, 3
Russian, 3
Korean, 2
Other, 17.8
Web Users (Language)
on the fact that businesses have been attempting to expand into Asia for decades
since Nixons trade mission in 1973, it will probably take a while. Keep in mind
that per capita income is still lower in those nations, and much of the infrastruc-
ture developed for the Western world is not yet in place in Asiaincluding ship-
ping and payment methods.
The increased competition created by internationalization and decentralization
requires corporations to be more fexible. Flexibility is needed to adapt products
to different markets, choose suppliers, adopt new production processes, fnd inno-
vative fnancing, change marketing campaigns, and modify accounting systems.
Firms that attain this fexibility can respond faster to market changes, catch op-
portunities unavailable to slower frms, and become more proftable.
Service-Oriented Business
Another trend facing industrialized nations is the move toward a service-oriented
economy. As shown in Figure 1.11, in 1920 the U.S. census showed 29 percent of
the employed were in farming. By 2000, that fgure had fallen below 1 percent.
In the early 1900s, people were afraid that this trend would cause food shortages
throughout the United States and the world. Improvements in technology in the
form of mechanization, transportation, growing techniques, chemicals, and crop
genetics proved them wrong.
A similar trend in manufacturing has produced the same consternation. Al-
though the number of workers employed in manufacturing has varied over time,
it is clear that the largest increase in jobs has been in the management, clerical,
and service sectors. In 2010, 11 percent of the jobs were in manufacturing, with
Figure 1.10
Internationalization of the Web. Number of users by geographic region. Source:
http://www.internetworldstats.com/stats.htm (2010).
Asia, 825.1
Europe,
475.1
North
America,
266.2
Latn
America/
Caribbean,
204.7
Africa, 110.9
Middle East,
63.2
Oceania/
Australia,
21.3
Million
Users
88 percent in service and management jobs. The largest increase in new jobs has
been in the management, clerical, and service sectors.
These trends represent changes in the U.S. economy and in demographics such
as age characteristics of the population. The importance of the management, cleri-
cal, and service sectors has to be considered when examining how MIS can ben-
eft a frm and its workers. The goal is to gain a competitive advantage through
better customer service. Even manufacturing companies are beginning to focus
their efforts around the concept of providing services to the customer.
Reengineering: Altering the Rules
Does technology alone improve a business? Many companies are
managed by rules and procedures. It would be virtually impossible to do other-
wisethe cost of an intense evaluation of every single decision would be over-
whelming. Hence, upper-level managers establish procedures and rules and an
organizational structure that automatically solve typical problems. More complex
problems are supposed to be identifed by managers and forwarded up the chain
of command for answers.
This type of management creates a fxed approach to operations and to solv-
ing problems. However, the business environment rarely remains constant. Over
time, new technologies are introduced, new competitors arrive, products change,
old markets shrink, and frms merge. At some point, frms that have been guided
by relatively static methodologies fnd their methods no longer match the market-
place. Hence, they decide to reengineer the company: beginning from scratch,
they identify goals along with the most effcient means of attaining those goals,
and create new processes that change the company to meet the new goals. The
Figure 1.11
Over time, Americans have moved from agricultural to manufacturing to service and
management jobs. Management and service jobs are often dedicated to collecting and
analyzing data. Just as the decline of workers in agriculture did not create a shortage
of food, the relative decline in manufacturing did not create a shortage of products.
0
20
40
60
80
100
120
140
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
M
i
l
l
i
o
n

W
o
r
k
e
r
s
U.S. Employment Paterns
Service
Management
Manufacturing
Farm
term reengineering and its current usage were made popular in 1990 by manage-
ment consultants James Champy and Michael Hammer. Many of the underlying
concepts have been in use for years.
Sometimes reengineering is undertaken by internal management as a means to
improve the company. For example, in the early 1990s, Compaq Computer al-
tered its strategy and reengineered its operations and management to cut millions
of dollars in costs and save the company. But in 2000, Dell Computers just-in-
time and made-to-order production system dominated the industry. Unable to alter
the company fast enough, Compaq was ultimately purchased by Hewlett-Packard.
Sometimes reengineering is forced on the company when it is taken over by
another corporation. In a few rare cases, managers continually evaluate the frm to
make several small changes instead of relying on a major overhaul.
Reengineering can be a highly complex process, requiring thousands of hours
of time to analyze the company and its processes. In addition to the complexity,
reengineering often faces resistance because it results in a change in the organiza-
tions structure, which affects the authority and power of various managers.
Like any management technique, reengineering is not guaranteed to work. A re-
port by CSC Index, a major reengineering consulting company, that surveyed 497
large companies in the United States and 124 in Europe, noted that 69 percent of
the American and 75 percent of the European companies have already undertaken
reengineering projects. Several of these projects have not been successful. CSC
Index notes that three factors are necessary for success: (1) overcome resistance
by managers who are afraid of losing jobs or power, (2) earn strong support by
upper management, and (3) aim high and go for major changes instead of small
rearrangements.
Often, the point of reengineering is to rebuild the company so that it can make
better use of technology. Simply placing computers into a frm, or just buying re-
placement computers, does not provide many advantages. The key to technology
is to restructure the operations and management to reduce costs and make better
Reality Bytes: Driptech with 20 Employees Goes International
Peter Frykman is the owner of Driptech, a small irrigation equipment company head-
quartered in Palo Alto, California. The frm has 20 employeesbut seven of them
are located in China and India. Mr. Frykman was part of a group of Stanford Uni-
versity graduate students who designed a method to make drip irrigation systems
inexpensively. The group tested the equipment in Ethiopia, formed the startup com-
pany, and raised $900,000 in funding. After a pilot project in India kicked off sales
in Asia in 2009, Chinese offcials became interested in the technology. Driptech set
up offces in Mumbai and Beijing, and Mr. Frykman soon expects half his employees
to be overseas. He notes that having offces in three countries is diffcult, but with
modern communication technology he is able to pass work around the world and
have engineers working continuously. He notes that If you get the rhythm right, you
can really be working around the clock as an organization. A few other companies
also work internationally, but the U.S. Census Bureau notes that only two percent of
small companies (fewer than 100 employees) sell their products in overseas markets.
Adapted from Justin Lahart, For Small Businesses, Big World Beckons, The Wall
decisions. For example, in the 1980s, replacing a secretarys typewriter with a per-
sonal computer had some benefts. But frms ultimately gained more benefts by
giving personal computers to managers and eliminating the secretary. Today, you
might consider simply updating those personal computers. However, building an
information system that provides easy-to-read up-to-the-minute data to top man-
agement makes it possible to eliminate the lower-level managers and still make
better decisions.
One of the challenges of reengineering is that the proposals can seem drastic.
What do you mean you want to eliminate half of the corporate-level employees!?
But being driven out of business by a leaner, lower-cost competitor is always a
worse situation.
Management and Decision Levels
How do you break businesses into smaller pieces to analyze
them? To understand management, reengineering, and information systems, it
helps to divide the organization into three decision levels: strategy, tactics, and
operations. Each level has unique characteristics, which use different types of sup-
port from information technology. These levels were explained by Robert Antho-
ny in 1965. In 1971, Gorry and Scott Morton added a detailed explanation of how
information systems at that time could support the various levels of management.
However, the terms and characteristics are largely taken from military terms.
Throughout history, in most nations the military has been the largest organization
in terms of the number of employees. Figure 1.12 is an updated picture of the
typical pyramid shape of most organizations involving operations and tactical and
strategic decisions. As is typical with most management models there are many
gray areas and the lines are not absolute.
Figure 1.12
There are three primary levels of decisions in business. Business operations consist
of tasks to keep the business operating on a day-to-day basis. Tactical decisions
involve changes to the frm without altering the overall structure. Strategic decisions
can alter the entire frm or even the industry. Information system tools exist to help
with each type of decision.
Operations
Tactics
Strategy
E
I
S
E
R
P
E
S
D
S
S
T
e
a
m
w
o
r
k
T
r
a
n
s
a
c
t
i
o
n
P
r
o
c
e
s
s

C
o
n
t
r
o
l
The power of the model is that it makes it easier to solve business problems.
With any problem, your goal is to identify the primary management level. Once
you know the level, it is easier to focus on the types of solutions that will be rel-
evant. For example, if a company is having basic problems with its day-to-day
accounting, you would focus on improving the data collectionand worry later
about strategic tools and problems with competition. As you read the cases and
Reality Bytes throughout this book, you should identify the primary level of each
problem.
Operations
The operations level consists of day-to-day operations and decisions. In your frst
job, you will typically concentrate on the problems that arise at this level. For
example, in a manufacturing frm, machine settings, worker schedules, and main-
tenance requirements would represent management tasks and decisions at the op-
erational level. Information technology at this level is used to collect data and per-
form well-defned computations. Most of the tasks and decisions are well struc-
tured, in the sense that they can be defned by a set of rules or procedures. For ex-
ample, a clerk at Wal-Mart follows the procedures in the guidebook to deal with
Reality Bytes: Fewer Workers More Output
Looking at the statistics, it is clear that employment at manufacturing companies in
the U.S. has been declining for 40 years. But due to increases in productivity, total
manufacturing output over that time has increased. U.S. companies are producing
more than twice the output they did 40 years ago. In the meantime, even in low-
wage countries such as China, a shortage of workers is driving wage increases of 15
percent or more. Christian Murck, president of the American Chamber of Commerce
in Beijing notes that Chinas low-wage advantage will disappear over the next fve
years. Supply chains are already being disrupted.
Adapted from John Bussey, Analysis: Will Costs Drive Firms Home? The Wall
Street Journal, May 5, 2011.2004.
Sector Operations Tactics Strategy
Production Machinesettings
Workerschedules
Maintenanceschedule
Rearrangeworkarea
Schedulenewproducts
Changeinventorymode
Newfactory
Newproducts
Newindustry
Accounting Categorizeassets
Assignexpenses
Producereports
Inventoryvaluation
Depreciationmethod
Financeshort/longterm
NewGLsystem
Debtvs.equity
Internationaltaxes
Marketing Rewardsalespeople
Surveycustomers
Monitorpromotions
Determinepricing
Promotionalcampaigns
Selectmarketingmedia
Monitorcompetitors
Newproducts
Newmarkets
Figure 1.13
Each functional area of management faces the three categories of decisions and
problems. Only a few examples are presented here.
typical operations. Common problems are anticipated, with actions spelled out in
the guidebook. Computer security is an increasingly important problemfor both
individuals and companies. Chapter 5 examines the major threats and tools avail-
able to protect your assets.
As summarized in Figure 1.13, managers in other disciplinessuch as account-
ing, marketing, or fnancealso face operational decisions. Personal productivity
tools, like spreadsheets, word processors, and database management systems help
managers collect and evaluate data they receive on a daily basis. The use of these
tools is reviewed in Chapter 2.
An important task at the operations level is to collect data on transactions and
operations; hence transaction processing systems are a crucial component of the
organizations information system. The data collected form the foundation for all
other information system capabilities. As discussed in Chapter 6, an important
characteristic of transaction processing systems is the ability to provide data for
multiple users at the same time. A special class of transaction processing soft-
ware designed for factory operations is called process control software. Chapter
7 shows how modern enterprise resource planning (ERP) software extends the
concepts of transactions across the organization.
Database management systems are increasingly used to control data and build
systems to share data. Their role is explained in Chapter 4. Chapter 3 shows how
communication networks are used to provide access to data throughout the organi-
zation. Increasingly managers work in teamseither with workers in the same de-
partment or across departments and sometimes companies. Sophisticated software
tools are being developed to help integrate data in these collaborative arrange-
ments. These integration tools and enterprise resource planning systems are
described in Chapter 7. Operational decisions are often the easiest to understand.
They deal with structured problems over relatively short periods of time.
Tactics
As you move up in your career to project leader or department manager, you
will encounter a different level of decision making, where the types of problems
will depend on your specialization, but some common features will stand out. At
the tactical level, decisions typically involve time frames of less than a year. As
shown in Figure 1.14, these decisions usually result in making relatively major
changes but stay within the existing structure of the organization.
Level Description Example Type of Information
Strategy Competitive advantage,
becomeamarketleader.
Long-termoutlook.
New product that will
changetheindustry.
External events, rivals,
sales,costsquality,trends.
Tactics Improving operations
without restructuring the
company.
New tools to cut costs
orimproveefciency.
Expenses, schedules, sales,
models,forecasts.
Operations Day-to-day actions
to keep the company
functioning.
Scheduling employees,
orderingsupplies.
Transactions, accounting,
human resource
management,inventory.
Figure 1.14
Each decision level affects the frm in different ways. Each level uses and produces
different types of information.
A manufacturing tactical-level decision might involve rearranging the work
area, altering production schedules, changing inventory methods, or expanding
quality control measures. These changes require time to implement and represent
changes to the basic methods of the frm. What distinguishes them is that they
can be made without altering the overall characteristics of the organization. For
example, in most cases, expanding quality control measures does not require the
frm to expand into new industries, build new facilities, or alter the structure of
the industry. Much of the information for making tactical decisions comes from
the transaction records that have been stored in the computer. Computer tools to
help analyze this type of data are called decision support systems (DSSs) and are
described in detail in Chapter 10.
Reality Bytes: Growing and Dying Industries
Industry Revenue Change
2000-2010
Revenue in 2010
($ million)
Voice over IP (VoIP) 194.0% 12,498
Internet Publishing 25.2% 32,573
Wind Power 16.9% 3,388
E-Commerce & Auctions 12.2% 95,005
Biotechnology 11.0% 86,971
Correctional Facilities 9.1% 34,373
Environmental Consulting 7.7% 18,153
Insurance Claims Adjusters 6.9% 57,530
Video Games 6.2% 38,622
Solar Power 2.7% 69
Video Postproduction -24.9% 4,276
Formal and Costume Rental -35.0% 736
DVD, Game, Video Rental -35.7% 7,839
Newspaper Publishing -35.9% 40,726
Mills (textile) -50.2% 54,645
Wired Telecomm. Carriers -54.9% 154,096
Photofnishing -69.1% 1,603
Manufactured Home Dealers -73.7% 4,538
Record Stores -76.3% 1,804
Apparel Manufacturing -77.1% 12,800
Adapted from Phil Izzo, Top 10 Dying Industries, The Wall Street Journal, March
28, 2011; and Phil Izzo, Top 10 Thriving Industries, The Wall Street Journal, May
16, 2011.
Other types of problems that involve more complex models occur in business.
For instance, diagnostic situations consist of spotting problems, searching for
the cause, and implementing corrections. Examples of these situations include re-
sponding to problem reports from operations to identify the cause of the problem
and potential solutions. For instance, a marketing manager might be asked to de-
termine why the latest marketing approach did not perform as well as expected.
Tactical-level decisions tend to involve specialized problems and can often be
solved with the help of an expert. Chapter 10 presents expert systems (ES) to
make this knowledge more accessible to an organization.
Strategy
The next step on the pyramid moves up the corporate ladder to executive-level
decisions. Although you may never be a CEO, you might be in a position to advise
upper-level management about strategic opportunitiesespecially in small busi-
nesses. Strategic decisions involve changing the overall structure of the frm to
give it an advantage over the competition. They are long-term decisions and are
unstructured. In other words, they are usually diffcult and risky decisions. Exam-
ples of strategic decisions in the manufacturing arena include building new facto-
ries, expanding to new products or industries, or even going out of business. Stra-
tegic decisions represent an attempt to gain a competitive advantage over your ri-
vals. Because of the complexity and unstructured nature of executives decisions,
it is diffcult to determine how information systems can help at the strategic level.
However, Chapter 11 explores information system techniques that frms have used
to gain a competitive advantage.
Rivalry Among
Existing Competitors
Bargaining Power
of Buyers
Bargaining Power
of Suppliers
Threat of New Entrants
Threat of Substitute
Products or Services
Figure 1.15
In analyzing strategies, Michael Porter focuses on the fve forces: threat of new
entrants, threat of substitute products or services, bargaining power of suppliers,
bargaining power of buyers, and rivalry among existing competitors. Competitive
advantage can be obtained by using these forces or altering the relationships between
these external agents.
An Introduction to Strategy
Why are strategic decisions so diffcult? How do you begin
searching for competitive advantage? In all industries, competition is
challenging. Firms are constantly searching for ways to gain an advantage over
their rivals. Finding these opportunities is hard: it requires extensive knowledge
of the industry, and it requires creativity. Managers also have to be willing to take
risks to implement strategic options. Strategic uses of IT often involve the use of
new technology and development of new software. Being the frst company to
implement a new idea can be risky. However, it can also bring substantial rewards.
Strategic uses of IT are discussed in detail in Chapter 11 because you need to
understand the technology before trying to solve diffcult problems. On the other
hand, to stimulate the imagination needed for creativity, it helps to begin think-
ing about the basic ideas right from the start. Many cases throughout the book
illustrate how frms have used technology to gain substantial advantages. These
examples should help you solve other problems. If you can recognize a pattern or
similarity between your problem and actions taken by other frms, the association
may help you create a solution.
Searching for Ideas
Michael Porter noted that often the frst step in searching for competitive advan-
tage is to focus on external agents, or entities that are outside the direct control of
your company. Porters Five Forces model in Figure 1.15 illustrates that typical
external agents are customers, suppliers, and rivals. You should also look
at the role of government as an external agent. Competitive advantages can be
found by producing better quality items or services at a lower cost than your ri-
Supply storeroom
Supply
Closets
Hospital
Warehouse
American Hospital
Supply
Supplier
Typical Supply Relationship
Supplier
Supplier
Figure 1.16
American Hospital Supply began as an intermediary that bought various medical
supplies and distributed them in bulk to hospitals. The hospital distributed supplies
throughout its facility and was responsible for maintaining its own inventory.
vals. Also, many frms have strengthened their positions by building closer
ties with their suppliers and customers.
It is hard to fnd revolutionary ideas that alter the entire industry. And every
frm in the world is constantly searching for that next big idea. Not only do you
have to come up with the best idea, but you have to do it before someone else
tries the same (or better) strategy. Then, because new strategies involve signifcant
changes, and often high costs, you have to be able to persuade the other managers
that your idea will succeed.
Strategy Example: Baxter Healthcare
Strategy is often easier to understand with examples or cases. A classic case in-
volving IT involves management of hospitals. For a moment, picture yourself as
the business manager of a medical center. Hospitals use a large amount of routine
supplies such as bandages and antiseptics. Originally, they purchased them from
various suppliers, held them in inventory, and distributed them throughout the
hospital as they were needed. This relationship is shown in Figure 1.16. Ameri-
can Hospital Supply (AHS) was one of these suppliers. Then, hospitals be-
came squeezed by initial governmental efforts to rein in and standardize medical
costs. One consequence of these early controls is that hospitals and physicians no
longer controlled the length of stay for patients. Consequently, hospital managers
were forced to fnd ways to reduce costs. To gain an advantage over their competi-
tors, AHS created a new system and made an offer to the hospital managers. AHS
Supply Closets
Hospital
Warehouse
American Hospital
Supply
Supplier
Baxter
Supplier
Supplier
AHS/Cardinal Health Computer Link
Computer
Monitor
Usage data
Deliver
Supplies as
Needed
Accurate usage data
Free space
Figure 1.17
American Hospital Supply changed the industry by providing a just-in-time inventory
delivery service. Supplies then were delivered directly to where they are used within
the hospital. AHS could offer this system only by maintaining a computer link
between supply usage and the local warehouse. The computer data also provided
summary reports to management. By purchasing AHS, Baxter Healthcare gained
immediate access to that sales data.
placed computer terminals in hospital locations where the supplies were used
(emergency, operating rooms, nursing stations, etc.). As shown in Figure 1.17,
these terminals were connected to the AHS computer.
As hospital personnel removed supplies, they recorded them on the terminals.
The computer kept track of the amount of supplies in each location. A list would
be printed at the warehouse, and drivers delivered the necessary supplies to each
location in the hospital. Monthly usage statistics were sent to the hospital.
The hospital gained because the facility did not need to maintain extra inven-
tory, which saved money and space. Fewer hospital employees were required, be-
cause the supplies were delivered directly to the needed locations. Additionally,
the hospital received detailed usage records.
To offer this service, AHS incurred higher costslargely the cost of creating
and maintaining the information system. What did AHS gain in return? As long
as it was the only company offering this service, AHS gained a competitive ad-
vantage by providing a new service. Hospitals were more likely to choose AHS
over the rivals. But what would happen if a competitor created a similar system?
Would the hospitals stay with AHS or switch to the rivals?
Although the answer depended on the prices, hospitals had a strong incentive to
stay with AHS. They would encounter various switching costs if they chose an-
other supplier. For example, daily operations would be disrupted while the system
was changed. Employees would have to be retrained to use the new system. Man-
agers who used the monthly usage reports would have to adapt to the new system.
A rival would have to offer strong price advantages to overcome these costs.
Supply Closets
Hospital
Supplier
Baxter
Supplier
Johnson
Supplier
Winning bidder
delivers suppiles
Internet
Daily Auction
Bid 1
Bid 2
Bid 3 << purchase
Figure 1.18
Moving to the Internet. Today, competition could be increased by connecting the
hospitals and suppliers through a B2B Internet auction. The daily hospital data could
be listed on a private auction site, and supplier computer system would automatically
bid for each job. The winning bidder would deliver supplies as usual, but different
suppliers might win the bidding each day.
In 1985, Baxter Healthcare, a large manufacturer of supplies, purchased AHS.
Of course, over time Baxter had an incentive to cut its costs to maintain higher
profts. In the process their delivery service might suffer. Some hospitals appar-
ently experienced problems and returned to in-house stock rooms to eliminate
shortages of basic supplies. In 1996, Baxter spun off Allegiance Medical Sup-
ply Corporation as a separate unit. Today, Allegiance is a subsidiary of Cardinal
Health, one of the three main health care distributors in the United States (Owens
& Minor and McKesson are the other two).
With the expansion of the Internet, as shown in Figure 1.18, the entire medical
supply chain industry is attempting to build an online Web service system. Ideally,
miniature electronic auctions would take place each day. The supplier systems
would automatically monitor the hospital needs and compete to resupply them.
Most of the monitoring and bidding could take place automatically on the Inter-
net. The best bid would win each day, and the hospitals would not be tied to a
single supplier. Yet all transactions and payments would be automated, holding
costs down for all parties.
Cloud Computing
As a consumer, you have certainly seen the increasing importance of Web-based
technologies, including sales, news and entertainment, communications, and so-
cial network interactivity. Many of these same concepts can be applied to infor-
mation technology within businesses. Instead of running applications on personal
computers, all of the data and services can be moved to Web-based servers. Man-
agers could access the data through portable devices including laptops, tablets,
and cell phones. Centralizing the Web servers provides the ability to control and
monitor the data and software. It also makes it easy to provide backups, duplicate
facilities, and robust Internet connections. In some cases, services might be pro-
vided by other companies. The Internet is often pictured as a cloud because it has
imprecise boundaries. So running operations from Web servers is often referred to
as cloud computing.
Many aspects of MIS can impact and beneft from cloud computing. Each
chapter in the book looks at various issues and examples of how Web-based serv-
ers can change the way companies deal with information systems.
Summary
Information technology is altering jobs, businesses, and society. Managers who
understand and use this technology will be able to improve companies and ad-
vance their personal careers. Studying technology means that you must also study
businesses and understand how they operate. Information systems consist of hard-
ware, software, people, procedures, and collections of data. These components
work together to provide information and help managers run the frm, solve prob-
lems, and make decisions. Studying information systems will also teach you to
analyze business operations and solve problems.
The role of a manager is changing, but at a basic level all managers spend time
organizing resources, planning, motivating workers, and communicating with oth-
er employees and managers. Several business trends will affect individual jobs,
business operations, and society. Important trends include specialization, manage-
ment by methodology and franchising, decentralization, the increased importance
of small businesses, the use of temporary workers and consultants, the growing
international scope of business, and the rise in service-oriented businesses. Infor-
mation technology is used to support these trends and provide new management
alternatives.
As is true of many problems, management and information technology can be
studied by breaking them down into smaller pieces. The three basic levels to man-
agement are operations, tactics, and strategies. The operations level is concerned
with day-to-day operations of the frm. Tactics involve changes and decisions that
improve operations but do not require a major restructuring of the frm. Strategies
are designed to give a frm a competitive advantage.
Strategy typically involves examining external forces: rivals (competitors with-
in the industry), customers, suppliers, potential new competitors, and potential
substitute products or services. Information technology can be used to strengthen
links strategically between customers and suppliers. It can also be used to create
new products and services and to improve the quality of the operations.
Key Words
B2B
B2C
cloud computing
data
database
decision process
decision support system (DSS)
diagnostic situation
dot-com
e-business
e-commerce (EC)
enterprise resource planning (ERP)
expert system (ES)
hardware
information
information technology (IT)
knowledge
management informa-
tion system (MIS)
people
procedures
reengineering
software
strategic decisions
structured decisions
switching costs
transaction processing system
wisdom
A Managers View
How do you manage and control a frm? In the 1950s, an army of back-of-
fce workers and managers were required just to record the basic data for the
frm. Today, a good information system helps you manage a small company
or a large empire with fewer workers. The workers that remain are those
who use the technology intelligently to solve business problems. Regardless
of your area of expertise, as a manager you have to be able to analyze and in-
terpret data. You also have to communicate and share your work with team-
mates. Information technology provides the tools you need to solve common
business problems.
Web Site References
General Searches
Ask www.ask.com
Bing www.bing.com
Dogpile www.dogpile.com
Google www.google.com
Yahoo www.yahoo.com
Yippy www.yippy.com
People and Businesses
Anywho www.anywho.com
Infospace www.infospace.com
Knowx www.knowx.com
ChoicePoint www.choicepoint.com
Securities and Exchange www.sec.gov
SuperPages www.superpages.com
Switchboard www.switchboard.com
Whitepages www.whitepages.com
Reference
Britannica (encyclopedia) www.britannica.com
CIA World Factbook www.cia.gov/cia/publications/
factbook
Dictionary www.dictionary.com
Encarta (encyclopedia) www.encarta.com
FedStats www.fedstats.gov
Translation dictionaries www.freedict.com
Translate/Google translate.google.com
Wiktionary www.wiktionary.org
Wikipedia www.wikipedia.org
Wolfram Alpha, math and science www.wolframalpha.com
Review Questions
1. What is the main purpose of MIS?
2. How is MIS different from studying personal productivity tools?
3. Describe the fve components of a management information system.
4. Why do students who are not MIS majors need to study MIS?
5. How important is the Internet in sales?
6. How do you know if you are buying the correct level of technology?
7. What are the roles of managers in a modern company?
8. Describe how seven basic trends in todays business environment are related
to MIS.
9. Why is re-engineering important in business?
10. What are the three main management decision levels, and why are they
important to know?
11. How does strategy involve external organizations?
12. What is cloud computing?
Exercises
1. Identify a job that has minimal use of information technology. Explain why
the job might not need technology or what type of technology would be
needed to improve productivity in the job.
2. Talk to at least three people and ask them how often they get a new cell
phone and a new computer. Compare the answers and comment on any
differences.
3. Choose a company and describe three decisions that must be made: one at the
operations level, one tactical, one strategic. Be specifc.
4. Choose a company and read its two most recent annual reports. Summarize
the companys strategy and goals for the coming year.
5. As an entrepreneur, you decide to open a fast-food restaurant. You can
purchase a franchise from one of the established corporations (as discussed
in the McDonalds case) or create your own restaurant. Compare the choices
by identifying the decisions you will face with each approach. What data will
you need to collect?
6. Assuming you run a small business, fnd two Web sites that enable you to
hire contract workers online. Provide a brief list of the types of tasks that are
offered on the sites.
7. Review business magazines, newspapers, and Web sites. Find two
organizations and identify a specifc business problem that each one faces.
Classify the problem as operations, tactics, or strategies.
8. Identify the most likely decision level for each of the following situations.
a. A manufacturing frm lays off 100 workers.
b. A restaurant completely changes it menu.
c. A Web site begins accepting Google payments.
d. A company builds an iPhone-based app to provide complete customer and
order data to its salespeople.
e. A California farmer tears out his felds of grape vines and plants olive
trees.
f. A large retail chain implements a new computer system to analyze sales by
time and day and uses it to alter prices in every store.
g. A manufacturer creates a computer system to evaluate workers in terms of
productivity and cost and uses it to give bonuses or fre workers.
9. Find one company that is using information technology in a way that
surprises you. Briefy explain what the company does and why it seems
different or new. (Check places you go, try the Wall Street Journal,
Computerworld, or CIO Magazine.)
10. Which geographic region of the world has the most Internet users and why?
Technology Toolbox: Searching
11. What was the name of the space shuttle that took the last fight?
12. What is the approximate ratio of cell phones to landline phones in Ethiopia?
13. How many singers have covered (recorded) John Lennons song Imagine?
14. How much proft did Exxon Mobile earn in 2010?
15. Who is the main executive producer of the TV show The Big Bang Theory?
Technology Toolbox: Government Data
16. Find monthly unemployment data for at least fve years and display it in an
appropriate chart.
17. Compare the most recent population by state.
Technology Toolbox: Spreadsheets
18. Create a spreadsheet to record current sales and forecast future sales as a 4
percent increase.
increase 4%
Region Sales Forecast
Midwest 132
Southwest 651
Northeast 478
Southeast 391
1652
19. If you have Excel 2010 or later, create the following spreadsheet and chart
that consists of fctional data evaluating three search engines. The weighted
average is the total of the response value in the cell times the point value
on the frst row, where the total is divided by the sum of the number of
responses.
5 4 3 2 1
SA A N D SD Wtd Avg
Google 71 37 12 5 1 4.365
Bing 65 43 27 8 4 4.068
Yahoo 32 55 45 32 12 3.358

Teamwork
20. Each team member should fnd and select a new computer that could be used
by a typical manager in a business. Combine the data from each person into a
single spreadsheet. Compute the average price, drive size, and memory size.
As a group, select one of the computers that could be used as the standard for
each person in the group.
21. Choose an industry. Have each team member select one company within
the industry and fnd the number of employees at that company for the two
most recent years. Compare the total number of employees to see if the total
has increased, and to see which frms are growing the fastest. Compare the
industry results to those in the rest of the class.
22. As a team, fnd an example of companies that have recently merged. Identify
potential reasons for the merger and how each company might beneft (or
suffer) from the merger.
23. As a team, choose a specifc company and identify at least one decision that
has to be made at each of the management levels. What data would be needed
to make the decision? Briefy explain why the chosen level applies to each
decision.
Rolling Thunder Database
24. Install the Rolling Thunder Bicycles database. Look through the various
forms. List each of the main forms and briefy describe the purpose of the
form.
25. Using the Rolling Thunder help fles or the description available on the
Internet site, describe the goals of the frm and outline the basic operations.
26. Using Internet sources, identify the competitors to Rolling Thunder Bicycles.
27. Using Internet, fnancial, and government sources, estimate the size of the
market (total sales and number of bicycles) for quality bicycles.
28. Locate at least fve sources for additional information about bicycles and
bicycle components on the Internet. List and briefy describe the sites.
Additional Reading
Anthony, Robert N. Planning and Control Systems: A Framework for Analysis.
Cambridge: Harvard University Press, 1965. [Early MIS]
Booker, Ellis. Baxter Gets PC Smart, Ousts Dumb Terminals, Computerworld,
April 3, 1989, p. 33. [Baxter Healthcare]
Gorry. G.A. and M. Scott Morton. A Framework for Management Information
Systems, Sloan Management Review, Fall 1971, pp 55-70. [Early MIS.]
Health-Care Guys Can Make Good on Retail IT. PC Week, August 21, 1995, p.
11. [Baxter Healthcare.]
Leavitt, Harold J. and Thomas L. Whisler. Management in the 1980s, Harvard
Business Review, November 1958, p 41-48. [Prediction of decline in middle
management.]
Luthans, Fred. Organizational Behavior: A Modern Behavioral Approach to
Management, New York, McGraw Hill, 1973. [Management.]
Mintzberg, Henry. The Nature of Managerial Work, Harper & Row, New York,
1973. [Management.]
Nash, Jim. Just What the Doctor Ordered, Computerworld, June 1, 1992, p. 79.
[Baxter Healthcare.]
Porter, Michael. Competitive Strategy: Techniques for Analyzing Industries and
Competitors, New York: Free Press, 1980. [Strategy.]
Sloan, Alfred. Adventures of a White-Collar Man. New York: Doubleday, 1941.
[Management.]
Cases: Te Fast-food Industry
Te Industry
What do customers want? Look at the sales for the major fast-food res-
taurants, and you see increases by many companies. Look at the Stan-
dard and Poors restaurant stock market index, and investors agree. In
2003, the index was up 37.5 percent, compared to a 24.8 percent rise
in the S&P 1500. Since an aging, wealthier population favors dining in
full-service restaurants, the casual-dining sector is gaining share from
fast-food restaurants. This trend is projected to continue as the popu-
lation ages. Many restaurant chains, especially those in the fast-food
sector, will increase their focus on healthy food initiatives to attract
customers and reduce the impact of obesity-related lawsuits. Increased
diversity in menus will help to reduce the dependence on industry price
discounting, enabling operating margins to further expand. The intro-
duction of new restaurants will likely slow in the over-stored U.S. fast-
food market. Most fast-food restaurants are looking to international
expansion to lead to growth. But, fast-food restaurants cannot stand
still. Many radically expanded their menus, both to target higher-mar-
gin threats such as coffee and smoothies, and to add healthier snacks.
And when the recession hit in 2007, fast-food restaurants were able
to provide meals at substantially lower costs than full-service restau-
rants. In the ensuing couple of years, fast-food chains increased sales
while many high-end restaurants closed down.
During the past two decades, the percentage of U.S. food dollars that has gone
to eating out has increased substantially. A greater percentage of people, particu-
larly women, are working more, leaving less time available to prepare food at
home. Overall sales have increased based on three factors:
The opening of new stores.
Higher contributions from older restaurants.
Acquisitions of other chains or selective sites.
-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
R
a
t
o
Net Income / Revenue
McDonald's
Burger King
Yum
Starbucks
Wendys
0
5
10
15
20
25
30
19941995199619971998199920002001200220032004200520062007200820092010
B
i
l
l
i
o
n

$
Annual Revenue
McDonald's
Burger King
Yum
Starbucks
Wendys
But over the past several years, fast-food sales gains have lagged those of the
full-service sector, due in part to ferocious competition, ferce price discounting,
and reduced same-store sales and proftability. In January 2003, both McDonalds
and Burger King changed management teams. Both companies have focused on
rebranding efforts to recast their dowdy image and less reliance on price discount-
ing as a means to drive traffc. In 2004, the CEO of McDonalds died of a heart
attack, but it is likely the changes he instituted will continue.
What Do Customers Want?
Most of the fast-food restaurants have tested a variety of options to fnd what
customers want. Originally, McDonalds was successful because it promised the
same level of service and quality regardless of where you traveled. But incomes
increased and health became an important issue. Do customers still want the same
things? How can the big fast-food chains identify what customers want? How can
they respond quickly enough?
New Concepts: Quick Casual.
The frst generation to grow up with fast food has now reached their mature, high-
income years. To meet their needs, quick casual, limited, self-service restau-
rants are geared toward adults. They feature upscale menus, with more health-
ful items such as gourmet soups, salads, and sandwiches. In some quick-casual
units, workers take orders from behind the counter as customers proceed in a line
toward the register. In others, they take orders at a counter where customers pay
for the food, which is then prepared and delivered to the customers table. Checks,
averaging between $6 and $9, are higher than in traditional limited-service units,
but lower than in full-service casual dining restaurants.
The most successful and most visible of the quick-casual chains is Panera
Bread Company. An operator of bakery/cafs, system sales now surpass $800 mil-
lion. The company plans to add another 100 to 150 units annually over the next
several years.
Focus on Health.
The American culture has become signifcantly more health-conscious and liti-
gious over the past decades. This culture has resulted in lawsuits focused on the
responsibility for obesity-related health problems faced by consumers, particular-
ly children. Plaintiffs have sought remedies such as menu changes, nutritional la-
beling, advertising restrictions, and monetary damages. In response to this strong
customer demand, many restaurant chains have begun to make signifcant chang-
es in their menu offerings. Applebees International signed a deal with Weight
Watchers International Inc. in July 2003 to develop a menu for diet-conscious in-
dividuals. In February 2003, Darden Restaurants opened its frst Seasons 52 unit,
a test concept offering low-calorie menu items.
The fast-food industry has introduced even more dramatic changes, perhaps be-
cause it has the most to lose from consumer perceptions of the healthfulness of its
food offerings and from potential lawsuits. In August 2003, Wendys announced
that it would promote four meal combinations from items already on the menu
that would have less than 10 grams of fat. Wendys also added another salad offer-
ing to its menu. Burger King and Jack in the Box have focused on salad, chicken,
and turkey offerings to revamp their menus.
McDonalds has developed a wide range of Healthy Lifestyle programs,
including the addition of menu offerings that the company believes will attract
health-conscious consumers. It has also developed new Happy Meals that include
yogurt, milk, vegetables, or fruit, depending on the end market. In 2002, McDon-
alds changed its cooking oils to reduce the amount of trans-fatty acids in its fried
foods. In June 2003, it phased out animal-growth-promoting antibiotics in its meat
supply. McDonalds has also sought to promote nutritional education and aware-
ness among its customers. In May 2003, the company formed its Global Advisory
Council on Healthy Lifestyles, consisting of experts such as doctors, educators,
and athletes in the areas of ftness, nutrition, and active lifestyles. The group is
commissioned to help guide the company toward activities that promote balanced,
healthy lifestyles among its customers. McDonalds has also begun to collaborate
with the World Health Organization and the U.S. Department of Health and Hu-
man Services to educate consumers on the importance of nutrition and ftness.
The company has educated consumers by printing brochures directing them to the
nutritional information on its corporate Web site.
Te Future
The restaurant industry generally depends on the economy. When people work,
they have less time and more money, so they vote for convenience. As long as
costs stay low, the industry should continue to do well. Costs are heavily depen-
dent on labor and food. Companies tend to lock in food costs for up to a year, so
they are less affected by short-term swings in prices. Labor costs depend on the
overall economy. The question of what customers want is diffcult to answer. And
the answer changes over time. Information systems can help identify sales pat-
terns. They can also help control costs and reduce order lags with suppliers. But,
analyzing the data requires a keen eye and experience.
Case: McDonalds
What do customers want? John Gusapari in The Customer Connection suggests
that one goal is to meet customers expectations. You accomplish this by creating
value. Creating quality is not just a matter of reducing defects but of providing
customers with something they value. McDonalds (ticker: MCD) has a clear def-
nition of customer needs and expectations. Its formula is QSC, quality, service,
and cleanliness (Band 1989). Transaction quality, defned from a customers per-
spective, means paying for a product or service in a way that makes the customer
feel good about doing business with a company.
According to Harry Beckwith (Beckwith 2003), customer defnition is essential
to success in marketing. Beckwith viewed McDonalds as an example of a classic,
but simple, excellent service model. However, he feels that they have recently tak-
en their eye off the ball, forgetting their formula for success. Beckwith feels their
mistake is that they have been thinking that fast food is a food business. Beckwith
focuses instead on the fast part of the defnition: It is a time and convenience
business. In Beckwiths opinion, McDonalds has made the menu choices too
complex costing customers too much time. Beckwith does not feel that people go
to McDonalds for the menu.
Corporate Summary
McDonalds Corporation serves more than 64 million customers daily from
32,737 fast-food restaurants in over 115 countries. In 2011 there were 14,000 res-
taurants in the United States and 17,614 in other countries. McDonalds maintains
its competitiveness through substantially uniform menus and standard operating
processes. The company has experimented with expansion through partner brands
in additional market segments including Boston Market, Donatos Pizza, and, (in
the UK) Pret A Manger. McDonalds sold its Chipotle Grill in 2006 for $300 mil-
lion. McDonalds operates all of its restaurants under joint venture agreements.
In 2010, McDonalds annual report showed the mix was (although only company
stores are recorded as revenue):
Restaurant Sales ($ billion) Number
Company-operated $16,233 6,399
Franchise + Affliated $61,147 26,338
Total $78,380 32,737
Systemwide sales were $64 billion in 2010, accounting for $24 billion in rev-
enue for McDonalds. The company is targeting sales growth rates of 3-5 percent
a year. U.S. sales increases were largely driven by breakfast items and coffee, and
the introduction of wraps as a menu item. International business contributed 66
percent of revenue in 2010 and 50 percent of operating income, compared to 45
percent in 2003. McDonalds 2006 Annual Report notes the existence of 550,000
restaurants in the U.S. with $365 billion in sales. McDonalds restaurants repre-
sent 2.5 percent of the total and 7.4 percent of the sales. Key factors in the perfor-
mance of the company include
The ability to forecast trends, demographic changes, and food preferences.
The selection of menu items and product mix.
The ability to improve day-to-day restaurant operations and hire workers.
The ability to identify new franchisees.
To focus on the restaurant business, in 2009 McDonalds sold its interest in the
Redbox company for $140 million. In 2008, it sold its interest in the UK Pret A
Manger company for $229 million. However, many McDonalds stores still have
Redbox movie rental kiosks.
Turnaround Under Way
Jim Cantalupo, one of the architects of McDonalds highly successful worldwide
expansion in the 1990s, was lured out of retirement to run the company. He as-
sembled a youthful and energetic management team, led by president and chief
operating offcer Charles Bell, to revamp the companys fortunes. Under this new
leadership, the company embarked on a strategic plan to return to positive sales
growth and income trends. To accomplish this goal, the company shifted its focus
away from expansion and toward the improvement of existing operations. Tragi-
cally, Jim died of a heart attack in April 2003.
McDonalds began to address concerns that its menu had become stale and ir-
relevant to todays consumer. McDonalds made new product introductions, in-
cluding salad offerings, the McGriddle breakfast sandwich, and increased Happy
Meal options. Even though management believes that excessive price discounting
has diminished the companys brand image, the company plans to maintain the
99-cent Value Meal offering. In relation to the brand image, the publics percep-
tion of the quality, service, and cleanliness at McDonalds units has suffered over
the last several years. The company has lagged behind its peers in consumer at-
titude toward these areas. To improve overall customer satisfaction, the company
has refocused on its quality, service, cleanliness (QSC) program. This program
gauges performance at each location through customer satisfaction studies and
mystery shoppers, who pose as customers and score each units performance in
various categories. The company has also taken a harder line with franchisers who
do not perform to expectations.
McDonalds must also reinvigorate the companys franchise base. Franchisees
run more than 70 percent of McDonalds 31,000 restaurants worldwide and ac-
count for a similar level of systemwide sales. In recent years, many franchisees
have been demoralized by declining profts caused by company-mandated dis-
counting, a proliferation of stores that has cannibalized individual store sales, and
the overall lack of a coherent national advertising program. Retaining hardwork-
ing, entrepreneurial store owners is paramount to reinvigorating the companys
health.
McDonalds began to reap benefts from its new strategies in early 2003. The
company had seen its proftability deteriorate in the United States during 2001 and
2002 because of operational shortcomings and sales trends that were signifcantly
below the industry average. In the second quarter of 2003, driven by new prod-
uct offerings and improved marketing effectiveness, same-store sales jumped 4.9
percent, year to year. While these sales results are only an important frst step in
reviving McDonalds image, the company must show tangible improvement in its
customer satisfaction scores to maintain momentum. New products may lure cus-
tomers through the doors, but high degrees of customer satisfaction are often the
key to keep them returning.
CEO Jim Skinner is pushing for innovation in new products, but is also focused
on maintaining fscal discipline and tight controls on company expenses (2006
Annual Report). 2006 was a good year for McDonalds with a 7 percent global
increase in sales and improved prices of the stock. However, McDonalds used the
proceeds from the sale of its Chipotle franchise to repurchase sharescontribut-
ing to the increase in stock prices.
Since 2006, McDonalds has averaged four percent increases in sales revenue
at U.S. stores. Despite the recession, McDonalds has growneven faster than
its nearest competitors. The company plans to push its advantageswith plans
to remodel or rebuild almost half of its 14,000 stores. Key elements in the rebuild
include doubling the number of drive-through lanes and redesigning interiors that
encourage people to stay longer (Jannarone 2011).
Enhancing Information Technology through Project Management
To improve the project success rates on information technology projects, McDon-
alds has developed an apprenticeship program for prospective project managers,
combining classroom theory, on-the-job learning, and support from mentors. Mc-
Donalds is accomplishing this program in association with the Computer Tech-
nology Industry Association (CompTIA), six other companies, and a $2.8 million
grant from the Department of Labor. The goal is to develop a National Informa-
tion Technology Apprenticeship System, aimed at building skills and credentials
around specifc business-technology functions.
The grant enables CompTIA to build four career tracks in areas that it feels are
prime for IT apprenticeships: IT generalist, project manager, security, and data-
base. The grant also provides funds to implement Web-based processing of ap-
plications and certifcation. Over the next fve years, CompTIA has committed
matching funds of nearly $3.8 million to develop the systems infrastructure, skill
standards and work processes, and marketing strategies to encourage large-scale
private sector adoption.
The Labor Department awarded CompTIA a grant of $550,000 through De-
cember 2002 to assess whether apprenticeships would work in the IT industry and
to develop one apprenticeship track for the IT generalist. Success with that led the
Labor Department to award a $475,000 second-round grant through December
2003 to develop additional apprenticeship tracks, including IT project manage-
ment, and to test them in pilot companies such as McDonalds. The latest round of
funding enables the Labor Department to expand on these original efforts.
Using Technology to Improve Operations
Day-to-day operations are critical to McDonalds restaurants. Hyperion, a soft-
ware company that sells analytical tools, observes that reducing service time by
six seconds increases sales revenue by one percent (Vance 2004). HyperActive
Technologies, a specialty company from Pittsburgh, created HyperActive Bob, a
tool that helps estimate fast-food sales for the next 10 minutes. The tool uses past
sales data and cameras to evaluate traffc entering the restaurant. The computer
display tells the grill cooks how many high-volume items to start cooking. When
the order arrives, it is pulled right off the grill. In one trial, the system noticed
fewer cars entering the drive-through than expected and it told the grill cook to
hold back. The cook, with three years of experience, wanted to ramp up produc-
tion for the lunch hour. The manager convinced him to listen to Bob and wait. It
turns out a ruptured gas line on the road blocked traffc, and the real-time informa-
tion analyzed by Bob was correct. In 2006, McDonalds rolled out a new point of
sale system to 8,400 restaurants. The system created improved order accuracy and
speed of service.
Sometime around 2001, McDonalds Corporation started the Innovate project
that was designed to build an intranet to collect detailed data from every restau-
rant franchise around the globe and transfer data to corporate headquarters. The
system would have integrated 30,000 restaurants in more than 120 countries. Pre-
sumably, it would have given McDonalds detailed records on sales and opera-
tions. According to SEC flings in 2003, the project was cancelled after spending
$170 million on consultants and initial implementation (McDougall 2006).
Dee Crawford runs fve McDonalds restaurants. She notes that Change is a
part of our business, to keep up with customer demands, and there have been a lot
of changes. In 2010, her restaurants had more than 100 items on the menufar
more than in the past. (Jargon 2010). The expanded menu has helped McDon-
alds increase same-store sales for 30 straight quarters since early 2003. Even dur-
ing the worst of the recession in 2008, same-store sales increased by 6.1 percent.
Introducing high-end coffee has increased sales in some stores and struggled in
others, and the machines cost $100,000 per store. The company has learned that
snack wraps and other items often eaten between traditional meals have become
the fastest selling items. Most stores have increased their hours to attract more
non-traditional traffc. With the larger menus, a corporate distribution center auto-
matically handles inventory and restocks stores two or three times a week.
The 2010 annual report notes that the company uses a strategic menu pric-
ing tool that optimizes price, product mix, and promotions. Several newer res-
taurants use electronic display boards to digitally display the menu and prices.
Among other things, the tools enable the restaurant to quickly alter prices. The
company also plans to introduce a new point-of-sale system beginning in 2011. In
some locations, particularly Europe, the company is introducing self-order kiosks.
In a few locations, drive-through productivity is improved by having employees
with hand-held terminals take orders by walking down the drive-through lane.
In 2010, McDonalds worked with Nintendo to develop a DS-based game to
help train new workers. Most restaurants have relatively high employee turnover,
so they are constantly training new workers. The system reportedly reduced train-
ing time from 45 to 24 hours and eliminated the need for human trainers for the
initial training (Perkins 2010).
Application of the Internet to Solving Business Problems
One of the ways that McDonalds has used the Internet is to provide information
to customers to deal with the fat content of their meals. While this is in some way
a response to attorneys and consumer groups that have criticized the industry for
fattening America, McDonalds is using Internet technology to better disseminate
information about the nutritional properties of their menu items. In 2003, McDon-
alds revamped its www.mcdonalds.com Web site to better explain the nutritional
values of the chains foods.
Among the most graphically sophisticated nutrition tools on the Internet, Bag
a McMeal enables users to drag up to fve McDonalds menu items from pull-
down lists into a virtual bag. Once completed, the users receive a cumulative nu-
trition profle covering calories, fat, cholesterol, and sodium, among other infor-
mation. A Customize an Item feature enables consumers to obtain a nutrition
profle for a special order, such as a McChicken sandwich without mayonnaise
or a Big Mac without cheese or sauce. Bag a McMeal uses multiple pull-down
lists and drag-and-drop functionality to generate a nutrition profle for a complete
meal of up to fve items. The goal is to provide additional educational resources to
demonstrate the range of options and service sizes available to make it easy to ft
McDonalds into a balanced diet.
According to McDonalds representative Lisa Howard, the Web-surfng public
has found value in the dynamic nutrition databases. There are from 160,000 to
200,000 unique visitors to the nutrition section of McDonalds.com each month,
Howard said. Weve seen it spike [upward] since the new Web tools were put in
place, she added. According to Howard, We continue to look at new and innova-
tive ways to communicate with customers, like in store kiosks (Brewin 2003).
Reaction to the Web-based nutrition information has been critical. If [the
chains] were really serious about doing a better job at giving consumers informa-
tion, diners would have that information at the point of decision in the restaurant,
said Jeff Cronin, director of communications for the Washington, D.C.-based
Center for Science in the Public Interest. Now the [nutrition] brochures are hard
to read and hard to fnd and sometimes altogether absent. Few consumers would
leave their place in the drive-thru or counter line to fnd and read a poster with
nutrition information before ordering, and many fewer are likely to go to a Web
site before making a decision. National Restaurant Association offcials and rep-
resentatives of some other trade groups have countered this argument by asserting
that restaurateurs provide what consumers want, not what restaurateurs think con-
sumers should eat. They believe menu board notices are unnecessary, since many
chains already provide interested consumers with printed brochures or Web site
pages containing nutrition information for regular menu items.
Given the debate about practicality and accessibility, information about nutri-
tion can be stored in large, centralized, and easily updated databases that many
consumers can tap through the Internet. Such data can be presented in ways that
enable Web surfers to personalize searches and drill down through multiple lay-
ers of details to get as little or as much insight into a topic as desired. Many res-
taurant companies have ignored such dynamic Internet presentation plans in favor
of merely presenting at their Web sites the same nutrition information tables con-
tained in printed brochures.
McDonalds may be considering ways to leverage its Web site technology in
such a fashion, company spokeswoman Howard said. We continue to look at new
and innovative ways to communicate with customersthat is certainly something
we might look at.
Of course, using the Web also opens up more holes for hackers. In December
2010, McDonalds reported that someone had broken into the companys data-
bases and stole information about customers (Perez 2010). Much of the data was
collected by Arc Worldwide, a company that develops and coordinates promotion-
al e-mail messages for McDonalds. The data was basically personal information
provided by customers and did not include SSN or fnancial data.
McDonalds also runs a Facebook page which it basically uses as a Web site
to provide basic data but also to collect comments from users. The McDonalds
site is also one of the frst to use Facebooks location feature. It enables people to
check in at a location and receive coupons or promotions for the nearest Mc-
Donalds restaurant (Gannes 2010).
Implementing Wireless Technology
Many restaurants view the public-access Wi-Fi hot spot technology as essential
for attracting customers. But what is still unclear is how much businesses can
charge customers to use the Wi-Fi linksor whether they should simply provide
the Internet and e-mail access capabilities for free with the hope that increased
sales of food, drinks, and other products will offset the cost of the technology.
McDonalds launched a Wi-Fi pilot project at 75 restaurants in the San Fran-
cisco Bay area in mid-2003 through a deal with Austin-based Internet access pro-
vider Wayport Inc. Mark Jamison, vice president of business strategy and devel-
opment at McDonalds, said the company would use the San Francisco trial and
similar ones in Chicago and New York to evaluate potential pricing models for the
service and Wi-Fi technologys ability to attract customers. The end result is that
McDonalds plans to equip several hundred restaurants in the United States with
Wi-Fi connections by 2004. McDonalds charges $4.95 for two hours of Wi-Fi
access at the San Francisco locations, but customers who buy a meal can use the
technology for free. By the end of 2006, 15,000 restaurants worldwide had wire-
less access.
But will customers really go to a restaurant because it has wireless access?
Consider the opinion of Matthew Nuss of the Valencia Group, a Houston-based
hotel operator: Wireless, in our opinion, is the next running water. It has become
part of the infrastructure of a hotel (Wright 2003). On the other hand, McDon-
alds is using an outside company (Wayport) to set up and run the Wi-Fi opera-
tions. McDonalds is also benefting from the technology. It is using the same
Internet connection to run its cashless payment and credit card systems. Basically,
it is turning communications with the stores into a revenue-generating operation
(Brewin 2004). David Grooms, VP of IT at McDonalds noted that It gives us a
platform to use wireless applications within the restaurant. It turns out that 25
percent of the Wi-Fi traffc is produced by gamers using Nintendo DS systems
(Mitchell 2007).
Seattle-based Starbucks Corp. launched Wi-Fi service in its U.S. cafes in Au-
gust 2002 and offers access in about 2,000 locations. Users have to sign up for the
service with Bellevue, Washington-based T-Mobile USA Inc., whose prices start
at $19.99 per month. Lovina McMurchy, director of Wi-Fi business and alliances
at Starbucks, said the company plans to stick with that approach. But she added
that Wi-Fi hot-spot deployment is a learning experience for businesses and said
its hard to tell how different pricing plans or free services will play out. At this
point, a lot of companies are still just dabbling in Wi-Fi through pilot projects,
McMurchy said.
In 2009, McDonalds in partnership with AT&T opened up the Wi-Fi and of-
fered it for free. The goal was to encourage customers to sit around longer and buy
more productsparticularly higher-proft margin drinks including coffee and fruit
smoothies. CIO David Grooms noted that Were becoming a destination and free
Wi-Fi just naturally fts. This is another long-term investment that we see helps
McDonalds stay relevant as a brand in the marketplace. McDonalds is now a
signifcant Wi-Fi provider in the U.S., delivering one out of six hot spots (Ziobro
2009).
Questions
1. How does McDonalds use technology to learn what customers want?
2. How does information technology reduce management costs?
3. Has the Wi-Fi implementation been successful?
4. Does the franchise model make it more diffcult to implement information
technology solutions? Explain.
5. Is it more important for information technologies to reduce costs or provide
new services?
Additional Reading
Band, William, Quality: I Know IT When I See It. Sales & Marketing
Management in Canada v30n2, (Feb 1989): p.36-38.
Beckwith, Harry, 10 Questions With... Harry Beckwith, Journal of Financial
Planning, Vol 16(10), October 2003, pp. 12-15.
Brewin, Bob, Computerworld, Vol 37(28), July 14, 2003, p. 1.
Brewin, Bob, McDonalds to Supersize Use of Wi-Fi Connections,
Computerworld, April 19, 2004.
Gannes, Liz, McDonalds to Be First Location-based Marketer on Facebook,
Gigaom, May 6, 2010. http://gigaom.com/2010/05/06/mcdonalds-to-be-frst-
location-based-marketer-on-facebook-report/
Jannarone, John, McDonalds Lucrative Shake-Up, The Wall Street Journal,
May 24, 2011.
Jargon, Julie, On McDonalds Menu: Variety, Caution, The Wall Street Journal,
December 27, 2010.
McDougall, Paul, 8 Expensive IT Blunders, Information Week, October 16,
2006.
Mitchell, Robert L., Would You Like Wi-Fi with that Burger? Computerworld,
April 13, 2007.
Perkins, Bart, Want a Job? Bring Your Game! Computerworld, November 23,
2010.
Perez, Juan Carlos, Hackers Steal McDonalds Customer Data,
Computerworld, December 11, 2010.
Vance, Jeff, You Will Have French Fries With That, CIO, December 15, 2004.
Wright, Maury, Combo, Please, EDN, Vol 48(16), July 24 2003, pp. 95-6, 98.
Ziobro, Paul, McDonalds to Offer Free Wireless Internet, The Wall Street
Journal, December 15, 2009.
Case: Burger King
Burger King (private) is the second largest fast-food chain in the United States,
with more than 12,000 locations 2010. For the year ending in June 2006, Burg-
er Kings annual report estimates that McDonalds, Burger King, and Wendys
account for 73 percent of the fast-food sales in the United States, with Burger
King accounting for 15. Burger King has encountered diffculties over the last
few years. According to Nations Restaurant News (NRN), the companys market
share among quick-service sandwich chains dropped from 15.03 percent in 2000
to 13.68 percent in 2002. The comparative revenue from the annual reports shows
slight increases for BK sales after 2003, but far below that of McDonalds.
In the fall of 2002, Burger King was sold to a private investment group for ap-
proximately $1.5 billion. Many analysts believed the change in ownership would
improve the chains management focus and vision, increasing its competitiveness
in the marketplace. Brady Blum, formerly a senior executive of Darden Restau-
rants Inc., was hired as CEO to engineer a turnaround. Burger King now seeks
to recast its brand image by focusing on its grilling processes and retooling its
menu, while downplaying its previous emphasis on discounts. The new strategy
has yet to provide the desired results, however. The chain has acknowledged that
negative same-store sales trends continued into the frst six months of 2003. From
2003 to 2006, the chain experienced a drop-off in franchisees renewing their 20-
year agreements. To renew a contract, franchisees have to pay a $50,000 fee and
upgrade their restaurants to current standardstypically paying around $325,000.
Fewer than 50 percent of the franchisees were able to or willing to pay the costs
for a full renewal. BKC also experienced problems collecting royalty payments
from many franchisees and the company initiated a program to assist strong stores
and help close the weaker ones. In 2006, BK went public again with an IPO. In
2010, BK returned to private ownership when it was purchased by 3G Capital
(Web site history). Less than a week after the takeover, Raj Rawal was replaced
as CIO by Heitor Goncalves (Betts and King 2010). Previously, Goncalves had
worked Anheuser-Busch InBev, both as director of Mergers and Acquisitions and
as a vice president of rewards and target setting.
Standardized Point-of-Sale Terminals
New point-of-sale (POS) terminals from NCR Corp. have provided more than 130
Burger King restaurants in Canada with a standard method of order taking and
data entry. Burger King Corp. has already installed NCRs Compris food service
software, support services, and RealPOS 7454 terminals in more than 600 com-
pany-owned restaurants in the United States. It is wrapping up deployment in its
259 company locations in Canada, the United Kingdom, and Mexico. The goal
of the deployment is to streamline equipment across corporate-owned restaurants
around the globe, according to Michael Lingswiler, director of technical services
with Burger King. The drive behind this is standardization. The NCR platform
provided the functionality and given where they are in the market are able to pro-
vide for future releases (Hilson 2003).
A standard interface for all restaurants enables changes in the menu or special
promotions to be easily linked with the corporations back-end systems. Training
can be standardized and deployed across all regions of the company. Simplicity,
durability, and environmental friendliness are important in these fast-paced envi-
ronments. Prior to the standardization effort, Burger King had a mixture of NCR
and non-NCR equipment in their restaurants. Typically, these terminals have a life
cycle of about three years. While franchise restaurants are responsible for their
own POS hardware/software selections, they often adopt the corporate-owned
model for ease in purchase and maintenance.
In its 2006 Annual Report, Burger King noted that it had selected three primary
POS systems as preferred providers, but it does not require franchisees to choose
a particular hardware and software package. Data collection from franchisees is
still handled manually. By 2010 (annual report), the new POS systems had been
installed in all company-owned restaurants and 57 percent of the franchises. All
restaurants are required to install the new POS before January 1, 2014. Until that
time, most of the restaurant are still reporting sales manually and BK has only
incomplete data on sales.
Computer Security
Given its worldwide status, Burger King wanted to use automation to help it ef-
fciently assign and manage the identities and network privileges tied to enterprise
and online initiatives. In our opinion it is a necessity, says Burger Kings chief
information offcer, Rafael Sanchez. Such tools are needed because every orga-
nization has legacy systems, and most legacy systems have their own security.
(Liddle October 2003) In 2003, the installation of Oblix NetPoint software for
enterprise Web access and identity management purposes was the latest devel-
opment in a sweeping Burger King information technology project. That under-
taking, for which Burger King involved consultant PricewaterhouseCoopers, is
aimed at improving internal network security and controls over fnancial state-
ments and enhancing the chains performance by better empowering employees
and improving relationships with franchisees, and vendors.
According to Burger King offcials and PricewaterhouseCoopers documents,
the latter two goals can be achieved by making available to feld personnel, fran-
chisees and suppliers certain business support applications and information once
accessible only by select headquarters or regional staff. Such information shar-
ing can take place through an Internet portal that might also beneft Miami-based
Burger King by acquiring additional higher-quality operations information from
franchisees, a PricewaterhouseCoopers case study of the project suggested.
Oblix NetPoint supports single sign-on for network users, or the consolida-
tion of user-ID and password information for multiple Web-based applications
or applications with Web front ends. It works with the Active Directory feature
of Microsofts Windows Server to streamline and automate several of the steps
needed to make changes to user identity information and access privileges. The
NetPoint software supports self-registration by users and the delegation of some
administrative duties to certain classes of end users, such as department heads. It
also ties access and privileges administration into a user organizations workfow
routine. That makes possible scenarios such as one in which network access is
immediately revoked for any employee subject to a termination notice from the
companys human resources department.
Under the former security plan at Burger King, different legacy applications
required or permitted administrators to create log-in IDs of passwords different
from those used for other programs. That, CIO Sanchez says, made it possible for
individual users to wind up with multiple log-on IDs or passwords or both. In the
Burger King information technology realm a person may have access to from
fve to 15 different applications, Sanchez explains. Those programs, he adds, in-
clude such things as basic network access and applications tied to sales, fnance,
and franchise-related matters.
Sanchez says he or someone else in his department had to change log-on IDs
and password information and access privileges for a variety of applications
whenever an employee joined the company, left the company, was given addi-
tional responsibilities, or was stripped of duties. Because disgruntled terminated
employees are a potential threat to company resources accessible via a network,
Sanchez says of the old security plan, I had to pray that whoever provided access
[to the employee in question] let me know about the termination. Creating an
identity management infrastructure allows you to manage the [network] environ-
ment a lot more effciently, Sanchez says. Everyone has one user ID and pass-
word for everything they use.
According to Sanchez, under the new system the primary user of a particular
application, such as a department head, assigns access and privileges to people
within his or her sphere of infuence. Such assignments, however, must be in
keeping with the protocols and security parameters established by the confgura-
tion of NetPoint and the underlying identity management infrastructure, he indi-
cates. Dealing with outside communitiesthat is where we will use the power of
Oblix [NetPoint] for self-administration, Sanchez remarks, referring to resources
he expects to save by delegating administration of identity and privileges to sup-
pliers and franchise groups with extranet access. The concept, he says, is that he
might say to a supplier, Ill give you access for up to 20 people, but the internal
assignment within your company is your responsibility.
Human resources can clue us [to personnel changes] in house, but when your
partners are outside, it becomes even more diffcult to determine who should
have access or who should have access terminated, Sanchez observes. Burger
Kings goal, Sanchez reports, is to reduce the time spent administering network
security, while increasing the companys ability to implement new security strate-
gies. To help achieve that goal, the Miami-based chain used PricewaterhouseC-
oopers to help build a modern directory and identity management infrastructure.
Among other things, Burger King will use its new identity management ca-
pabilities to help support portal access to the companys SAP R/3 suite of appli-
cations. Offering a wide range of end users access to business support software
through a portal with solid identity management underpinnings allows us to ex-
tend the applications to the outer edge of the Burger King universe, Sanchez
says. Technologies like identity management that drive costs down and increase
employee, franchisee and [business] partner satisfaction give companies in the
restaurant industry a competitive edge, Sanchez states. Because the use of Net-
Point will reduce to one the number of log-on IDs and passwords used by Burger
King employees, it should be easier for those workers to remember that informa-
tion. And that, Sanchez says, should help reduce costs, or at least the workload,
associated with the chains information technology help desk.
Te Internet
Burger King sometimes runs off-beat advertising campaigns. In 2009, they cre-
ated a new campaign that involved Facebook. The company offered people free
hamburgers if they deleted 10 online friends, calling it the Whopper Sacrifce.
Facebook took offense at the campaignpartly because the BK site sent messag-
es to the people being deleted telling them they were sacrifced for a free burger
(Gaudin 2009). About 60,000 people had taken advantage of the free burgers in the
frst week, but BK decided to stop the campaign after discussions with Facebook.
Questions
1. Why is security so important to Burger King?
2. Why is technology standardization so critical?
3. Do all franchises need to worry as much about security, or just the fast-food
industry?
Additional Reading
Betts, Mitch and Julia King, CIO Watch, Computerworld, November 3, 2010.
Gaudin, Sharon, Facebookhas Whopper of a Problem with Burger King
Campaign, Computerworld, January 15, 2009.
Hilson, Gary, Hardware Fit For A King: Restaurant Chain Upgrades POS
Terminals Across Canadian Locations In Quest For Standardization,
Computing Canada, Vol 29 (3), February 14, 2003, p. 19.
Liddle, Alan J., Operators Chew On Ways To Offer In-Store Internet Access,
Nations Restaurant News, Vol. 37(32), August 11, 2003, pp. 1,69.
Case: Wendys
Driven by expansion and new product offerings, U.S. annual sales gains at Wen-
dys (ticker: WEN) Old Fashioned Hamburgers outpaced those of McDonalds
from 1998 through the end of 2002. According to Nations Restaurant News, Wen-
dys market share in the quick-service sandwich sector rose to 11.4 percent in
2002, from 10.3 percent in 2000. While the companys expansion enabled it to
continue capturing market share in 2003, same-store sales at its restaurants were
down, year to year, through July. This decline was due to a change in comparisons
from 2002 and competitive pressures, particularly from a recovering McDonalds.
Systemwide sales for 2006 totaled $7.8 billion. Same-store sales increased by less
than one percent, but proft margins at company-operated restaurants increased
slightly.
The company has sustained its image as selling high-quality products in the
quick-service sector and continues to enhance its reputation for offering a diverse
selection of sandwiches. The company has also maintained industry-leading cus-
tomer satisfaction scores. Future plans include an increase in national advertis-
ing spending targeting the growing number of Hispanics and late-night customers.
Though the company has stalled for the near term, its strong brand image has
prepared it for the longer term. In 2006, the company spun off the Tim Hortons
and Baja Fresh subsidiaries, and approved the sale of its Caf Express stores. Ap-
parently copying McDonalds, the company used the funds to repurchase more
than $1 billion in shares, hoping to prop up the stock price. According to Wendys
annual report, the company intends to expand its breakfast offerings to more store
in an attempt to grow revenues in this newer segment. The company recognizes
that it needs to work to rebuild its image. It also needs to reduce expenses, partly
through selling some of the company-owned stores. In 2007, Wendys executives
foated the news that it was willing to be purchasedpossibly by a private equity
frm.
In 2011, Wendys sold off the Arbys chain for $130 million to a private equity
frm. The Arbys chain had suffered during the 2008 recessionlargely because
its sandwiches are more expensive than those at other chains. Wendys will ini-
tially retain an 18.5 percent share of the new company.
Customer Cards
Under the direction of Scott McClenahan, Wendys in Redwood City, Utah, is
collaborating with Visa USA in a payment card acceptance pilot test. Wendys is
using a Verifone Omni 3200 customer-activated, credit-card-payment terminal on
the front counter and a Verifone Everest terminal at the menu board in the drive-
through. In confguring the Everest for the drive-through, Verifone developed a
weatherized case for the terminal and attached a MagTek dual-head dip card
reader to minimize instances of incorrect insertion that could slow transactions.
The drive-through Everest terminal served as a remote link to the countertop
Omni model. The two units were connected by a cable that ran underground from
the restaurant to the outdoor terminal at vehicle-window height near a red LED
order-confrmation board.
To keep transaction times as low as possible, most quick-service chains have
limited their initial tests of card acceptance to credit cards and check cards that
can be processed without requiring card users to sign receipts or enter personal
identifcation numbers. As a result, drive-through point-of-order payment termi-
nals often came without numeric keypads. The latest card-acceptance platforms
combine a payment terminal supporting PIN-based debit transactions with LCD
order-confrmation systems. A 2.4-gigahertz, wireless transmitter with encryption
capabilities from AeroComm is now being used to transfer payment terminal data
to the POS system.
Drive Trough
Drive-through orders add up to 75 percent of a fast-food restaurants sales. But
most restaurants have only one order-taking station and that slows down the pro-
cess. You have undoubtedly sat in the lines waiting for the person in front of you
to stumble through an order. Miami Management, Inc., a company that owns 16
Wendys franchises places two order stations in the drive-through lane. All of
them are connected to an order center in Lexington, KY through voice-over-Inter-
net technology. The orders are entered into the point of sale system and displayed
at the appropriate restaurant. Brian Fields, director of operations noted the system
increased peak lunch hour performance from 117 to 137 cars per hour (Mitchell
2006). But all was not perfect. In at least one restaurant, the manager reported that
the call center employees sometimes got orders wrong or failed to greet custom-
ers (Mitchell 2007). Part of the problem was that restaurants chose to use lower-
priced DSL phone lines instead of dedicated T1 connections. And in many loca-
tions, the communication quality was inconsistent or poor.
Questions
1. Why has Wendys been more successful at identifying customer desires than
Burger King?
2. How do most customers pay for fast food? How much are they willing to pay
in service charges to use ATMs in the restaurant?
Additional Reading
Jargon, Julie and Annie Gasparro, Wendys Parts with Arbys, The Wall Street
Journal, June 14, 2011.
Mitchell, Robert L., Would You Like Wi-Fi with that Burger? Computerworld,
April 13, 2007.
Mitchell, Robert L., Derailed by Defcient DSL, Computerworld, November
26, 2007.
Case: Yum! Brands
After several years of closing unproftable stores and selling company-owned
units to franchisees, Yum! Brands Inc. (ticker: YUM) has shown renewed vigor.
The owner of the KFC, Pizza Hut, and Taco Bell brands acquired the Long John
Silvers and A&W restaurant brands in 2002 as part of its multibranding restaurant
strategy. The company is aggressively expanding internationally as well, particu-
larly in China, Mexico, Korea, and the United Kingdom. The companys diversi-
fed portfolio has enabled Yum! Brands to successfully grow its business despite
diffculties that may occur at any one of its restaurant brands. Through late August
2003, strength at Taco Bell had helped to offset diffculties at KFC, while over-
all proftability was aided by a weaker dollar. Today, Yum! Brands, with 33,000
restaurants in fve major chains and 2006 revenue of $9.6 billion with worldwide
total system sales of $31.1 billion, is the worlds largest restaurant operator in
terms of units (34,595 in 2006). Yum has focused heavily on international growth,
including KFC in China. The potential sales to over 1 billion customers is appeal-
ing to any company, but KFC has had to weather challenges with supplypartic-
ularly with SARS, the avian virus. According to the companys 2006 Annual Re-
port, Pizza Hut is already the leader in casual dining in the UK and KFC and Pizza
Hut are the number 1 quick-service brands in mainland China. In 2011, Yum! an-
nounced that it was selling the A&W and Long John Silvers chains. The company
also announced plans to expand in China by acquiring the Little Sheep chain. Yum
already owned 20 percent of the company, and profts from China were driving
proftability for the entire company in 2010 (annual report).
The restaurants are trying to lift traffc and sales around the world by adding
ambience, quality, and service to a business that largely ignores such niceties.
Emphasis is placed upon training at the counter and upgrading the restaurants in
terms of equipment and food. The average checks in the restaurants are $3 to $4 at
Taco Bell and $5 to $7 at KFC.
While the burger chains duke it out on price, KFC, Pizza Hut, and Taco Bell are
adding and emphasizing better food, which, in the case of KFCs roasted chicken,
at least, also means lighter fare. Yum restaurants are pushing higher-priced items
to bolster dine-in business. Among them: Pizza Huts Chicago Dish pizza ($13)
and Taco Bells popular Southwest Steak Border Bowl ($3.50).
The CEO, David Novak, began an overhaul in 2000 in response to customer
feedback. Customers bluntly told the chains their service was shoddy, their food
subpar, and their restaurants, in some cases, shabby (Wells 2003). Novak, aged
50, still fumes when he recalls answering a call on Taco Bells toll-free complaint
line in early 2001. A woman named Michelle complained bitterly about getting
the wrong $3.60 order and a lot of attitude. I can assure you Michelle would
be justifed to tell everyone she sees about how poorly she was treated, Novak
exploded in an e-mail to restaurant managers after the call. This is the kind of
word-of-mouth that kills us.
David Novak hopes that acting on this feedback will make the company stron-
ger. Employees at all the chains around the world now attend training four times
a year at which customer-service initiatives are hammered home. They also get
evaluated, in part, on how they treat customers and react to problems when they
arise. Novak is pushing to slice service times, particularly at Taco Bell and KFC.
At Taco Bell a timer installed at drive-through windows beeps after 60 seconds.
This is the time in which an employee is supposed to spend taking and flling an
order. Novak blames some of the lingering problems at the restaurants on for-
mer owner PepsiCo. He feels Pepsi emphasized marketing at the expense of qual-
ity food, service, and atmosphere. It wasnt our schtick, PepsiCo President and
Chief Financial Offcer Indra Nooyi answers. The restaurant business wasnt our
schtick.
David Novak is orchestrating this dramatic shift to civilize the fast-food experi-
ence. At Taco Bell, where food is delivered to its 6,444 U.S. restaurants premade,
new $1,450 grills are being installed to cook new menu entries on-site. Restau-
rants in the Mexican-themed chain are also paying $16 million more a year for
better-quality beef, tortillas, and beans. The company is developing higher-end
ideas, such as Yan Can, the name of four new Asian-themed restaurants in Cali-
fornia created with Arthur Ho, a Hong Kong-based franchisee of KFC, and chef
Martin Yan of PBS Yan Can Cook fame. Average sales at these restaurants exceed
$35,000 a week, compared with just under $20,000 a week at a typical Taco Bell.
Yum! Brands is also combining Pizza Hut with a fast-casual chain called Pasta
Bravo in test markets.
Some Taco Bell franchisees argued against paying more for the new ingredi-
ents. I didnt think our customer base was that discerning, says Ned Kirby, a
franchisee in Noblesville, Indiana But they noticed the better food and didnt
resist the higher prices (Wells 2003). Not every trial restaurant is successful. Bell
Grille was an experimental restaurant the company opened in Garden Grove, Cali-
fornia to test higher-end menu items, such as smoothies. According to Emil Bro-
lick, president of Taco Bell, it was short-lived, because the food did not ft in with
the restaurant, which looked like a regular Taco Bell.
Novak is also pushing to pair the different Yum! Brands restaurants in multi-
branded units, which offer unit sales that are 20 percent higher than that of stand-
alone restaurants. With just 1,870 twofers open, the goal is to have 6,000 by 2007.
Novak has focused on new designs for the multibranded units, emphasizing what
traditionally makes fast-casual chains popular. The new designs include inviting
lighting with sleek sconces that are conducive to reading or hanging out. Whimsi-
cal murals fuse the personalities of two restaurants when they are combined into
one unit. Fast-casual chains make a statement. They spend little on advertising
but a lot on the dining experience. If we apply this kind of thinking to our cat-
egory we think it will give us some edge (Wells 2003).
Given the weakness in the market sector and the focus on fast-food compa-
nies vulnerability to new lawsuits, Yums stock has recently fallen. Since 1997,
revenue has sagged from a high of $9.7 billion. Since that time, operating income
has more than tripled to $891 million in 2001. Debt has been slashed from $4.6
to $2.1 billion. Emphasis continues to be placed on refurbishing the restaurants
themselves in a concentrated program.
Yum has experienced problems with its supply chain, from SARS in China to
E. coli in lettuce in New England in December 2006, costing the company at least
$20 million in proft (Annual Report). Sales growth in the U.S. was fat in 2006,
but 4 percent worldwide. Almost all of the growth was through new stores. Pizza
Hut has to compete against all of the fast food chains, the national pizza chains,
and the local pizzeria. Dominos CEO David Brandon said that our industry has
seen a shift where all of the national brands have lost market share and seen sig-
nifcant reductions in traffc growth at the expense of regional and local pizza
shops (Gibson 2007). Papa Johns CEO also notes the lack of product innova-
tion. On the other hand, Papa Johns is promoting online ordering in an attempt to
build brand loyalty.
In 2005, Alinean, a research company, awarded Yum! Brands the number 3 spot
in the retail industry in terms of return on IT investment. Dividing economic value
added by its IT spending gave it a value of 201 percent, not to far below the 290
percent of the number 1 company (Next Group PLC) (Betts 2005).
One of the challenges of Yum! Brands is that the company actually consists of
multiple concept restaurants. Finding common ground and getting 1.6 million
global employees to work together is a challenge. In 2011, Dickie Oliver, head of
global IT at Yum! Brands emphasized a four-point strategy for coordinating work-
ers. He created the iChing social network to enable employees to post documents,
ask questions, and collaborate. Similarly, an enterprise search system from Coveo
enables workers to search for answers and data. An online learning system from
Saba makes it easier to distribute training. And a high-defnition videoconferenc-
ing system from Tandberg lets employees hold virtual meetings to reduce travel
time and costs.
Questions
1. How does Yum! Brands use information technology to improve effciency?
2. With the mix of restaurants, how could Yum! Brands use IT to determine the
best store combinations and selection of items to sell?
Additional Reading
Betts, Mitch, Top Retailers for IT ROI, Computerworld, November 22, 2005.
Betts, Mitch, How One IT Chief is Building a Know-How Network, April 18,
2011.
Gibson, Richard, Pizza Chains Cook Up Ways To Battle Old, New Rivals, The
Wall Street Journal, May 30, 2007.
Wells, Melanie, Happier Meals, Forbes, Vol. 171(2), January 20, 2003, pp. 76-
78.
Case: Starbucks
Since 1985, Starbucks (ticker: SBUX) is committed to offering the highest qual-
ity coffee and the Starbucks Experience while conducting its business in ways
that produce social, environmental, and economic benefts for the communities
in which it does business. In addition to its retail operations, the company pro-
duces and sells bottled Frappuccino coffee drinks, Starbucks DoubleShot coffee
drink, and a line of super-premium ice creams through its joint venture partner-
ships. The companys brand portfolio provides a wide variety of consumer prod-
ucts. Tazo Teas line of innovative premium teas and Hear Musics exceptional
compact discs enhance the Starbucks Experience through best-of-class products.
The Seattles Best Coffee and Torrefazione Italia Coffee brands enable Starbucks
to appeal to a broader consumer base by offering an alternative variety of coffee
favors. Some of these products are now sold directly to consumers as Starbucks
VIA Ready Brew and through national food distributors SYSCO Corporation and
US Foodservice.
An interesting feature of Starbucks is that it does not franchise its stores. Star-
bucks has grown to 8,833 corporate stores plus 8,025 licensed retail stores inter-
nationally in 2010 from only 125 stores in September 1991. Starbucks purchases
green coffee beans for more than 50 blends and varieties from coffee-producing
regions worldwide. All green coffee beans purchased are of the Arabica species,
which is of higher quality than the Robusta species typically found in supermarket
coffees. Starbucks custom roasts the coffee beans to exacting standards. To add
sales margins, the company stores offer a wide selection of coffee-making equip-
ment, accessories, pastries, and confections.
In fscal year 2010, retail stores accounted for 84 percent of net sales. Stores are
typically clustered in high-traffc, high-visibility locations in each market. This in-
cludes offce buildings, downtown and suburban retail centers, and kiosks placed
in building lobbies, airport terminals, and supermarkets. In fscal year 2002 the
retail store sales mix by product type was 77 percent beverages, 13 percent food
items, 6 percent whole bean coffees, and 4 percent coffee-related hardware items.
Starbucks has expanded its retail business by increasing its share in existing
markets, and opening stores in new markets in which it sees an opportunity to be-
come the leading specialty coffee retailer. The company opened a net total of 810
company-owned stores in fscal year 2006 and planned to open a similar number
in fscal year 2007. But in the U.S. it closed 474 stores in 2009 and 57 in 2010.
Starbucks has tried to use its Specialty Operations, which accounted for 16 per-
cent of total revenues in fscal year 2003, to develop the Starbucks brand outside
the company-owned retail store environment. Starbucks has licensing agreements
(35 percent of specialty revenues) with prominent retailers in North America,
Central America, Europe and Asia. The company has about 5,600 food service
accounts (27 percent), where it sells whole bean and ground coffee to various cof-
fee distributors, hotels, airlines, and restaurants. Starbucks has a licensing agree-
ment (13 percent) with Kraft, Inc., which markets and distributes the companys
whole bean and ground coffees in the U.S. grocery channel. In addition, the com-
pany sells its coffee products through warehouse club accounts (13 percent), and
through mail order and online (7 percent) (Standard and Poors, Starbucks annual
report). The companys long-term plans are to run 20,000 stores in the U.S. and
another 20,000 worldwide, with an emphasis on China. Although the Annual Re-
port observes the increasing diffculties in managing that many stores, comparable
store sales grew by 7 percent in 2006; but hints are that the growth rate has slowed
in 2007. The company also spent $1 billion buying back its stock to prop up the
price (Blumenthal 2007). From 2007 through 2009, Starbucks was hit hard by the
recession where sales growth slowed and fell in 2008 and 2009 (annual report
2010). Howard Schulz returned to the CEO position in 2008 to help organize the
response to competition and the recession.
In the fourth quarter 2003, Starbucks acquired Seattle Coffee Company, which
includes Seattles Best Coffee and Torrefazione Italia coffee brands, for $72 mil-
lion. Negotiating and controlling the supply of its products is a key factor in the
Starbucks operation. The company also spent $6.5 million in 2006 on research to
create and improve products as well as operations (Annual Report).
When the top management was replaced in 2008, Starbucks also named Chris
Bruzzo to be acting CIO and asked him to create innovative ways for Starbucks
to connect with our customers and build loyalty programs (Wailgum 2008). At
the time, the company had grown too rapidlyunfortunately just as the economy
was entering the recession. The frm was focused on growth through adding new
stores, so the technology focused on evaluating real estate, location, and demo-
graphics. Instead, the company needed to fnd a way to encourage each customer
to spend more money.
In 2009, Stephen Gillett was given the job of CIO at Starbucks. He has an IT
staff of 1,000 people. The size of the staff is almost a problemit is diffcult
to fnd and hire good workers. And he still faces one key area: business intelli-
gencefnding people and tools that can analyze the customer data (Mann 2009).
Stored Value Card
Brian Cyrnes, senior vice president and chief information offcer, is proud of the
coffeehouse chains stored-value, customer loyalty-enhancing Starbucks Card and
Duetto. The new Duetto offering is a combination Starbucks Card and Visa credit
card. This was accomplished by keeping an eye toward offering the best service
and value, while targeting competitive advantages, speed-to-market issues, and
enterprisewide integration. The launch follows the success of the reloadable Star-
bucks Card, of which more than 11 million have been activated since the card
began in November 2001. A Starbucks Visa credit card, incorporating the stored-
value feature, was launched in association with Bank One in March 2003.
The Starbucks Card program was run by an outside vendor, ValueLink, a First
Data Corporation. First Data is an electronic payment service specialist. Custom-
ers buy the cards in dollar amounts to be redeemed like cash at the checkout line.
ValueLink processes the transactions and manages an offsite database for the sys-
tem. The program is already a big loyalty winner. Starbucks redeemed 11 mil-
lion cards worth $41 million in the recent second quarter. Starbucks outsourced
the processing because it felt it lacked the knowledge to develop a complex cus-
tomer service program on its own.
Wi-Fi technology
Based on its experience at thousands of Starbucks Coffee outlets, Starbucks is
convinced that providing Internet access to guests will build customer loyalty and
sales. As a result, Starbucks is forging ahead with deployment of wireless local
area networks, as other food service chains accelerate rollouts or expand tests of
the technology. Starbucks Corp. launched Wi-Fi service in its U.S. cafes in August
2003 and now offers access in over 2,000 locations. Users have to sign up for the
service with Bellevue, Washington-based T-Mobile USA Inc., a unit of Deutsche
Telkom, whose prices start at $19.99 per month. The T-Mobile HotSpot service
is backed by reliable high-speed T-1 connections that can accommodate the full
spectrum of applications from checking e-mail to viewing rich multimedia and
video. T-Mobile HotSpot window signs are visible near the entrance of all partici-
pating Starbucks locations. A complete list of stores can be found by visiting the
Starbucks store locator on www.starbucks.com/hotspot and selecting Wireless
HotSpot Stores as the Store Type.
Today the chain has fee-based hot spots managed by strategic partner T-Mobile
International at about 2,600 of its 3,854 company-operated coffee bars. The goal
is to install hot spots in about 70 percent of the retail sites operated by the compa-
ny. Size, seating capacity, and the demographics of a market all determine which
stores will incorporate the technology. Starbucks looked at a number of alterna-
tives in terms of how to provide in-store Internet access, but decided wireless was
the way to go to provide access without disrupting the other customers. The chain
did not want to become known as an Internet caf littered with cabling and
other hardwired network paraphernalia.
The wireless service is already paying dividends at Starbucks. The company
charges customers about $6 an hour for the T-Mobile service. The typical wireless
user stays for 45 minutes. Of the nonwireless users, 70 percent spend 5 minutes or
less at the store; the remaining 30 percent linger for about 20 minutes. We cer-
tainly believe that means they buy more coffee and food. The real exciting thing
for us about the hot-spot service is that it is bringing people into our stores at dif-
ferent times of the day, and hot-spot users are staying longer, Nick Davis, Star-
bucks spokesman, remarked. (Liddle August 2003). Typically, a Starbucks unit is
busiest from 6 a.m. to 9 a.m., so building traffc later is a positive development.
About 90 percent of hot spot users arrive after 9 a.m., and that group, on average,
spends about 45 minutes online and in the stores.
In the attempt to leverage its wireless network installations further, Starbucks
in 2004 began testing possible synergies between its hot spot strategy and the dis-
tribution of product by its music industry holding, Hear Music. To spur additional
hot spot registrations and publicize Hear Musics Artists Choice compact disc
of Sheryl Crows favorite songs by other recording stars, Starbucks hot spot us-
ers were given Web access to audio clips of the songs and Crow interviews on the
CD. They were permitted to download three complete songs from the compilation
product.
By 2010, partly pressured by McDonalds, Starbucks opened the Wi-Fi to free
and unlimited access. Other coffee shops fnd Wi-Fi to me more challenging.
Freeloaders might buy a cheap cup of coffee and then take up space for hours
making it hard for paying customers to fnd a place to sit (Elgan 2010). Starbucks
will also offer free access to major newspapers such as The Wall Street Journal
and New York Times.
Starbucks and the Sundance Film Festival
In addition to being the offcial coffee at the 2004 Sundance Film Festival, Star-
bucks worked closely with the Sundance Institute to sponsor the 2004 festival, as
well as bring the excitement of the Sundance Film Festival experience directly
to customers in its stores. Throughout the month of January, customers who vis-
ited their favorite participating Starbucks location with a Wi-Fi (802.11b)-enabled
notebook computer or Tablet PC experienced the passion and inspiration of the
2004 Sundance Film Festival frsthand. By simply launching the Internet browser
on their device, customers watched exclusive flmmaker interviews, behind-the-
scenes video clips, and flm trailers for free prior to logging on to the T-Mobile
HotSpot wireless broadband Internet service.
Customers who logged on to the T-Mobile HotSpot service received a free day
pass to the Sundance Online Film Festivalthe online extension of the 2004 Sun-
dance Film Festival. The free day pass enabled Starbucks customers to access the
Sundance Online Film Festival, where they could cast a vote for their favorite ani-
mation and short subject flms while in the comfort of any participating Starbucks
location. Independent flm afcionados and consumers did not have to be in Park
City to experience the festivalthey only needed to fnd the nearest T-Mobile
HotSpot-enabled Starbucks to access the exclusive content.
Starbucks sponsorship of the flm festival enabled them to add value to the
T-Mobile HotSpot service in their stores by giving their customers the ability to
interact directly with digital entertainment. Since Starbucks has sponsored flm
festivals large and small, this next step enabled the company to share their enthu-
siasm for independent flm directly with their customers. Starbucks expanded its
involvement in the entertainment industryeven creating a separate subsidiary.
In 2006, it promoted a movie to its customers, but the effect was minimal. In
2007, Starbucks continued its sales of music CDs with a special solo album from
Paul McCartney. Starbucks also backed another movie in 2007 (Adamy 2007).
Analysts are mixed over whether the entertainment options attract customers and
encourage sales, or if they just distract management from selling coffee.
Phones
Like many other companies, Starbucks sells customers a prepaid card that can be
used for future purchases. In 2011, Starbucks released an app that lets custom-
ers store their prepaid card data on their smart phone. Cashiers simply scan the
bar code from the phone and deduct the amount from the balance (Segall 2011).
Because the Starbucks card is linked to a credit card, the payment card can be re-
loaded. The smartphone app includes the ability to easily add more money to the
card (Hamblen 2011).
In a different twist, Starbucks in 2010 became one of the frst companies to pay
Twitter to deliver Promoted Tweets, or advertising over Twitter. The marketers
will be charged per thousand impressions of their ads that are displayed on their
Web site alongside search results. Starbucks already publishes Twitter notes about
its promotions, but the ads will always show up near the top of the page, making
them more visible (Miller 2010).
Questions
1. Has Wi-Fi access improved sales and profts at Starbucks?
2. Given the huge number of stores, how could Starbucks use information
technology to determine where to open new stores and which ones to close?
Additional Reading
Adamy, Janet, Starbucks Sticks with Film-Promotion Plan, The Wall Street
Blumenthal, Karen, One Stock, One Year: What I Learned, The Wall Street
Journal, July 1, 2007.
Elgan, Mike, Why Starbucks Will Win the Wi-Fi Wars, Computerworld,
August 14, 2010.
Hamblen, Matt, Starbucks Extends Mobile Payments to 7,500 Stores,
Computerworld, January 19, 2011.
Liddle, Alan J., BK Seeks Simplifcation of Complex Network, Online Identity
Management Process, Nations Restaurant News, Vol. 37(40), October 6,
2003, p. 36.
Mann, John, Starbucks CIO Tells How He Picks New Hires, Computerworld,
January 14, 2009.
Miller, Claire Cain, Twitter Unveils Plans to Draw Money from Ads, The New
York Times, April 12, 2010.
Segall, Laurie, Your Mobile Phone is Becoming Your Wallet, CNN Online,
January 19, 2011.
Wailgum, Thomas, How IT Systems Can Help Starbucks Fix Itself, CIO,
January 25, 2008.
Summary Industry Questions
1. What information technologies have helped this industry?
2. Did the technologies provide a competitive advantage or were they quickly
adopted by rivals?
3. Which technologies could this industry use that were developed in other
sectors?
4. Is the level of competition increasing or decreasing in this industry? Is it
dominated by a few frms, or are they fairly balanced?
5. What problems have been created from the use of information technology
and how did the frms solve the problems?
1
Part
Infrastructure
How is information technology changing and how do you evalu-
ate and choose technologies for business? How do you share in-
formation and collaborate? How do you store and retrieve data?
These are fundamental questions that any manager needs to be
able to answer. Information technology forms the foundations of
management information systems. As a manager, you need to un-
derstand what technologies are available today, and how they are
likely to change tomorrow. Although you do not have to be able
to build your own computer, you need to understand the trends
and limitations of computing equipment.
Networks are critical in most organizations today. The Internet
continues to expand into professional and personal lives. Manag-
ers should understand how wired and wireless technologies can
be used to share data and communicate. Particularly in terms of
the Internet, you need to be able to evaluate emerging technolo-
gies and evaluate costs.
Database management systems are probably the most im-
portant tools in business today. Managers can no longer rely on
programmers to answer business questions. Instead, you need
to know how to retrieve data and build reports from central
databases.
Chapter 2: Information Technology Foundations
Chapter 3: Networks and Telecommunications
Chapter 4: Database Management
What are the main features and tools of information systems?
65
What types of computers are needed for business applications?
What are the basic objects that computers process?
How do computers handle music and video?
What are the main components of a computer?
Why is the operating system so important?
How does the Internet change the role of computers?
What are the main software applications used in business?
Is it possible to handle all computing tasks with online services?
Chapter Outline
Foundations
2
Chapter
Introduction, 66
Types of Data, 69
Object Orientation, 70
Numbers and Text, 72
Pictures, 74
Color, 78
Style Sheets, 79
Sound, 81
Video, 82
Size Complications, 82
Data Compression, 83
Hardware Components, 84
Processors, 86
Input, 92
Output, 95
Secondary Storage, 95
Operating Systems, 97
Computers in e-Business, 99
What Is a Browser?, 100
What Is a Server?, 102
Virtual Machines, 104
Application Software, 104
Research: Databases, 105
Analysis: Calculations, 105
Communication: Writing, 107
Communication: Presentation and
Graphics, 107
Communication: Voice and Mail, 108
Organizing Resources: Calendars and
Schedules, 108
Augmented Reality, 109
The Paperless Offce?, 110
Open Software, 110
Summary, 113
Key Words, 114
Exercises, 116
Cases: The Computer Industry, 121
66 Chapter 2: Information Technology Foundations
Dell
Introduction
What types of computers are needed for business applications?
Figure 2.1 shows the challenge of the choices. A wide range is available from
portable devices like cell phones and tablets to large-screen desktops to enterprise
servers and super computers. The interesting aspect is that at each level, the per-
formance and capabilities have continued to improve. So todays desktops have
the power of super computers from a few years ago. You are probably familiar
Should you buy new computers? How is information technology changing and how
do you evaluate and choose technologies for business? These are diffcult questions
to answer for any business. Buying one new computer is not the problemreplacing
hundreds or thousands of computers is expensive. One of the keys to changing tech-
nology is that buying a computer is not a one-time event. Organizations need to plan
on buying new machines on a regular basis. Of course, you also have to think about
how to dispose of the old computers.
The computer industry is driven by change and growth. An interesting feature of
computers is that they have a short shelf life. A computer that sits unsold for a few
weeks quickly becomes obsolete as more powerful or cheaper machines are intro-
duced. From a manufacturers perspective, this pattern of change is diffcult to han-
dle. It is even harder when you realize that people often want to try out a computer
frst. At one time, Dell tried to sell computers through retail stores, but had to give
up. Dells solution has become a driving force in the industry. Dell tightly manages
the supply chain and builds each machine to order. Components such as disk drives,
processors, and cases are purchased in bulkoften with long-term contracts. The
majority of the components are delivered in small just-in-time batches to manufac-
turing facilities. The individualized machines are built and formatted for each spe-
cifc customer. Dell holds costs down by minimizing inventory and producing the
exact products ordered by customers. Dell risks losing sales to retail stores, but the
company knows it will make a proft on every machine.
Which options should you choose on a computer? Do you need the fastest proces-
sor, a huge amount of RAM, and a monster disk drive? Where can you cut perfor-
mance slightly and save money? These questions are hard enough for buyers to an-
swer, but cause headaches for vendors. Again, Dells made-to-order process reduces
the risks. But vendors like Hewlett-Packard, Lenovo, and Acer have shown that with
accurate predictions, it is possible to capture a substantial slice of the retail market.
The key is to have the combination of features that are demanded by customers, at
the right price.
Getting rid of old computer equipment could involve almost as big a decision as
purchasing a new computer. California and Massachusetts have banned computer
equipment from landflls. Europe and Japan have relatively strict recycling laws.
Vendors are struggling to create recycling programs. Dell has committed to increas-
ing the recycling rates by offering new programs that embed the cost of recycling
into the new equipment. Other vendors will take back old equipment for a fee of $30
to $50.
with common user devices such as laptops, cellphones, and tablets. These devices
are also important in business, but today they are essentially commodity items that
are purchased in bulk and replaced every couple of years. Servers require more at-
tention because they are more expensive and handle important tasks for the entire
company.
Some basic aspects of computing hardware have remained the same for many
years. However, the continued improvements have led to powerful changes in ca-
pabilities and physical size. Small, portable devices now have huge processing
and communication capabilities. These changes are leading to new applications,
new ways to build information systems, and new ways to interact with comput-
ers. Some of these effects are explored throughout this book. In terms of hardware
choices, the main impact in this chapter is the question of whether hardware se-
lection matters anymore. For example, most people in the U.S. essentially lease
their cell phones paying a small up-front price and paying for the device with
higher monthly contract prices. Consequently, it is relatively painless to replace
the equipment every two years. As more applications move to cell phones and
browsers, the choice of hardware becomes less important every year. Still, it will
be a few years before all hardware reaches this state. So, it is still useful to under-
stand some of the foundations of information technology hardware and software.
Trends
The frst computers were simple pieces of hardware. Like all computers,
they had to be programmed to produce results. Programming these machines
was a time-consuming task, because all of the pieces were connected by
individual wires. Programming the computer consisted of rearranging the
wires to change the way the pieces were connected. As computer hardware
improved, it became possible to program the processor without having to
change the internal wiring. These new programs were just a list of numbers
and short words that were input to the machine. These lists of commands
were called software, because they were easier to change.
Programmers soon learned that some tasks needed to be performed for
most programs. For example, almost every program needs to send data to
a printer or retrieve information stored on a disk drive. It would be a waste
of time if every programmer or user had to write a special program to talk
to the printer. By writing these common routines only once, the other pro-
grammers could concentrate on their specifc applications. As a result, every
computer has a basic collection of software programs called an operating
system. The operating system handles jobs that are common to all users and
programmers. It is responsible for controlling the hardware devices, such as
displays, disk drives, and printers.
The Internet changed everything by focusing on networks and sharing
data. The Web changed everything by making the browser the most impor-
tant display device. Increasingly, software and data are stored on Web serv-
ers. Users rely on browsers installed on a variety of clientsfrom desktops
to handheld computers, tablets, and cell phones.
Hardware trends have been astonishing compared to older industries.
Performance, size, and price have continued to improve each year. These
changes have created new, portable products and pushed advanced comput-
ing into everyday products.
Keep in mind that the demands for data storage and computational tasks con-
tinue to increase. So someone has to buy more computing power and data storage
on a continuing basis. Most businesses purchase information technology at mul-
tiple levels: devices for users, servers for business tasks and data storage, long-
run backup, and sometimes intensive analytical tasks. Each of these categories
requires different types of computers, and it means someone has to monitor the
investments and changes to technology.
Why do you care how computers work? After all, it is easy to use a photocopy
machine without understanding how it works. Automobiles cost more than most
computers, yet you can buy an automobile without knowing the difference be-
tween a manifold and a muffer. You can also make telephone calls without under-
standing fber optic cables and digital transmissions.
On the other hand, when you buy an automobile you need to decide if you
want options such as power windows, a turbo charger, or a sunroof. Almost every
product you purchase has a set of technical elements, and you can fnd an expert
(techie) who is more concerned about details than others. But, does it truly matter
to you? Perhaps not, but if you do not understand the options, you might not end
up with the car, telephone service, photocopier, or computer that you need. Or you
could end up paying extra money for services that you will not use. To choose
among the various options, you need to know a little about how computers work.
Computers are typically discussed in terms of two important classifcations:
hardware and software. Hardware consists of physical items; software is the
logical component such as a set of instructions. Of course, many functions can
Figure 2.1
Information technology changes constantly. Change makes it diffcult to choose
products and control costs as you upgrade over time. As a manager, you have to
evaluate the technology features and choose options that provide the functions you
need while minimizing costs.
Desktop: $400-2,000
Laptop:$600-2,000
Tablet:$500-2,000
Workstation: $2,000-7,000
Super computer: $1,000,000+
Enterprise Server:
$10,000-$1,000,000
Cell phone:$200-700
HP
Apple
Motorola
Sun (extinct)
Cray
be provided in either software or hardware, and a computer user often cannot tell
which has been used, and most often does not care. The one main difference is
that it is easier to make changes to software than to hardwareespecially since
software patches can be transmitted across the Internet. Before looking at the vari-
ous hardware components and software options, it helps to examine the primary
data types. You care most about the tasks a computer can perform, and the data
types are the fundamental blocks within any information system.
Types of Data
What are the basic objects that computers process? To a computer
user, the most important aspect of the computer is the data that it handles. Today,
computers are used to process fve basic types of data: numbers, text, images,
sound, and video. Because of limited speed and storage capacity, early comput-
ers could handle only simple text and numbers. Only in the past few years have
computers become fast enough and inexpensive enough to handle more complex
sound and video data on a regular basis. As computers continue to improve, these
new capabilities will alter many aspects of our jobs and society.
Single bit: one or zero (on or of)
8 bits = 1 Byte: 10101010
1 Byte holds values from 0 - 255
2
0
= 1
2
1
= 2
2
2
= 4
2
3
= 8
2
4
= 16
2
5
= 32
2
6
= 64
2
7
= 128
2
8
= 256
2
9
= 512
2
10
= 1024
2
20
= 1,048,576
Bytes bits Power of 2
1 8 256
2 16 65,536
3 24 16,777,216
4 32 4,294,967,296
8 64 18,446,744,073,709,551,616
Figure 2.2
Byte and bits. Binary counting imposes some limitations based on the size of the
counter. Note that 32-bit hardware and software can only address 4 GB of memory.
Windows 7/32-bit is limited to 3 GB. To use more RAM, you need 64-bit hardware
and software.
Object Orientation
One of the most important concepts to understand with computers is that all data
are represented as binary data. The computer processor can handle only binary
data in the form of bits. Each bit can be on or off or mathematically, zero or one.
Processors operate on groups of bits at a time. The smallest group of 8 bits is
called a byte. The industry often refers to the maximum number of bits that a
processor can handle. Common machines use 32-bit processors, but the industry
is shifting to 64-bit processors. A 64-bit processor can process 8 bytes of data in
one operation.
The number of bits (or bytes) determines the size of individual data that can be
handled or addressed. Figure 2.2 summarizes some of the values of binary num-
bers that commonly arise in computing. In particular, note that a 32-bit system has
a limitation of about 4 gigabytes (GB). Consequently, a 32-bit Windows operating
system can only address 3 gigabytes of RAM. If you want to use more memory,
you need to install 64-bit versions of the operating system and other software.
It can be diffcult to deal with huge numbers. So instead of trying to write out
strings of zeros and commas, people create names for different levels of numbers.
Figure 2.3 lists the main prefxes in use today and some that might be reached in a
few years. Notice that there are some differences between decimal and binary val-
ues because binary is based on 1024 instead of 1000. These differences were mi-
nor at small numbers but become larger with higher powers. Consequently, there
is some confusion when you see a number expressed as gigabytes or terabytes.
For instance, disk drive vendors commonly count storage space in decimal values
(because it sounds bigger); but the Windows operating system reports drive space
in binary values so a 512 GB drive (in decimal) only contains about 488 gigabytes
of space counted in binary. Drive vendors would probably be happier if operating
system programmers reported data in decimal counts instead.
Figure 2.3
Terms and prefxes. The IEC (electrical) has proposed using different names for
powers of two instead of decimal. The new versions are sometimes used in Europe
and have been adopted as standards. The goal is to reduce confusion.
Term Name Power
of 10
Power
of 2
IEC
term
Binary value
Kilo Thousand 3 10 Kibi 1,024
Mega Million 6 20 Mibi 1,048,576
Giga Billion 9 30 Gibi 1,073,741,824
Tera Trillion 12 40 Tebi 1,099,511,627,776
Peta Quadrillion 15 50 Pebi 1,125,899,906,842,624
Exa Quintillion 18 60 Exbi 1,152,921,504,606,846,976
Zetta Sextillion 21 70 Zebi 1,180,591,620,717,411,303,424
Yotta Septillion 24 80 Yobi
1,208,925,819,614,629,174, 706,176
In common business usage, the prefxes tera and peta are becoming relatively
common. Even desktops and laptops can have terabyte-sized drives. Some large
companies collect petabytes of raw data, which are a thousand times larger than
terabyte systems. A couple of companies have reached the exabyte scale for spe-
cialized data. Think about a million terabyte drives for a minute and you will be-
gin to see what happens as companies collect increasing amounts of data.
One of the key points in terms of data storage is that companies avoid delet-
ing data. It can be diffcult to know exactly which data might be valuable in the
future. Given falling prices for storage, many companies simply decide to keep
everything.
Beyond counting, humans do not deal well with bits or bytes. Real-world data
consists of analog objects such as numbers, text, images, sound, and video. Figure
2.4 illustrates the fve basic data types. Computers need input devices to convert
these objects into bits. Output devices then change this data back to a form that
humans can understand. Understanding these types makes it easier to see the in-
dustry trends and to understand applications.
Each of the basic data types has its own characteristics. Numbers have a preci-
sion and usually a scaling factor. Text has a typeface and display size, as well as
appearance attributes such as bold, underline, and italic. Pictures can be described
by their resolution and the number of colors. Digitized sound is based on a speci-
fed sampling and playback rate, and fts into frequency and amplitude ranges.
Video combines the attributes of pictures with a frames-per-second defnition.
Along with the attributes, several predefned functions can operate on each data
type. Basic functions dealing with numbers include totals, calculations, and com-
parisons. Standard methods to manipulate text include searching, formatting, and
Figure 2.4
Five basic types of data. Because processors only deal with binary data, each format
must be converted or digitized. Computer peripheral devices are used to convert the
binary data back to the original format.
Numbers
Images
Sound
Video
Input Process Output
12 + 8 = 20
000001100
000001000
---------------
000010100 20
0010000000000000000
0100000000000001001
0110000011000011011
0111111111111001111
1111111111111011111
1111111111100011111
000001000 000001001 000010100
pitch or
volume
Time
00101010111
11010101010
01010101010
11110100011
00101011011
8 9 0 5
00101010111
11010101010
01010101010
11110100011
00101011011
00101010111
11010101010
01010101010
11110100011
00101011011
00101010111
11010101010
01010101010
11110100011
00101011011
00101010111
11010101010
01010101010
11110100011
00101011011
Text
This is a test
84 104 73 115
This is a test
spell-checking. Methods dealing with pictures involve artistic alterations such as
color or lighting changes, rescaling, and rotation. Existing functions for sound and
video are often limited to recording, playback, and compression/decompression.
Most application packages create their own objects as combinations of the ba-
sic data types. For example, graphs consist of numbers, text, and pictures. As a
result, all of the attributes, and functions that apply to the base types also apply
to derived graph objects. Hence, graphing packages can perform calculations on
numbers, select fonts for displaying text, and enable you to set colors and rotation
for the graph. Other applications, such as slide shows, provide additional func-
tions, like controls for how one slide is replaced by another (e.g., fades, dissolves,
and wipes). The point is, once you understand the properties and functions of the
basic data types, you can learn how to use any new application by concentrating
on the new objects, their attributes and functions. This approach is especially true
since many applications are designed and built using object-oriented methods. In-
creasingly, operating systems are designed around these data types. They have
search functions and applications to fnd and use fles that contain specifc types
of data.
Numbers and Text
Numbers and text are still the most common types of computer data used in busi-
nessfor example, each cell in a spreadsheet can display a number or text (la-
bels). Most business reports consist of tables of numbers with supporting text.
As shown in Figure 2.5, a key aspect of numbers is that you must be careful with
round offparticularly within spreadsheets. Consider the frst column of numbers
with three decimal digits. What happens when the format command tells a spread-
sheet to display those numbers in the third column with only two decimal places?
The computer will perform the addition frst and round off the displayed result,
which means the total will appear to be incorrect. If you really want only two dig-
its of precision, you should use the Round function to convert each number before
it is added, as shown in the second column. Precision of calculations is impor-
tant in business. For example, the 1999 European Union standards require that all
monetary conversions be computed to six digits to the right of the decimal point.
Since processors can handle only binary data, how can they deal with text? The
answer is that they assign a number to each character in the alphabet, which raises
Reality Bytes: RAM: Smaller, Faster, Less Power
In early 2011, Samsung announced the production of the frst DDR4 RAM module.
Based on 30 nanometer (nm) lithography the module doubles the access speed from
DDR3s 1.6Gbits/sec to 3.2Gbit/sec. Samsung also noted a 40 percent reduction in
power consumption when used in notebook computers. At the same time, prices for
existing DDR2 and DDR3 RAM was continuing to decline. The market research
frm iSuppli noted that in the last six months of 2010, the contract price for 2GB
DDR3 dropped from $44.40 to $21. Mike Howard, a principal analyst at iSuppli pre-
dicted that DDR3 prices might drop as low as $1 per gigabyte, which would be less
than the marginal cost of production.
Adapted from Lucas Mearian, Samsung Announces First DDR4 DRAM Module,
a very important, but tricky, question: How many letters are there in the alpha-
bet? The question is tricky because you need to frst ask: Which alphabet? Until
the 1990s (or later), most computer systems were based on an English-language
assumption that there are only 26 letters in the alphabet (plus uppercase letters,
numbers, and punctuation). The key to numbering the alphabet is that all hard-
ware and software vendors have to agree to use the same numbering system. It is a
little hard for everyone to agree on a system that only displays English characters.
Consequently, most computers today support the Unicode system. Unicode is an
international standard that defnes character sets for every modern language and
even most dead languages.
Figure 2.6 shows that a key feature of Unicode is that it can handle oriental
languages such as Chinese and Japanese that use ideograms instead of characters.
The challenge with ideograms is that each language contains tens of thousands of
different ideograms. Character-based languages like English used to be handled
by storing one character number in one byte. But one byte can only hold 256 (2
8
)
numberswhich is not enough to handle ideograms. Consequently, Unicode uses
two bytes to count each character or ideogram, enabling it to handle 65,536 (2
16
)
Precision ROUND( ) Format
5.563 5.56 5.56
0.354 0.35 0.35
6.864 6.86 6.86
12.781 12.77 12.78
Time Languages Character Sets
Early U.S.andEngland ASCII and EBCDIC
127 characters, 7 bits/1 byte
1980s Latin-based characters: tilde,
accent,umlaut,...
, ,
Code pages and extended character
sets
255 characters, 8 bits/1 byte
1990s Asian ideograms, plus any
language

Unicode
All modern langauges and most dead
languages
1 character, 2 (or 3) bytes
Figure 2.5
Numeric precision. If you tell a spreadsheet to display two decimal-digits, it performs
the additions in the full precision of the dataand displays the data in the right
column. If you really want two-digit precision, you should use the Round(x, 2)
function to roundoff the data before addinggiving the results in the middle column.
Figure 2.6
Character sets. Standards defne how alphabets and ideograms are converted to
numbers. Early systems were based on the English language and extended to handle
Latin-based characters. Today, Unicode is the preferred way to encode text because
it can handle any language. However, it takes twice as much space to store Unicode
data.
ideograms in each language. It also has the ability to switch to three bytes if even
more characters or ideograms are needed.
Why do you care? Remember the increasing importance of international trade.
If you build a computer system or database that needs to be used around the
world, make sure that you use the Unicode character set so you can collect data
in multiple languages. Unfortunately, some systems by default still use the older
character systems, and you have to deliberately change the settings to ensure you
can handle other languages.
Figure 2.7 illustrates one of the more important properties of text: its typeface.
You can choose from several thousand typefaces to achieve a desired style. Be
careful with your choices. Some choices are obvious; for example, do not use an
ornamental (script) typeface to write a business report. Also, serif fonts are typi-
cally used for printed material, and sans-serif fonts are used for very small or very
large presentations. Other choices are more subtle and may require the assistance
of a professional designer. You will also have to choose font sizes. The basic rules
of thumb are that most printed text is displayed at 10 to 12 points. It is also useful
to know that letters in a 72-point font are approximately one inch tall.
Pictures
Pictures, graphics, and icons are used to display graphs and charts. Pictures are
sometimes used as backgrounds or as icons in presentations that are used to help
the audience remember certain points. In fact, almost all of the computer work
that you do today is based on a graphical perspective. Video screens and print-
ers are designed to display everything, including text, in a graphical format. The
graphical foundation provides considerably more control over the presentation.
For example, you can change the appearance of characters or combine fgures and
graphs in your reports.
Figure 2.7
Typefaces fall into two main categories: serif and sans serif. Serifs are little curls
and dots on the edges of letters. They make text easier to read; however, san serif
typefaces are useful for overheads and signs because the added white space makes
them easier to see from a distance. Ornamental typefaces can be used for headlines.
Size of fonts is measured in points. Characters in a 72-point font are about 1 inch tall,
and most books and newspapers use a font between 10 and 12 points.
Arial 20
Times New Roman 21
Century Schoolbook 16
Typeface Classification
Sans serif
Serif
Ornamental
Juice ITC 18
Script MT Bold 18
Garamond 24
Courier 18 (monospace)
A
72 points,
1 inch
leading
Two basic ways to store images are bitmap and vector formats. The easiest
to visualize is bitmap (or raster or pixel) format. A bitmap image is converted
into small dots or pixels (picture elements). If the picture is in color, each dot is
assigned a (binary) number to represent its color. This method is often used to
display photographic pictures where subtle changes in color (such as blends) are
important. Bitmap pictures are evaluated in terms of the number of colors and
resolution of the picture. Resolution is often measured in dots per inch (dpi).
You will encounter certain problems with bitmap images. Consider what happens
if you create a bitmap picture in a square that has 50 pixels per side. How do you
make the image larger? Use larger dots? Then the lines are very rough. How can
you stretch the picture into a box that is 100 by 200 pixels? Although these opera-
tions are possible, they do not usually produce good results.
Historically, it took several years, and the Web, for companies to standardize
on the types of image fle formats. Today, although individual software packages
still have their own proprietary formats, most of them can save and edit fles in
the three major formats: GIF, JPEG, and PNG. All three are bitmap formats and
have slightly different benefts. The PNG format is newer and combines most of
the benefts of the other two formats, so it is a good default choice. However,
JPEG (Joint Photographic Experts Group) is still commonly used by most digital
cameras.
When you need to change the size of pictures and keep fne lines and smooth
edges, it is better to store pictures in vector format. Vector images consist of
mathematical descriptions instead of dots. In most cases, they take up less storage
space than bitmaps. For example, a line is described by a statement such as: line
from (0,0) to (10,10) in yellow. Do not worry: you do not have to be a mathema-
tician to use vector images. Most of the current drawing packages store their work
in vector format. Web browsers are beginning to support a new Internet vector
Reality Bytes: Print Your Own Parts
Have you ever tried to track down a replacement part for a faucet, a toy, a lawn
mower, printer, or some other device? You might be able to track down the manufac-
turer and order the part. Today, many of the manuals are online so it is feasible, but
you still have to wait. What about designing and building new objects? If you are an
inventor and have a new product idea, you need a fabrication shop with expensive
tools to create prototypes. What if you could build any small-size part you need us-
ing a 3D fabricator that essentially prints the part in plastic or metal powders? Man-
ufacturers could put design specifcations for replacement parts online, you could
download the design, run the fabber and have your part built from scratch in a few
minutes or a couple of hours. The really cool part? You can build your own fabber
for $2400 in parts! The Fab@Home Web site (fabathome.com) has a complete list of
parts and directions on how to build the fabber. It also has examples of the projects
they have built. The device is computer-controlled and works like a printer but uses
syringes to deposit silicon, epoxy, or cake icing in layersbuilding up the product
one layer at a time.
Adapted from Tom Simonite, Desktop Fabricator May Kick-start Home Revolu-
tion, New Scientist, January 9, 2007.
image (SVG: scalable vector graphics) format that provides for faster image trans-
fers and scalable images.
Graphics devicesincluding displays, printers, digital cameras, and scan-
nersare evaluated largely in terms of resolution. Resolution is commonly mea-
sured by the number of pixels (or dots) per inch. Figure 2.8 shows the importance
of resolution. Doubling the number of pixels per inch delivers much fner lines
particularly for diagonal lines. But doubling the resolutions results in four times
the number of total pixelswhich requires substantially more memory and pro-
cessing time to handle.
Figure 2.9 shows how to compute the data size of an image. A photo taken with
a resolution of 100 dots per inch and a size of 6 x 4 inches has 600*400 or 240,000
pixels. How many bytes are needed to store one pixel? The answer depends on the
number of colors. How many colors do you need? The human eye can distin-
32
24
16
12
32/24 = (8/8)*(4/3)
total pixels: 24*32 = 768
768 = 4*192
16/12 = (4/4)*(4/3)
total pixels: 16?*12 = 192
Figure 2.8
Resolution. The image on the left has twice the resolution as the image on the right.
In total, the higher resolution image has four times (two squared) the number of
pixels, and provides more precise linesparticularly diagonal lines.
100 dots per inch
6 inches
4 inches
6*100 = 600 dots per line
4*100 = 400 dots per column
400*600 = 240,000 pixels
Figure 2.9
Resolution and size. Multiple the number of rows by the number of columns to get
the total number of pixels. Multiply by the number of color bytes per pixel (usually
3) to get the total raw size (720,000 bytes).
guish at most about 16 million colors. Three bytes of storage is 24 bits and 2
24

is 16,777,216, so many systems use a three-byte color coding. Consequently, the
raw data for the image requires 3*240,000 or 720,000 bytes of storage. But, 100
dpi is a weak resolution, 200 dpi would be better for printing. Doubling the reso-
lution results in a total size that is four times larger (2-squared), or slightly under
3 megabytes. A few systems use 32-bit color for a broader color gamut and faster
processing.
Most digital cameras express resolution in terms of total pixels instead of dots
per inchlargely because the dpi changes depending on the size of the print.
Most of the time, you care about the total number of pixels available. More pix-
els means you get a higher-quality print. Figure 2.10 shows resolutions for some
common devices. Notice how the resolution of a photograph declines when it is
printed in a larger size. Print resolutions less than 300 pixels per inch (ppi) are
grainy and jagged. A similar effect occurs with video screens. On the other hand,
almost all operating systems draw text characters with a set font size. Increasing
a monitors resolution results in smaller text. For example, if you have a monitor
with a physical diagonal measure of 16 inches, an SVGA resolution produces 80
pixels per inch, so the dot on an i is 1/80 of an inch. Increasing the resolution to
UXGA creates 125 ppi, so every character and dot shrinks to a little over half the
original size (64 percent). Microsoft documents have suggested Windows might
someday be rewritten to solve this problem by separating the size of the fonts
from the resolution. Screen displays will behave the way printers do now, so that
increasing the resolution will leave the text the same height but provide sharper
characters with fner details. The main Microsoft tools including Offce and Inter-
net Explorer support this approach by providing a zoom feature that enables you
to increase the display of the font without changing the font size.
The aspect ratio is another topic in graphics that has become important
largely because it has been changing. The aspect ratio is the ratio of the horizontal
width to the vertical height of an image. It is usually expressed in terms of two
numbers, such as 4:3. Early flm projectors were limited by the physical size of
Figure 2.10
Resolution examples. Video screens use the 4/3 aspect ratio from the older television
standard. HDTV uses a wider 16/9 ratio and about 2.5 times the resolution of the
original VGA/TV screen. Final resolution values depend on the physical size of the
screen or paper.
Video Print Camera
Video Pixels Method Pixels/Inch Print
Size
Pixels/Inch
VGA 640 x 480 Fax 100-200 3 x 4 768
XGA 1024 x 768 Ink Jet 300-700 4 x 6 512
SXGA 1280 x 1024 Laser 600-1200 8 x 10 307
UXGA 1600 x 1200 Typeset 2400 pixels 3072 x 2304
WSXGA 1680 x 1050
HDTV 1920 x 1080
the flm so the 4:3 ratio was selected to ft the flms properties. Photographs often
had a similar ratio. Television adopted the same ratio. Movie studios then adopted
a widescreen format with a 1.85:1 or 2.40:1 ratio. When computer displays ar-
rived, they followed the standard television 4:3 ratio; eventually leading to resolu-
tions of 640 x 480 and 1600 x 1200. Then HDTV was introduced, partially with
the goal of displaying higher-quality images for movies. The standard aspect ratio
adopted for HDTV was 16:9 (or 1.78:1). Eventually, computer monitors and lap-
top screens adopted the same aspect ratio. But, a tremendous variation still exists
in aspect ratios among monitors, laptops, tablets, and cell phones. As a designer
you cannot count on a specifc resolution or aspect ratio. Which often means it is
necessary to create images and layouts in multiple formats; or to choose a size that
will ft reasonably well on the most common displays.
Color
Color is an important component of images and video. It is also important in Web
sites and graphics design of print pages. To date, most organizations rely on black-
and-white for basic internal reports and memos. However, as color printing costs
decline and as organizations move towards electronic documents, color is becom-
ing popular in many reports. To begin to understand color, you need to study art
and graphics design. In particular, color and design convey emotion, so you need
to be careful about overall color schemes. Fortunately, most software packages
support templates, so you can choose a template that fts your needs without
needing to be a design expert. However, for complex designs and documents that
are published, you should get expert advice.
Figure 2.11 provides a brief introduction to three of the basic color models.
These models are used to identify colors, but do not do much to help you choose
colors appropriate to your job. In the computer world, RGB (red, greed, blue) is
the most common model. You should use a color picker to explore how the colors
work so you can fnd or adjust colors when you need them. Think of each color
byte as a spotlight. Increasing a single value increases the brightness of that color.
CMYK (cyan, magenta, yellow, key) is commonly used by print shops, so if you
create large print jobs that need to be sent to a print shop, you should work with
the CMYK model. Each color number represents the percentage strength of the
specifed color. The challenge to printing is that your display screen rarely shows
Reality Bytes: SSD Performance Surpasses Hard Drives
In the frst quarter of 2011, Micron Technology, the leading U.S. manufacturer of
RAM, began shipping the RealSSD C400 fash drive. The 1.8 and 2.5-inch drives
connect to laptops using SATA 3 which supports a data transfer rate of 6 G bit/sec-
ond. The drives are created from NAND fash chips using a 25 nanometer lithog-
raphy technology. The largest drive capacity is 512 GB and offers sequential read
speeds up to 255 MB/sec and write speeds up to 215 MB/sec. These speeds are faster
than hard drives and the half-terabyte capacity puts it in the range of many existing
hard drives. On the other hand, the $900 consumer price (for 512 GB) is substan-
tially higher than a hard drive cost; but a signifcant drop for SSDs.
Adapted from Lucas Mearian, Micron Releases Half-terabyte Laptop SSDs, Com-
puterworld, January 4, 2011.
colors the same way they will appear in print. Different lighting strongly affects
the color and brightness of printed images. It also affects your display, but usually
to a lesser degree. If you need to see colors as they will appear in print, you have
to run a gamma correction program on your monitor. It can adjust the display
to match the desired output characteristics. High-end graphics packages, such as
those produced by Adobe, automatically install the gamma correction software
and provide instructions for adjusting it. A related problem you will face with
printing is that some colors you see on your monitor cannot be reproduced by
printers. For example, a pure high-intensity blue is usually printed in a darker
shade. The better graphics packages will show you the constrained color palette.
You probably do not need to become a color expert, but these concepts explain
why your system will sometimes display colors differently than you expected.
Style Sheets
When working on large documents, Web sites, or portfolios, consistency becomes
a critical design factor. Actually, even a small document quickly becomes ugly
and impossible to read if someone uses multiple typefaces, font sizes, and dis-
cordant colors. But the challenge with large documents is that it can be diffcult
to remember exactly which fonts were used for each element. It is even more dif-
fcult if the document or Web site is created by multiple people. Style sheets were
designed to solve these problems and ensure consistency across a collection of
documents. Style sheets have been used for many years in the publishing industry,
but computer tools have made them more powerful and easier to use.
Figure 2.11
Colors. RGB and CMYK are the primary color models in use. RGB for computer
colors created by generating light, and CMKY in printing for refected light. HSL is
used predominantly in the video world. Most software enables you to choose a color
model and enter values the way you prefer.
RGB: Red Green Blue, 1 byte each (0-255 values).
Visualize as lights:
255, 0, 0 is all red.
0, 128, 0 is half green.
255, 255, 0 is yellow.
0, 0, 0 = black.
CMYK: Cyan Magenta Yellow Key
Used for printing (Key is black).
Expressed as a percentage of pure color.
0, 0, 0, 0 = no color (white page).
HSL: Hue, Saturation, Luminosity
Used in video/television.
x, 0, 0 = black
Hue
Saturation
Luminosity
The basic goal is to predefne all of the styles that will be needed in a document
and store these defnitions in one location. When creating a document or Web
page, you simply write the content and assign the appropriate style from the list.
For example, frst-level titles might have a Heading 1 style. The typeface, font,
color, and other attributes are defned within the style. By assigning the style to a
section of text, that text is displayed correctly. The power of a style sheet is that
a graphics designer can tweak the entire layout and design simply by modifying
the style defnitions in one location. Then every page automatically picks up the
new style. Without a style sheet, making a design change requires searching every
single page and editing multiple locations.
If your job consists of creating contentwriting articles, presenting data, and
so onyou simply have to use the predefned styles. But you must always be
careful. Any special font or emphasis that you want to add should never be made
directly to the text. Every design element has to be defned within the style sheet
and then attached to specifc objects. Any change that is made directly to a docu-
ment, such as setting a word or letter to boldface, will be almost impossible to
fnd and change later. Unfortunately, existing tools such as word processors make
it relatively easy to select a word and set it to bold or italicwhich bypasses the
style sheet. You should avoid this temptation and instead open the style sheet list
and assign a style instead of assigning bold or italic.
Reality Bytes: Choosing the Right Tools at Thule
In computer programming terms, the term spaghetti code represents logic that
wanders all over the placemuch like a plate of intertwined spaghetti. This code is
hard to read and diffcult to debug or change. Programmers are taught to write clearer
structures that are easier to follow. But, most people who use spreadsheets learn just
the basic concepts of the tools and then go and build complex monsters. Bicyclists
and Skiers know Thule as a company that makes high-end bike and ski racks for
cars. The Thule Vehicle Solutions North America division has 400 people and gener-
ates 15 percent of the corporate revenue. Vice President of Finance, Mark Cohen,
notes the importance of the data, particularly for budgeting and forecasting: Were a
seasonal company, so we have to have product ready to go or the sale will go to our
competitor. (The main competitor is Yakima.) Cohen noted that this data was main-
tained with spreadsheets. We had one person in accounting who kept all the data
as up to date as possible, including departmental expenses that outlined the costs of
running the factory. However, it wasnt unusual for it to take a week to consolidate
cross-company information, and it wasnt unusual to discover that what managers
were looking at wasnt the fnal spreadsheet. In 2008, Cohen threw out the spread-
sheet approach and went with an online system from Host Analytics Budget that
provides real-time collaboration and mobile access to the data and projections. Indi-
viduals are now responsible for entering their own data and everyone gets immediate
access to current values. The problem with the spreadsheet is that it was the wrong
tool for the job. Yes, it could generate forecasts, but the bulk of the requirements
were to collect and share current datasomething spreadsheets do poorly.
Adapted from Sandra Gittlen, Why Some Companies are Ditching Their Spread-
sheets, Computerworld, January 19, 2011.
Sound
Digitized sound is increasingly important for computer users. For example, you
can use music to add emphasis and entertainment to a presentation. More impor-
tant, you can use your computer to send voice mail messages and to store notes
with your e-mail and documents. Tools exist to convert text to voiceso you
can have your computer read your e-mail messages over the phone. Increasingly,
voice input is being used both to control the computer and for dictation. Increased
storage capacity and declining costs make it easier for sound to be stored in digital
format.
As shown in Figure 2.12, sound consists of two basic components: volume
(amplitude) and pitch (frequency). These two values are changed over time to pro-
duce words and music. To digitize sound, volume and pitch are measured several
thousand times per second. The resulting binary numbers are stored in sequence.
Synthesizers convert the sampled numbers back into music or speech and play
them through amplifers and speakers. The challenge is to sample the source often
enough so no important information is lost.
Most audio applications store sound by digitizing the sound waves. These fles
can become very largedepending on the quality. If you tell the computer to sam-
ple the sound more often, you will get a higher quality recording, but it can take
considerably more storage space. Sound-recording systems generally allow you
to specify the sampling parameter to control the initial quality. From that point,
Figure 2.12
Sound sampling. At a specifc point in time, the system measures frequency and
amplitude using 16-bit numbers. The quality of the sample is largely determined by
the number of times per second the audio is measured. The CD standard is 44,100
times per second.
time
time
frequency (pitch)
amplitude (volume)
440.01
37.15
Frequency: (hertz) cycles per second
lower / higher
Amplitude: height of the wave
compression methods can be used to reduce the fle size; but some of them result
in a loss of quality.
An even more effcient method of storing music is to fle it in the musical in-
strument data interchange (MIDI) format. MIDI is a method of specifying music
based on the musical notation and instrument, so it is substantially more effcient.
A version of this method was designed for the MP4 standard, but this standard has
not been widely adopted.
Video
Computerized video has dramatically increased in the past few years, even in
business. However, most of the video is relatively low quality, in terms of low
resolution and low frame rates. Some video applications are better than others, but
it is still challenging to transmit high-quality video. Yet, as transportation costs in-
crease, there is considerable incentive to use video communication for some types
of meetings.
Converting motion picture and television signals to binary form generates a
tremendous amount of raw data. The process is similar to that for single pictures,
except standard movies display 24 frames (images) per second, while U.S. televi-
sions display 30 frames per second. Video can also be completely computer-gen-
erated. These tools are commonly used in marketing and engineering to display
prototypes and product features.
It is relatively easy to display videoeven cell phones have enough processing
power to display high-resolution video. High-defnitionTV(HDTV) has become
popular, and most devices have the ability to play standard HDTV fles. Even rela-
tively inexpensive cameras have the ability to capture images in HDTV format.
The challenge arises in trying to edit videoparticularly HDTV fles. Editing re-
quires that the fles be decompressed and edited in raw format, which is a large
amount of data. The next sections illustrates some of the issues with data size.
Size Complications
To understand the importance of the fve types of data, it helps to examine the
size of each type. For many years, computers predominantly handled numbers and
limited text. Over time, usage has gradually changed to include images, sound,
Figure 2.13
Size complications. Lossy compression results in substantially less data but at a
lower quality. Text fle is compressed with a Zip folder. Image is compressed with
JPEG from high to low quality. Sound is compressed with a WAV fle (44.1 kbps) and
WMA (64 kbps). Video is compressed with the DV standard and WMV at 6383 kbps.
Object Raw Compressed Lossy
Text and numbers 5 KB/page 2.3KB/page N/A
Image (300 dpi, 24-bit color, 4 x 6
in.)
6.32MB 2.4MB 78-245 KB
Sound(44.1KHzstereo) 170 KB/sec 100 KB/sec 0.01KB/sec
Video (NTSC 30 fps, stero sound) 25 MB/sec 3.7MB/sec 1 MB/sec
HDTV (1080p: 1920 x 1080) MP4 6.8GB/min 1.5MB/sec
and video. The more complex data types require much greater storage space, pro-
cessing power, and transmission capacity.
Consider a typical single-spaced printed page. An average page might contain
5,000 characters. A 300-page book would hold about 1.5 million characters. Now,
consider a picture. At 300 dots per inch, a full 8.5- by 11-inch page would require
a little over a million bytes if it were scanned in black and white. A photograph
in a base resolution with 16 million colors (24 bits per pixel) would require 18
megabytes (million bytes) of storage space.
Sound and video require considerably more storage space. Remember that
thousands of numbers are being generated every second. A typical CD holds 700
megabytes of data, which can store 72 minutes of stereo music or almost 10 mega-
bytes per minute. At lower quality, the current standard for digitizing telephone
conversations generates 64 kilobits per second, almost half a megabyte per min-
ute. Video generates approximately 3 megabytes of data every second, while the
highest resolution HDTV generates about 113 megabytes per second. Figure 2.13
summarizes the basic size characteristics for the standard data types.
Data Compression
An individual photograph at high resolution takes a large amount of space in raw
form (typically over 20 megabytes). Storing and handling raw video creates sub-
stantially larger fles. The answer in many cases is data compression. Look at the
photograph in Figure 2.14 for a second. Consider the line of data from the lower
left corner. To store it as raw pixels, you would have several lines of black pixels.
Instead of recording every one of these pixels, data compression builds up pat-
terns. It takes less space to encode a note that a section contains a large set of
black pixels, than to store every pixel. Some systems, such as the JPEG method,
support lossy compression. In these cases, patterns are matched as long as the
pixels are close to an existing match. When you compress the data, you get to
select how close of a match to allow. For example, with a tight match, the system
might allow a very dark gray pixel to count as black, but might not allow a dark
blue one to match a black pattern. Lossy compression methods are commonly
used to transmit photographs, audio fles, and videos.
Figure 2.14
Data compression. Storing every single pixel requires a huge amount of space.
Compression looks for patterns. For example, instead of storing 1,000 black dots in a
row, it is much shorter to store a note that says 1,000 black dots come next.
Hardware Components
What are the main components of a computer? As a manager, you
need to understand the basic features of hardware. You do not have to be a com-
puter-repair expert, but you need to understand the fundamental structure and
characteristics of the computer components. The most important frst step is to
recognize that declining prices rule this industry. Because of technological ad-
vancements and standardization you know that you can buy a new computer next
year that is faster than the one you just bought today. Instead of worrying about
your new computer becoming obsolete, just accept that you will be able to buy a
faster computer next year, which raises the questions: How often should you buy a
new computer and what features should you buy? The answers to these questions
change over time and depend on your specifc uses, but as long as you under-
stand how to evaluate the major components, you will be able to make a rational
decision.
Along with declining prices, one of the most important trends in hardware is
the shrinking size. But regardless of the size, computers contain four primary
types of components to handle: input, output, processing, and secondary storage.
Of course, each category has hundreds of options so that you can tailor the com-
puter to your specifc needs. One trend that you have to remember is that the hard-
ware industry changes rapidly, especially for small systems. Most computers that
you buy today will have a short economic lifeperhaps only three years. At the
end of that time, each component will have changed so much that typically you
will want to replace the entire computer.
Figure 2.15
Computer performance and capabilities are highly dependent on the peripheral
devices. Most computers will use several devices in each of the four major
categories. Technological progress in one area often results in changes to all four
types of components.
Input Process Output
Secondary
storage
Video monitor
Printer
Plotter
Process control
Voice output
Music synthesizers
Other computers
Magnetic Disk
Floppy Disk
Optical Disk
Tape Drive
USB Drive
Processor
RAM
Device controllers
Keyboard
Mouse
Optical scanner
Voice input
Bar code
Touch screen
Light pen
MICR
Magnetic strips
Card reader
Other computers
seconds - milliseconds
nanoseconds seconds - milliseconds
milliseconds
The relationship among the four components is summarized in Figure 2.15.
Note that the process subsystem consists of random access memory (RAM) and
at least one processor. In many ways, the processor is the most important com-
ponent of the computer. It carries out instructions written by programmers. It uses
RAM to temporarily store instructions and data. Data and instructions that need to
be stored for a longer time are transferred to secondary storage.
One of the interesting aspects of computers is that the various components op-
erate at different speeds. The processor is the fastest of the group and its perfor-
mance is measured in nanoseconds (billionths of a second). Processors are even
faster than RAM. As a result, processors contain separate memory caches embed-
ded in the chip. When the processor asks for data items from RAM, it loads a large
chunk of data at one time so that it can fetch related items from its internal cache
to save time. Secondary storage devices are substantially slower. Because most
disk drives are mechanical, performance time is measured in milliseconds (thou-
sandths of a second). Divide those numbers, and you see that the processor is a
million times faster than the disk drive! Of course, humans are even slowermost
people type a couple of characters per second. Why do these speed differences
matter? It means the processor can be told to do multiple tasksapparently at
Figure 2.16
Basic computer components. The motherboard has the connectors to integrate all of
the components. Expansion cards can be added for graphics processors, TV tuners,
and other add-on tools. The processor has a large heat sink and fan to cool it. RAM
is added to the board on strips. Disk drives are generally connected by serial (SATA)
connectors, but older drives might use parallel IDE cables. The motherboard also
has connectors for front-panel items such as lights and USB ports (hidden under the
wires in this photo).
Power
supply
Processor
--under the
fan and
heat sink
Keyboard, video,
and other connectors
Disk drives
RAM
Expansion
slots
Motherboard
IDE
SATA
Graphics
onboard and
external
the same time. While you are typing, the processor can take a fraction of its time
to store your keystroke, then retrieve some data from the disk drive, then send an
item to the printer, then add some numbers together, and fnally return to get an-
other keystroke.
Figure 2.16 shows the basic components of a PC. The motherboard contains
connectors for the processor, RAM, drives, and expansion cards. Many boards
also contain built-in graphics processors; however, add-on graphics cards gener-
ally provide better performance. The processor is relatively smallabout the size
of a 50-cent coin, but it is covered by a large metal heat sink and the processor
fan to remove heat. RAM is added by installing strips. RAM styles change over
time to provide better performance, so you need to get the right type of RAM for
each board. Disk drives require a power connection and a data connection. Serial
(SATA) connectors are commonly used now because they can move data faster
than the older parallel IDE connections. Video connections and other data con-
nections for networks and USB devices also change over time as new devices and
technologies are introduced.
Warning: You should avoid taking apart your computer; and if you dobe
sure to unplug it frst! You also have to ground your body before touching any
computer component. Static electricity can instantly destroy almost any electronic
item. But, if you are careful, it is relatively easy to install new disk drives or even
RAM into a computer. But always remember that you can never force a part to ft.
Computer parts are designed to ft together with almost no effort. And remember
that static electricity is your enemy.
Processors
Processors come in a wide variety of formats, but most of the existing mainstream
processors (Intel and AMD) can run the same code sets. Increasingly, manufactur-
ers are placing multiple processors on a single chip. So a single processor chip can
contain 2, 4, 6, or more independent processing cores.
Processors are relatively complex today and include many components. Cache
memory is a factor that infuences the overall performance. The processor runs
faster than any other device in the computer, including RAM. Any time there is a
large speed difference performance can be improved by adding a memory cache
to the faster device. The processor chip contains several megabytes of high-speed
internal memory. When it retrieves data from RAM, the processor grabs the data
in large chunks. It then works from the internal (level 2 or level 3) cache, which it
periodically reflls.
The processor is the component that does the hard work of calculations and
executing the software code. You can check on the status of your processor and its
various tasks by opening the Task Manager. Right-click on the Windows menu bar
and select the Processes tab, but be careful not to end any of the processes or your
computer might crash. Most of the time the System Idle Process will be the main
userwhich means the processor is simply waiting. You can also watch the per-
formance chart to see the total demands placed on your processor. Windows has a
sidebar gadget to let you monitor your processor all the time. If you fnd that the
processor is consistently in heavy use, you should consider buying a new machine.
However, unless you are running huge computational tasks or video-intensive ap-
plications, it is more likely that something is wrong with your computer and you
should check it for viruses and spyware software described in Chapter 5. Typical
users will fnd that they have plenty of processing power for almost all tasks.
Te Evolution of Processors
Figure 2.17 shows an approximate measure of Intel processor performance over
time, as indicated by an index value (estimated SysMark). The exponential in-
crease in performance arises because the industry has been able to follow Moores
Law: the number of components placed on a chip doubles every 18 months. But
three potential problems are facing the chip industry: (1) light waves are too wide
to draw increasingly smaller lines; (2) when the components shrink to the size of a
couple of atoms, the electronic properties no longer behave the same (quantum ef-
fects); and (3) the power requirements and heat generation increase dramatically.
For at least two decades, the IT industry has lived with and relied on this in-
creasing performance (and reduction in costs). Eventually, this pattern will stop
possibly within 10 or 20 years. With luck and research, the industry will develop
entirely new technology that will produce even greater performance. But in any
case, managers need to monitor technical developments and be prepared for sub-
stantial changes.
The industry is switching to new chemicals to improve the electrical properties
of the chips to buy some improvements. Several researchers are experimenting
with optical computing (using light instead of electricity), and more esoteric con-
cepts such as quantum computing and biological processors.
In practical terms, the biggest change to processors has been the switch to
multi-core processorswhere each chip contains multiple independent proces-
sors. Dual core processors gained importance in 2006, and the quad-cores made
it to standard production in 2007. Experimental designs have been built with doz-
ens, and potentially hundreds of processor cores on a chip. These features are dis-
cussed in the section on parallel processing.
Figure 2.17
To assist buyers, Intel provides a measure of its own processors. The rating is an
index that measures relative performance. A processor with a rating of 200 would
generally be about twice as fast as a processor with a rating of 100. Of course,
your other devices also infuence the overall system speed, so just buying a faster
processor does not guarantee that your machine will run faster.
0
50
100
150
200
250
300
1
9
8
9
1
9
9
0
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
2
0
0
7
2
0
0
8
2
0
0
9
2
0
1
0
2
0
1
1
SysMark 2007 Intel Processor Performance
Multi-core
Random Access Memory
As processor speed has improved, RAM has become a crucial factor in system
performance. Because disk drives are mechanical, they are the greatest bottleneck
in a computer system. Hence, modern operating systems try to hold as much data
as possible in RAM. Figure 2.18 shows that RAM price has dropped substantially
over the last decades. In relative terms, RAM is virtually free. For a couple hun-
dred dollars, you can easily buy enough RAM to hold several applications and
their data in RAM at one time. Operating systems, particularly Windows, take ad-
vantage of RAM to cache as much data and code as possible. Disk drives are no-
toriously slow, so system performance is dramatically improved by adding RAM.
High-speed RAM enables the processor to pull data from RAM faster, leading to
improved performance, but it is relatively more expensive.
Parallel Processors
In the past, when processors were more expensive, designers used only one pro-
cessor in a machine. Today, many computers contain multiple processors. Al-
though it can be a desirable feature, you must be careful when evaluating parallel-
processing machines. If a computer has four processors (or four cores), it is tempt-
ing to say that the machine is four times faster than a computer with only one
processor. Indeed, many computer companies advertise their computers this way.
Can a computer with four processors really do your job four times faster? The
answer is that it depends on your job. Consider an example. A computer with two
processors has to add two sets of numbers together. Each processor works on one
pair of numbers and fnishes in half the time of a single processor. Now, the same
two computers have to work the problem in Figure 2.19.
Notice that the second calculation depends on the outcome of the frst one. The
second one cannot be computed until the frst one is fnished. Even if we assign
one processor to each calculation, the parallel-processing machine will take just
as long as the single processor. On the other hand, if you commonly run multiple
Figure 2.18
RAM prices over time. RAM prices have dropped at exponential rates as
manufacturing processes have evolved, even as RAM speeds have increased.
0
50
100
150
200
250
300
350
400
2000 2005 2006 2007 2008 2009 2010
$
/
G
B
Cost of RAM
Technology Toolbox: Voice Input
Problem: You are a slow typist or need to use your hands for other tasks.
Tools: A voice input system is built into Microsoft Offce. You need a decent micro-
phone to use it, and a headset microphone is recommended because it minimizes ex-
ternal noise. The system is relatively easy to set up, but it takes a half hour to train so
that it recognizes your voice. Remember that no voice system is perfect, so you will
still have to edit the text that it enters. You can fnd information on speech recogni-
tion in the Offce help system by searching on the speech key word.
You can install the speech recognition software within either Word or the Micro-
soft Windows Control Panel. Within Word, select Tools/Speech. From the Control
Panel, select Add/Remove Programs, fnd Offce and Change it. Use Add or Remove
Features and fnd the option to add Speech under the Alternative User Input. Once
the software is installed, you have to train it. Follow the installation instructions to
perform the basic training. To improve recognition accuracy, you should also take
the time to read at least one of the additional training documents. Once the system is
installed and trained, you can dictate text into an Offce document. Remember that
the system works better if you dictate complete sentences.
The voice toolbar enables you to turn the microphone off, so that you can speak to
someone else or cough. Notice that it also enables you to switch between Dictation
and Voice Command modes. The help system provides additional information about
command mode and the additional things that you can say to control the computer
with voice. For example, in Voice Command, you can say the word font followed
by the name of the font to change the font in use. You can also start other programs
and select items on the menu. The help documentation includes information on com-
mands you can use during dictation. Some of the common commands are in the
table.
Command Result
period or dot
comma
new line
new paragraph
open paren or parenthesis
close paren or parenthesis
force num, pause, digits
spel it or spelling mode
microphone
correct that
scratch that
go to top
move up
backspace
select word
.
,
enter key/new line
Enter twice
(
)
numbers (for several numbers in a row)
spell out a word
turn microphone on or of
change or delete the last phrase entered
delete the last phrase entered
move to top of the document (or use bottom)
move up one line (also accepts down, left, and right)
delete one character to the left
select a word (or use phrase with several options)
Quick Quiz: Use the help system to fnd the commands for the following:
1. !, ?, #, $
2. Make a word boldface or italic.
3. Print the current page.
applications at the same time, multiple processors will improve performance be-
cause each processor can handle the applications separately.
Massively parallel machines can include thousands of processors. They are
used for some highly specialized applications. For example, governments use
them to break codes; physicists use them to simulate large-scale events such as
nuclear explosions and weather patterns; computer artists use them at special-ef-
fects studios, such as Industrial Light & Magic, to create movies. This concept has
been extended with the use of grid computing, where thousands or millions of
separate computers are assigned portions of a task. The search for extraterrestrial
intelligence (SETI) project is the most famous example using a grid approach.
Users are encouraged to download a small program that runs in the background
on the computer. The program retrieves portions of radio-astronomy data from
an Internet site and scans them looking for patterns. The millions of participating
computers provide a huge (free) resource to process data that could not be handled
by any other method. Using similar software, the grid concept can be applied to
business and scientifc applications. A company could use existing clerical com-
puters or, since personal computers are so inexpensive today, could buy hundreds
of machines and tie them together to handle large databases.
Todays computers take advantage of multiple processors in more subtle ways.
Most computers today utilize a separate graphics processor to draw items on the
screen. It is a sophisticated processor with its own RAM that handles complex
three-dimensional calculations. This additional processor is particularly important
for 3D graphics, such as those used in games. However, Windows Vista takes
advantage of the processor for all display chores and offoads the drawing tasks
to the graphics processorcreating faster graphics and enabling special effects
within applications.
Connections
Even if two computers have the same basic components, it is still possible for one
machine to be substantially faster than the other. The reason is that the compo-
nents need to exchange data with each other. This communication typically uses a
short-distance electrical connection. Most computers have special slots that form
the connection between add-on boards and the processor bus. Various manufac-
turers make boards that ft into these slots. The processor can exchange data with
these other devices, but it is constrained by the design of the bus. For example,
even if a processor can handle 64 bits of data at a time, the bus might not have the
connections to allow it to transfer that much data at a time. Standards enable users
to exchange cards and devices from one computer to another. The problem is that
a bus designed for todays computers will eventually become obsolete as the pro-
Figure 2.19
Some computations must be performed in sequence, so there is little to be gained
with multiple parallel processors. In this example, the second computation (yyy)
must wait for the frst one to fnish.
23 xx
+ 54 + 92
xx yyy
cessor improves. At some point, the standards need to be changed. In the personal
computer market, standards for the bus have gradually evolved. For several years,
the industry has relied on an Intel-sponsored design known as the Personal Com-
puter Interconnect (PCI) bus. PCI was also designed to make it easier for users
to set up their computers. Over time, manufacturers have increased the transfer
speeds of PCI, but it is increasingly a limitation in computer performance.
External components also need to be connected to computers, such as thumb
drives, external disk drives, cameras, and video cameras. For example, the IEEE
1394 or frewire standard is commonly used to transfer data from video equipment,
but can also be used to connect disk drives to the computer. USB 2.0 is a similar
standard used to connect peripherals including disk drives and RAM drives. Fiber
channel and serial ATA are used to connect disk drives at high speeds. Ultra-wide-
band wireless (UWB) might eventually be used to connect devices at high speeds
without using cables. For example, you could tell your video camera to send the
signal to your television simply by setting it next to the TV and pushing a button.
Figure 2.20 shows the common connection methods available and their maximum
transfer rates. Be cautious with the transfer rate numbers. No system ever reaches
those numbers, but they do impose a speed limit. For example, hard disk drives
are limited by their write speed (perhaps 400 mbps), so a SATA 3 connection will
provide no improvement over a SATA II cable. But you would see better perfor-
mance if you connect that drive through USB 3.0 or eSATA than with USB 2.0.
With overhead transfer costs, USB 2.0 is probably restricting the data transfer.
The basic rule is that when you purchase newer, faster peripherals, you need to
connect them through faster cables. So do not expect to be happy connecting a
new high-speed device to a three-year old computer running USB 2.
Figure 2.20
Common methods to connect peripheral devices to computers. The max speed is
never achieved but it can reveal bottlenecks. Hard drive transfer rates are often
limited by drive write speeds. But the newer methods (SATA 3 and USB 3.0) will
improve the performance of large data transfersparticularly with large solid state
drives.
Method Max Speed Primary Purpose
PCI-e 2.0 x16 500 M Bytes/s * 16
64 G bits/sec
Connect peripherals, particularly graphics
cards
SATA II 3 G bits/sec Disk drives
SATA 3 6 G bits/sec Disk drives
Fibre Channel 20 G bits/sec Data center storage network
Firewire 2.0 800 M bits/sec Video, drives
HDMI 3.4 G bits/sec HDTV video
USB 2.0 480 M bits/sec External drives
USB 3.0 4.8 G bits/sec External devices
Intel
Thunderbolt
10 - 100 G bits/sec External devices
LAN/gigabit 1 G bits/sec Computers, drives
Input
A variety of input devices are available because of the need to handle the many
data types. Hundreds of variations exist because of the challenges of adapting the
devices to human needs. People have different preferences and styles, so manu-
facturers are constantly improving their products. The purpose of an input device
is to convert data into electronic binary form. Keyboards are the most common
method of entering new text and data. Note that you can purchase different types
of keyboards to suit individual users. For example, some keyboards enable you to
change the layout of the keys. Keyboards have their own feel; some individuals
prefer sensitive keys requiring a light touch, while others like stiffer keys to sup-
port their hands.
Ergonomics is the study of how machines can be made to ft humans better.
One of the main conclusions of this research in the computer area is that individu-
als need to be able to adjust input (and output) devices to their own preference.
Forcing people to adapt to rigid devices can lead to complaints and even physical
injuries. Since the mid-1980s, some workers have encountered a disabling condi-
tion known as repetitive stress injury, which some people claim results from ex-
tended use of tools that do not physically match the worker.
Although there is limited scientifc study of these injuries and causes, some
people have found relief after ergonomic changes to their work environment.
Complaints typically involve back pain, eye strain, headaches, arm and shoulder
pain, and fnger or wrist pain due to carpal tunnel syndrome. Common ergonomic
suggestions include adjustable chairs, footrests, armrests, adjustable keyboards,
high-resolution low-ficker monitors, and improved lighting. Most of these adjust-
ments are relatively inexpensive today, but can add up if new chairs and desks are
needed for thousands of workers.
Reality Bytes: People Cant Memorize Computer Industry Acronyms
The computer industry lives on acronyms. Any time a new product is created, a new
acronym is developed so that everyone can use shorthand letters to refer to it. TLAs
(three-letter acronyms) are best, but sometimes they are longer. Actually, every in-
dustry uses acronyms to save time when communicatingbut the computer and net-
work industries change so quickly that they continually develop new products and
new acronyms. But, ordinary people have considerable diffculty memorizing the
acronyms. Read any computer ad and try to fgure out the specifcations if you do not
know the acronyms. Mike Molloy, USAF, Ret. and his wife, Susan Ebert, have come
to your rescue. They have created a Web site, www.AcronymFinder.com, that lists
any acronym you can imagine and the various interpretations. Many acronyms have
multiple defnitionsdepending on which industry or feld is using it. If you spend
hours a day writing text messages, you might think you are an expert on acronyms,
but remember that new ones are introduced every day. As a central source, where all
of the acronyms are validated by Mr. Molloy and his wife, the site can save you time
in your research.
Adapted from Barry Newman, BTW, if You Need Info About C4ISR, Read This
ASAP, The Wall Street Journal, January 13, 2007.
Pointing Devices
With the increased use of graphics and pictures, it is common for computers to
use pointing devices for input. A mouse is the most popular device in use today,
although pressure-sensitive pens, touch screens, and digitizer tablets are heavily
used in some applications. Touch screens are commonly used for displays that
involve customers or other atypical computer users, but they are also common on
tablet PCs. Many tourist bureaus, hotels, and shopping areas use computer dis-
plays with touch screens to give directions to visitors. Besides the fngerprints, the
biggest problem with touch screens is that the tip of your fnger is often too large
to be a useful pointer. For more detailed use, an engineer who is designing a wir-
ing diagram for an automobile would use a digitizer tablet with a special pen to
draw fne lines and select individual points that are close together.
Scanners
When you are dealing with pictures, it is often helpful to have a scanner convert
a paper-based image into digital (bitmap) form. For certain types of images (line
drawings and text), vector tracing software can convert the bitmap image into vec-
tor form. Most digital cameras also have the ability to take photographs of paper
documents. The beneft of the camera approach is that the camera is much faster
at capturing the image. Some photocopiers use a similar process to enable workers
to scan documents and e-mail them to a computer.
The quality of a scanner is measured by the number of pixels per inch that it
can distinguish as well as the number of colors. Most scanners can read at least
600 dots per inch. More dots mean you get fner lines and sharper pictures.
Scanners also can be used to input text and data into a computer. The scanner
frst converts the page into a picture of dots. Then optical character recogni-
tion (OCR) software examines the picture and looks for text. The software checks
each line and deciphers one character at a time. Although OCR software is im-
proving, it is not 100 percent accurate. Some systems automatically look up each
word in a dictionary to spot conversion errors and improve accuracy. Even then,
users report about a 95 percent accuracy ratewhich is highly dependent on the
quality of the original document.
Digital cameras can also function as scanners. The principle is the samethe
camera has a sensor that captures individual pixels. Most cameras have a setting
optimized for taking pictures of black-and-white pages of text. It can be useful for
quick captures of a couple of pages, or for storing notes taken on a board.
Sound
Sound is initially captured with a microphone that converts sound pressure waves
into electrical signals. A sampler then converts these signals into numeric data that
can be stored on the computer. Musical synthesizer technology is used to con-
vert the numbers back to electrical signals that can be heard with speakers. Sound
boards can be purchased for personal computers that contain both the sampler
and synthesizer technology. Digital sound conversion occurs on almost every long
distance telephone call you make. Your voice is converted to 1s and 0s to be sent
across fber-optic telephone lines.
Speech Recognition
As long as computers have existed, individuals have dreamed of being able to talk
to them and have their words translated into text. Today, computers can digitize
and record speech, and they can convert spoken words into computer text. Some
problems still exist, and the systems need to be trained to a specifc user. Common
problems include the use of homonyms, variations in speech patterns or dialects,
and the effects of punctuation on meaning.
Initially, speech recognition systems were adopted by occupations that require
note taking along with two hands to do the job. Quality control inspectors and
surgeons use them regularly. As performance continues to improve, we will see an
expanded use of speech recognition systems among all users. Ultimately, speech
recognition will be a key element in dealing with computers. Keyboards do not
work well in a wireless environment. They are too large to be portable and hard
to use.
Today, speech recognition is built into the Windows operating system. Speech
recognition generally requires you to train the system to recognize the way you
pronounce various words and phrases. If you enunciate clearly and learn to dictate
in complete sentences, speech recognition is relatively accurate. You can also use
voice commands to start applications and issue simple commands to the computer.
However, the industry still has a long path ahead for speech recognition to begin
replacing keyboards.
Video Capture
As technology improves, companies are increasingly adding video clips to pre-
sentations and Web sites. Digital video transmissions are also being used for com-
munication. Today, most television monitors can be used as computer monitors.
The HDMI connector that is common on HDTVs is becoming a standard way
of connecting video equipment, including computers. On the fip side, capturing
or displaying television signals on your computer requires additional hardware
specifcally a tuner or capture card. A common television digital video recorder
(DVR) is really just a computer with a capture card and a large hard drive. The
challenge is that the capture card has to be matched to the type of input: over-
the-air digital broadcast, cable-TV digital signal, satellite signal, or video camera
input. Throw in the complication that most digital broadcasts are encrypted, and
it becomes diffcult to build your own capture system. Interestingly, it is actually
Reality Bytes: People Really do Want Digital Content
For several years, book publishers have resisted shifting to digital books. Part of
the reason is because several early e-book readers were weak, had limited support,
and were not adopted by customers for those reasons. After Amazon introduced the
Kindle e-book reader, and then dropped prices with the second and third versions,
people began buying them in large numbersalthough Amazon does not report the
actual number of Kindle sales, which are estimated to be in the millions. Several oth-
er companies, including Apple through its iPhone and iPad apps also provide reading
devices for digital text. At the end of 2010, Amazon reported that it had sold 15 per-
cent more Kindle e-books than paperbacks. By the middle of 2011, Amazon reported
that sales of e-books exceeded the sales of all print books. And that is not counting
downloads of the free, public domain e-books.
Adapted from Matt Hamblen, Amazon.com Touts More e-Book Sales than Paper-
backs, Computerworld, January 28, 2011.
easier to deal with over-the-air HDTV broadcasts in 2011. The quality is generally
good (often better than cable TV), the signal is not encrypted, and it is free.
Output
Most people are interested in two types of output: video and paper copy. Video
output is created by a video card and displayed on a monitor. Computer projec-
tion systems for meetings and presentations use high-intensity light to project an
image onto a screen. In addition to resolution, they are evaluated by the intensity
of the light, measured in lumens. Other display technologies have been developed
including heads-up displays for cars, personal projection systems that project an
image from special glasses, and electronic ink that displays text and images on
fexible displays.
The other common output device is the printer. Printers come in many differ-
ent forms. The most common formats are laser and ink jet, where the output is
created by printing dots on the page. Common resolutions include 600 dpi and
1200 dpi lasers. In contrast, standard typesetters, such as those that are used to
print books, operate at resolutions of at least 2,400 dots per inch. Again, higher-
resolution devices are more expensive. Also, the increased amount of data being
displayed takes longer to print (for the frst copy).
Laser printers operate much like photocopiers. In fact, newer copiers include
hardware to connect to a local network so the departmental copier can function
as a high-speed printer. From your desktop, you can tell the printer to make mul-
tiple copies and collate and staple themat speeds of over 70 pages per minute.
As noted in Figure 2.21, the initial cost is usually only a minor component of a
printers price. It is more important that you look at the cost per pageparticu-
larly for color copies.
Secondary Storage
Why are there secondary storage devices? Why not just store everything in high-
speed RAM? The main reason is that standard RAM is dynamicwhich means
that the data disappears when the power is turned off. The second reason is cost:
a terabyte hard drive costs less than $100. Even at todays relatively low prices,
the equivalent amount of RAM would cost several thousand dollars. On the oth-
er hand, because disk drives are mechanical, they are slow and can fail. Some
companies, notably Google, have invested in huge RAM drive storage. All of the
Figure 2.21
Printer evaluations. Printers are evaluated in terms of initial cost, cost per page,
resolution, and speed. There are many types of printers, led by laser, ink jets, and dot
matrix printers. Prices vary depending largely on speed and resolution. Technological
changes are leading to new varieties of printers that can produce full color at a cost of
around 5 to 10 cents per page.
Printer Initial Cost
(dollars)
Cost per Page
(cents)
Quality
(dots/inch)
Speed
(pages/min)
Laser: B&W 300 - 20,000 0.6 - 3 600 - 1200 4 - 8 - 17 - 150+
Laser: Color 500+ 5 - 75 600 - 1200 1 - 30
Ink jet 50 - 800 5 - 150 300 - 1200 1 - 20
search indexes are stored in RAM, resulting in substantially faster searches in-
stead of relying on disk drives.
Hard disk drives are the most common method of secondary storage. The pri-
mary drawback to disk drives is that data transfer is limited by the physical need
to spin the disk and to move the read/write heads. Besides being slower, mechani-
cal processes are sensitive to movement and deteriorate faster than pure electron-
ic systems. Consequently, several manufacturers have developed static memory
chips that hold data for portable deviceslike Sonys memory stick. The USB
drives are a popular method for transporting data and have quickly replaced fop-
py disk drives.
A more recent variation is the solid state drive (SSD) which has even faster
read/write speeds and is packaged similar to a standard hard drive. In particular,
SSDs for laptops plug into the same slot as a standard hard drive. Todays USB
and SSD drives are as fast or faster than hard driveseven for write speeds. So,
the biggest problem is price. SSDs are roughly ffty times more expensive than
hard drives. But, this price difference has been declining, and as vendors focus on
improving the chips and the production capacities increasethese prices should
continue to drop.
Except for prices (declining) and capacity (increasing), typical hard drives have
changed little during the last few years. Secondary storage is needed to hold data
and programs for longer periods. Secondary storage is evaluated by three attri-
butes: capacity, speed, and price. Figure 2.22 shows the values for different types
of storage.
With the increasing importance of data, companies are searching for ways to
prevent loss of data by providing continuous backups. One straightforward meth-
od is to install an extra disk drive and keep duplicate copies of all fles on this mir-
ror drive. If something goes wrong with one of the drives, users simply switch
to the other drive. A second method of protecting data is known as a redundant
array of independent drives (RAID). Instead of containing one large drive, this
Figure 2.22
Secondary storage comparison. Hard drive prices and speed make them the best
choice for secondary storage. Tapes make good backup systems because of the cost
per gigabyte and 50-100 gigabyte storage per tape, but the sequential storage makes
it harder to retrieve data.
Drive Capacity
(gigabytes)
Speed
(Write MB/s)
Initial Cost
(dollars)
Cost/GB
(dollars)
Magnetic hard 80 - 3,000 60 - 200 65 - 200+ 0.07
SSD 16 - 512 60 - 320 200 - 900 1.76
USB drive 2 - 64 25 - 150 10 - 115 1.80
Tape 250 - 2,000 20 - 120 300 - 5,000+ 0.50-1.00
CD-ROM 0.70 2 - 8 50 0.18
DVD 4.77(8.5DL) 2 - 21 50 0.04
Blu-Ray 25 (50 DL) 4.5 - 36 80 0.12
Blu-Ray BDXL 128
system consists of several smaller drives. Large fles are split into pieces stored on
several different physical drives. At the same time, the pieces can be duplicated
and stored in more than one location. In addition to the duplication, RAID sys-
tems provide faster access to the data, because each of the drives can be searching
through its part of the fle at the same time. Essentially, RAID drives use parallel
operations to dramatically improve the storage and retrieval of large documents.
On personal machines, even if you do not want to spend the money for a RAID
solution, it can be a good idea to buy two hard drives instead of one large drive.
Particularly for data-intensive applications like video editing, it is wise to put the
operating system and application software on one drive and place the data fles on
the second drive. Two drives are better than one because the computer can retrieve
the data simultaneously from both drives. In fact, video and large database appli-
cations often beneft from using three drives: one for the system, one for the main
data, and one to hold temporary fles. Putting everything on one drive forces the
computer to trade between tasks and wait for the single drive to spin to the proper
location.
Magnetic tapes are also used to store data. Their biggest drawback is that they
can only store and retrieve data sequentially. That means you have to search the
tape from the beginning to fnd any data. A single tape can hold 20 to 400 gi-
gabytes of data. Compression can double or triple the capacity of a single tape,
driving the cost per gigabyte from $0.50 down to $0.20 or lower. Because of these
two features, tapes are most commonly used to hold backup data. Just be careful
to store the tapes safely, and stay away from ultra low-cost drives that might fail.
With the increased use of images, sound, and video, there is a need for substan-
tially greater storage capacity for data fles. Optical disks have provided substan-
tial capacity at relatively low costs. The optical (or magneto-optical) disk drive
uses a laser light to read data stored on the disk. CD-ROM stands for compact
disk-read only memory, the format used to store music CDs. The ROM portion
of the name means that you can only read data from the disk. But, actual storage
depends on the type of blank disk and your CD burner. Standard disks can be writ-
ten once, or you can use rewritable disks. However, rewritable disks might not be
compatible with other systems. The biggest drawback is that saving data to a CD
is relatively slow. And, 700 megabytes does not go far these days.
The DVD (digital video disk or digital versatile disk) is a signifcantly better
technology. DVD was created to distribute digitized video. Compared to CD, the
strengths of DVD are (1) increased capacity, (2) signifcantly faster access speeds,
and (3) standards for audio, video, and computer data. A basic DVD holds 4.77
GB of data. However, the commercial DVDs for movies can hold as much as 8.5
GB of data by using an advanced writing technique (dual layer).
Blu-Ray is the most recent format created for storing high-defnition digital
video data. Sony, the leader in the Blu-Ray consortium, reports that a dual-layer
Blu-Ray disk can hold 50 GB of dataenough for about four hours of high-def-
inition video recording. You can also purchase Blu-Ray recorders. The burners
have dropped in price, but the blank disks are still expensive. Probably too expen-
sive to use as backup devices. But, the prices should drop over time and it would
be useful to be able to store 50 GB at a time on a backup disk.
Operating Systems
Why is the operating system so important? Computers follow instruc-
tions called software that are written by programmers. Every computer needs one
special type of software known as the operating system. The operating system
(OS) is software that is responsible for communication among the hardware com-
ponents. The operating system is also a primary factor in determining how the
user deals with the machine.
As shown in Figure 2.23, the OS handles the interaction with other devices
through device drivers, which are small software programs that provide hooks
that are called by the OS to execute commands on the device. For example, a
printer driver provides a standard set of commands to the operating system, and
handles the details of drawing text and objects on the specifc printer. These soft-
ware drivers are routinely updated by the vendors to fx bugs and add new fea-
tures. The power of this approach is that any vendor can create a new device. By
creating an appropriate driver, even older computers can take advantage of the
new device.
Historically, each computer manufacturer created its own operating system tai-
lored for that specifc hardware. Proprietary operating systems caused many prob-
lems; in particular, changing vendors typically required purchasing new applica-
tion software. AT&T researchers began to solve this problem when they created
a hardware-independent operating system. It is known as UNIX and was designed
to work with computers from many different manufacturers. However, UNIX is
not a complete standard, and application software must generally be rewritten or
at least recompiled before it can function on different computers.
Today, fewer operating system choices exist. Personal computers generally run
a version of Microsoft Windows. The other main alternative is a variation of the
UNIX operating system. Several years ago, much of the UNIX code was essential-
ly released to the public (although a few lawsuits remain over exactly who owns
parts of the code). Linus Torvalds used the code to create an inexpensive and rela-
Figure 2.23
Operating system and device drivers. Every peripheral device uses device driver
software to connect to the operating system. The software uses defned interfaces to
transfer data and monitor status of the device.
Operating System
Device
driver Device
driver
Device
driver
Device
driver
tively popular version for personal computers. His base versions are called Linux
and form the core of several open source systems. The Linux system is available
in several varieties for personal computers. It is also embedded in many devices
such as eBook readers. Apple computer also adopted the UNIX foundation with
its OS X operating system. Most midrange computers and many servers also run
an operating system derived from UNIX. Notably, IBM (AIX and Linux) focuses
on UNIX versions for its servers. IBM also continues to support some older, pro-
prietary operating systems. For servers and midrange systems, you still typically
use the operating system provided by the hardware manufacturer.
The battle over operating systems continues today. At least some portion of
Microsofts market strength is tied to its ownership and control of the desktop op-
erating system (Windows). The standardization of the operating system has made
it possible for vendors to create thousands of useful software applications. The
relatively open platform has encouraged the development of thousands of new
hardware devices that can be easily installed in millions of computers worldwide.
Yet the common operating system has also made it easier for hackers and viruses
to attack millions of machines. And many smaller companies are concerned about
the ability to compete against a company as frmly entrenched as Microsoft.
The operating system battle now extends into cell phoneswhich are essen-
tially small handheld computers. For years, cell phone manufacturers used their
own proprietary operating systems. As smart phones became more popular, the
operating system became more important and vendors began to choose among
a few standard versions. Apple phones use their iOS system; many other phones
use the Android operating system developed and maintained as open source by
Google. Microsoft also provides a Windows operating system for cell phones, and
some vendors (notably Blackberry) use their own systems.
To a user or manager the choice of an operating system is probably not impor-
tant. However, it does affect how you use the device, and it might limit the appli-
cations or software available for the device. In general, you pick a device based on
reliability, capabilities, and the availability of the software you want to use.
Computers in e-Business
How does the Internet change the role of computers? The Internet
and the Web changed computing and business in several ways. Beyond the issues
of connectivity, the key feature of the Web was the introduction and acceptance
of the browser. The reason the browser is so important is that it has become the
standard display mechanism. Increasingly, applications are being built that rely
Reality Bytes: Where Are You Going to Put Everything?
Most large corporations are reluctant to trust their data to anyone else. A survey of
247 Fortune 1000 frms by TheInfoPro in 2011 revealed that only 10 percent were
even considering the public cloud for even the lowest tier of data. Most companies
(73 percent) are relying on Fibre Channel storage area networks (SANs), even over
network-attached storage (NAS). Although companies are unwilling to trust data to
public cloud providers, they were willing to store data on private clouds.
Adapted from Lucas Mearian, Fortune 1000 Firms Shun Public Cloud Storage,
Computerworld, May 4, 2011.
on the services of the browser. When the browser becomes the most important
display device, then nothing else matters. To emphasize the point, in 2011, Google
introduced an operating system for simple hardware that is based completely on
their Chrome browser. All operations are designed to work across the Web. Al-
though this approach might seem extreme to those used to working on traditional
PCs, it is similar in many respects to mobile computing on cell phones and tablet
computers.
A Web-based approach with a stripped down operating system has several po-
tential benefts. The hardware can be simpler and cheaper. The computer and start
up and shut down much faster. All data is automatically stored on central Web
servers, which provide backup and access from any location. Software is upgrad-
ed and replaced on the servers instead of each PC.
In terms of drawbacks, the most important one is that you need good access
to an Internet connection. Without the Internet, you might as well be carrying a
brick. You also rely on the serversboth in terms of uptime and maintenance and
in terms of pricing. Services that are free or low-cost today might suddenly be-
come expensive if the service providers decide to charge higher prices. Your data
is also no longer in your controlyou run slightly higher security risks in terms of
data transmission and data storage.
Many of these issues refect ongoing decisions between centralization and de-
centralization, which are covered in more detail in Chapter 12. In the past, the de-
cisions applied primarily within companies. The Web now makes them pertinent
to everyone.
What Is a Browser?
At heart, browsers are simply software display devices. They read incoming fles,
recognize the data type, and display the data as instructed. The data could be text,
images, sound, or even video. However, the sound and video are usually handled
Reality Bytes: How Safe is Your Data?
Check out an ad for a disk drive and you will see many specifcations. The mean
time between failure (MTBF) is an important one because it indicates the reliabil-
ity of the drive. Manufacturer data sheets often indicate MTBF between 1 and 1.5
million hours. However, a Carnegie Mellon study coauthored by Garth Gibson of
100,000 disk drives in data centers run by Google indicate that actual failure rates
are substantially higher. The posted MTBF numbers would indicate annual failure
rates of at most 0.88 percent. In reality, typically 2 to 4 percent of the drives failed,
with some systems having failure rates of 13 percent. Professor Gibson pointed out
that not all of the drives had actually failed. Sometimes it is diffcult to identify the
cause of a problem and drives are removed in the testing process. Vendors report that
as many as 50 percent of returned drives actually work fne. The dirty, hot, vibration-
prone environment of a data center is also harder on drives than a clean-room testing
center. Regardless of the failure rates, disk drives are mechanical so they are less
reliable than pure electronic equipment. Consequently, you must always assume that
a drive can fail and ensure that you have proper backups to recover your data.
Adapted from Robert L. Scheier, Disk Drive Failures 13 Times what Vendors Say,
Study Says, Computerworld, March 2 ,2007.
Technology Toolbox: Creating Effective Charts
Problem: You need to create a chart that users can understand.
Tools: Spreadsheets contain powerful tools for creating charts; however, you must
be careful when using the tools to ensure that the features are used to make the chart
more informative and easier to read.
The frst step in creating a chart is to enter the data correctly and perform any
needed transformations. The second step is to choose the correct type of chart to
emphasize your message. The third step is to clean up the chart and make sure it
contains the proper title and labels. The table shows the main purpose for each chart
type and the common mistakes made by novices.
Chart Type Purpose Most Common Mistakes
Bar or
column
Showcategoryvalues. Too many series
Unreadable colors
Not zero-based
Pie Comparecategorypercentages. Too many observations/slices
Unreadable features/3-D
Poorly labeled
Line Showtrendsovertime. Too many series
Poor or missing legend
Not zero-based
Scatter Show relationship between two
variables.
Poor choice of variables
Not zero-based
One major mistake that is common to all chart types is to use excessive orna-
mentation. As Edward Tufte points out in his book The Visual Display of Quantita-
tive Information, graphing software offers many temptations that should be avoided.
Make sure every option you choose highlights the main purpose of the chart. Avoid
cluttering your graph with excess lines, images, or garish color schemes. The goal of
any design is to strive for elegance.
Creating a chart in Excel is straightforward. In this example, export the sales by
model by month from Rolling Thunder Bicycles. Open the generated spreadsheet
and select all of the data cells. Click the chart button and build the line chart. To sim-
plify the chart, remove the Hybrid and Tour models.
Quick Quiz: Create the following charts:
1. Use the export data form in Rolling Thunder Bicycles to generate sales by state.
Create a column chart and a pie chart for this data. Briefy explain why one chart
is better than the other.
2. Using Bureau of Labor Statistics data, plot the unemployment rate and the hour-
ly wage rate over three years.
0
100,000
200,000
300,000
400,000
500,000
600,000
700,000
800,000
900,000
1,000,000
1
9
9
4
-
0
1
1
9
9
4
-
0
8
1
9
9
5
-
0
3
1
9
9
5
-
1
0
1
9
9
6
-
0
5
1
9
9
6
-
1
2
1
9
9
7
-
0
7
1
9
9
8
-
0
2
1
9
9
8
-
0
9
1
9
9
9
-
0
4
1
9
9
9
-
1
1
2
0
0
0
-
0
6
2
0
0
1
-
0
1
2
0
0
1
-
0
8
2
0
0
2
-
0
3
2
0
0
2
-
1
0
2
0
0
3
-
0
5
2
0
0
3
-
1
2
2
0
0
4
-
0
7
2
0
0
5
-
0
2
2
0
0
5
-
0
9
2
0
0
6
-
0
4
2
0
0
6
-
1
1
2
0
0
7
-
0
6
2
0
0
8
-
0
1
2
0
0
8
-
0
8
2
0
0
9
-
0
3
2
0
0
9
-
1
0
2
0
1
0
-
0
5
2
0
1
0
-
1
2
2
0
1
1
-
0
7
Rolling Thunder Bicycle Sales
Mountain
Mountain full
Race
Road
by an add-in component for the browser because standards are still evolving. To
add more interactivity, browsers also have an internal programming language (Ja-
vaScript). Developers can include program code that gives detailed instructions to
the browser and reacts to changes users make. For example, when users move the
cursor, the code can highlight an object on the browser. Browser code is also used
to check user data. This code runs entirely on the local machine, but can send data
in batches to the Web server.
The beauty of the browser approach is that it standardizes the way data is dis-
played, so that everyone is free to choose any hardware and software platform
they prefer. Since browsers are relatively easy to implement and do not require
huge amounts of hardware, they can be built into smaller, portable devices. Com-
bined with the wireless Internet possibilities, these new devices have the ability to
change the business world.
As technology and the Internet become more important in our lives and busi-
nesses, portability becomes increasingly critical. Even the portability offered by
laptops is useful. For example, you might need to pull up sales data in a meeting
or take some work home with you. In other cases, even more portability is needed.
Tablet PCs and PDAs provide the ability to access the Internet and your corporate
servers as long as you can fnd a network connection. Wireless connections are
available in many locations, but still have limited range. Newer cell phones offer
greater connectivity to the Internet from any major city or transportation area in
the world. As people demand greater continuous access to the Internet, portable
devices become more important. At the same time, the servers that store the data
and provide the major processing will also become more important.
What Is a Server?
Do you want to run a Web site? Being a participant in the Internet and running a
browser is one thing. Running a Web server is completely different. Establishing
an e-business requires that you either run a server or pay someone to run it for
you, so you should understand some of the main issues in Web servers.
Reality Bytes: GPU Processors Go High Performance
Gamers and other users of high-end graphics know that Nvidia is one of the largest
providers of high-powered graphics cards. The cards are based on a graphics process-
ing unit (GPU) architecture where each car contains many small high-speed proces-
sors. Although a high-end graphics card might appear to be expensive, on a per-GPU
basis, the costs if considerably lower than for traditional processors. Consequent-
ly, Nvidia also builds specialty cards for supercomputers that use thousands of the
GPUS to handle massively parallel computations. In 2011, a single TeslaM2090 card
could produce up to 665 Gigafops of computations. The M2090 has 512 Cuda
cores and GDDR5 memory cards. Transferring heat off the cards is a major design
consideration and the cards come complete with liquid-cooled heat sinks. The cards
support code written in common languages including C++, FORTRAN, Java, and
OpenCL. The M2090 can be plugged into relatively standard computers through an
arrangement with HP. Nvidia also sells plug-in modules for supercomputers, includ-
ing those from Cray Research.
Adapted from Jonathan Angel, Nvidia Taps GPUs for Worlds Fastest Parallel
Processing, WindowsForDevices.com, May 17, 2011.
Technically, almost any reasonably up-to-date PC can function as a Web server.
A Web server is essentially a piece of software that monitors the full-time Internet
connection and delivers the requested pages. But to perform e-business tasks, the
server also needs to evaluate programmed Web code and interact with a database.
In most cases, the primary issue with Web servers is scalability, the ability to in-
crease the computing power without changing the application. The goal is to build
a server inexpensively enough to make the application proftable, yet capable of
expanding to handle increased demands of the future.
In terms of hardware, the primary characteristics that you want in a Web server
are (1) scalability, (2) easy backup, and (3) easy maintenance. One solution to
all of these problems is splitting the major computer components into separate
pieces. For example, server farms use multiple small computers instead of one
large machine. A storage area network (SAN) can be built using fber channel con-
nections to a set of external disk drives.
High-end servers have built-in backups for all critical components. For exam-
ple, they have dual power supplies, network cards, and hot-swappable disk drives
in a RAID confguration. If a power-supply or disk drive fails, the other copy au-
tomatically picks up the load. In most cases, the failed component can be replaced
without turning off the computer (hot-swapped). More sophisticated servers can
be built into server farms that contain multiple computers that distribute the com-
puting load and have access to the same databases. If one computer fails, the oth-
ers automatically handle the incoming tasks. More importantly, the system can be
scaled up by adding more computers.
Reality Bytes: International Notations
Most applications today have the ability to use characters that are not found in the
U.S. alphabet. For instance, in France or Mexico, you might need to use an acute
mark (). However, different software packages handle the characters differently, so
you might have trouble converting a document from one word processor to another
or to a different computer. For example, if a French subsidiary is using WordPerfect
and the Canadian headquarters is using Microsoft Word, they can both print reports
using the special characters. However, the document might change when the Cana-
dian users attempt to retrieve a French document electronically.
Furthermore, if you work for an international organization, remember that peo-
ple in different countries write dates differently. For example, 5/10/93 means May
10, 1993, in the United States but would be interpreted as October 5, 1993, in Eu-
rope. Most word processors enable you to choose how automatic dates should be
displayed.
Numbers are also handled differently in European nations. The use of commas (,)
and points (.) is reversed from the U.S. version where commas separate thousands
and the decimal point delineates the fractional component. (the number 126,843.57
in the United States should be denoted as 126.843,57 in Europe.)
You also need to be careful with currencies in spreadsheets. When you transfer
documents to other languages or fonts, be sure to check any currency symbols. A
few systems will automatically change the symbol to the local units (e.g., change $
to ), but unless the numbers are converted by exchange rates, these changes would
be incorrect.
Virtual Machines
With multiple-core processors and tons of memory available, it is common prac-
tice today to set up computers as virtual machines. A virtual machine (VM) is the
process of running a separate computer in software on top of one set of hardware.
Generally, the process is used to run multiple virtual computers on a single set of
hardware. Basically, one physical set of hardware is confgured to run multiple
operating systems at the same time so it appears to be several separate computers.
Figure 2.24 illustrates the basic concept of a virtual machine. One physical com-
puter is used to host three separate computers. The hosted VMs are often servers,
but the technique can also be used for personal computers. The goal is to allocate
hardware resources to the VMs that need the additional computing power. The
technique can save money, space, and electrical power, by reducing the amount
of physical hardware. It is also useful for separating tasks. Some server software
can interfere with other tools and it is often easier to run them separately. VMs
also make it signifcantly easier to backup and restore the entire computer. Once
the VM is created, it is stored as a software fle on the disk drives. If anything
goes wrong, the fle can be reloaded and the entire VM is recreated immediately
without needing to reinstall the operating system and all of the applications. On
personal computers, the VM technique can be used for creating development and
test environments. One system could be used for day-to-day work, and another for
testing experimental software. If the experimental software causes problems, the
entire VM can be deleted with no effects on the production computer.
VMs can be useful, but keep in mind that you usually need separate software
licenses for each VMboth the operating system and the application software.
If the systems are based on open-source software, or the organization has a site
license these costs are minimal.
Application Software
What are the main software applications used in business? The
main reason for buying a computer is its application software. Regardless of your
job, you will always perform certain tasks. As a manager and decision maker, you
Figure 2.24
Virtual machine. One set of computer hardware is confgured to run multiple,
independent operating systems. Particularly useful for servers, the goal is to allocate
the hardware to the tasks that are busy.
One physical Computer
Multiple core processor
Shared Memory
Shared network
Allocated
disk
space
VM1: Windows Server
1 processor, 4 GB RAM
VM2: Linux Database Server
2 processors, 8 GB RAM
VM3: Windows PC
1 processor, 2 GB RAM
frst have to gather data (research). Then you analyze the data and determine al-
ternatives (involving calculations). Next you will make decisions and implement
them (requiring writing, communication, and organizing resources). Each of these
tasks can be supported by computer resources. The trick is to use the appropriate
tools for each task. The objective is to use the tools to make you more productive
in your job.
The catch is that productivity is a slippery problem in businessespecially
with respect to computer tools. One measure of productivity involves effciency:
Can you perform tasks faster? A second, more complicated measure of productiv-
ity involves effectiveness: Can you make better decisions? Early uses of comput-
ers focused on improving effciency by automating manual tasks. The tools were
often justifed on the basis of decreased labor costs. Today, managerial uses often
focus on effectiveness, which is harder to measure.
An important concept to remember with application software is that it was cre-
ated by teams of designers and programmers. In the old days, when software
was custom-written for specifc companies and users, users could tell the design-
ers how the software should behave. Today, most companies rely on commercial
off-the-shelf (COTS) software that is created to support the needs of millions of
different users. In creating the software, designers had to make thousands of de-
cisions about what the software should do and how it should appear. Sometimes
their fnal choices might seem strange, and software does not always work the
way you might prefer. Some issues can be resolved by customizing the software
to your particular situation. Other times, just remember that you acquired the soft-
ware for a tiny fraction of the price you would have paid for custom software.
Research: Databases
Almost any job involves workers searching for information. This search could
involve any of the fve basic types of data. Computers have made it substantial-
ly easier to fnd, compare, and retrieve data. Two important strengths of a data-
base management system (DBMS) are the ease of sharing data and the ability to
search for data by any criteria. In terms of productivity, a DBMS can both make
you more effcient and improve your decisions. It improves effciency by provid-
ing easier and faster data retrieval. By providing more complete access to data, a
DBMS helps ensure that your decision incorporates all relevant data.
One complication with research is that you must frst determine where the in-
formation is located. It could be located on your personal computer, the groups
networked server, the companys central computers, a computer run by the gov-
ernment, or one purchased from another company. Unless all of these databases
are connected, you must frst determine which database to search. Chapter 5 fo-
cuses on the use of database management systems. Today, along with numbers
and text, DBMSs can handle large text fles, pictures, sound, and video. DBMSs
form the foundation of almost all business applications.
Today, almost all data in a company is stored in a DBMS, and the things that
are not stored in the DBMS really should be there. Even data outside the company
such as government data and data from suppliers is often stored in a DBMS. Re-
search typically involves fnding and retrieving the relevant data from a DBMS.
Analysis: Calculations
Almost everyone performs computations in a job. Spreadsheets were initially cre-
ated to overcome the limitations of simple calculators. They are useful for com-
plex calculations and for creating charts. Most people fnd spreadsheets useful
because their disciplines began with models on paper that used columns and rows
of numbers. For instance, most accounting systems and fnancial models were de-
signed for ledgers in this way. Whenever software mimics the way you already
work, it is easier to learn.
Spreadsheets have many additional features. Graphs can be created with a cou-
ple of mouse clicks. Most packages enable users to modify the graphs by adding
text, arrows, and pictures. Spreadsheets also perform various statistical and math-
ematical analyses. You can perform basic matrix operations such as multiplication
and inversion. Statistics capabilities include multiple regression to examine the re-
lationships among different variables. Linear programming can be used to search
for optimum solutions to problems. These additional features are designed to help
you make better decisions by providing more powerful decision-evaluation tools.
The problem with spreadsheets is that people use them to store data. It is also
too easy to make mistakes on spreadsheets that are diffcult to spot. What often
happens is that someone develops a spreadsheet to analyze some data, then other
people use the spreadsheet and modify the data and formulas. Fairly quickly, there
Reality Bytes: Keyboards are Faster than Mice
Key Task
Ctrl-A Select all
Ctrl-C Copy
Ctrl-V Paste
Ctrl-P Print
Ctrl-S Save
Ctrl-X Cut
Ctrl-Z Undo
Alt Menu, then number or letter
Alt-Tab Switch application
Windows-# Start applications in taskbar
Windows-D Show desktop (and restore)
Windows-L Lock the computer
Windows-P Projector/switch display
Windows-R Run box
Ctrl-F4 Close document
Alt-F4 Close application
Ctrl-F6 Switch document within application
Shift-Select Click or arrow, select contigous items
Ctrl-Select Click, select multiple discrete items
Adapted from Patrick Miller, Windows Toolkit: 50 Super Software Secrets, Com-
puterworld, July 2, 2010.
are multiple versions of the spreadsheet, out-of-date data, and errors in the calcu-
lations. And managers want to use these to make decisions. Yes, it is possible to
improve the way spreadsheets are built, and Chapters 3 and 8 show ways in which
spreadsheets can be shared. But, ultimately, spreadsheets are often better used as
short-term tools. They are useful if you need to analyze a small set of data or make
one-time calculations. Any decision that requires large, changing, data sets, mul-
tiple people, and critical analyses should be built as a specifc application. Chapter
9 examines some of the more powerful analytical tools available that can be used
for complex decisions.
Communication: Writing
The primary gain from word processing is increased effciency. Word processors
improve communication by making it easier to revise text, fnd writing errors,
and produce legible reports. Word processors today also include a spell-checker,
a thesaurus, and a grammar-checker. Although they are not the same as having a
human editor correct your writing, these are all useful tools for writers. Grammar-
checkers use standard measures to estimate the reading diffculty level of your
writing. For instance, if you write an employee policy manual, you want to make
sure that an average employee can understand it. Most word processors also have
outline tools that help you organize your thoughts and rearrange a document, im-
proving the communication.
The proliferation of word processors creates additional advantages. Today, all
documents should be electronic. It should be possible to search electronically for
existing reports, manuals, and so on.
Communication: Presentation and Graphics
In many cases, the difference between a good report and an outstanding report is
the presence and quality of the artwork. Graphs and pictures are used to commu-
nicate information and feelings. Charts and graphs are easy to create, store, and
modify using graphics software. Even if you are not an artist, you can buy clip art
that was created by someone else and use it in your reports or charts. By using ex-
isting art libraries, you can create reports and presentations in a few hours. In the
past, reports and presentations took days or weeks to fnish by a staff of artists. By
improving the appearance, a well-designed graphic can also improve communica-
tion and decision making.
To create or modify your artwork, you need a graphics package and an input
device such as a mouse that enables you to draw on the computer screen. Most
commercial artists use scanners so they can draw the original on paper and con-
vert it to computer form. The digitized form enables you to make very precise
changes, since you can zoom into a specifc area. Zooming is helpful if you need
to force lines to meet exactly or you want to make changes to small items, such as
eyelashes on a person.
Although you do not have to be an artist to incorporate artwork into your re-
ports and documents, you do need an element of artistic sensibility. The goal is
to use art to enhance your presentation, not clutter it. Knowing what looks good
and using restraint are signs of artistic talent. Remember that faster does not al-
ways mean better. Use some of the time savings to put more thought into your
presentations.
Communication: Voice and Mail
All jobs require communicationwith coworkers, managers, and customers or
clients. Word processors help with reports and memos, but much of our commu-
nication is less formal. Everyone is familiar with answering machines for tele-
phones. Businesses have taken this concept a step further by using voice mail
systems. Voice mail systems record messages much like an answering machine,
but they store the messages in digital form on computers. They usually give the
caller control over where the message is sent. Most people in the United States
have dealt with systems that direct you to press buttons on the telephone to make
choices. Some voice mail systems enable you to send the same message to several
people. Some systems also enable you to skip messages or fast-forward to the end.
Many are now tied to e-mail systems so your messages can be integrated into a
single site. This process also makes it easier for companies to back up fles and
store data for later use. In particular, e-discovery systems can be used to archive
all message data for easy search and retrieval during a lawsuit.
Networked computers can be used to send messages directly to other users.
These electronic mail (e-mail) systems are typically used to send written notices
to various people. They can be used to send pictures, facsimiles (faxes), or even
voice messages if the computers have sound boards and speakers. Instant mes-
saging (IM) is also relatively popularoften in the form of text messages sent by
cell phone. It is nice to have several options for communication. The challenge is
to choose the appropriate method for each conversation and each person. Some
people prefer the immediate response of voice communication, others prefer the
asynchronous nature of e-mail, so they can schedule communications around their
work.
The basic problem with any communication system is that sooner or later it
becomes cluttered with junk mail. One of the advantages of text e-mail is that the
recipient can have the computer scan the messages to search for items that are im-
portant or interesting. With the appropriate mail flters, junk mail can be discarded
automatically. Messages also can be retrieved and stored for future reference or
forwarded to other people.
Organizing Resources: Calendars and Schedules
An important task of managers is to organize company resources so that they are
used most effectively. An important function is scheduling workers. Schedules in-
volving line workers entail making sure that there are enough employees to get
the job done, but no extra employees. Schedules involving managers typically in-
volve meetings and require trade-offs between competing demands for a manag-
ers time. Several software tools are available to improve both types of scheduling
and make more effcient use of the human resources.
Most managers spend a considerable amount of time in meetings. In fact, it
becomes diffcult to keep track of appointments, set up new meetings, and re-
schedule meetings. The process is even more diffcult when some participants
are traveling or are based in another city or country. Several software packages
store appointments and schedules on electronic calendars. Electronic calendar and
scheduling software enables data to be shared with other people. For instance,
each person in a department would use the electronic calendar to keep track of his
or her personal, departmental, and corporate appointments. To schedule a meeting
with departmental members, the manager selects an approximate time, specifes
the priority of the meeting, and flls in the participants, location, and subject infor-
mation. The calendar software then searches each personal calendar for the best
time for the meeting. It overrides lower-priority meetings and places the complete
notice on each persons calendar in a matter of seconds. Employees have to check
their calendars periodically to see whether meetings have been added or changed.
Augmented Reality
Augmented reality is a specialized tool is useful in some business operations.
So far, it has minimal uses in management, and the applications are still evolv-
ing. Augmented reality consists of overlaying computer-retrieved data on top of
images of real-world items. One of the more useful applications was developed
several years ago for mechanics working on highly-complex engines, such as jet
engines for the military. Workers with special glasses would look at engine com-
ponents and the computer would display an overlay of the schematics and the
part numbers. The part numbers enabled the mechanic to easily order the correct
replacement item for worn parts.
Today, people are more familiar with examples of cell phones, where the com-
puter (phone) overlays information on images taken through the phones cam-
era. The technique combines location data with image recognition and retrieves
matching data from online databases. A variation that has been proposed is to add
facial recognition software and tag images with peoples names. It might be useful
for executives who meet hundreds of people a day, enabling them to add a per-
sonal touch by talking to people by name and retrieving background information.
Or, the technology might lead to creepy versions of stalking.
Figure 2.25
Paper (and paperboard) consumption appears to be declining in the U.S. and
possibly in the world as well. As video screens improve and people become more
adept at handling data online, there should be less need for paper. Raw data: UN
Food and Agriculture Organization, http://faostat.fao.org/site/626/DesktopDefault.
aspx?PageID=626.
0
2
4
6
8
10
12
14
16
18
Paper Consumption: Kg/Person/Year
World
USA
Te Paperless Ofce?
You might think that with increased use of electronic data, there would be less
need for paper. For many years, the opposite was true. People tended to waste
paper when they knew they could easily reprint a report. However, as display de-
vices have improved, and people have become better at adjusting to electronic
screens, it appears that the use of paper is decreasingat least in some countries.
Figure 2.25 uses UN data to show consumption of paper per person per year in the
U.S. and the world. The world fgures show a slight decline, but the drop for the
U.S. is relatively strong over the last decade.
Some issues exist with long-term storage of electronic data. In terms of paper,
libraries have stored books for hundreds of years. It is not clear that electronic
data can survive even a few decades. First, the data might not physically last that
long. Magnetic drives, Tapes, and even CDs and DVDs might lose their data over
time. There are also unresolved issues about future compatibility. The hardware
and software to read them may disappear in a short time.
Open Software
It is diffcult and time consuming to create software. Commercial vendors spend
billions of dollars researching customer needs, developing software, and updating
and revising the code. For a few decades, major software development has been
performed by large corporations that hire thousands of programmers and sell the
software for a proft. Microsoft alone spends over $6 billion a year on research
and development. On the other hand, the marginal cost of a software product is
almost zero. Almost all of the expenses are front-end fxed costs, except for mar-
keting, distribution, and dealing with customers.
In the past few years, led by visionaries such as Richard Stallman with GNU
and Linus Torvalds with Linux, a new approach to software development has
arisen. Complex programs have been created by using the volunteer talents of
thousands of programmers around the world. The result is software specifcally
developed to be distributed free of charge for anyone to use. The operating system
Linux is a popular example, but several application packages exist as well.
As a manager, you will eventually have to answer the question of whether you
should use open-source software or continue to rely on commercial packages. The
answers are diffcult, and discussions are often tinged with religious fervor. Al-
though free seems like a good price, how much do you really pay for software?
Today, the Windows operating system price is largely embedded in the cost of
buying a new computer. On the other hand, more expensive packages such as the
Offce suite, graphics editor, or database management system can represent large
sums of money.
Other issues in open-source software remain unresolved. Often, the open-
source packages have minimal support and can be harder to use. Certainly, nov-
ices fnd the Linux/UNIX operating system diffcult to learn. Issues about security
are unresolved. With thousands of programmers working on millions of lines of
code, it might be possible for someone to sneak in a tiny, but nasty, routine. On the
other hand, with thousands of people looking through the code, the problem might
be caught early. You also need to consider whether the open-source programs will
be around for a long period of time. With no real income stream, can they survive?
Will the leaders remain in charge? These are important questions to ask any soft-
ware vendor. But the risks are slightly higher with open-source programs. On the
other hand, you do get copies of the source code, so in theory, you could edit and
maintain the software indefnitely by yourself. But how many managers or com-
panies have that ability?
In the end, the choice of software comes down to the features and the cost of
support. Commercial software is generally supported and updated with the costs
covered by the initial purchase price, and sometimes annual license fees. With
software such as Microsoft Offce, millions of people use the software and gener-
ally need minimal training. Open-source software usually has no up-front cost,
but often carries annual costs for help, installation, and sometimes updates. Sever-
al companies, such as Red Hat and Oracle, make money selling support for open-
source software.
Cloud Computing
Is it possible to handle all computing tasks with online servic-
es? When computers were frst invented, they were large, expensive, and needed
large numbers of support personnel. Consequently, all applications and data were
run in a single location. Eventually, slow networks were created and display ter-
minals were installed in various locations in a company. These terminals were
basically keyboards with simple text displays but they enabled people to work
on computer applications from their offces. All of the data and applications re-
mained on the central computer. As minicomputers and personal computers were
developed, computer applications and data moved outside the central computer.
Figure 2.26
Spreadsheet in Google Docs. http://docs.google.com. Google Docs is a cloud service
that is free for individuals or $50/user/year for companies that need more services.
You can create documents, spreadsheets, presentations, and drawings. The features
are limited compared to standard Microsoft Offce products but can be created using
only Web browsers and are easy to share.
Gradually, computers got smaller and more powerful. At the same time, network
speeds increased dramatically, making it possible to connect handheld devices at
high speeds from almost anywhere in the world. These networks are explored in
Chapter 3 and the consequences of data management are covered in Chapter 13.
But, the key for this chapter is that these capabilities exist now, which provide a
method to change the way people use computers. Today, central Web servers can
hold all of the data and run applications, while users access them through power-
ful, but relatively inexpensive portable devices such as cell phones, tablets, and
laptops.
Cloud computing is the popular term for storing applications and data on Web
servers. The Internet is sometimes referred to as a cloud (and drawn that way) be-
cause the servers are interconnected and it does not matter where they are located.
Often, cloud computing costs are based on variable prices with low or no fxed
costs. Sometimes costs are low enough to be supported by advertising revenue
so services are provided free to users. Most people are familiar with the social
network services (Facebook, YouTube, and Twitter), as well as the search engine
services (Google, Bing), and other specifc applications such as photo sharing and
editing sites (Picasa, Flickr, Photoshop). Thousands of cloud computing applica-
tions and sites exist and it seems that more are being created every day.
Figure 2.27
Spreadsheet in Microsoft Web Apps. http://offce.microsoft.com/en-us/web-apps. The
user interface is similar to the standalone Offce tools with slightly fewer features.
It is easy to upload and download documents to the standalone version and easy to
share with other users. Online documents can be created and edited in a browser even
on cell phones and tablets.
Think about cell phones and tablets for a second. For the most part, these de-
vices use special-purpose operating systems (often Apple iOS or Google An-
droid). These systems were not designed to run the commonly used business tools
such as word processors, spreadsheets, and presentation software. But, do you re-
ally want them to run that software? And if they do, where are you going to store
all of the data? Instead, it makes more sense to use the devices to display the
applications, but leave the applications and data on a Web server. The two main
cloud computing sites that provide these services are Google Docs and Microsoft
Web Apps. Figure 2.26 and Figure 2.27 show a simple spreadsheet created on
the Google and Microsoft services. The Microsoft applications are similar to the
standalone Offce applications, and it is easy to transfer the documents between
the Web and standalone software. Both Google and Microsoft provide free access
for individualsbut the advertising can be distracting. Google also sells a version
for businesses that integrate a few additional features (such as a shared calendar)
and eliminates the advertising for $50 per user per year. With both the Google
and Microsoft cloud services, documents can be created and edited using a basic
browser, so they can be accessed by relatively simple devices include cell phones
and tablets. But, most people would fnd it tiring to write a long document on a
touch-screen keypad, so it would be useful to have a separate keyboard in many
cases.
One of the strengths of cloud computing is the ability to share documents with
other people. As long as the people sign up on the same service, it is straightfor-
ward to give them access to your documentsto view or even edit the document.
Generally, the cloud providers handle backups and keep the servers running with
high reliability. However, cases exist where documents have been lost. On the oth-
er hand, many other companies provide storage space online, so it should be pos-
sible to make your own backup copies using multiple storage or e-mail services.
So, is it possible to run most common applications online using cell phones and
tablets? Possibly; but it would be tiring to create large documents with a touch-
screen keyboard. Most people would fnd it easier to type using a separate key-
board. Still, the main documents could be created using a laptop or perhaps a
tablet, then they will be available for viewing and small edits using a cell phone.
Summary
One of the original purposes of computers was to make it easier to perform basic
tasks. Over time, as computers have become more powerful, they have come to
support increasingly complex tasks. Today, in addition to increasing effciency,
computers can help you make better decisions. One major change is in the type
of data routinely processed. The fve major types of data are numbers, text, imag-
es, sound, and video. To handle more sophisticated data and more diffcult tasks,
computer hardware and software have grown increasingly complex.
To choose a computer that best meets your needs, you must evaluate the four
basic hardware components: input, processor, output, and secondary storage de-
vices. Each component is measured by slightly different characteristics. Input de-
vices are selected based on the specifc task (such as a keyboard for typing, mouse
for pointing, or a microphone for voice input). Processors are often selected based
on speed and price. Output device quality is appraised by resolution and color
capabilities as well as initial and ongoing costs. Secondary storage is evaluated
based on speed, capacity, and price.
Application software is the primary source of improved productivity. Packages
exist to assist in research, analysis, communication, and organizing resources. Da-
tabase management systems are used for research and data sharing. Spreadsheets
and other analytical tools assist in calculations. Word processors, drawing pack-
ages, voice mail, and e-mail are used for communication. Electronic calendars and
scheduling software are used to help organize human resources.
With the presence of global high-speed networks, it is now possible to store
most applications and data using cloud computing on the Internet. Companies and
individuals now have the choice of carrying smaller, simpler display devices that
access everything stored on Web servers. Some limitations exist today, and cost,
security, and reliability questions remain but this approach can simplify the man-
agement of fundamental computing.
Key Words
aspect ratio
augmented reality
binary data
bitmap image
byte
cache memory
CD-ROM
clip art
cloud computing
CMYK
commercial off-the-shelf (COTS)
device drivers
dots per inch (dpi)
DVD
e-discovery
electronic mail (e-mail)
ergonomics
grid computing
hardware
high-defnition TV (HDTV)
icons
input
instant messaging (IM)
megabytes
milliseconds
motherboard
nanoseconds
operating system (OS)
optical character recognition (OCR)
output
pixels
processor
random access memory (RAM)
redundant array of inde-
pendent drives (RAID)
resolution
RGB
scalability
software
solid state drive (SSD)
style sheets
synthesizer
templates
Unicode
vector images
virtual machine (VM)
voice mail
A Managers View
Technology constantly changes. New features are added every day to hard-
ware and software. How do you know which features to buy and which to
ignore? The key is to evaluate the features of the main components and to
understand how you will use the computer. Given the declining prices of
processors, RAM, and secondary storage, it is relatively easy to buy a rea-
sonable computer today for most common business applications. When you
get to issues of portability or Web servers, you need to evaluate the choices
more carefully.
Web Site References
Free News Sources
Associated Press www.ap.org
CNN www.cnn.com
ESPN espn.go.com
Fox News www.foxnews.com
Internet News www.internetnews.com
MSNBC www.msnbc.msn.com
News.com www.news.com
News Map newsmap.jp
USA Today www.usatoday.com
United Press International www.upi.com
Wired www.wired.com
ZDNet www.zdnet.com
Almost any Magazine or Newspaper (some charge)
Fortune www.fortune.com
Te Economist www.economist.com
Wall Street Journal wsj.com
Washington Post www.washingtonpost.com
Review Questions
1. List and describe the fve basic types of data.
2. What are the two main color models? What are the color codes for yellow
and green in each of the models?
3. What is lossy data compression and why does it create smaller fles?
4. What are the primary hardware components in a computer?
5. How do you know which components to upgrade to improve your computers
performance?
6. Why are style sheets important for large documents and Web sites?
7. Why are SSDs better than traditional hard drives? What are the drawbacks?
8. What are the advantages and drawbacks to using removable disk drives as
backup devices?
9. How does the operating system defne the way a computer works?
10. How does the use of Web browsers change the way applications operate?
11. How does a virtual machine reduce the costs of large server farms?
12. Briefy describe fve tasks you expect to perform in your job as a manager
and list the application tool you will use.
13. What will it take for people to adopt a paperless offce?
Exercises
1. What are the current best prices of disk drives ($/GB) and RAM ($/GB)?
How much does it cost for the fastest processor you can fnd (not counting
exotic supercomputers)?
2. Using a common Web site or retail store, fnd the price of a mid-level laptop.
List the primary specifcations. Identify the cost of doubling the amount
of RAM on the laptop at the time of purchase. Use a discount Web site
to compare the price of purchasing the RAM separately and installing it
yourself.
3. Find or create a color chart that identifes and displays at least the 27 primary
colors for the RGB color model. Use three levels (0, mid, max) for each of
the three colors.
4. Find a high-resolution digital photograph. Using photo editing software,
save several copies of the photo in JPEG format with different levels of
compression. At what point do you begin to notice the photograph quality
degrade? Compare the fle sizes as well. Print the photo (a black and white
print is fne) and compare the results to the original.
5. What is the highest capacity common SSD available? What is a typical price?
How fast is the write speed?
6. How much would it cost to purchase the current release of Microsoft Offce
for a group of 10 people? How much would it cost to use similar software
from an online provider? What are the drawbacks to going with an online
provider?
7. Estimate the storage space (number of bytes) required for each of the
following items:
a. A textbook of 700 pages. Create two single-spaced pages with a small
PowerPoint drawing on one page. Convert it to PDF format and estimate
the size. Compare this number to other sources.
b. A full-color image from a 7-megapixel camera. Both in raw form (TIFF)
and an estimate of the size in high-quality JPEG format.
c. A 15-minute lecture recorded with the PC audio recorder. Make and
save a 30-second recording and use the fle size to estimate a 15-minute
recording.
d. If you wanted to store your favorite half-hour television show in digital
form, how many bytes of storage would it take? Extra credit: How much
space would it take if you remove the commercials? (Hint: Time the
commercials.) How much space would it take in HDTV format?
e. A Unicode document of about three pages of single-spaced text.
8. You have a photo from a 7-megapixel camera at 3072 x 2034 pixels. Your
image editing software defaults to 72 pixels per inch, and claims the image is
42.667 x 32 inches. You want to convert to a standard 10-inch wide photo but
keep the aspect ratio. You need to fnd a new resolution that sets the proper
dpi and the height.
9. Use research to determine the percent of computers running the Linux
operating system.
10. Operating systems defne drive space in terms of binary numbers where
1024 = 1 kilobyte, and 1 gigabyte is 1024*1024*1024. But disk drive
manufacturers advertise storage space in decimal values, using 1,000 instead
of 1,024. If you buy a drive advertised as 200 GB, how many bytes will the
operating system recognize on it?
11. Identify at least two competitors to Microsoft Offce software and fnd
reviews from people who have used them to comment on their viability for
business use.
Technology Toolbox: Speech Recognition
12. Set up speech recognition on your computer and train it. Choose a paragraph
of text and dictate it into Word. Do not make corrections to the text as
you dictate it. When you have fnished, copy the paragraph and make the
corrections to the copy. Count the number of mistakes and hand in both
paragraphs.
13. Do a quick survey of other students (not in the class) to fnd out how many
have used speech input. Of the ones who have tested it, how many use it on a
regular basis? Report the results and comment on the low usage rates.
14. How are the speech recognition systems used in other devices, such as cell
phones, different from those used in basic personal computers?
Technology Toolbox: Charts
15. Identify the best chart to use for the following datasets, and give a brief
justifcation for your choice:
a. Sales by department for the last fve years.
b. Sales by employee for the last month.
c. Production data for output quality and percent of carbon.
d. Share of sales to fve nations.
e. Total customer billings by employee for the last 24 months.
16. Find at least two data series from the government (try www.fedstats.gov)
and plot them. Briefy explain any patterns or trends.
17. Enter the following sales data into a spreadsheet. Create two charts, one
showing total sales over time and one showing the changing percentage sales
over time by product type. Briefy comment on the two charts and describe
the difference in the information content.
Year Soft Goods Hard Goods Movies
2007 425 632 265
2008 531 789 378
2009 618 865 425
2010 692 897 581
2011 753 933 792
2012 829 989 882
Technology Toolbox: Spreadsheets
18. A company is evaluating the purchase of a machine that costs $50,000. The
company will have to pay money each year for maintenance costs. However,
the machine will increase profts by several thousand dollars each year. Enter
cost and proft data into a spreadsheet. Since the costs and profts occur over
time, you have to discount them to a single point in time. Use the net present
value (NPV) function to compute the present value of the costs and profts.
Initially, assume the discount rate is 4 percent, but build the spreadsheet so
managers can change the value easily. Use the Goal Seek tool to fnd the
discount rate at which the investment breaks even. That is, search for the
discount rate that sets the computed net value (net profts-net costs) to zero.
Year Cost Proft
2008 -50,000 10,000
2009 -2,000 12,000
2010 -2,000 13,000
2011 -2,000 14,000
2012 -5,000 15,000
19. Create a new spreadsheet for a company that has sales of two products (shoes
and hats) in two regions (East and West). Put data for the East division in one
worksheet and data for the West division in a separate worksheet. Create a
third worksheet to display the totals of the two divisions. Chart the changes
over time.
Year Shoes Hats
2008 456 110
2009 571 98
2010 632 87
2011 771 69
2012 780 120
East
Year Shoes Hats
2008 1024 234
2009 1305 198
2010 1525 178
2011 1598 165
2012 1652 208
West
Year East West Total
2008
2009
2010
2011
2012
Total
20. You need to evaluate your employees. You have a list of their performance
on four tasksmeasured in terms of time. The time is the number of minutes
they fnished behind the fastest person. So smaller numbers represent better
performance. Two of the employees were injured at different points and did
not complete all of the tasks. Enter the data into a spreadsheet and analyze it
by computing the mean and standard deviation for each stage. Then compute
four new columns that contain the individual Z-score for each employee in
each stage: Zi = (Xi average)/Std Dev. Compute the average Z-score for
each person and rank them. Remember that smaller (negative) numbers are
better. Chart the fnal averages. One extra credit point if you can identify the
employees.
Employee S0 S3 S5 Sf
Basso 0.23 3.05 1.05 7.23
Cruz 0.53 14.85 3.13 31.80
Danielson 0.17 9.92 1.33 11.50
Davis 0.32
Hincapie 0.13 3.05 0.67
Leipheimer 0.00 0.00 0.00 0.00
McCartney 0.32 0.07 0.42 0.90
Vandborg 0.30 14.85 3.03 25.30
Teamwork
21. Compare your cell phones. Build an Excel table and compare them based
on screen resolution and camera resolution. Indicate if the phone has speech
recognition and rate how well it works (on your own phone). What data entry
methods are supported? How well does the Web browser work (if it has one)?
22. Each person should research a current version of a tablet computer and write
a short list of its strengths and weaknesses. The group should choose one
tablet that could be used by business managers, and list the main tasks for
which it would be useful.
23. Find a good quality photograph or take one yourself. Have one person
save the photograph to at least three copies using different levels of JPEG
compression. Hint: Adobe Photoshop gives complete control, but if it is
not available, use Microsoft Picture Manager or try an online site such as
imageoptimizer.net. Have each team member look at the compressed photos
and try to rank them from highest-to-lowest quality. Compare the rankings
with the actual order. What do the results tell you about how you should store
and display photographs?
24. Have each person on the team fnd an example of a good chart and a bad
chart using newspapers and the Internet. Combine the results from each
person into a document so that each person can vote on the best and worst
charts.
25. Set up free accounts on one of the online application sites (such as Google).
Create a short document to describe a recent movie or television show. Have
each person add a couple of lines of comments about the movie. When
fnished, comment on the benefts and drawbacks of using the online system
to write documents versus the traditional method of using separate copies and
e-mailing them.
26. Using the Export Data form, copy the data to a spreadsheet and create graphs
for the following situations. (Choose the type of graph you feel is best suited
to present the data.)
a. Sales by model type.
b. Sales by month.
c. Sales by model type for each month.
d. Sales by state.
e. Sales by employee by month.
27. Using the existing forms, enter data for a new bicycle order.
28. Find at least two other bicycles (e.g., on the Internet or from a dealer). Create
a spreadsheet comparing the features and costs with a similar bicycle built by
Rolling Thunder Bicycles.
29. Using the Export Data form, copy the data to a spreadsheet and compute
the average proft margin for each type of bicycle for one different year.
Comment on any differences that you fnd.
30. Assume you have to give a presentation to the marketing manager. Create a
slide show to compare the sales of each model type for the last two years.
Additional Reading
Adams, S., R. Rosemier, and P. Sleeman Readable Letter Size and Visibility for
Overhead Projection Transparencies, AV Communication Review, 1965, 412-
417. [An early discussion of creating good presentations.]
Science: The Numbers Game, Time, February 20, 1988, p 54-58. [Short history
of computers.]
Simonds, D. and L. Reynolds, Computer Presentation of Data in Science: A Do It
Yourself Guide, Boston: Kluwer Academic, 1989. [Ideas for presentations.]
Cases: Te Computer Industry
Te Industry
How often should you replace a personal computer? Hardware and software are
continually changing. To run new software and take advantage of new peripher-
als such as Blu-Ray players or high-speed drives, you often need to purchase new
computers. Typically, personal computers are replaced after three years. Yet frms
can often delay purchasing new computers in a weak economyto save money.
All of these decisions affect the industry, and overall sales can skyrocket one year
and die down the next. The industry is also driven by brutal competition. Between
Dells quest for low prices and effcient production, and relentless price cuts by
Chinese, Taiwanese, and Korean manufacturers, PCs have largely become com-
modity items.
Where do you buy PCs? Some people like to physically see the items they
buy and be able to take them home immediately. So there is some incentive to
sell computers in retail stores. But remember that PCs have a limited shelf life.
A model can become obsolete in a few weeks, and its value can drop in a matter
of days. How can a retailer stock items when they depreciate so quickly? Stores
can try to forecast demand, but if the economy drops, people may stop buying and
the store will be stuck with obsolete computers. Yet selling strictly by mail order
means that a manufacturer might miss out on sales to a relatively large market.
These conficts have caused considerable grief to computer frms over the last
few years. Dell, the leading mail-order/Internet vendor, at one time tried retail
sales and gave up. Gateway tried to ride the middle line by running its own retail
stores. It, too, lost money and closed the stores. Apple has opened retail stores
and been relatively successful. But Apples target has largely been consumer, not
0
20
40
60
80
100
120
140
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
$

B
i
l
l
i
o
n
Annual Revenue
HP
IBM
Dell
Apple
Sun
Acer
Lenovo
-5
0
5
10
15
20
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
R
a
t
o
HP
IBM
Dell
Apple
Sun
Acer
Lenovo
businesses. As HP as increased sales through stores such as Best Buy, Dell has
countered by offering a few models through Wal-Mart.
After several years of dismal growth, demand for personal computers increased
in 2003 and 2004. PC unit shipments in both the second and third quarters of 2003
were higher than expected. Consumer buying remained strong, and a transition
from desktop computers to notebooks has helped to fuel growth in the category.
As the average age of the existing machines increased, companies began replacing
their computers. Aggressive price competition is a signifcant factor in stimulating
demand. Vendors have reported disappointing PC revenues due to price compe-
tition. Competitive pricing pressures continue to drive pricing decisions. In ad-
dition, business pressures have not recovered. Consumer demand for notebook
computers has also driven demand.
International sales are an increasingly important element in growth for the ma-
jor computer vendors. Growth in Europe and the Middle East has seen an 18.7
percent year-to-year increase in PC shipments for the third quarter of 2003. This
growth was built on the robust sales of portables, which rose 51 percent in con-
sumer and business demand. Aggressive price competition was a signifcant factor
in stimulating demand.
The top frms continue to fght for dominance in the industry, but the weaker
frms are attempting to gain ground by consolidating. In 2004, Gateway purchased
eMachines. In the process, Gateway fred most of its existing managers and re-
placed them with managers from eMachines. Gateway eventually fled bankrupt-
cy. In 2006, Dell claimed that it made 1/3 of the PCs sold in the U.S. (2006 An-
nual Report). By 2009, HP had overtaken Dell as the leading PC manufacturer.
For 2009, the worldwide rankings were: 1. Hewlett-Packard, 2. Dell, 3. Acer, 4.
Lenovo, 5. Toshiba (Whitney 2010). By 2011, the argument about ranking was
whether or not tablets should be counted as PCs. If so, Apples market share
would be signifcant.
Te Consumer Electronics Market
The PC vendors are attracted to the appealing growth projections for consumer
electronics markets. For a while, fat-panel TVs helped Dell and Gateway improve
profts. By 2007, Gateway/e-Machines had dumped all of its stores and stopped
selling televisions. Dell still listed televisions, but the market had been stolen by
retailers engaged in vicious price-cutting (including Wal-Mart, Best Buy, and Cir-
cuit City).
Apple Computer had success with its retail outlets in 2003. Its stores are locat-
ed in high-traffc areas, such as upscale malls, near sizable Apple customer bases.
Apples concept has been to create a setting where salespeople can demonstrate
some of the new digital camera, digital music, and digital video products, and cus-
tomers can work with and experience these offerings in the store. The hope is that
the consumer will buy not only a new PC but also all the software and peripherals
needed for the complete digital solution.
Given the trend toward maturing growth rates in the PC industry, the strategy
of broadening product lines to tap higher-growth opportunities appears to be rea-
sonable. However, the margin benefts to the vendors remain in question, given
the highly competitive market dynamics of the consumer electronics industry and
the uncertainty of growth projections and product life cycles.
In late 2010 and 2011, manufacturers were focused on tabletsled by Apples
success with the iPad. Yet, early tablets were largely consumer devicescapable
of displaying movies, retrieving fles, playing music, and handling small e-mail
messages and Facebook updates. But, with low resolution (typically 1024 x 768)
and no keyboards, they were not replacements for full-blown laptops because it
was diffcult to enter data and create content on them. Continued evolution and
features might change these relationships.
Case: Dell
Dell posted rapid sales growth since July 1994, when it discontinued retail chan-
nel sales to focus exclusively on its direct business. After entering the market for
Windows-based workstations in 1997, the company quickly became the worlds
leading supplier. As of the frst quarter 2003, Dell was the worlds largest PC mak-
er, retaking the title from Hewlett Packard Company, which with its May 2002
acquisition of Compaq Computer became Dells largest rival. Desktop computer
systems made up the largest segment (53 percent) of the business. Dell ranked
frst in the United States and second worldwide in desktop shipments in calendar
2002. At the end of fscal 2006, Dell had net revenue of about $56 billion. By
2007, Dell was rapidly losing ground to Hewlett-Packard. PC shipments increased
by 13.1 percent in the frst quarter of 2006 but Dells sales grew by only 10.2 per-
cent (ZD Net 2006). By the frst quarter of 2007, Hewlett-Packard surpassed Dell
as the worlds largest producer and Dells worldwide shipments fell by 6.9 percent
in the frst quarter of 2007 and U.S. shipments fell by 14.4 percent leaving Dell
barely in the lead with 26.8 percent of the market compared to HP at 25.5 percent
(Darlin 2007). In 2009, HPs share of shipments was 19.7 percent, Dells was
12.9, and Acer was close behind at 12.7 percent.(Whitney 2010).
The companys higher margin notebooks and portables include the Latitude
product lines, for corporate users, and the Inspiron line, for consumers and small
businesses. Dell offers services that include professional consulting, custom hard-
ware and software integration, and network installation and support. Manufactur-
ing sites include Austin, Texas; Limerick, Ireland; Penang, Malaysia; and Xiamen,
China.
Dell is aggressive in pricing its products. As a direct seller of PCs, Dell benefts
because it can pass lower component costs through to customers faster than its
rivals. As such, it can lead the competition in price cuts. The company strategy
includes the ability to gain market share by leveraging its position as the low-cost
producer in the PC industry. Dells market share gains over the past year refect
the successful implementation of this strategy. This strategy is cramped, however,
by the rapid growth in component prices.
In February 2007, Michael Dell, fred Kevin Rollins who had been CEO since
2004, and Mr. Dell took the job back in an effort to turnaround the company.
Dells sales model was good at holding down costs, but sales were shifting. Dell
had been focused on business and educational customers who were more than
willing to purchase direct from Dell. But sales to consumers surpassed sales to
businesses and it turns out that individuals preferred to buy computers from retail-
ersnotably Best Buy, Circuit City, CompUSA, and Offce Depot. Markets that
H-P had locked up several years earlier (Lawton and Lublin 2007). Mr. Dell con-
tinued the takeover by removing or replacing a third of its executive board.
In May 2007, Dell announced a deal with Wal-Mart to sell two desktop models
directly to consumers. Bob Pearson, Dell spokesman, said our customers are ask-
ing us for additional ways to purchase our products and we plan on delivering on
a global levelTodays announcement with Wal-Mart represents our frst step.
Stay tuned. (Lawton 2007). Perhaps Dell will recognize that two desktops are not
going to change the world, and that many people prefer to buy laptops. But, if you
put Dell laptops side-by-side with those from H-P, Sony, Apple, and IBM, will
consumers reject them because of a lack of style and engineering?
Keys to Success
Much of Dells success has been credited to the built-to-order model. Nothing is
manufactured until someone buys it. This reduces inventory holding time to three
to fve days. Dell also applies this just-in-time approach to its marketing efforts.
Advertising dollars and resources are focused on the segments that are growing
the most. For example, the fourth quarter is always focused on the holiday gift
season for consumers. Quarters 2 and 3 are focused on the public sector when
schools and the government begin their buying cycles.
While Dell has always sold to the education market, in 1996 the sector became
more interesting, thanks to the federal governments E-Rates program that pro-
vided from $2.5 billion to $3 billion per year to subsidize schools technology
budgets. Apple had previously addressed this market by engaging in emotional
marketing, pitching warm and fuzzy computing to teachers in the hope that they
would lobby their schools. In contrast, Dell had realized that computers in the
schools had moved from freestanding desktop tools to nodes in an increasingly
complex web of technology that included wireless networks, handheld devices,
software, and servers. While Apple advertised personal creativity, Dell marketed
standardization and commoditization. Dell did not go to its education buyers and
say, Well sell you desktops and laptops. They offered solutions including edu-
cation software, packages, and IT services. They addressed storage and the lever-
age of the existing technology. This approach proved to be more compelling for
people with limited budgets and no resources or staff.
Perhaps Dells most important weapon was its huge customer database, the re-
sult of Dells sales model. This provides a direct relationship with the customer.
Since it sells directly, Dell can get immediate feedback from customers before,
during, and after the sales process. Dell courts such feedback, maintaining a Plati-
num Council of buyers to offer their insight at the executive level, and also invit-
ing customers to work with product development teams in a direct feedback loop.
Due to tight supply chain management, Dells made-to-order model is also a
fnancial gold mine. Dell orders and receives components as they are needed to
build a machine. The computer is shipped within a few days, and the customer is
charged. But Dell does not have to pay the suppliers until 30 days later. So Dell
earns interest on the revenue for a couple of weeks before it pays suppliers.
Database Applications to Better Service Customers
Valerie Hausladen knew that she could get Dells customer database to work hard-
er. I used to be more of a traditional marketing person, she says. But I learned
that unless I could measure and show ROI on marketing, my budget kept getting
cut. I realized at Dell, you have to be able to show results (Kuchinskas 2003).
Hausladen set up a pilot program to test whether better customer relationships
could be established in the interactive channel. Smaller Dell accounts were initial-
ly targeted in three areas: health care, K-12 schools, and colleges and universities.
Dell divides its accounts into the three categories of acquisition, development,
or retention. Development accounts are those that the company thinks could grow,
or that are not spending all their budgets on Dell products. The tactical goal of the
campaign was to get the name of one buying decision maker for each account.
The strategic goal was to become more actionable on how the company served
customers with right information at the right time. The implementation goal was
to send laptop information to customers who were interested in laptops and server
info to server buyers.
Dells database group went to work. They analyzed Dells customer list to es-
tablish benchmarks. To fnd out if Dell knew a decision maker at an organization,
they analyzed job titles, asked salespeople, and even called the customer to ask.
They also began outreach via e-mail and direct mail, renting mailing lists and
building and refning the Dell customer database through opt-in methods.
To implement the program, Web portals were built for each of the customer
verticals, offering tools and information to aid them in their technology decisions.
This was aided by the fact that the three Knowledge Centersfor K-12, institu-
tions of higher education, and government purchaserswere similar in their ap-
plication. The visual identity and content for the Web pages were crafted to serve
as a virtual account executive for each market.
The direct mail campaign focused on two goals. First was the desire to alert
customers to deals and sell more products. Second was the effort to improve infor-
mation in the customer database and turn names into relationships. When custom-
ers registered on the Knowledge Center sites to read content or request e-newslet-
ters, they also selected products they were interested in from a list. The next time
Dell communicated with them, it focused on the items they had checked. The Dell
team set up a control group to measure the effects of their marketing and defne
subsequent benchmarks. Dell began to see results from this database marketing
effort in 2002. The short-term gains included the maximization of tactical things
like response rates. This included the establishment of 5,000 new accounts in eight
months. The Web interface lets the marketers compare different campaigns, dif-
ferent media, and calculate the current ROI. Three years into the program, Dells
information-rich database contains at least one key decision maker for each ac-
count. The goal is to expand this list to the names of fve key decision makers for
each location in the database.
Dell addresses the budget-conscious public sector market by talking to them
about getting what they need, supporting them, and saving them money. In direct
mail, e-mails, and on the Web, it uses a straightforward style and packs its com-
munications with information. The postcards are not fancy. They include a couple
of generic hardware close-ups and text. They follow a simple formula: one close-
up of the product, another that shows the entire case, a photo of smiling custom-
ers, and lots of text, with prices prominently featured.
All this leads back to the Knowledge Center Web page, where Dell is building
a deep resource that offers quick links to shopping, customer service, and tech-
nical support backed by articles that can help Dells customers evaluate, plan,
and use technology more effectively. They are also integrated with more offers,
sweepstakes, and discounts. Every time a customer clicks within a newsletter,
navigates around a Knowledge Center, or responds to a postcard electronically or
via a call center, the choices are recorded in Dells customer database. Over time,
the Knowledge Center begins to appear differently from the one encountered by
another education decision maker. This system results in a more interesting inter-
action with Dell each time the customer comes to the Web page.
Dell has Rerouted its US Support Calls to the United States
In November 2003, Dell rerouted calls from its U.S.-based users to a U.S.-based
call center instead of to its facility in India. Some customers had complained
about the quality of the help they had received from the offshore center. Sup-
port operations for Dells Optiplex line of desktop PCs and its Latitude notebooks
were moved back to facilities in Texas, Idaho, and Tennessee.
Ned May, an analyst at the market research frm IDC, said he had heard criti-
cisms from corporate IT managers about the support they were getting from Dells
call center in Bangalore, India. Barry French, a Dell spokesman, said a portion
of the calls from U.S. PC users would still go to India, where support has been
handled since mid-2001. He indicated that Dell may shift all of the support back
offshore at some point. We are increasing the number of people we have provid-
ing support in India, he said. Dell has a number of call centers around the world,
and we will continue to optimize those to provide the best customer experience
(Brewin 2003).
In 2011, Dell and HP both began buying companies to expand offerings for
business servers. Dell paid $800 million for Compellent, which provides data tier-
ing. A network attached storage box contains multiple levels of drivessome fast
and expensive (fash drives), and some slower but large and cheap (SATA). The
software automatically moves lesser-used data to the cheaper drives (Mearian
2011).
Also following HP, in 2009 Dell purchased Perot Systemsone of the large
outsourcing vendorswhich happens to be located in Texas (Niccolai 2011). As
part of its new services unit, Dell is focusing on providing private cloud-based
computing to business users. Dell partnered with SAP and several other compa-
nies to be able to host SAP applications on its servers (Kanaracus 2011). With its
traditional emphasis on business sales, Dell has been building server farms to pro-
vide private cloud-computing to businesses that need high-speed facilities. Dell
also expanded its server offerings including blade servers with up to 96 processor
cores. Dells moves into enterprise servers and cloud-based hosting helped pro-
vide large increases to net income in 2011.
Questions
1. Is the pace of technological change slowing? Can a business wait longer now
to refresh computers?
2. How is information technology critical to Dells success?
3. What is the current status of Dells recycling program and is it successful?
4. Why is Dell expanding into consumer electronics and corporate networking
products?
5. Is there a limit to the number of computers Dell can sell through its current
methods? Should it reconsider retail sales?
Case: Gateway and Acer
Gateway was the third largest manufacturer of PCs in the United States, but it
struggled for several years. It tried selling computers, and plasma TVs, through
standalone stores, but struggled to make profts. Gateway gained market share in
the PC industry until the second half of 2000. Since that time, its rapid growth
has reversed. In 2002, Gateway estimated its market share to be 6 percent, but
the company still believed it had a strong presence in the U.S. consumer market.
In early 2001, management reviewed its operations and discontinued unproftable
revenue streams beginning in the second quarter of 2001 with roughly $200 mil-
lion per quarter. In the third quarter 2001, it exited its international operations. In
March 2003, management planned to save $400 million annually from workforce
cuts, cost programs, and a reduction in the number of its stores. In March 2007,
the former CFO and controller of Gateway were found guilty of manipulating
earnings and revenue statements. Gateway also acquired a new CEO in late 2005.
Sales in 2006 were just shy of $4 billion with a net income of about $9 million.
But the company has to grow profts considerably to overcome its half-billion
dollar loss in 2004. Slightly over half the sales value was for desktop computers.
Most of sales ($2.7 billion) were through retail outlets (2006 Annual Report).
In the beginning, Gateway focused on providing high-end personal computers
at relatively low prices. The company led the industry consolidation through the
1990s and dominated PC sales for a time. Based in South Dakota, the company
featured a cow motif as a means to create a brand identity. Eventually, headquar-
ters was moved to San Diego. In 1997, Gateway introduced an innovative exten-
sion of its traditional business model. Instead of relying on developing customers
solely through the Internet, Gateway opened 37 Gateway Country stores in which
customers could test-drive PCs. One of the goals was to expand sales of consumer
electronics goods such as printers, cameras, and eventually plasma televisions.
After several years of expansion, the company began closing certain stores. As
of December 31, 2002, Gateway had 272 retail stores in the United States, down
from 327 at the end of 2000. In the frst quarter of 2003, the company closed
another 80 stores. In 2004, Gateway acquired eMachines largely in exchange for
stock. Ted Waitt, founder and chairman, quickly replaced Gateway management
with eMachines managers. In April 2004, the new management shut down all of
the remaining Gateway retail stores (Zimmerman 2004).
In 1997, Gateway also launched its Internet service. In October 1999, the com-
pany announced a strategic alliance with America Online, calling for Gateway
to package AOL service into all its computers. AOL was to invest $800 million
in Gateway over two years, including $150 million in AOL stock. In December
2001, Gateway extinguished its convertible note to AOL through the issuance of
50,000 shares of Series C redeemable convertible preferred stock. This resulted in
an extraordinary gain of $4.3 million, net of tax.
Realizing the price competitiveness in technology, Gateway was among the
frst PC vendors to shift its focus to beyond the box revenues. These revenues
incorporate the sales of software and peripherals, Internet access, fnancing, and
warranty and training revenue. In 2002, these sales were $692 million, down from
$1.2 billion in 2001 and $1.9 billion in 2000. These sales have wide margins.
Several categories offer recurring revenue streams. Gateway reported a net loss of
$0.95 in 2002. By 2006, revenue from non-PC sales had dropped to about $650
million.
The merger with eMachines appeared to represent a refocusing on the value
end of the computer spectrum. Originally, eMachines succeeded by building stan-
dard computers in China and beating the cost of equivalent Hewlett-Packard ma-
chines by $150. As PCs become a commodity industry, people are less interested
in the raw technology. Instead, certain features become important, such as more
memory or larger disk drives. John Hui, the owner of eMachines, explained that
they created what we called a value formula that Wayne [Inouye] developed at
Best Buy. They attach value to every single component of a PC. They can at any
time tell you if you use, lets say, a DVD RW versus a CD ROM how much more
the consumer is willing to pay. Now that has nothing to do with the cost. Because
(its) the value that the consumer will attach to every single component. For ex-
ample, how much is the consumer willing to pay for a 512MB system versus a
256? They can attach value to it, because the consumer knows they can easily buy
the memory and stick it in themselves. But the memory prices go up and down
every day. So they have a certain perceived value. And when that perceived value
is high and the cost is low, thats when eMachines would put those components in
there. Inouye became CEO of Gateway after the merger (Zimmerman 2004). By
August 2007, Gateway had failed and the name was purchased by Acer (Lemon
and Nystedt 2007). Acer, a Taiwan-based manufacturer quickly became one of
the leading PC manufacturers. A large part of the growth came through selling
low-priced netbookssmall laptops designed for basic Web browsing and e-mail
tasks.
By 2011, Acer was being squeezed on the low end by tablets, particularly the
Apple iPad. People chose to purchase the small, lightweight tablets instead of
buying the small netbooks that had increased Acers growth. Later in 2011, Acer
decided to re-enter the server market. Servers often have higher proft margins,
but reliability is critical and manufacturers build in duplication in case compo-
nents die (Thibodeau 2011). Acer plans to use third-party makers to put the serv-
ers together. Todd Mottershead, senior manager for servers and storage noted that
I want to make sure that we can build very, very quickly and deliver much faster
than everyone else. One interesting, but unanswered, question is whether Acer
will go with a proprietary server format. In late 2010, Facebook began pressuring
vendors to develop a generic server. The company created design specifcations
down to the cabinet and motherboard layouts. Up to this point, server manufac-
turers built incompatible systems. Although they used off-the-shelf components
for drives, RAM, and processors, it generally was not possible to even transfer a
motherboard from one case to another. Standardization will make it easier to sub-
stitute parts; which ultimately increases competition and drives down prices. But
it is not clear yet if vendors will be willing to build to the standards.
Questions
1. Why did the Gateway stores fail?
2. Can smaller vendors like Acer and Toshiba survive?
3. Are tablets PCs and will they replace laptops?
Case: Sun
Sun was the No. 3 manufacturer in server technology, with 13.5 percent of the
market, and its market share and revenue were dropping in the early 2000s. Sun
focused on the UNIX market where it was the No. 1 vendor. Founded in 1982, Sun
Microsystems invented the workstation. It relied on the concept that the network
is the computer. Sun was a leading supplier of networked computing products, in-
cluding workstations, servers and storage products that primarily used Suns own
Scaleable Processor Architecture (SPARC) microprocessors and its Solaris soft-
ware. Computer systems accounted for 55 percent of net revenues in 2003, net-
work storage products 14 percent, support services 25 percent, and professional
services 7 percent.
Suns workstations were primarily used for engineering applications (CAD/
CAM), desktop publishing, software development, and other applications that
need a moderate amount of computing power and high-quality graphics capabili-
ties. In terms of computing power, workstations fall between PCs and mid-range
computers. UNIX has been the most common operating system for workstations,
but Microsofts Windows has posed a formidable challenge. Suns workstations
ranged from low-cost UltraSPARC-based workstations to high-end, multiproces-
sor color graphics systems. Sun servers can be used for fle sharing, letting us-
ers access data distributed across multiple storage devices and networks, or as
computer resources, to distribute computer-intensive applications across multiple
processors.
In June 2000, Sun shifted its focus to include storage, launching its StorEdge
T3 line, as well as software and services to enhance its storage solution offerings.
In 2002, Sun introduced Solaris 9, the latest upgrade of its popular UNIX operat-
ing system. Its features include identity management and enhanced security and
manageability.
Sun has concentrated on software development as well. It invented the Java
object-oriented programming language. Java has attracted signifcant interest in
the software development industry because of its portability; software created in
Java can run on any type of system. As a result, it is a popular tool for designing
software for distribution over the Internet.
Sales at Sun were weak through the early 2000s as customers realized that
Intel-based personal computers could perform as effectively as workstations and
midrange servers, at a substantially lower upfront cost. In 2004, Sun received $1.6
billion in a settlement with Microsoft over the use of Java. Yet Sun lost $0.75
billion in just one quarter of that year and was forced to lay off 9 percent of its
workforce. In 2004, Sun began selling Intel- and AMD-based workstations that
could run the Windows and Linux operating systems. But Sun could not compete
in the low-end and midrange markets. Suns 2006 Annual Report shows increas-
ing sales revenue (to $13 billion in 2006), but it sustained net losses from at least
2002-2006. A huge portion (15 percent) of Suns sales were to GE. Demand for
high-end data center servers declined in 2006 in favor of smaller systemswhere
Sun faces greater competition.
For years, Sun had relied on proprietary processors. By 2006, most production
had shifted to high-end processors available from Intel and AMD. The company
also expanded into the data storage market, buying StorageTek in 2005, in an at-
tempt to provide a market for storage applications. In 2006, revenue from data
management products and support services exceeded revenue from computer sys-
tems products. On the high-end, Sun lost out to IBM and Cray Research in a bid
to provide supercomputers for the U.S. government. According to Suns 2006 An-
nual Report, most HRM and IT functions for the company have been outsourced
and are provided by other companies.
In 2009, Oracle, the database vendor, paid $7.4 billion to acquire Sun (Clark
and Worthen 2009). The general presumption was that Oracle wanted to gain con-
trol over Java, which is an important element in developing systems in the Oracle
database. Sun still produces some servers, but the workstation market had largely
disappeared by 2011. Sun still markets the Sun Ray system which is one of the
original thin-client approaches. All of the computing work is handled on a server
and users interact with basic graphics terminals and a keyboard. The primary ob-
jective is to keep the software and data in a central location to make it easier to
install and maintain. The problem is that the client terminals cost almost as much
money as a complete PC and the servers are expensive. The original goal was to
reduce management costs by centralizing the software and data, but with network-
based controls available today, there is little reason to spend the extra amount of
money on the server. In any case, Sun was struggling for years to sell servers. The
business continued to decline due to uncertainty after the purchase by Oracle. But,
in the frst quarter of 2011, Sun saw an increase in the sales of servers hitting $773
million in the frst quarter (Niccolai 2011).
Questions
1. Can Sun survive? How will it compete against Linux and generic PCs?
2. Since much of the world has accepted Java as a Web-based programming
language, can Sun make any money from it?
Additional Reading
Boslet, Mark, Sun Microsystems Posts Big Loss in 3Q As Sales Slump, The
Wall Street Journal, April 16, 2004.
Brewin, Bob, User complaints push Dell to return PC support to U.S.,
Computerworld, Vol 37(48), December 1, 2003, p. 6.
Clark, Don and Ben Worthen, Oracle Snatches Sun, Foiling IBM, The Wall
Street Journal, April 21, 2009.
Darlin, Damon, Dells Share of PC Market Slumped Further in First Quarter,
International Herald Tribune, April 19, 2007.
Guth, Robert A., Microsoft Settles Another in String of Antitrust Suits, The
Wall Street Journal, April 20, 2004, p. B5.
Kanaracus, Chris, SAP, Dell Partner on In-Memory and The Cloud,
Kuchinskas, Susan, Data-based Dell, Adweek Magazines Technology,
Marketing, Vol 23(6), September, 2003, p. 20-23.
Lawton, Christopher and Joann S. Lublin, Dells Founder Returns as CEO as
Rollins Quits, The Wall Street Journal, February 1, 2007.
Lawton, Christopher, Dell to Sell PCs at Wal-Mart, The Wall Street
Journal, May 24, 2007.
Lemon, Sumner and Dan Nystedt, Acer to Acquire Gateway in $710 Million
Deal, PC World, August 27, 2007.
Mearian, Lucas, Dell to Add Compellents Tiering Technology to its Storage,
Computerworld, February 22, 2011.
Niccolai, James, Dell Services Chief to Leave Company, Computerworld,
January 13, 2011.
Niccolai, James, Oracle Returns Growth to Sun Server Business: IDC,
Plamondon, Scott, PC Recycling Made Easier, PC World, December 29,
2003.
Skillings, Jonathan, Dell To Get Green With PC Recycling, News.Com, May
16, 2002.
Thibodeau, Patrick, Acer Returning to U.S. Server Market, Computerworld,
February 21, 2011.
Whitney, Lance, Acers PC Market Numbers Add Up, CNet, March 10, 2010.
http://news.cnet.com/8301-1001_3-10466669-92.html
ZD Net, Dell Could be Losing its Shine, April 19, 2006.
Zimmerman, Michael R., Why the eMachines Model Is Paying Off, eWeek,
April 18, 2004.
adopted by rivals?
sectors?
132
What is the value of a single computer?
Why are computer networks so important in todays businesses?
What components do you need to install to create a network?
How can multiple users share a single network?
How is it possible that you can connect your computer to a network at the offce, at
home, or while on the road, even overseas?
What is the Internet, how is it controlled, and how does it work?
Are personal computers necessary anymore?
What problems are you likely to encounter if you need to connect to a supplier in a
different country?
Chapter Outline
Networks and
Telecommunications
3
Chapter
Introduction, 133
Network Functions, 134
Sharing Data, 135
Sharing Hardware, 142
Sharing Software, 144
Components of a Network, 145
Computers, 145
Transmission Media, 147
Connection Devices, 157
Network Structure, 159
Network Packets, 159
Shared Connections, 160
Managing Shared Networks, 163
Standards, 166
The Need for Standards, 166
A Changing Environment, 167
Internet TCP/IP Reference Model, 167
The Internet, 171
How the Internet Works, 172
Video and Heavy Use Complications,
174
Telephones and The Internet, 175
Internet Addresses, 177
Domain Names, 179
Internet Committees, 181
Internet 2, 181
Wireless Networks and Mobile
Commerce, 183
Global Telecommunications, 186
Technical Problems, 186
Legal and Political Complications, 187
Cultural Issues, 188
Summary, 188
Key Words , 190
Exercises, 191
Cases: Wholesale Suppliers, 196
133 Chapter 3: Networks and Telecommunications
Owens & Minor
Introduction
What is the value of a single computer? This question almost sounds
like the old Buddhist Zen koan: What is the sound of one hand clapping? You do
not need to debate the complexities of Zen to understand that the value of com-
puters is multiplied by connecting them and sharing information. Today, many
applications exist only on the networkthey are built to use servers to store and
process data and they are accessed across the Web from any location in the world.
And communication ranging from text, to voice calls, to video conferencing has
largely moved to Internet-based formats.
As shown in Figure 3.1, networks are useful on two levels. Internal compa-
ny networks are used for teamwork and sharing data; the Internet connects you
to external organizations, including suppliers, customers, and service providers.
Actually, you use many other networks in your daily life, including cell phones,
television, and even electricity. Today, almost everything you do in business re-
lies on computer data networks. You use them for communication, sharing data,
team work and collaboration, and simpler things such as backing up your data and
printing.
Even though you will probably not be responsible for creating or maintaining a
network, knowledge of the underlying technologies makes it easier to understand
some of the problems that can arise, and helps you see how the technology is
likely to change in the future so you can be prepared. Network technologies have
experienced some of the most rapid changes in the past few years, and new tech-
nologies are being deployed now. You need to plan ahead so you can create new
organizations and new methods of handling work.
The Internet has changed many aspects of daily life, and has the capability of
fostering even more changes. To understand these opportunities and recognize the
challenges that need to be overcome, you need to learn a little about how the In-
ternet is designed and how it works. The Internet is an interesting organization
because it is governed purely by committees and individuals and works because
of standards. These organizations face diffcult questions and the answers will
dramatically infuence your life and the role of businesses in the years to come. If
nothing else, as an informed citizen, you need to be aware of these discussions.
Many times the decisions hinge on social, economic, and political issues instead
of purely technical ideas.
How do suppliers connect to customers? Owens & Minor sells thousands of prod-
ucts to hospitals. It is one of three large companies in the market, competing with
Cardinal Health (which owns the system built by American Hospital Supply). Like
other wholesalers, Owens & Minor exists in a highly competitive environment with
low margins. Whether you call it logistics or supply chain management, buyers con-
tinually search for ways to reduce their purchasing costs. Companies like Owens &
Minor use information technology to reduce inventories throughout the purchasing
system. By building links with customers and suppliers, Owens & Minor can im-
mediately react to changes in demand and plan the most effcient method to deliver
supplies.
The global economy makes life more interestingparticularly in terms of net-
works. Sharing data, or even using cell phones, across international borders raises
many complications. Because technology continually changes, nations implement
the technology at different points in time. Compatibility across borders is an even
greater challenge when nations have differing goals and income levels. As man-
agers, you need to be aware of some of the issues, both to communicate and to
travel.
The objective of a network is to connect computers transparently, so that the
users do not know the network exists. The network provides access to data on cen-
tral computers, departmental computers, and sometimes other workers personal
computers. Networks can create shared access to fax machines, modems, printers,
scanners, and other specialized hardware. To the user, it does not matter where
these devices are located. The network makes them work the same as if they were
on any desktop. The Internet expands these capabilities across the world. Wireless
makes the services available to you wherever you travel in major cities.
Network Functions
Why are computer networks so important in todays business-
es? Most companies did not seriously begin installing networks until the early
1990s. The Internet expanded into the commercial world in the mid to late 1990s.
Trends
In many ways, the communication systems available today began with
the telephone. Originally, sounds were converted into analog electrical sig-
nals, and calls were connected through physical switches. As computer tech-
nology evolved, it became possible to convert everything into digital data
and the computerized switches simply transfer digital packets to a desired
destination.
A key aspect of communication systems is that some connection medium
is required to transmit data. Most homes have two major connections: tele-
phone and TV cable. The phone and cable companies have been expanding
their services by providing better and faster connections to houses. Thanks
to the long-distance phone companies, hundreds of thousands of miles of
national and international fber optic cable connect cities and nations around
the world. These backbone connections carry almost all of the Internet data,
voice, and video connections.
But, the biggest trend in the past few years has been the switch to cellu-
lar phones including smart phones used as computers to access the Internet.
People are increasingly giving up wired connections in favor of mobile con-
nections that operate over radio waves. As nations move to provide more
bandwidth for cellular service, data transmission capabilities for mobile
phones is approaching the speeds available for fxed-line connections.
As connections and usage have grown, so have the applications and data
available. Consequently, it is increasingly important to be able to connect
to the Internet from anywhere at any time. Which leads to an increasing cy-
cle: As connection capabilities improve more applications and data become
available, which means connection capabilities need to improve even fur-
ther, which leads to new applications, and so on.
Perhaps the better question is: Why did it take so long? The answer: cost. Think
about that time frame for a minute. The digital world as you know it is only about
15 years old, and it has already had a huge impact on business and society. But,
the world is still only at the beginning of the network evolution. New opportu-
nities in both communications and applications will continue to arise. And you
will face new questions about choosing technologies and applications. To under-
stand the value of networks, you need to look at the business applications and new
procedures that have been created with networks. Then, you can start thinking
about the future networks and new applicationsparticularly wireless and mobile
applications.
Sharing Data
Sharing data is one of the most obvious uses of networks and it can make pro-
found changes in the way an organization works. Managers can see customer and
marketing data immediately as it is collected. Employees in one department can
easily share data with other departments. A network facilitates the use of teams. In
particular, it enables informal teams to spring up throughout the company to solve
problems as they arise. Instead of waiting for a higher-level manager to appoint a
team, employees can use the network to ask questions, notify others involved, and
fnd in-house experts. A local area network (LAN) is commonly used to connect
computers and share data within a company. A LAN uses the same technologies as
any other type of network, but is recognized separately because all of the compo-
Figure 3.1
Networks. Managing an organization requires communication and teamwork.
Networks facilitate teamwork and create opportunities for new organizational
structures. The Internet makes it easy to connect to customers and suppliers and
access new services and Web-based applications.
Teamwork
Communication
Scheduling
Sharing
Internet
Suppliers
Customers
Banks
Internal
External
Services
Applications
Research
Hosting
nents of the network are owned by your company or organization (or household).
Since you own everything, you get to make all of the decisions, and you do not
have to pay anyone else to transmit the data.
Transactions
One of the most important reasons for connecting computers is the ability to share
data. Point of sale (POS) systems were some of the frst networks installed in
businesses.
Consider a retail store with fve checkout registers. Each register is actually a
computer. If these computers are not connected, it is diffcult to compute the daily
sales for the store. At the end of the day, someone would have to manually collect
the data from each computer and enter it into another computer. Also, imagine
what would happen if a customer asked a clerk to determine whether a product
was sold out. The clerk would have to check with each of the other clerks or per-
haps call the storeroom.
As shown in Figure 3.2, with e-commerce transactions can take place across
a wider network, with the Internet as the network and customer browsers as the
client computers. The product data and sales transactions are stored in the central
database connected to the Internet. Using a central database provides inventory
data to customers. When a customer asks whether an item is in stock, the Web
site can provide the answer. Managers can get daily sales fgures from any loca-
tion with a Web browser and an Internet connection. Payments and bills can also
be handled directly online. The same diagram can be used for traditional in-house
Figure 3.2
Network for transaction processing. Networks are often used to collect data in a
central database. From there, the data can be queried and analyzed by managers.
E-commerce sales represent transactions across the Internet.
Database Management
System and Web Server
Or Point-of-Sale System
Internet
POS systems by replacing the Internet with a simple LAN. The main point is that
a central server holds all of the data and handles the actual transactions.
Decisions and Collaboration
Many types of data need to be shared in a company. But beyond simple data shar-
ing, people often work on teams. Teamwork entails developing reports together. It
requires searching for data, editing documents, and sharing comments. A portion
of a network for making decisions and sharing work with team members is illus-
trated in Figure 3.3. Centralized data and shared documents ensure that everyone
has the same version of the data and the same fnal documents.
In any job, it is rare for one person to work alone. Most businesses are arranged
as teams. Within the teams, individual people are given specifc assignments, and
each team member contributes to the fnal product. For instance, the marketing
department might have to prepare a sales forecast for the next six months. Each
person could work on a specifc sales region or product category. These individual
pieces would then be merged into a single document. If the computers are net-
worked, the managers computer can pull the individual pieces from each of the
individual computers. Also, each team member can be given access to the others
work to ensure that the reports use the same data and assumptions.
Groupware includes software that enables several people to work on the same
document. Each individual computer has access to the master document. When
one person makes a change to the document, the change is highlighted for every-
one to read and approve. With existing international networks, each person might
be located in a different country. E-mail systems can work as simple groupware
Team Document
File Server and Database
Report and
Comments
Figure 3.3
Network for decisions and collaboration. The fle server holds basic data and
software tools. Managers retrieve data and create reports. The reports can be shared
with other managers. With collaborative software, revisions are automatically tracked
and combined to form the fnal document.
tools by routing copies of fles to everyone on the team. More sophisticated tools
are included in Web sites built by systems like Microsofts SharePoint. These
tools are discussed in greater detail in Chapter 8.
Messages
Most people are familiar with electronic mail, or e-mail. With e-mail, you can
send a message to any person with a mail account. Many people have come to
prefer e-mail contacts to phone calls. As shown in Figure 3.4, e-mail messages
are asynchronous since the sender and recipient do not have to participate at the
same time. A mail server holds the message until the user logs in and retrieves the
e-mail. Moreover, users can create mailing lists and send a message to many peo-
ple at one time. Messages can be stored, retrieved, and searched for later usage.
In most systems, the computer can automatically notify you when the recipient
reads your message. You never have to worry about whether your message was
received. Voice mail systems, which resemble answering machines, have some of
these same advantages. However, e-mail takes less space to store. More impor-
tant, e-mail can be scanned electronically to search for key topics.
E-mail has some drawbacks. For one, some people do not check the computer
often enough to keep up with their mail. Another problem is that in 1991, the U.S.
courts ruled that public transmission systems such as e-mail are not subject to the
same legal protections for privacy as paper mail and phone calls. Unless the laws
are rewritten, therefore, it is legal for employers (and almost anyone else) to read
everyones e-mail messages. Of course, the fact that it is legal does not make it
ethical.
Examine the fgure for a few seconds and another aspect of e-mail becomes
fairly obvious. How can you delete e-mail (once it has been sent)? The answer
is that it probably cannot be done. Even though some systems appear to give you
Figure 3.4
Networks for communication. With e-mail, a server holds the mail until the recipient
logs on and can receive the message. Because several servers might be involved and
each makes backup copies, e-mail messages are hard to delete.
Internet
1. User creates e-mail message.
2. Message
transferred to
account on
server.
3. Transferred
via the Internet
to the destination
account.
4. Message received
when user checks
e-mail.
the option to delete or retract e-mail that has been sent, it generally will not work.
Copies of the e-mail can exist in several locationsthe original computer, the
senders e-mail server, the recipients server, and the recipients computer.
E-mail, and the fact that it is diffcult to delete, has become an important as-
pect in legal casesboth private and those involving government offcials. In
court cases, many lawyers automatically ask for copies of all relevant e-mail in
discovery motions. Consequently, many companies build e-discovery databases
that contain copies of all e-mail messages sent through their servers. These sys-
tems have keyword search facilities so administrators can retrieve specifc mes-
sages when required. The systems also have rules-based systems to handle reten-
tion policies. For example, company policies might specify that routine messages
should be deleted after six months. The system can enforce retention and deletion
policies automatically. Many of these systems also integrate voice mail messages
with e-mail messages, so they are subject to the same retention, deletion, and dis-
covery rules. The main point to remember is that any digital data might be stored
for an unlimited time. So, although it is easy to think of electronic text and mes-
sages as ephemeral and feeting comments, you should always remember that the
potential exists for any message to be retrieved later.
You need to pay attention to two key aspects to any communication method:
(1) Whether it is synchronous or asynchronous, and (2) The level of interruption it
creates and the preferences of the people you work with. For example, a personal
visit and phone call are synchronous, and the other person has to stop working
on a task to talk with you. IM or texting is also synchronous, but users often feel
more comfortable delaying these communications. E-mail is asynchronous, so
each person works on his or her own schedule, and deals with mail during breaks.
As the message sender, you have to choose the appropriate method of communi-
cation for each message and person. You also have to learn to allocate your own
communication time, to leave suffcient time to fnish your own work.
Reality Bytes: Worldwide Broadband Rankings
Akamai is an Internet company that provides services to many other companies.
It specializes in distributing data around the worldputting it on servers closer to
where it will be needed. This positioning gives the frm the ability to measure data
fows and speeds around the world. An interesting statistic is the connection speeds
of the people in each country. In 2010, the nations with the highest average connec-
tion speeds and their average speed in Mbps, were: 1. South Korea (17), 2. Hong
Kong (8.6), 3. Japan (8.0), 4. Romania (6.8), 5. The Netherlands (6.5). Note that In
South Korea, over 75 percent of the population connected to the Internet at speeds
over 5 Mbps. The U.S. ranked 16 on the list of average speeds at 4.6 Mbps, and only
30 percent of the population had access to speeds over 5 Mbps. It is easier to build
high-speed networks in nations with small physical space and large populations such
as South Korea and Japan. Similarly, population areas in the U.S. such as the North-
east have more people with faster connections. But, ultimately, the speed available to
the average person begins to matteras companies rollout Web-based applications
with sound and video.
Adapted from Karl Bode, Akamai: U.S. 33rd Fastest Broadband Country, www.dsl-
reports.com, July 9, 2009. Updated data: http://www.akamai.com/stateoftheinternet
Web Site Basics
Web sites have evolved considerably, with several well-known sites providing
services and applications to address many new tasks. However, many sites still
exist in the early formatproviding information and data. As a step up, some sites
provide limited interaction, such as sales, feedback, discussion, and fle sharing.
At a higher level, a few sites provide greater interactivityincluding games and
online applications. The frst decade of 2000 also saw the expansion of social net-
working sitessuch as Facebook and Twitter. Social networking sites facilitate
communications among groups of users. These types of sites and other capabili-
ties are covered in more detail in Chapter 7 on e-business.
A key element of communication that is emphasized with the Internet and the
Web is that communication only exists through the adoption of standards. Stan-
dards are simply agreements among the major users to handle tasks and data in a
specifc way. In terms of the Internet, standards exist for connecting networks and
transmitting packets through routers. Common Web standards include hypertext
markup language (HTML) which is the standard format for telling browsers
how to display a page of text and images. These same standards can be used for
other types of networks, including intranets and extranets. Intranet sites employ
Internet technologies but use security methods to restrict access to internal users.
For example, employees probably use an intranet to retrieve and update personal
data in the human resources database. Similarly, extranets are sites specifcally
designed for companies that partner with your organization. For example, your
suppliers can log into a special section to check on your production schedule, pick
up technical specifcations, or bid on jobs. Intranets and extranets use the same
technologies as the common Internet, but are limited to special groups of users.
Several common Web tools can be used to support communication. Initial Web
sites were designed primarily for one-way communication. They are still used
for posting centralized data that is needed by many other people. However, in-
teractive tools have also evolved to support feedback and communication within
groups. Tools such as instant messaging (IM) can be used to send short messages
to friends or coworkers. The messages are delivered instantly and can provide a
way to ask questions or share ideas with team members who are online. Of course,
cell phone texting services operate the same way, but the interaction requires a
cell phone, or preferably a smart phone with a decent keyboard. The computerized
IM tools can also be used to see when someone is online. If you are working late
a night, you can quickly see which of the other team members are also online, so
you know that you can ask a question without having to call everyone just to see
if they are available.
A Web log or blog is an interesting variation of a Web site. The tools were cre-
ated to make it easier to create and post content online and facilitate comments
and feedback. Several sites provide public blogs enabling you to write comments
on any topic. The software is also readily available for installation on internal
company servers so managers can create blogs on topics that can be targeted to
employees or to external partners. For instance, a blog could be used by managers
to keep a log of the daily issues, problems, and solutions. Other managers in the
company could skim the logs to see if problems arising in one area might cause
problems in their own sections. Similarly, if a manager encounters a problem, he
or she could search the blogs for similar problems, ideas on how to approach it, or
even solutions that worked in the past. In essence, the blogs become a knowledge
base that is accessible to other managers. Of course, the sites would be secured
so that only managers within the organization could read them. Some people are
so dependent on blogs that they want to be notifed whenever the information
changes. Really simple syndication (RSS) feeds can be established so that your
browser or reader automatically checks for new information and displays it auto-
matically. It is even possible to connect your reader to your desktop so that any of
your subscriptions are automatically displayed on the desktop when information
changes. However, this approach is likely to be a major distraction to your work.
With the introduction of the Apple iPod, the concepts of blogs and RSS feeds
was extended into audio notes that can be downloaded to the iPod and played
later. These podcasts can be useful if you have time to listen to voice notes, but in
business, it is faster to read and search written notes.
Wikipedia, the online user-supported encyclopedia has led to a new form of
Web-based communication. You can use the same software to create in-house wi-
kis where workers on a team can write their own descriptions, edit work by other
people, and add comments. The entire system contains a search engine, so other
employees can search the system for problem descriptions, work documents, and
approaches or solutions used in prior cases. It is a powerful and inexpensive sys-
tem when you want to record case-based data.
Calendars and Scheduling
Managers spend a great deal of time in meetings. Yet sometimes the greatest chal-
lenge with meetings is fnding a time when everyone can get together. Several
software packages use computer networks to solve this problem. As shown in Fig-
ure 3.5, managers enter planned meeting times and scheduled events into their
personal electronic calendar fle, where each event is assigned a priority number.
For example, a time allotted for a haircut would be given a low priority; a meet-
ing with a supervisor would receive a higher rating. If the CEO wants to set up a
meeting, the CEO tells the computer which people are to be included, sets a pri-
Figure 3.5
Sharing calendars. With the appropriate software, you can open your calendar to
other members of your team. A team member can check the calendar and have the
software fnd a common open time for a meeting.
8:00 Mgt meeting
8:30 (open)
9:00 Staff meeting
9:30 Staff meeting
10:00 new meeting
ority level, and gives an approximate time. The computer then uses the network
to check everyone elses schedule. When it fnds an open time (overriding lower
priority events if needed), it enters the meeting into each persons calendar. These
systems can be useful when the calendar on your cell phone is connected to the
shared calendar.
Sharing Hardware
Networks also enable workers to share expensive hardware. For example, net-
works are used to provide users access to special output devices, such as high-
speed printers, plotters, or color printers. Networks can be used to give people
access to special computers, such as when an engineer needs to use a high-speed
supercomputer. Figure 3.6 shows some of the hardware devices commonly shared
using a LAN: printers, tape drive backups, server processing, and external access
to the Internet.
Printers
A common use of networks is to give users access to high-speed, high-quality
printers. Even if each computer has a small personal printer attached, some jobs
need to be handled by a high-speed laser printer. Many modern copiers now func-
tion as network printers and can collate and staple large quantities of documents.
Similarly, even at $1,000 each it would be expensive to buy color laser printers for
everyone who might need one, yet it might be reasonable to buy one for a depart-
ment to share. With a network, users can choose from among two or three differ-
ent printers by selecting them on the screen.
Another advantage is that if one printer breaks down, users can send their jobs
to another printer on the network. Think about what happens if there is no network
and your printer breaks down. You have to copy the fle to a USB drive and in-
terrupt someone elses work to borrow their computer to send the fle to another
printer. What happens if you are using a special software package that no one else
has on his or her computer? You will probably have to physically move a printer
from another computer desk to yours, connect the hardware, print your document,
Figure 3.6
Networks for sharing hardware. The workstations use the server to perform backups.
Files are picked up by the server and transferred to tape. The LAN administrator can
reload a tape and restore fles as needed. Networks are often used to share printers
and storage devices. Networks can be used to share access to supercomputerseven
if they are in a different city or different country.
Corporate or
external computer access
Server
Shared Printer
Workstations
tape drive
(backup)
and return the printer. When you are on a network, you simply select a different
printer from a list displayed on your computer and go pick up the output.
Several e-commerce printing companies enable you to send your large print
jobs over the Internet and have the boxes of papers shipped to you. All of the
setup, pricing, and payment can be handled over the Internet.
Storage Devices
The arguments used for network printer sharing can be applied to sharing storage
devices. If you have huge data fles that you want to share across the organization,
it is best to put them in a central location and provide network access to everyone
who needs them. The central location makes it easy to upgrade the drives, provide
suffcient capacity, and control the access rights and monitor security. A special-
ized storage area network (SAN) is often used to provide vast amounts of fex-
ible storage. The disk drives are separated from the computers and connected by
a high-speed network using fbre channel or high-speed LAN connections. Physi-
cally separating the secondary storage from the computer box makes it easy to
expand the capacity, provide redundancy, and move the drives to safer locations.
Because of the high transfer speeds across the SAN, the drives appear as simple
local devices to the computer, so no software changes are needed.
Backup
Another important reason for sharing data over computer networks is that most
people are not very good at maintaining backup copies of their workespecial-
Reality Bytes: Network Switches for Servers hit 10Gbps
As the amount of data on Internet and corporate servers continues to increase, net-
work vendors are pushing new technologies to improve the transfer of vast amounts
of data. In 2010, 10 Gigabit Ethernet is pushing into data center network design.
Most data centers are built as vertical racks of servers, where each rack has a net-
work switch at the top. In the past, these top-rack switches were connected to a mid-
dle-layer switch that then connected to the primary Internet backbones. But, hard-
ware servers are increasingly used to host multiple virtual serversand each virtual
server pushes data across the network. Replacing the top-rack switches with 10 Gbps
speeds and eliminating the middle layer speeds the data to the Internet connection.
Robin Layland, an advisor to companies and vendors, notes that Over the next few
years, the old switching equipment needs to be replaced with faster and more fexible
switches. This time, speed needs to be coupled with lower latency, abandoning span-
ning tree and support for the new storage protocols. IEEE standards are also being
ratifed for new switches to handle 40G and 100G Ethernet in the future. The speed
increases are important in a cloud-based world where Layland notes that applica-
tions require less than 10 microseconds of latency to function properly, compared
to existing 50-100 microseconds. Jayshree Ullal, CEO of Arista Networks, a com-
pany that makes some of the new high-speed switches, notes that most of the [exist-
ing] switches are 10:1, 50:1oversubscribed. Ultimately, companies save money by
reducing the number of network switchesboth in the cost of the equipment, main-
tenance, power, and air conditioning.
Adapted from Jim Duffy, 10G Ethernet Shakes Net Design to the Core, Network-
world, September 14, 2009.
ly on personal computers. If each computer is attached to a network, automatic
back up can be confgured two ways for individual personal computers. Individual
workers can save their data fles to a central fle server. The network manager then
makes daily (or hourly) backups of the data on the central server. A few compa-
nies even provide this service over the Internet. For a monthly fee, you can trans-
fer your fles to their server, giving you a backup copyplus they keep backup
tapes for the server. The second approach is to have backup software running on
the server query each users computer to obtain changed fles and save them au-
tomatically. This approach requires that the client computers remain connected to
the server and remain running, so it does not work well with laptops.
Special Processors and Grid Computing
Special computers that are relatively expensive can be attached to a network to
save costs and make it easier for users to access these machines. Parallel-process-
ing computers and other supercomputers can perform calculations thousands of
times faster than ordinary computers, but they are expensive. Consider a small
engineering company. For the most part, the engineers use their workstations to
develop their designs. They occasionally need to run simulations or produce de-
tailed graphics images. Both of these tasks could take many hours to perform on
individual computers. With a network, each workstation can be connected to a
supercomputer. When an engineer needs to perform high-speed calculations, the
job is sent directly to the supercomputer. The results are returned via the network
to the workstation so they can be added to the fnal report. More likely, a univer-
sity could own the supercomputer, and the frm would lease time to run each job.
If the network is designed properly, it makes no difference where the machine is
located.
Grid computing extends the concept of parallel processinginstead of having
multiple processors in one computer, you simply attach multiple computers to-
gether across a network. Remember that personal computers are relatively fast and
inexpensive. When you can buy a new computer for $500, it becomes possible to
build a huge amount of computing power by purchasing hundreds or thousands of
computers and spreading the work across all of them. With special software, the
job is split into multiple pieces and assigned to spare time on each computer. SETI
@home is probably the best-known example of this type of computing. However,
the same tools can be used within a single companytaking advantage of ma-
chines running overnight. Just keep in mind that running computers 24-hours a
day substantially increases the power consumption.
Sharing Software
Networks have been used at different times to share software. When disk space
was expensive, it was cheaper to put one copy of the software on a server and
download it to each computer as it was needed. Today, it is possible to run soft-
ware applications directly from a server, where client computers use Web brows-
ers to connect to the server and run the software. This centralization of software
and data leads to new ways to work and to manage companies. However, it raises
issues of reliability and security. Some of these are covered in more detail at the
end of this chapter. But the overall topics of management, centralization, and se-
curity arise in several other chapters.
Components of a Network
What components do you need to install to create a network?
As networks have become more important, the connection components are in-
creasingly built into the computers. However, you still face many decisions about
which technologies to use and how to solve problems, so all managers should un-
derstand the basic elements of a network shown in Figure 3.7. Each of these com-
ponents (computers, transmission medium, and connection devices) is discussed
in greater detail in this section.
Computers
Virtually any type of computer device can be connected to a network. The earliest
computer networks consisted of one computer with several terminals and printers
attached to it. These networks were fairly simple, because the one computer per-
formed all of the work. Substantially more problems are involved in connecting
several computers together. For starters, each computer needs to know that the
others exist and that they have a specifc location (address) on the network. The
computers need to know how to send and receive information from each other.
Just to connect to the network, they need LAN cards. Most desktop computers
have built-in LAN cards and connectors for wired networks. Most laptops also
have wireless network cards; some even have cards for cellular networks. For
now, most companies run a combination of both wired and wireless networks.
Most homes rely solely on wireless networks because no one wants to run cables
through the house.
Computers attached to networks tend to perform one of two functions: serv-
ers or clients. Servers are computers that store data to be used by other comput-
ers attached to the network. Clients are computers used by individual people.
Figure 3.7
Network components. Networks require a transmission medium and each device
connected must have a network card to make the connection. A switch connects the
devices in the LAN, and a router connects the LAN to the Internet. A frewall is often
used to flter out certain types of packets or messages so they do not transfer from the
Internet to the LAN.
LAN card
LAN card
LAN card
LAN card
Shared Printer
Server
Personal Computer
Personal Computer
Router
Internet
Firewall
Switch
Sometimes a computer is used as both a client and a fle server. Networks where
many of the machines operate as both clients and servers are called peer-to-peer
networks.
Servers
A wide range of servers exists todayfrom a simple PC to huge, expensive spe-
cialized computers. The main questions you face in choosing a server are the op-
erating system and the issue of scalability, and the two questions are intertwined.
Scalability is the ability to easily move up to greater performancewithout hav-
ing to rewrite all of your existing applications. Figure 3.8 shows two common
methods used to provide scalability: (1) a vendor-provided range of servers from
low-cost machines to handle small loads, to midrange, to high-capacity computers
that can handle millions of users simultaneously; and (2) integration technology
that enables the workload to be distributed across hundreds or thousands of small
servers, known as a server farm.
Both approaches have their benefts. The single server is easier to confgure and
administer. The server farm can be expanded easily and cheaply without disrupt-
ing the existing applications. The operating system software is crucial to making a
server farm work effciently. Several vendors sell enterprise versions of software
that assigns applications to the least-busy server and makes it easier to manage the
server farm.
Figure 3.8
Server scalability. Two common methods of providing easy expansion of
performance are (1) to purchase a faster performance server within the same product
family, and (2) to build a server farm where the workload is automatically distributed
across machines and new low-cost servers can be added at any time.
IBM Blue Gene/L
IBM PS700 Express
HP
Increasing
performance
within a
product family.
Server farms distribute
the workload. Add
more computers for
more power.
Rack mount
server farm.
IBM PS702 Express
(multiple blades)
https://asc.llnl.gov
Client Computers
The networked computers could be any type of machine. Because individual peo-
ple at their own desks typically use these computers, they are often called client
computers. These computers need to access the network and be able to send infor-
mation to at least one other computer. A network interface card (NIC) (or LAN
card) is installed in each computer. These cards are connected together by some
transmission medium (e.g., cable). In addition, these computers might have to be
confgured to connect to the network and set security parameters. Client comput-
ers today include laptops and PDAs that can be connected with wireless networks
to enable workers to maintain connections and share data while they move around
the building.
Many of the mobile devices have limited capabilities and essentially run brows-
ers, e-mail, and calendars. Hence, more of the processing is done on the server,
and the mobile device is only responsible for displaying data and basic user inter-
face tasks. This type of environment is considerably different from the situation
where all workers have desktop machines capable of handling large amounts of
the processing. The thin-client cell-phone applications rely on the server, while
the desktop applications simply use the server for basic data sharing. Develop-
ing applications for this new environment requires some major changes in de-
sign, programming, and securityissues that will be explained in detail in other
chapters.
Transmission Media
All communication requires a transmission medium. As illustrated in Figure
3.9, common methods include electric wires, light waves, and radio waves. Each
Figure 3.9
Signals can be sent via several types of media. They can be carried by electricity,
light waves, or radio waves. All of these methods have advantages and disadvantages.
Fiber optic cabling offers the fastest transmission rates with the least interference, but
because it is relatively new, the initial cost tends to be higher.
Radio or Micro Waves
Example:
Cellular phones
glass or plastic
Fiber Optic Cable
Example:
Long distance phone lines
antenna
Twisted Pair
Example:
reflective cladding
Coaxial
Example:
Cable TV
Local phone lines
method offers certain advantages and disadvantages, so they are designed for spe-
cifc applications. Installation costs for all types of cables are high, and most or-
ganizations keep cabling in place at least 10 years. Consequently, you have to be
careful to research the technologies when installing new cables to ensure you will
be able to support future needs.
Electric Cables
The two basic types of electric cables are twisted pair and coaxial. Twisted pair
is the oldest form of electrical wiring. Since electricity must travel in a closed
loop, electrical connections require at least two wires. Twisted-pair wires are
simply pairs of plain copper wires. Telephone cables are the most common ex-
ample of twisted-pair wires in households. Because of the cost, most businesses
have already installed twisted-pair wirestypically a specifc version known as
Cat 5 or Cat 6. These standards defne four pairs of wires. Older, slower networks
(100Base-T Ethernet) uses only two of the pairs (4 wires), but higher-speed net-
works (gigabit Ethernet) require all eight wires. Because of the extensive use of
Cat 5 cables, the network industry has invested considerable research into maxi-
mizing the data that can be carried over that type of cable. However, newer high-
speed transmissions systems might require better cables. The Cat 5e (enhanced)
standard would be the minimum for installing new cables, and depending on price
and location, you might consider Cat 6 or Cat 7, which is heavily shielded. But
check the prices. Cat 7 cable can be fve times more expensive than Cat 6
The biggest drawback to twisted-pair wires is that they are not usually shielded.
Consequently, the data signal is subject to interference from other electrical appli-
ances or cables. In particular, never run data cables next to power lines or electri-
cal motors. Additionally, the data cable length for most LANs must stay under 90
Reality Bytes: Excess Capacity
Partly because of crazy predictions of Internet growth, partly because of economies
of scale, and partly because of improved technologies, several long-haul data-trans-
mission companies overbuilt when installing fber-optic cables. In late 2002, only
2.7 percent of the fber was being useda whopping 97.3 percent was dark. Over
time, economics prevailed and bandwidth prices droppedat a rate of 65 percent per
year. Many of the large telecommunications frms fled for bankruptcy. The bank-
ruptcy flings generally erased the debts of the companiesenabling them to charge
even lower prices when returning. Prior to 1995, telecommunications companies
could send the equivalent of 25,000 one-page e-mails per second over one fber-optic
line. By 2002, advances in technology enabled them to send 25 million messages
over the same linean increase of one thousand times. The new technology called
dense-wave division multiplexing (DWDM), effectively splits light into multiple
colorsenabling the companies to send different messages on each color frequency.
By 2011, demand had started to cut into some of the excess capacity, but not enough
to begin installing signifcant new cables.
Adapted from Dreazen, Yochi, J., Wildly Optimistic Data Drove Telecoms to Build
Fiber Glut, The Wall Street Journal, September 26, 2002; and Berman, Dennis K.,
Technology Races Far Ahead Of Demand and the Workplace, The Wall Street
Journal, September 26, 2002.
meters (about 300 feet). You can extend the length by adding switches, but you
will be amazed at how quickly you can run into the limit when a cable needs to
turn corners and travel multiple foors. Despite these drawbacks, twisted-pair wir-
ing is the most common LAN cabling method in use today. Largely because of the
cost and the research that has led to huge performance capacities. Standard Cat-5e
wiring is capable of carrying data at 1,000 megabits per second (gigabit).
Coaxial cables were designed to carry more information than twisted pairs, with
lower chances of interference. Coaxial cable (often shortened to coax) consists of
a central wire surrounded by a nonconductive plastic, which is surrounded by a
second wire. The second wire is actually a metallic foil or mesh that is wrapped
around the entire cable. This shielding minimizes interference from outside sourc-
es. Cable television signals are transmitted on coaxial cables. Coax is capable of
carrying more information for longer distances than twisted pair. But, it is more
expensive and is rarely used outside the television industry.
Fiber Optics
A relatively recent invention (early 1970s) in communication uses light instead
of electricity. Because light generally travels in a straight line, it could be diffcult
to use for communication. Fiber-optic cable allows light to travel in straight lines
but still be bent around corners. A fber-optic cable consists of a glass or plas-
tic core that is surrounded by a refective material. A laser light is sent down the
cable. When the cable bends, the light is refected around the corner and continues
down the cable. Fiber-optic cable provides the advantages of high capacity with
almost no interference. The limitation in using fber is the higher cost of the cable
and the cost of the connectors and interface cards that convert computer electri-
Reality Bytes: LightSquared v. GPS
LightSquared is a startup company with a new technology to create a national wire-
less broadband system. The plan would entail installing 40,000 antennas around the
country. In early 2011, the company signed agreements with Best Buy and Leap
Wireless to enable them to sell wireless Internet access. Adding a new provider like
LightSquared would ultimately provide more options for consumers and provide
competition to ISPs such as the cable and phone companies. It would also provide
competition to mobile phone carriers. However, the frequencies LightSquared wants
to use are near the frequencies used for satellite signalsparticularly low-powered
GPS signals. In testing, the LightSquared devices were found to interfere with high-
end GPS devices, particularly those used by aviation, police, and agricultural appli-
cations. For instance, John Deere, the tractor company, said the devices created se-
vere interference as much as 20 miles away and complete loss of service between
four and 22 miles. LightSquareds executive vice president for regulatory affairs and
public policy, Jeff Carlisle, noted that LightSquared and BPS can and will be able
to coexist peacefully. Were committed to identify and resolving the issues through
this process. Ultimately, the FCC will decide whether to move forward and what
changes might have to be made. LightSquared later announced that it was moving to
a different frequency farther away from the GPS signals.
Adapted from Amy Schatz, LightSquareds Wireless Network Interferes With
GPS, The Wall Street Journal, June 2, 2011.
cal signals into light. For example, NICs for coaxial or twisted-pair cables can be
purchased for around $10, whereas NICs for fber-optic lines that run directly to
personal computers range from $50 to $400 (in 2011). The cards are harder to fnd
and more expensive, because it is rare to use fber lines to the desktop. A study by
Partha Mitra and Jason Stark showed that even fber-optic cables have limits. The
theoretical capacity of a single fber-optic cable is at least 100 terabits per second
(The Economist June 28, 2001 or Nature June 28, 2001). In 2011, two researchers
achieved this speed in demonstrations reported by New Scientist. Although it is
possible to run fber optics directly to your desktop, it is rarely worth the expense.
Fiber-optic connections are used for long distances, in areas with electrical inter-
ference, or when connecting buildings. In the mid-2000s, Verizon, the local tele-
phone company (not the cellular phone company) began installing thousands of
miles of fber cables directly to customer homes. The FiOS service is designed to
provide high-speed Internet connections (20 mbps or more), and ultimately video
on demand and other data-intensive services. Installing the fber into homes is
challenging because although fber cable does not have to be perfectly straight, if
it is bent too sharply or too many times, the light will bleed out at the corners and
the signal will disappear. Corning has an answer with a new fber that contains
nano-particles to allow fber cables to be wrapped around objects as small as a
pencil and still carry the light signal. However, it might take a couple of years to
create a factory that can produce the cable in large quantities (Mehta 2007).
Radio, Micro, and Infrared Waves
Radio, microwave, and infrared transmissions do not require cables. These com-
munication methods are called broadcasts. Any receiver or antenna that lies in
its path can pick up the signal. However, infrared transmissions and some mi-
crowaves require clear line-of-sight transmission. The major advantage of broad-
cast methods is portability. For example, computers can be installed in delivery
vehicles to receive information from corporate headquarters. Or individuals can
carry around laptops and cell phones and remain connected to the network. These
computers can communicate with each other and with a central database via a
broadcast network. For example, physicians in hospitals can carry small comput-
ers that automatically receive information for each patient when the physician en-
ters the room. Any instructions can be sent directly from the physicians computer
to the nursing station. In the business world, you can carry a cell phone or tablet
that maintains contact on the Internet to retrieve e-mail or scan Web sites while
you are in meetings, other offces, or a clients offce.
Broadcast media have two primary drawbacks. First, it is more important to
provide security for the transmissions. Second, broadcast transmissions are lim-
ited to a defned frequency range, which might be subject to weather, echoes, or
other interference and distance problems. Consequently, because the frequency
space is shared, it can become overloaded when too many people attempt to trans-
fer large amounts of data at the same time in the same location.
The problem of limited capacity arises because only a small number of radio
frequencies can be used to carry data. Most of the radio and television frequencies
are already being used for other purposes. Figure 3.10 shows some of the major
frequency allocations in the United States. The Federal Communications Com-
mission (FCC) allocated the personal communication service (PCS) bands in late
1993 for use by personal communication devices such as laptop computers and
personal digital assistants (PDAs). To provide these frequencies, the FCC had to
take them away from existing users. Imagine what would happen if computers
suddenly started sending information over the same radio frequency as that used
by your favorite radio station. You would not be able to hear the voices on the ra-
dio, and the computers would miss most of their data because of the interference.
All governments allocate the frequency spectrum for various uses, such as ra-
dio, television, cellular phones, and garage door openers. The PCS frequencies
were auctioned off to the highest bidders in 1994, raising more than $65 billion.
The frequency problem is even more complicated when the signals might cross
political boundaries. As a result, most broadcast telecommunications issues are
established with international cooperation. Some of the overcrowding problems
are being mitigated through the use of digital transmissions that cram more calls
and more data into the same amount of frequency space.
Despite these problems, an increasing amount of business communication is
being carried over radio networks. The International Telecommunication Union
reports that in 2002, both Taiwan and Luxembourg averaged about 105 cell phone
users per 100 people! In comparison with broadband Internet connections, in 2010,
The Netherlands led the way where 37.8 percent of the population had broadband
Internet access. The United States was ranked 14 at 27.1 percent; below Belgium,
Canada, the United Kingdom, several European nations, and Korea. (source: http://
www.oecd.org/document/54/0,3343,en_2649_34225_38690102_1_1_1_1,00.
Figure 3.10
Electromagnetic frequency spectrum. Communication techniques are essentially the
same for all media, because all waves physically have similar properties. However,
the different frequencies affect the communication performance. Shorter wave
lengths (higher frequencies) can carry more data. Some waves can travel longer
distances. Others are more susceptible to interference. In any case, there are a fnite
number of frequencies available for communication. Hence, the frequency spectrum
is allocated by governmental agencies. For complete details, see http://www.ntia.doc.
gov/osmhome/allochrt.pdf.
ELF VLF LF MF HF VHF UHF Microwave Optical
N
a
v
y
/
s
u
b
m
a
r
i
n
e
A
M
:

5
5
0
K

-

1
6
5
0

K
H
z
T
V
:

5
4
M

-

2
1
6

M
H
z
F
M
:

8
8
M

-

1
0
8

M
H
z
T
V
:

2
2
0
M

-

5
0
0

M
H
z
P
u
b
l
i
c

S
a
f
e
t
y
:

1
5
0
M
-
1
6
0

M
H
z
P
u
b
l
i
c

S
a
f
e
t
y
:

4
6
0
M
-
5
0
0

M
H
z
C
e
l
l

p
h
o
n
e
s
:

8
0
0

M
H
z
P
e
r
s
.

C
O
m
.

S
y
s

(
P
C
S
)

1
.
8
5
G
-
2
.
2
G
H
z
P
C
S

E
T
:

2

G
H
z
100 1K 100K 1M 10M 100M 1G 10G Hertz
html). On the other hand, advertised broadband download speeds in Japan and
Portugal greatly exceed those in the other nations; running at slightly over 100
megabits per second, versus about 14 megabits in the United States (OECD data).
The OECD data show monthly prices in Japan are about equal to U.S. prices. In
US dollars, from $20 55 per month.
Although wireless Internet connections are not yet as fast as broadband con-
nections, it is substantially less expensive to reach people with wireless connec-
tions. Already, in many nations, the number of cell phone subscribers exceeds the
number of fxed telephone subscribers. According to the OECD, Korea tops the
national list in terms of percentage of the population that has wireless broadband
access (95 percent). In the U.S. 44 percent of the population had broadband wire-
less data plans.
Wireless is an increasingly important method for connecting computers and
equipment. Historically, cables were the most common method of connecting ma-
chines, but people keep acquiring more devices and the tangle of cables is getting
out of control. In addition, people are placing more value on portability and the
need to access networks and data constantly. As a result, Figure 3.11 shows that
multiple levels of wireless connections are being developed. In terms of wider net-
work connections, you can use the cell phone system or WiMax which give you
a range of 2-10 miles from an antenna. These frequencies are shared and usually
regulated, so you have to purchase access through a commercial carrier (phone
company). Availability is generally limited to major cities or highways, but are
Reality Bytes: Please Stop Watching Video So I Can Send e-Mail
From late 2010 to 2011, cell phone companies began rolling out fourth-generation
(4G) services. Or they called them 4G service. Some, such as Verizons LTE, have
substantially faster data transmission rates. Early reports showed download speeds
of 12 Mbps or more on Verizons LTE networkalthough minimal competition ex-
isted so speeds might drop. Through more effcient use of the available bandwidth,
LTE can provide as much as 20-times the capacity of older 3G networks. The only
problem is that mobile data traffc is growing even faster. Cisco expects traffc to in-
crease by an average of 92 percent every year from 2011 through 2015. Most of that
growth will be due to video. Ciscos estimate might actually be low. Over the frst
four years that AT&T sold the iPhone, data traffc on its networks increased 8000
percent. Patrick Lopez, chief marketing offcer for Vantrix, noted that By the time
4G-LTE is mass market, all it will have done is allow wireless companies to keep up
with the problem. The FCC notes that the wireless spectrum surplus of 255 MHz
in 2011 will become a defcit of 275 MHz by 2014. The government is committed to
freeing up an additional 500 MHz of spectrum for wireless communication. Wireless
spectrum and effciency is a critical issue to handle traffc between towers and users
cell phones. Backhaul, or the connections from the towers to the main Internet con-
nections, is another problem. The big problem is that all of these changes take years
to implement. In the short run, wireless providers are likely to advertise high-end
video capabilities, and then charge a fortune to download the amount of data needed
for high-quality video.
Adapted from David Goldman, 4G Wont Solve 3Gs Problems, CNN Online,
March 29, 2011.
acceptable for most business uses. WiMax is a relatively new technology that is
designed to transfer wireless data over relatively long distances. Two primary fre-
quency ranges are usedthe most common one requires a license in the U.S. but
not in Europe, so it is best used by commercial providers.
Most people are familiar with WiFi technologies because access points are
relatively inexpensive and commonly installed in homes. WiFi enables multiple
computers to connect to an Internet connection and is particularly useful for en-
abling laptop users to wander around a house or small offce and remain con-
nected to the network. Early versions were relatively slow and could support only
a few users. The newest version (802.11n) can have data rates of 100-200 mbps or
perhaps higher. The key defning feature of WiFi is that it has a maximum range of
about 300 feet (outdoors). This range reduces interference from other devices, but
means businesses have to install multiple access points to cover large buildings.
After several years of marketing by IBM, the Bluetooth wireless standard is
becoming more popular. It has a range of 10-30 feet and version 2 has a maxi-
mum speed of about 2 mbps. The proposed version 3 increases the speed to 24
mbps. Today, Bluetooth is commonly used for connecting wireless headsets to
cell phones. It could also be used to connect cell phones to each other or to other
devices, but there are almost no other applications that use this technology at this
time. The distance limitation and the built-in security provisions make it useful for
short-distance, temporary connections.
From a technology perspective, ultra-wideband (UWB) is a more interesting
short-range wireless technology. Unlike traditional wireless systems, UWB sig-
nals use a huge amount of frequency bandwidth (more than 500 MHz). Normally,
this approach would interfere with other devices allocated to those frequencies,
Figure 3.11
Wireless technologies. Competition for bandwidth causes tradeoffs. Wide-area
technologies like cell phones and WiMax have limited bandwidth to reduce
interference. Wi-Fi is a medium-range local area network technology. Bluetooth and
Ultra-wideband are personal area technologies used for replacing cables for short-
range connections.
Cellphone or WiMax
2-10 miles
128 kbps - 3mbps
Wi-Fi
50-200 feet
11 mbps - 200 mbps
Bluetooth
10-30 feet
1 mbps
Ultra-wideband
10-30 feet
1 gbps
but UWB devices use low power levelsso that they appear as minor noise to
any other device and are ignored. With the low power levels, they have a range
limited to a few feet. On the other hand, UWB signals can carry huge amounts of
data, in the range of 500 mbps to 1 gbps. This speed provides the ability to elimi-
nate short cables for many devicesincluding video. For example, your digital
camera or game machine could use UWB to connect to a television set without us-
ing cables, or your laptop could connect to the conference room projector without
wires. At the moment, UWB is most commonly used for its side applicationas
short-range radar that can see through walls. Police and military organizations use
this technology to locate people in crisis situations.
Figure 3.12
Transmission capacity. Fiber-optic cables have the greatest capacity and they are
immune to most traditional interference. However, they are expensive to install and
repair. Most frms use twisted-pair or wireless connections within buildings and fber-
optic cables to connect buildings. You can purchase almost any Internet connection
speed that you are willing to pay for. Leased line rates are negotiable and depend on
distance and degree of local competition.
LAN Name Format Speed (mbps)
10Base-T Twisted pair 10
100Base-T Twisted pair 100
Gigabit Ethernet Twisted pair 1,000
Wireless LAN 11b, a,g Wireless 11-54
Wireless LAN 11n Wireless 250
fber FDDI Fiber optic 100
fber ATM Fiber optic 155
fber high-end Fiber optic 100,000,000 (100 terabits)
Internet Connection Format Speed (mbps) Estimated Cost
Dial-up Twisted pair 0.05 $20/month
DSL Twisted pair 1.5+down/0.5up $50/month
Cable modem Coaxial 6+ down/1 up $50/month
Satellite Microwave 1.5down/0.25up $50/month
Wireless/WiMax Microwave 1.5-6down/0.25+ $40/month
T1-lease Twisted pair 1.544 $400-$700/month
T3-lease Fiber optic 45 $2,500-$10,000/mo
ATM Fiber optic 155 $15,000-$30,000/mo
OC-3 Fiber optic 155 $16,000-$20,000/mo
OC-12 Fiber optic 622 $20,000-$70,000/mo
OC-48 Fiber optic 2,488 $50,000-?/mo
OC-192 Fiber optic 9,953
OC-768/future Fiber optic 39,813
Transmission Capacity
As shown in Figure 3.12, transmission capacity is often measured in millions of
bits per second. Each transmission medium has a different maximum capacity.
Twisted pair can be relatively fast for short distances, but fber optic is substan-
tially better for longer distances. Getting a faster Internet connection is primarily
an issue of cost. Fiber-optic cable might also carry a high installation cost if there
are no fber-optic connection points near your offce. Today, twisted pair lines are
commonly used for local area networks, but each connection run must be less than
95 meters (about 300 feet). Fiber optic lines are used between buildings and for
long-distance connections.
The effect of the transmission capacity is shown in Figure 3.13. For small text
and data fles, the speed is not critical. Even slow dial-up lines can transfer a full
page of text in a short time. The problem arises when you want to transfer more
complex data like photos or even video. This fgure shows why designing Web
pages carefully is still so important. When over 75 percent of your clients are
using dial-up lines, you need to limit pages to around 50,000 bytes, which takes
at least 8 seconds to download. Most people will not wait more than 15 seconds
for a page to load. Now you can see why Internet video is so diffcult even with
marginal quality. Even with video at broadband speeds of 1.5 mbps, site designers
have to restrict the video to small sizes (one-fourth of a TV screen or smaller), use
slower frame rates (as low as 15 frames per second), and employ lossy compres-
sion. These actions cut the video size by at least 1/8 (1/4 size * 1/2 frames). At
that size, the 10 seconds of video could be sent in 10 seconds (80/8). Compression
lets the developer specify a bit rate, and sites like YouTube tend to stay under 256
kbps, which reduces the quality.
Transmission delays are another important issue for some types of network
uses. Latency is the time it takes for your signal to reach its destination. It is af-
fected most by the speed at which the signal can travel across your connection
medium. Fiber optic signals are fastest because they travel at the speed of light
Figure 3.13
Importance of transmission capacity. Text is not a problem even for slow dial-up
lines, but images and video can be slow even over relatively high-speed Internet
connections.
Transmission speed Text Image Video (10 sec.)
Bytes 10,000 500,000 15,000,000
Bits (Bytes * 8) 80,000 4,000,000 120,000,000
Seconds
Dial-up 50 kbps 1.6 80 2400
DSL1.5mpbs 0.05 2.67 80
LAN 10 mbps 0.008 0.4 12
LAN 100 mpbs 0.0008 0.04 1.2
Gigabit 1 gbps 0.00008 0.004 0.12
Technology Toolbox: Creating Web Pages
Problem: You need to share information with others on the Internet.
Tools: Several tools exist to create Web pages, but at heart, Web pages are simply
text fles. The pages are written in the hypertext markup language (HTML). HTML
consists of a few dozen tags that tell the browser how to display a page. A simple
page can be written as:
<HTML>
<HEAD><TITLE>Sample HTML Page</TITLE>
<BODY>
<H1>Section One</H1>
<P>Tisisasampleparagraphonasamplepage.</P>
</BODY>
</HTML>
You can memorize the various tags, or you can use an editor to simply type the
text and let it generate the tags. However, if you are creating pages for the Internet,
avoid using document editors like Microsoft Word because the additional material it
inserts makes the pages considerably larger than necessary.
Web pages display images using the <IMG SRC=myfle.jpg> tag. You must be
careful when creating image fles to store them in a standard format: (1) graphics in-
terchange format (GIF), (2) joint photographic exports group (JPEG), or (3) portable
network graphics (PNG). You must be even more careful to watch the size of the fle.
A modern digital camera can create photographs that are 15 megabytes! Think about
how long it would take a browser to download a fle that large. In general, you must
keep the total size of a page below 100 kilobytes.
Links are created with the anchor tag: <a href=annual.html>Annual Results</
a>. The text or image between the starting and ending tags will be displayed to the
user. Clicking on the link opens the fle shown in quotes.
When you are creating an entire Web site, you need to use a style sheet. A style
sheet contains a list of styles (fonts, sizes, colors, margins, and so on) that will be ap-
plied to the various elements on a page. For instance, you could specify that the main
heading style (H1) would use an Arial typeface at 14 points in blue. The power of
the style sheet is that all styles are defned in one place. By changing the style only
once, every page on your site that is linked to that style sheet will automatically be
displayed with the new style.
HTML is easy to use, but if you need more precise control over the page lay-
out, you should use Adobes portable document format (pdf). You can buy software
that saves documents in this format. Browsers can download the Acrobat reader free
from Adobes Web site. This method is commonly used to distribute detailed docu-
ments such as t ax forms and posters.
Quick Quiz: Search the Web, or create a document and View the source to do the
following:
1. Display a word or phrase in boldface.
2. Link a style sheet to an HTML page.
3. Display a table with three rows and four columns.
4. Display a numbered list of fve items.
5. Display an icon in GIF format with a transparent background.
and require fewer repeaters. Electrical signals are also relatively fast, but the over-
all latency can be higher if the signal needs to travel through multiple repeaters
or switches. Microwave signals tend to be the slowest and subject to interference
from the weather. Satellite connections tend to have the most latency because the
signal has to travel from the ground, across 23,000 miles of space to a satellite and
another 23,000 miles back down to the base station. This delay can easily reach
a half-second or more and makes it diffcult to use satellite connections for some
purposessuch as multiplayer games. Note that latency is independent from the
bandwidthso an ISP can advertise that it provides a 1 mbps satellite connection,
but the delay due to latency might cause problems with how you want to use the
network.
Connection Devices
To reduce overall traffc, larger organizations often fnd it benefcial to build the
corporate network from a set of smaller networks. Both large and small compa-
nies use similar techniques to connect their networks to the Internet. Figure 3.14
shows a common confguration. Computers within a building or smaller area are
linked into a hub, switch, or router. This interconnection device is then linked to
the backbone, which is typically a fber-optic line. A specially-confgured router
then connects to the Internet.
If you look at the physical box, it can be diffcult to tell the difference between
a hub, router, and switch. However, internally, they function quite differently, and
you need to understand those differences to decide how a network should be con-
fgured. Hubs are the simplest connection devices. They essentially act like a gi-
Figure 3.14
Connecting networks. Several challenges arise if you build an enterprise network.
An enterprise network often connects many smaller networks established in different
departments, buildings, or even nations. The hardware and software components
must follow standards so they can communicate. An MIS team has to manage the
overall structure to maximize effciency, avoid duplication, and solve problems.
Backbone fiber optic
Hub
Switch
Hub
Radio-based network
Internet
Routers or Switches
ant junction box. Any device connected to a hub shares all of the lines with the
other devices. It is just a mechanism to plug several computers together. Switches
and routers actually examine every packet that passes through them and decide
where to send each packet. They are actually specialized computers that can be
programmed to identify network problems and intelligently route traffc to the
fastest route.
Routers and switches are crucial to improving effciency in large networks. In
many respects, a router works like a post offce. It examines the destination ad-
dress of each packet of information and selects the best way to send that packet.
Routers and switches improve performance by choosing the path of the message
and segmenting large networks into smaller pieces. For example, router segments
might be assigned to each department of a large company, where each department
has its own server. Most of the messages stay within the specifc department, so
they do not take up bandwidth across the entire company. Only messages that
truly need to be sent to other areas will be transmitted outside the departmental
segment.
The main differences between switches and routers consist of how each one
identifes other objects on the network. Switches use physical addresses that are
assigned to each network interface card. Routers use a logical address that is eas-
ier to change. Routers also contain more intelligent routing algorithms. They can
analyze traffc patterns and communicate more data with other routers to fnd the
best transmission path. Switches are commonly used to isolate traffc within a lo-
cal area network, and routers transfer data across longer distances and more com-
plex networks such as the Internet. It is possible to build a network entirely based
on routers, but it is usually cheaper to use switches where possible. Just keep in
mind that most switches cannot transfer data beyond seven hops, so if you need
lots of connections, you will have to add routers.
Figure 3.15
Network packets. All data is converted to digital form and broken into small packets.
Each packet contains a destination and source address, along with sequencing
instructions. The packets can be sent over different routes or mixed with other
packets. At their destination, the original message is restored by the network.
Voice
Computer
Sent as packets: 1 2 3 4 5
Sent as packets: A B C D E
Chicago
New York
Dallas
Atlanta
E
4
C
B
2
A
1
5
D
3
Network Structure
How can multiple users share a single network? Transferring data
between two devices is relatively easyyou just fnd a way to physically con-
nect the two computers and install some software to handle the communication.
You might be able to build a network that works the same wayby connecting
each computer to every other computer. But eventually the cabling costs would
be huge. Instead, at some point communications travel over a shared link. Think
about a network between two cities (say Los Angeles and Boston). Computers
within each city might be connected to each other, but they all share one long-dis-
tance line to transfer data between the two cities. Almost all networks are built us-
ing shared communication links. Several methods exist to share data connections,
and the method can affect the overall performance of the network. But, it is also
important to remember that the number of people and their usage of the network
will affect your performance. Shared networks require careful managementpar-
ticularly when a few bandwidth hogs exist.
Network Packets
A key feature in sharing networks is that data is transmitted in small batches. All
data, including voice and video, is converted to digital form. The data is split into
smaller packets that contain a source address, destination address, and a sequenc-
ing method. The packets tend to be a uniform size (a couple thousand bytes) and
Figure 3.16
Shared connections. Some networks rely on sharing the transmission medium with
many users. Sharing means computers must take turns using the network. Sometimes
one or two highly active users can slow down the network for everyone, so you will
not really get the listed transmission capacity.
With shared connections, machines have to take turns, and congestion
can slow down all connections.
With switched connections,
each computer has the full
bandwidth of the connection
at all times. Performance
depends on how fast the switch
can handle connections.
prevent one user from monopolizing the connection. They also make it easy for
a single network to handle data, voice, images, and video. The network simply
focuses on routing the packets to the appropriate destination. Figure 3.15 shows
an example of voice and computer data being sent over the same network. The
switches and routers choose the best path to send the data when each packet ar-
rives. The network delivers the packets to the appropriate receivers in the proper
sequence.
Today, most networks use the protocols defned in the Internet standards. The
Internet is capable of carrying any type of data packet; however, one catch with
the standard Internet is that there is no service guarantee. Your packets could be
delayed or lost along the way. Most applications have a method for recognizing
lost packets but it takes time to retransmit the packet. This issue is minor for Web
sites and e-mail, but can become annoying if you want to use the connection for
phone calls. Newer versions of the Internet protocols provide support for guaran-
teed levels of service. But, so far network companies have not devised methods to
implement and charge for quality of service (QOS). Part of the problem lies with
the need to communicate across networks owned by multiple companies. There
is also a question of whether people are willing to pay extra for a defned level of
quality. However, it is possible that someday you will be able to conduct video
meetings by reserving a certain level of speed at a set time.
Shared Connections
Figure 3.16 shows that individual machines can be connected to a shared medium
or they can be connected individually to a switch. However, the switch uses a
shared medium to connect to other switches and routers; so ultimately, connec-
tions are shared at some point on almost any network. Wired local area networks
commonly use switches to separate traffc within departments or physical areas.
Wireless (and cell phone) networks require all computers to share the same radio
frequencies. Some of the most advanced sharing technologies have been created
to handle the huge increases in traffc experienced with cell phone networks.
Figure 3.17
Time division. Only one machine or person can send data at a time. Data is limited to
short packets to prevent one communicator from monopolizing the time.
A B
D C
time
A C A C A
In theory, all communication systems that share a communication medium
share the same problems. The solutions to the problems are similar regardless
of the communication technology. In practice, engineers point out that the tech-
niques require considerable modifcation depending on the physical connection
method, but you will have to study the engineering books to understand those
details. For now, think about the protocols that people have created to handle a
common shared-communication environment: the classroom.
What happens if everyone tries to talk at the same time in a classroom? Is it
possible for more than one person to talk at one time? The common answer is
that only one person is allowed to talk at a time. A signaling mechanism (rais-
ing your hand or being recognized by the leader or instructor) is used to identify
which person gets to speak. Typically, the leader also limits the time any person
is allowed to speak. Figure 3.17 illustrates the basic process of time division
multiplexing (TDM). Only one machine or person can communicate at a given
time. Rules or protocols determine who gets to send a packet, what to do if two
machines try to communicate at the same time, how to address the packets, and
the length of the packets.
If you think about the problem for a whileand engineers have been thinking
about the problems for many yearsyou might fnd a second solution. If every-
ones hearing is good enough, it would be possible for two people to carry on a
conversation at the same time in the same room. Simply ask one pair to com-
municate at one frequency (sing soprano), and the other pair to communicate at a
different frequency (sing bass). In electronics terms, it is relatively easy to share a
communication line by using this frequency division multiplexing (FDM). Fig-
ure 3.18 shows the basic processit has been in common use for decades. Radio
and television stations are assigned separate channels (frequency ranges) so they
can broadcast at the same time without interfering with each other. The telephone
network was an early adopter of this approachlimiting the frequency range of
conversations to about 3000 Hertz and shifting connections to different frequen-
Figure 3.18
Frequency division. Each pair communicates at a different frequency (pitch). Radio
and television stations work this way by using assigned frequency ranges (channels).
f
r
e
q
u
e
n
c
y
3500 Hz
A
C
A B
D C
cies so they could be simultaneously transmitted across expensive long-distance
lines.
With a little more thoughtand a focus on packetsyou will see that it is pos-
sible to combine both frequency and time division into direct spread spectrum
(DSS). As an interesting historical footnote, the original patent for the technique
was granted to actress Hedy Lamar and composer George Antheil in 1942, during
World War II. Hedy recognized that a piano-roll mechanism could be used to shift
radio frequencies at random to prevent opponents from intercepting or jamming
radio transmissions.
Figure 3.19 illustrates the process. When a computer or cell phone wants to
transmit a packet, it checks for an open frequency slot. At each time interval, that
frequency might shift, so the receiving computers listen to all available frequency
slots and check the destination addresses of each packet. Today, the key beneft
lies in its ability to fully utilize the available bandwidth. Allocating fxed frequen-
cies to everyone is ineffcient because many of the frequencies will be unused at
any point in time. Splitting the communication by time and frequency enables
each transmitter to choose any available open frequency. Remember that commu-
nication frequencies are tight in every nation. Yet, researchers have shown there
is plenty of bandwidth availableif all transmissions were converted to spread
spectrum. Because of the technical and political issues, this change is unlikely to
happen soon. However, almost all new applications use spread spectrum technolo-
gies within a defned frequency range.
One cellular phone manufacturer (Qualcomm) invented a radically new method
of sharing a communication medium. Today, the Verizon network uses this tech-
nique to provide high-speed voice and data communications to its cell phone cus-
tomers. Newer proposals for cell phone networks have discussed using similar
technologies. The method begins with DSS, but every transmission also incorpo-
Figure 3.19
Spread spectrum. Radio networks share the scarce airwaves by breaking
communications into small packets that are sent at different frequencies and time
slots. In a given time slot, if one frequency is not available, the transmitter shifts to
the next frequency. If all frequencies are being used, the transmitter must wait for the
next time slot.
f
r
e
q
u
e
n
c
y
time
A B
D C
rates a unique ID code. By adding some mathematical calculations to every device
(including the cell phone), multiple devices can transmit at the same time on the
same frequencyusing the math to recover separate data packets. This code-divi-
sion multiple access (CDMA) method is effcient, but patent and licensing battles
have made other providers reluctant to adopt it.
Managing Shared Networks
As the number of users increase on a network, it is likely that the network will
encounter congestion problems. For example, the AT&T cell phone network was
widely criticized when it was the sole provider of the Apple iPhone. In some ma-
jor cities including New York and San Francisco, many users reported problems
with dropped phone calls. Similar problems can happen with any shared connec-
tion when traffc reaches the bandwidth capacity. For example, most cell phone
networks face problems if a major crisis hits a particular location and thousands or
millions of people all try to use their phones at the same time. Related problems
happen with data networks. Most Internet providers provide data capacity based
on the typical usage rates. For example, as shown in Figure 3.20, an Internet
connection might have a data capacity of 1.5 mbps, and that line might be con-
nected to 10 households. As long as each household uses typical Web browsing
at random times, the data will transmit quickly enough that interference will be
low and users will mostly see the full transfer rate. But, what happens if three
Figure 3.20
Managing shared networks. Some network applications use signifcantly more
bandwidth or capacity. If a small number of advanced users eat up the shared
capacity, does that make them bandwidth hogs?
Phone or Cable
Company/ISP
Shared: 1.5 mbps
1. Each person views simple data/Web pages.
Capacity is not pressed. Usage is even.
2. One person views 1mbps streaming video.
Capacity is pressed. All traffic slows down.
Is this person a bandwidth hog?
users start watching hours of streaming video that requires 500 kbps per person?
These three people will effectively use up the entire capacity of the shared line.
Streaming video can be a problem on many networks. The BitTorrent data transfer
program can also be a problem because it tries to utilize as much bandwidth as
possible and is often used to transfer huge fles.
Internet and cell phone providers (along with your university network manag-
ers) have faced congestion issues for years. As new network applications have
emergedparticularly videothe capacity demand has continued to increase. In
any shared network, some early adopters are likely to use the new applications
more than other people. Consequently, their bandwidth or capacity usage will be
higher than other users. Is that bad? Some commercial networks have approached
the problem by labeling them bandwidth hogs and imposing penalties for anyone
who uses the network more than others.
Managing a commercial network is more complicated than it appearsand it
already seems hard enough. In addition to installing lines and keeping the network
switches and routers running, the commercial providers have to determine pricing
and fnd ways to encourage people to choose their network, yet keep costs low
enough to earn a proft. And that does not begin to deal with any of the security is-
sues covered in Chapter 5. To obtain customers, many companies offered all-you-
can-eat data plans for a fxed monthly fee. But, new applications were developed
including video, large-fle sharing (BitTorrent), and Web video conferencing; and
these new applications require considerably more bandwidth than simple e-mail
and Web browsing. Some people adopted the new applications quickly, with the po-
tential to cause congestion for everyone. Ideally, network companies would expand
their capacity and provide higher speeds for everyonebut that increases costs,
and some people are likely to consume bandwidth no matter how much is available.
RealityBytes:VideoTraffcfromNetFlix Dominates the Internet
For many years, peer-to-peer fle sharing dominated the traffc fow on the In-
ternetparticularly in the United States. Many people used BitTorrent to exchange
large data flesparticularly video, and often illegally. By May 2011, NetFlix, with
its online subscriptions, had taken over the number one position in terms of total data
transferred on the Internet. The traffc account for 22.2 percent of all U.S. broadband
traffc. BitTorrent was still a close second at 21.6 percent. The percentages are based
on the actual usage or volume of traffc, not on the bandwidth capacity. The main
reason for the increase is the relatively low price of $9 per month for a subscription.
Experts estimate that it costs NetFlix between 2-10 cents to deliver a movie online,
based on server and bandwidth costs. The licensing costs are higher but the exact
cost is not available.In some ways, todays Internet business world is like the early
days of businessyou simply did not trust anyone you did not know. As economies
grew, states learned to control fraudulent businesses by requiring them to register.
Today, legitimate businesses are registered with at least one state government and
must fle annual forms and provide legal contact information. Most of this data is
accessible online. Is it really any different to require people to fle legitimate contact
information for online registrations?
Adapted from Ryan Singel, Most Content Online is Now Paid For, Thanks to Netf-
lix, CNN Online (Wired), May 18, 2011.
Figure 3.21 outlines some of the options available to manage traffc on net-
works to reduce congestion. Most commercial and university networks use a
combination of the choices. In some ways, prioritizing traffc can be the most
neutral. Network devices can monitor the type of data carried by each packet and
determine how quickly to send the packet. High-bandwidth, low-value packets
are delayed when network usage is high. For example, e-mail is often given high
priority, while online games and fle sharing are given low priorities. The packets
eventually get delivered, but users of those applications see substantially slower
data rates. A few providers have implemented plans to reduce the data speeds for
specifc users. Once a person exceeds a certain amount of data transfer for the
month, their future usage will be slowed down for a month.
Cell phone providers have become the most creative in terms of pricing. An
economist would observe that congestion pricing mechanisms are more effcient
than arbitrary limits. The catch is that marketers want to sell phones and services
by offering more open-ended plans, and customers get nervous about data caps.
Most vendors have implemented simple tiered data plans, where customers pick a
plan by estimating monthly usage levels.
However, data caps are often set relatively low and they can be easy to exceed.
For example, consider a monthly cap of 5 gigabytes (GB). Then consider a 4G
Figure 3.21
Options for managing network traffc. Commercial networks are trying several
methods to reduce the congestion caused by advanced users. Prioritizing certain types
of traffc can be a good solution, but it could also be abused. Pricing based on usage
makes the most sense economically, but it can be harder to implement and harder to
sell to customers.
Prioritize Traffc
- Slow down some usersperceived hogs.
- Slow down based on type of traffc.
. Packeteerexamine packets to identify.
. Connection port (rare).
- Sell quality of service (rare yet).
Pricing Mechanisms with Data Caps
- Overage fees.
- Differential pricing.
- Time-of-day pricing (rare yet).
- Potential problems as speed increases (4G cell).
. Medium quality is at least 1 GB/movie.
. Common data cap of 5 GB/month.
. At 5 mbps, 5 GB cap reached in 133 minutes.
Net Neutrality Proposal
- What if a commercial network tries to slow down traffc from a
competitor?
- Example: Comcast owns NBC (2011/01)
- But can neutrality be defned to still allow networks to man-
age traffc?
cell phone that can download data at 5 megabits per second. That phone would hit
the data cap after only 133 minutesor about one high-quality movie. (Calcula-
tion: 5 billion Bytes * 8 bits/Byte / 5 million bits/second / 60 seconds/minute).
Before you make any hasty decisions about the managers of networks, think
about the problem from a business perspective for a minute. What happens to
the network if ten people want to stream a movie at 5 mbps at the same time? Ef-
fectively, that generates 50 mbps demand. How much capacity can be built into
the network? What if 100 people do the same thing? That amounts to 500 mbps
simultaneous demand. How are you going to build a network that can handle that
level of traffc? And you are still looking at only 100 people on the network. Digi-
tized voice is only 64 kbps, so the high-defnition movie takes about 80 times the
bandwidth as a phone call. Either you need a massive increase in network capac-
ity or some way to discourage people from using high-bandwidth applications.
Although, you might want to talk to the marketing people about toning down the
advertising telling people they can watch movies on their cell phones.
Standards
How is it possible that you can connect your computer to a net-
work at the offce, at home, or while on the road, even over-
seas? Standards are agreements among vendors, customers, and nations. If ev-
eryone follows the standards, equipment can be connected, data can be shared,
and you can connect your computer to the network anywhere in the world. But
evolving technologies and competition signifcantly complicate these tasks. Set-
ting standards and moving to new ones is always a challenging task. However, the
Internet exists because of standards.
Te Need for Standards
Standards are important with networks. Many different types of computers and
connection devices exist. Each computer and network company has its own idea
of which methods are best. Without standards, there is no way to connect comput-
ers or networks produced by different vendors. Standards are also supposed to
protect the buyers from obsolescence. If you install network equipment that meets
existing standards, you should be able to buy products in the future that will work
with it.
Unfortunately, the process is complicated because several standard-setting or-
ganizations exist. Each major country has its own standards organization such
as ANSI, the American National Standards Institute, plus several international
organizations exist, such as ISO and the ITU (International Telecommunications
Union, renamed from CCITT), charged with defning computer and communica-
tion standards. In addition, manufacturers of computers and telecommunications
equipment try to defne their own standards. If one companys products are chosen
as a standard, it gains a slight advantage in design and production.
It is not likely that typical managers will be involved in the issues of setting
and choosing communication standards. Yet, as a consumer, you need to be aware
that there are many standards and sometimes variations on the standards. (In this
industry, a standard does not mean there is only one way to do something.) When
you are buying telecommunications equipment, the goal is to purchase equipment
that meets popular standards. It is not always easy to decide which standards will
become popular and which ones will be abandoned.
A Changing Environment
Why are there so many standards? It would be far simpler if everyone could agree
to use one standard and build products that are compatible. The problem with this
concept is that technology is continually changing. Thirty years ago, phone com-
panies never considered using digital transmission over fber-optic cables, which
is the dominant form of long-distance transmission used today. When the Internet
was frst created, the technology of the time would not handle the billions of ad-
dresses we need today.
As each technology is introduced, new standards are created. Yet we cannot
discard existing standards because it takes time for frms to convert to the new
technology. Besides, as manufacturers gain experience with a technology, they
add features and fnd better ways to use the products. These alterations usually
result in changes to the standards. An additional complication is that many com-
panies are modifying their products at the same time. It is hard to determine in
advance which changes are worthwhile and should be made standards.
The net result is that standards can be useful, but managers have to be care-
ful not to rely too much on a standard. First, remember that even if two products
support a standard, they still might not work well together. Second, if you choose
a standard for your department or company, remember that technology changes.
Corporate standards should be reevaluated every year or so to make sure they are
still the best solution. Global standards are required to ensure compatibility of net-
works and effcient routing of data.
Internet TCP/IP Reference Model
The Internet transmission protocol standard is an important standard. It also helps
you understand how networks operate and how the Internet works. As shown in
Figure 3.22, the TCP/IP (transmission control protocol/Internet Protocol) refer-
ence model breaks the process into four layers: application, transport, Internet,
and physical. Breaking the process into separate layers is critical to building large
networks. Each layer can be handled independently of the others. For example,
at the physical layer, replacing a wired connection with a wireless one should not
affect any of the higher layers. The physical layer devices simply have to provide
the same functionality to the Internet layer.
Figure 3.22
TCP/IP reference model. The model illustrates how data from an application like
e-mail is turned into packets at the transport layer, routed to the destination at the IP
layer, and physically transferred as bits at the physical layer.
Message
Header 3 Trailer 3 Message
Header 3 Trailer 3 Message Header 2 Trailer 2
Header 3 Trailer 3 Message Header 2 Trailer 2 Header 1 Trailer 1
4. Application
3. Transport (TCP)
2. Internet (IP)
1. Physical
Technology Toolbox: Transferring Files on the Internet
Problem: You need to transfer data fles to a Web site.
Tools: It is relatively easy to transfer fles from your computer to another computer
connected on a local area network. Assuming that you have the proper security per-
missions, you can see the other computer as if it were another disk drive and drag-
and-drop the fle. But no company is going to open its network to allow you to do the
same thing across the Internet. So, how do you transfer fles to a Web server if it is
not located on your LAN?
Connection Strengths Weaknesses
FTP Inexpensiveandeasytouse. Nosecurity.
FrontPage Very easy to use with secure
login.
Limited availability and users have to
purchaseclientsoftware.
WebDAV Internet standard and can
besecure.
Web server security is weaker by allowing
directorybrowsing.
VPN Verysecure. Need to purchase server and client
softwareandrequiresextrasetupsteps.
Surprisingly, given how frequently people need to transfer fles, this question does
not yet have a good answer. An older method still in use is the fle transfer protocol
(FTP). Your Web browser probably supports drag-and-drop with ftp. You can enter
an address like ftp://myserver.com/www, and the browser will connect and display
a list of fles in the folder. Then you can drag-and-drop your fles or an entire folder
onto the folder window and it will be transferred to the server. The drawback to FTP
is that passwords are sent in the clear and might be intercepted by hackers. Most
companies today require that you transfer fles using secure FTP tools that automati-
cally encrypt login and fle data. These systems require your client computer to have
special software that matches the servers encryption.
The Internet community has developed the Web distributed authoring and ver-
sioning (WebDAV) service to help solve the problems of sharing fles. As an open
standard, it is available on many servers. If you have an account on a server that has
WebDAV enabled, you can use the Internet Explorer to connect and transfer fles.
Just be sure to use the File/Open command, enter the name of the site, and check the
box to Open as a Web Folder. WebDAV is an extension of the HTTP protocols, so
it supports secure login and can take advantage of common encryption tools.
The other approach to transferring fles to a corporate server is to establish a virtu-
al private network (VPN) connection frst so that all communications are encrypted.
Then you are free to use FTP or any other common protocol inside the VPN tunnel
because no one can intercept and decrypt the messages. But you need control over
the server or a special server to confgure VPN.
Most hosting companies use Web fle uploads, which can be secure if Web se-
curity is installed on the server. But, transferring fles has been cumbersome and
generally only one fle can be uploaded at a time. The new HTML5 defnition might
improve this process because it supports fle upload with a drag-and-drop feature
built into the browsers.
Quick Quiz:
1. Which methods can you use to transfer fles to a university server?
2. Why is FTP considered a security threat?
3. What other objections exist to FrontPage?
Notice that moving down, each layer takes the data from the prior layer and
adds header and trailer information. This additional data is necessary for each
layer to perform its function, but it means that more data must be transferred.
For example, even if your physical connection can transmit data at 10 mbps, a
10-megabit fle cannot be transferred in one second. Depending on the application
and the network details, the overhead from the layers can be 20 percent or higher.
Note that there is a competing seven layer network model that defnes the ISO-
OSI view of networks. The model is similar in concept to the TCP/IP model, but
it breaks level 1 into two layers, and level 4 into three layers. Since this is not a
networking book and the Internet protocol dominates, there is no reason to cover
it here. You can fnd details on the Internet if you ever need them.
Subnet/Physical Layer
The purpose of the subnet or physical layer is to make the connection between
two machines and physically transfer bits of data. It is directly related to hard-
ware. Standards exist to specify constraints on voltage, type of wire, frequency
of signals, and sizes of physical connectors. Raw data bits are transferred at this
stage. Many different technologies exist, including wireless, wired, and fber-optic
lines.
The network interface card is a critical component of the physical layer. In ad-
dition to handling the data transfer, each card has a unique media access control
(MAC) address. This address is a globally unique number that is assigned to the
card by the manufacturer. Switches use this number to identify every device on
the network and send packets to the correct device. In fact, the technical differ-
ence between switches and routers is that switches work at Level 1 and routers at
Level 2.
Internet/Network Layer
The Internet layer is concerned with routing messages to the appropriate location.
In particular, it selects the appropriate path for a message in networks where there
is a choice. Its main advantage is that it separates the higher layers from the physi-
cal transmission of data. The network layer handles connections across different
machines and multiple networks. The Internet Protocol (IP) is the standard used in
routing packets on the Internet network. With IP, each packet is treated indepen-
dently of the others and each can follow a different route to the destination. Each
machine must have a globally unique address, so a mechanism is established to
assign numbers to machines. It would be diffcult to use the physical MAC ad-
dresses because they are randomly assigned. A logically assigned IP address is
better because the number can be used to help route each packet by using portions
of the number to segment the network. Plus, if you need to change a physical
network card, you can keep the same IP address and the Internet will continue to
function correctly. The current version of IP (IPv4) uses a 32-bit address, which is
beginning to cause some problems because IPv4 supports a maximum of about 2
billion addresses, and the world is beginning to run out of numbers.
The newer IPv6 standard supports 128-bit addresses, but it will take time to
phase in the new system. Newer servers and operating systems support the IPv6
protocol, but it will take time to update all of the routers on the Internet. In the
meantime, most companies are using intermediate steps to allocate the IPv4 num-
bers. Windows 7 automatically activates IPv6 alongside IPv4 so it can support
both types of connections.
Transport Layer
The transport layer is responsible for dividing the application data into packets
and providing logical connections to the network. The transport control protocol
(TCP) is commonly used on the Internet to handle these connections. TCP sup-
ports multiple applications at the same time by creating numbered ports. For ex-
ample, e-mail is usually transferred through port 25, and Web data through port
80. TCP sends the data packet to the specifed port on the desired machine. TCP
on the host machine listens to these ports and sends the incoming data to the ap-
propriate application server. TCP also monitors the packets to see if any are lost
in transmission. If so, the recipient machine can request that the missing packet be
re-sent, providing a highly reliable connection between two machines.
The Internet also supports the user datagram protocol (UDP), which is a highly
simplifed transport method. Most important, it does not guarantee that a pack-
et will be transferred. Generally, users do not get to choose between TCP and
UDP. This choice is made by the software developer at the network level. But why
would anyone choose UDP when there is no assurance that the packets will be de-
livered? The main reason is speed. Because UDP is so simple, it adds only a tiny
overhead to each packet, which makes it useful for large transfers of data, such as
large fles and streaming multimedia. If necessary, the application can check at the
end to ensure that all data was transferred.
Reality Bytes: NEC Lab Sets Fibre Speed Record
In March 2011, NEC Laboratories in Princeton, New Jersey announced they had
pushed more than 100 terabits of information per second through a single optical f-
ber. That amounts to transferring roughly 250 dual-layer Blu-ray discs in one second.
Tim Wang from the Lab noted the feat marks a critical milestone in fbre capacity.
By comparison, Tim Strong of Telegeography noted that total capacity between New
York and Washington D.C., one of the worlds busiest Internet routes, is only a few
terabits per second. But, he also noted that traffc has been growing by about 50 per-
cent per year. Typically, optical fber capacity is increased by splitting transmissions
into multiple frequenciesor colors. Additional capacity can be gained by using dif-
ferent polarities, amplitudes, and phases of light. Dayou Qian of NEC reported send-
ing 101.7 terabits per second through 165 kilometers of fber. His team used 370
separate lasers and several kilowatts of power to accomplish this feat. At the same
conference, Jun Sakaguchi of Japans National Institute of Information and Commu-
nications Technology in Tokyo also reported reaching 100 terabits per second. His
team used a special fber with seven internal light-carrying cores. Using a completely
different method, Wolfgang Freude and his team at the Karlsruhe Institute of Tech-
nology in Germany achieved transfer rates of 26 terabits per secondusing a single
laser. Although somewhat slower than the 100 terabits record, the cost of implement-
ing Freudes method is substantially lower. His team used orthogonal frequency divi-
sion multiplexing (multiple colors) and a fast Fourier transform to separate the col-
ors on the receiving end. Each color arrives at slightly different times and the process
could be embedded onto a silicon chip.
Adapted from Jeff Hecht, Ultrafast Fibre Optics Set New Speed Record, New Sci-
entist, April 29, 2011; and Jason Palmer, Laser Puts Record Data Rate Through
Fibre, BBC Online, May 22, 2011.
Application Layer
The application layer consists of tools and services for the user. Typical Internet
applications include e-mail, fle transfer (FTP), and Web browsing with the hy-
pertext transfer protocol (HTTP). These systems work because developers have
agreed to follow basic standards.
With the TCP/IP reference model, applications are responsible for incorporat-
ing authentication and compression. Not having a standard underlying method for
handling security has caused some problems with TCP/IP. Few applications actu-
ally have any security, several incompatible variations of security systems have
been created, and hackers have been able to write programs that attack the under-
lying, unprotected layers. Security is one of the main problems being addressed in
IPv6 and Internet2.
Te Internet
What is the Internet, how is it controlled, and how does it work?
The Internet is a loose collection of computer networks throughout the world. It
began as a means to exchange data among major U.S. universities (NSFnet of
the National Science Foundation) and connections to various military organiza-
tions and U.S. defense suppliers (Arpanet of the Advanced Research Projects
Agency). No one knows how many computers or networks are currently con-
nected by the Internet. The numbers have been increasing exponentially since the
early 1990s, so any number is likely to be wrong within a few days. To give you
some idea of the Internets astounding growth, in January 1993, there were 1.313
million host computers. In January 1994, there were 2.217 million hosts located
in more than 70 countries. In 2000, Telecordia estimated that the Internet had ex-
ceeded 100 million hosts. By 2007, Internet Systems Consortium counted over
400 million host computers. Billions more people connect to these hosts using
computers, tablets, cell phones, and so on. In 1994, over 20 million people had
access to at least e-mail services. As of mid-1994, commercial use (as opposed to
university and government use) accounted for 50 percent of the Internet usage. By
2004, over 500 million people worldwide had access to the Internet. Measuring
the Internet is diffcult, since machines are not always connected. Most studies
use some type of survey to estimate the size. However it is measured, usage of the
Internet continues to grow.
What exactly is the Internet? At heart, the Internet is just a communication sys-
tem for computers. It is defned by a set of standards that allow computers to
exchange messages. The most amazing aspect of the Internet is that there really is
no single person or group in charge. Anyone who wishes to connect a computer to
the Internet simply agrees to pay for a communication linkvia an Internet ser-
vice provider (ISP)and to install communications hardware and software that
supports the current Internet standard protocols. The person or company is given
a base address that allows other computers to identify users on the new computer.
Standards are defned by a loose committee, and addresses are controlled by an-
other small committee. The committees are convened purely for the purpose of
speeding the process; all decisions are up to the organizations connected to the
network. Participation in the Internet is voluntary, and there are few rules, just
standard practices and agreements. From a business or consumer viewpoint, there
are two primary aspects to the Internet: establishing a connection and using the
Internet.
How the Internet Works
The Internet is a communication system; it is a method of connecting computers
together. So the frst step in determining how the Internet works is to understand
how your computer connects to others. As shown in Figure 3.23, the Internet has
a hierarchy of service providers. Individuals pay a local Internet service pro-
vider (ISP) for access to the Internet. In turn, local ISPs pay an upstream network
service provider (NSP) for access to their systems and features. Each connection
must be made over a communication link. Local links are typically made over
telephone wires, but cable companies also provide service over their coaxial lines.
A few companies provide satellite connections. Some wireless providers exist, but
availability, features, and pricing vary by location. Most ISPs also utilize phone
company lines to connect to their NSP, but they lease dedicated, full-time lines
that provide faster service. The largest NSPs (Tier 1) also provide backbone ser-
vice. That is, they route communications over their own fber-optic lines that are
installed across the United States. Increasingly, NSPs are also phone companies.
Some started as phone companies and expanded into the Internet; others started
with the Internet and gradually offered voice services.
If you search for an ISP or look at the market for Internet connections, you
will fnd several providers using different technologies. The Internet backbone has
been relatively stable for several years, but the industry battles have been taking
place in the last mile of the Internetbuilding connections directly to customers.
The primary providers have been the local phone company and cable TV pro-
viders. They are important because they have existing lines into customer houses
Figure 3.23
Internet connections. Each computer must be connected to others. The Internet has a
connection hierarchy. Companies and individuals typically use phone company lines
to connect to an ISP. The ISP connects to an NSP, which routes data over the high-
speed backbone network to the destination NSP, down to the other ISP, and to the
fnal computer. Each step may involve several computers.
Individual
Internet
service
provider ISP)
Phone
Company
Network
service
provider (NSP)
Backbone
network
Phone
Company
Company
Web site
Dial-up 33.3-56 Kbps
ISDN: 128 Kbps
DSL: 256 Kbps - 2 Mbps
Cable: 1.5 Mbps - 6 Mbps
Cable
Company
T1: 1.544 Mbps
T3: 44.736 Mbps
OC3: 155.52 Mbps
OC12: 622 Mbps
and businesses. Some technical differences exist between a digital subscriber line
(DSL) and cable modem broadband connection, but ultimately customers make
a choice based on availability and secondary considerations. For example, both
tend to require that you purchase additional services, such as local phone service
or cable-TV. The electricity industry has also been trying for several years to en-
ter the ISP marketbecause they provide the third set of wires into households.
However, transmitting data over noisy (and old) power lines has presented many
technical obstacles. In 2006, the industry claimed to have found reasonably low-
cost solutions, but it is not clear that the industry knows how to compete with the
other providers. A few cities (and companies such as Google) have attempted to
cover entire cities with WiFi Internet accesseffectively offering free Internet
connections. However, WiFi access points have a limited range of a few hundred
feet. Although each access point is relatively inexpensive, it costs money to run
cables to all of the points needed to cover a wide area.
A greater challenge to the dominance of the phone and cable companies is
wireless access via fourth generation (4G) technologies. Sprint was the frst to
market with WiMax service provided by Clear. In late 2010, Verizon began of-
fering their 4G cell phone service long-term-evolution (LTE). With few users and
minimal congestion, early adopters reported speeds of 6-10 mbps for downloaded
data. Upload data rates are substantially slowereven less than 100 kbps in many
cases. But with these numbers, cell phone service might be usable as a full-time
Internet connection. In fact, most companies sell WiFi add-on services that enable
cell phones to operate as WiFi access pointsproviding Internet access to nearby
computers and tablets.
You should understand the foundations of the Internet, because someone has
to pay for each connection. Current pricing policies are to charge for the initial
communication link and for the point-of-contact Internet service. For example,
Backbone Providers Network Service Provides
AT&T Sprint AT&T Qwest
Level 3 MCI (UUNet) Cable & Wireless Sprint
Qwest IBM MCI
Phone Companies Cable Companies
Regional Bell operating companies (3 RBOCs) Comcast
Competitive local exchange carriers (CLECs) Cox Communications
Regional
Leading ISPs Satellite
AT&T Direct Satellite
Comcast Wild Blue
Verizon Starband
America Online
Figure 3.24
Leading Internet providers. There are thousands of ISPs and cable companies. This
list provides only some of the large companies in each category.
an individual pays the phone company for the local phone line and pays the ISP
for basic services. The ISP pays the phone company for the next link and pays the
NSP for access services. Figure 3.24 lists some of the largest providers in each
category. You can check with them for current prices and services.
The charging mechanism is similar for companies that wish to establish Web
sites. The catch is that the costs are higher because the company needs faster com-
munication services. The phone company charges more money for a faster link
(e.g., $500 to $1,000) per month for a T1 line). The ISP also charges more money
for the increased traffc because it needs faster equipment and faster connections
to the NSP.
The Internet service connection business is completely based on economies of
scale. The high-speed fber networks (OC3 and OC12) can handle a vast number
of transmissions, but they carry a high fxed cost. The backbone providers make
money by selling smaller increments of bandwidth to the ISPs, which incorporate
a suffcient proft. Many of the NSPs are backbone providers and increasingly
they also offer ISP services.
Video and Heavy Use Complications
The problem of video is relatively complex. Perhaps you have a high-speed Inter-
net connection and could receive high-quality video, say 1 mbps. But, think about
what will happen if as few as 100 people try to view that same high-resolution
video at the same time. The providers server has to stream 100 copies of that
video simultaneouslyeating up 100 mbps. Hence the server (such as YouTube)
needs to have at least an OC-3 Internet connection at $20,000 per month just to
handle 100 simultaneous viewers. The same problem can arise if millions of visi-
tors suddenly fnd a new Web siteas happens when a national radio or television
program highlights a site. How is it possible to handle that much traffc?
One answer is that several specialty hosting companies have evolved with mas-
sive Internet connections. They have built data centers with multiple OC-48 (or
better) Internet connections. You can contract with them to co-locate (colo) your
servers in their buildings. For a monthly fee, they provide the space, power
with UPS, and air conditioning, along with the high-speed Internet connection.
Figure 3.25 lists some of the biggest providers with multiple data centers. Larger
companies might contract directly with these frms; smaller companies will work
through intermediaries who already have co-location arrangements with these
companies. The specialty providers, such as IBM, tend to be more expensive and
negotiate individual contracts to perform specifc jobs.
Figure 3.25
Leading data center providers. The general data centers simply provide space,
electricity, cooling, and high-speed Internet connections. The specialty providers
usually sign contracts to provide individual services.
General Data Centers Specialty Hosting Companies
Equinix
Cybercon
Savis
Telecity (Europe)
IBM
AT&T
EDS
Tousands of small, regional providers
Even if you have a high-speed connection to the Internet, it is still diffcult and
expensive to handle thousands or millions of Web site visitors. A second answer
is provided by Akamai, a company that was created specifcally to reduce the load
on Internet servers. The company has installed thousands of servers in over 70
countries that hold duplicate copies of your content. As shown in Figure 3.26, in-
dividual users automatically retrieve data from the nearest server. Users see only
your site but get a faster response. More importantly, the load on your companys
server and Internet connection are signifcantly reduced because the individual
requests are handled at multiple points. Distributed network servers are the only
way to handle the distribution of video content, because even a few hundred us-
ers could overwhelm the network capacity of a single server. Amazon provides a
similar service that smaller companies can use for smaller tasks. This service (S3)
is described in more detail in Chapter 7.
Broadcast streams are slightly different from simple video fles. You might
have an event that you want to broadcast to multiple users at the same timesuch
as a ball game. Your server could generate a single feed, and this feed could be
duplicated and split to thousands of users. If handled correctly, your server and In-
ternet connection generate only a single data stream. This stream is then split and
duplicated at routers along the way to multiple clients. Some companies, notably
MLB, are working on these technologies, but problems still arise when too many
people access the streams.
Telephones and Te Internet
All phone calls today are converted to digital signals and sent over a network. So,
why should you pay for long distance phone calls? Of course, most people have
cell phones which rarely charge extra for long-distance calls within a country. But,
they still use plan minutes and generally require higher payments for international
calls. You already pay a fxed monthly fee for Internet access, and the Internet is
Figure 3.26
Distributing content through Akamai. By distributing your content to servers at the
edge of the Internet, customers retrieve data from multiple points, reducing the
load on your server and Internet connection.
Internet
Content
Video
Company Server
Akamai Servers
worldwide, so you could just use the Internet connection to handle voice calls.
Voice over IP (VoIP) is a technology that converts your phone call directly to
Internet packets and ships them to the destination. The technology is straightfor-
ward and relatively inexpensive. However, Figure 3.27 shows an important com-
plication. To use your VoIP phone to connect to people with regular or cellular
phone service, you have to contract with a company that can convert the Internet
packets back into the plain old telephone system (POTS). The process is particu-
larly effective for international calls because the Internet packets can be converted
as close as possible to the destinationavoiding international and long distance
charges. In the U.S., Vonage played that role until 2007 when Verizon, one of
the phone companies, won a major patent-infringement case against Vonage. The
phone companies were already pressured by the consumer switch to cell phones;
losing long-distance revenue to third-parties like Vonage presents a major compet-
itive threat. On the other hand, if the person you wish to call has Internet access,
you can simply use a direct PC-to-PC connection and skip the phone company
and third-party completely. If you can directly connect to the other computer, you
can use voice or even video connections for no additional chargeanywhere in
the world. Skype is a newer company that now dominates VoIP callspartly be-
cause they offer free calls for computer-to-computer access, including video chat
if you have Web cameras. Skype also offers relatively inexpensive monthly fees
even for international calls to land lines and cell phones. You can even get your
own phone number for your Skype/computer account.
Internet
Voice to IP
Cable modem
VoIP Provider
Telephone Co.
Figure 3.27
Voice over IP. The phone, or a separate device, digitizes the voice and sends it as IP
packets on the Internet. A third-party company converts the Internet packets back to
telephone packets and sends them across the phone network.
Internet Addresses
In any network, each computer must have a unique address. At the local level,
network cards contain a media access control (MAC) address. This number is
used by switches to route local traffc. On the Internet, every computer is assigned
a logical IP number.
Currently, these addresses are 32-bit numbers, typically written as 4 bytes sepa-
rated by dots. For example, your machine might be assigned 138.9.15.2 as an ad-
dress. However, 32-bit numbers will identify a maximum of 4 billion machines.
In practice, perhaps 80 percent of those numbers are usable because some values
cannot be assigned and others are assigned in bulk to companies but not in use.
Hence, an Internet committee designed a new numbering system consisting of 128
bits, which allows for several millions of numbers to be assigned to every person
likely to live on the planet. The new system is known as IPv6 and is being phased
in gradually.
Servers need to be assigned IP addresses that do not change, or seldom change.
Client computers can be assigned numbers from a local router or computer using
the dynamic host control protocol (DHCP). The DHCP server has a pool of ad-
dress numbers and gives one to each computer as it connects to the network.
In most organizations, the process of assigning addresses is more complex.
Large companies and universities were the frst to encounter problems with the
Reality Bytes: More Data More Speed
Check any wireless indicator and it is obvious that many households have installed
wireless routers. Wireless access is popular for almost all computers, both for homes
and businesses. The initial wireless transfer rates were only 11 mbps on the 802.11b
standard. Then the 802.11g standard was created to support 54 mbps transfer rates.
After a few years and some arguments over technology, the 802.11n standard sup-
ported a top rate of 450 mbps. Although 802.11n routers are widely available in 2011,
most operate at about a third of that peak speed or about 150 mbps. All transmission
methods have overhead and do not actually transfer data at the peak rates. Typically,
10 percent or so is used for overhead such as packet identifers and addresses. But,
wireless is relatively ineffcient, more often operating at 70 percent of the peak trans-
fer rates. Rates are even lower when multiple devices are heavily using the same
router. But, technology development continues. The 802.11ac standard defnes trans-
fer rates of 1 gigabit per second (gbps) and the 802.11ad standard will support a
peak rate of 7 gbps. Both are expected to reach the market in 2012 or 2013. The new
technologies will probably operate in the 60 GHz frequency band, compared to the
2.4 GHz band used by older devices. This range provides more data capacity, but
the radio waves do not travel very far and are absorbed by walls, trees, and water.
Why would anyone need such high transfer rates? The simple answer is: video. The
7 gbps speed is fast enough to transfer uncompressed high-defnition video. Which
means almost all devices could operate without wires. Wireless does have a problem
with latencythe delay between sending a signal and receiving a response. Latency
is not an issue with movies or basic displays, but it does cause problems for games or
similar interactions, where responses need to be almost instantaneous.
Adapted from Stephen Shankland, Coming to a Network Near You: Faster Wi-Fi,
CNet News, May 31, 2011.
limited number of IPv4 addresses. Consequently, they purchased Internet routers
that implemented network address translation (NAT), which is commonly used
by businesses and even home users today. As shown in Figure 3.28, with NAT,
your house (or offce) might be assigned one Internet address that connects to a
router. The router then assigns completely different numbers to all of the machines
in the home or offce. When a local computer requests a page or service from the
Internet, the router tags the address so it can identify which machine made the
request. All replies are returned to the same port, and the router directs the con-
nection to the original machine. Essentially, the router translates or changes the
internal addresses into the single external number. Client computers are assigned
internal IP numbers that cannot be used outside the NAT. Typically they start with
10 or 192.168. Essentially, one real IP address is multiplied to work with several
local devices.
NAT also creates an additional security level because no one outside this local
network can directly access any of the computers. However, sometimes you actu-
ally want to connect to other computers. In particular, it causes complications if
you want to use groupware or peer-to-peer systems across the NAT router. If you
and the person you wish to contact are both running computers behind different
NAT connections, you will not be able to make a direct connection. In this situ-
ation, you will need the use of a third-party server. Cisco WebEx and Microsoft
LiveMeeting perform this role for online conferences. Skype performs this func-
tion for voice phone calls that use the Internet, and SlingPlayer does the same for
watching television remotely. Both parties frst establish a connection to the serv-
Figure 3.28
Network address translation. Internally, machines are assigned numbers that cannot
be used outside the local network. Requests sent to external servers are temporarily
assigned a real IP address by the router. Any external attempt to reach an internal
computer is automatically blocked.
Web Server
10.1.30.15
Translate:
10.1.3.15=138.9.1.15
Disallow
incoming
peer-to-peer
er. The server has a fxed IP address, grabs the currently-assigned IP addresses of
the two parties and connects them to each other. Companies can also run in-house
servers with fxed IP addresses to handle the initial connection. Using in-house
computers is more secure because the communications do not have to go through
a third party.
NAT also makes it considerably more diffcult to track down individual users
on the Internet. If you receive a message (spam) from someone, or want to fnd the
true owner of a Web site, you begin with only the shared IP address. This address
will point you to the company that hosts the original user. You can fnd the hosting
company by entering the IP address at one of the registry sites, such as www.arin.
net. But to fnd the specifc person, the hosting company (or university) will have
to provide you with detailed records of who was using the computer at the specifc
time. Most of these organizations will reveal the information only when presented
with a court order (if they kept historical records). This process makes it more dif-
fcult and expensive to track down fraud and illegal activities on the Internet.
Domain Names
Numbers are diffcult for most people to remember. So the Internet utilizes a sys-
tem where a domain name system (DNS) server converts names to numbers.
Anyone can apply for a name and pay a nominal fee (perhaps $20 per year) to use
that name. Of course, names must be unique, so sometimes disagreements arise
over popular names. Names are followed by a suffx such as .com, .edu, or .gov.
Several newer suffxes were added in 2002 and 2003, but the .com and .net suf-
fxes are still the most coveted. One interesting feature of the names is that each
country is given a specifc suffx. Some of the countries sell names using their
unique suffx. For example, the small nation of Tuvalu allows companies to use its
suffx (.tv) for a fee.
Internet World
Real World
DNS
Registration
207.46.250.222
www.microsoft.com
Microsoft Corporation
One Microsoft Way
Redmond, WA 98052
US
Internic.org
ISP
Net Range
207.46.0.0 -
207.46.25.255
Microsoft-
Global-Net
Arin.net
Figure 3.29
Domain Name Registration. Internet entities are anchored to the real world through
the DNS registration and their ISP. But, some frms lie in their registration, and some
ISPs are controlled by unethical people.
As a business, you will need to register your domain name to make it accessible
to other users. Several companies act as registrars and you pay them to offcially
link your name to your IP address. In terms of data, the connection is ultimately
performed with the DNS system, and you can either run your own DNS com-
puter or pay another company to handle the updates. Figure 3.29 shows that DNS
registration is a critical element to electronic commerce. Registration provides a
tie between the Internet world and the real world. A second tie is provided by
the connection of the frm to its ISP. In a legal situation, the ISP can be forced to
reveal the true identity and address of its client. A third anchor is created when a
company installs a digital security certifcate. The certifcate authority verifes the
real-world identity of the client and binds it to the certifcate.
You can use the Whois facilities of the domain registrars (e.g., NetworkSolu-
tions, or Internic.org) to look up the registration data. You can use the Internet
registries to identify ISPs based on the IP address. These registries are organized
by geographical location and include ARIN (North America), RIPE_NCC (Eu-
rope, the Middle East and Central Asia), APNIC (Asia/Pacifc), LACNIC (Latin
America), and AfriNIC (Africa). However, some ethically-challenged people lie
on their DNS registrations. A few even more ethically-challenged people operate
ISP services specifcally to support spammers and hackers. As a consumer, the an-
swer is relatively easy: never, ever send money to a company that lies on its DNS
registration!
If you want to run your own Web site, you will want to create and register your
own domain name. The exact IP address will be given to you by your hosting ISP,
but you have to register for the domain name that will be linked to the address.
This registration information is kept in global whois databases. For example, you
can use the networksolutions.com database to enter a Web site name and retrieve
the registered owner of the site. When you register a site name, you are supposed
to provide accurate information on your name, address, and phone number. How-
ever, at this point in time, the registrars have been weak at verifying the data. The
resulting misleading information makes it diffcult to track down spammers and
other owners of fraudulent sites. In a sense, the Internet today is like the business
world in the 1800sany fy-by-night scammer can set up a business Web site.
In the business world, this situation was partly solved by requiring businesses to
obtain licenses and register with the government. If you have any doubts about the
validity of a company, you should check to ensure that the business is legally reg-
istered with the state government. Most states have searchable online databases
and provide free access through the offce of the secretary of state.
Domain names can consist of several parts separated by dots. The names are
read from right to left. The text on the far right is the top-level domain (TLD).
The most familiar TLDs are the six that were created frst: com, net, org, edu,
gov, and mil. Several others have been added, such as info and biz, and new ones
are proposed and debated each year. There are also hundreds of country codes
(such as us, ca, uk, and tv). Around 2010, the Internet committees introduced the
concept of generic TLDs (gTLD). In this case, companies or individuals will be
able to purchase their own unique TLD (such as .ibm or .houses). The purchaser
of a gTLD will act as its own registrar and be able to assign or sell domain names
within that TLD. However, the initial cost (evaluation fee) for a gTLD is $185,000
and costs increase from that point; including the ability to run your own registry.
The second name is the name you must pay for and register with a DNS regis-
trar. It usually is the name of a company or service, but could be almost anything.
Common examples include microsoft.com, google.com, pacifc.edu, and sec.gov.
Anything to the left of the registered domain can be created within your organiza-
tions network. Many times it is used to represent a department or specifc server
within the organization. However, there are no fxed rules so the names could
represent almost anything. For example, you might create marketing.mycompany.
com for the marketing department server. Why do you care? If you need to setup
networks, you have to learn more of these details, and you have to design a nam-
ing convention that is easy to use. But, even as a typical user you should be famil-
iar with the basic structure so you can recognize bad or dangerous domains. For
example, accounts.citibank.com.xi92lai.293aafa.com is most defnitely not an ad-
dressed owned by Citicorp. When you see that address in an e-mail message, you
immediately know it is a fakemost likely created to steal your bank password.
For years, the Internet committees have realized that the domain name system
is based on the English language and that many companies would like to create
domain names in local languages. However, a name system based on multiple
languages presents several technical challengesnotably, ensuring that names in
different languages or alphabets are never reduced to an existing name in a dif-
ferent language. Another problem is that the DNS database was designed to hold
only Latin-based characters. Most of these issues have been resolved and by 2011,
several TLDs became available in non-English languages including Arabic and
Chinese.
Internet Committees
The Internet works because of the adoption of standards. Everyone connecting a
device to the Internet agrees to follow the standards. But, these standards need to
evolve to handle new technologies. The standards are maintained and discussed
by several Internet committees. The Internet Corporation for Assigned Names
and Numbers (ICANN), the Internet Engineering Task Force (IETF), and the
Internet Assigned Numbers Authority (IANA) are publicly run organizations in
charge of establishing many of these standards. These organizations are heavily
dependent on volunteers and rely on public comments to design new standards.
The IETF is the most technical group and it deals with most of the routing and
HTML standards. IANA controls the allocation of IP address segments, TLDs,
and other assigned numbers. It is largely concerned with technical issues.
In some ways, ICANN is a more interesting group. It was not founded until
1998. It coordinates the Internet naming system. In actuality, it is the political
committee where everyone gets to argue over various directions for the Internet.
For example, it is the focus of discussions for whether new TLDs should be in-
troduced, or when internationalized domain names should be allowed. Originally,
these tasks were handled by U.S. organizations. ICANN is a not-for-proft cor-
poration that is not directly controlled by the government, but it is founded in
California. At various times, several nations have suggested that ICANN should
be run as an international bodyperhaps even within the United Nations. U.S.
politicians have resisted these attempts to change control of ICANN.
Internet 2
Originally, the U.S. government funded much of the Internet design and develop-
ment. By 1995, the U.S. government had discontinued almost all funding, and the
Internet was largely fnanced and controlled by private organizations. From 1994,
the commercial use of the Internet increased exponentially. In 1996, 34 university
participants decided that they needed faster connections (the number of partici-
pants expanded to 100 in 1999, 205 in 2003, and 381 in 2007 with 58 international
participants). In early 2011, the list of paying participants included 260 organiza-
tions and 147 sponsored participants according to the lists on www.internet2.edu.
With the support of the government and industry, they began creating Internet 2
(http://www.internet2.edu).
The two most important proposed features of Internet 2 are high-speed con-
nections and quality-of-service provisions. The overall objective is to provide a
transmission network that can support full-speed video and other high-bandwidth
applications. To understand the change, consider that most existing high-speed
Internet connections are in the range of 1 mbps to 50 mbps. The Internet 2 calls
for gigabit connection points and a minimum connection of 155 mbps. The cost
of connection depends on the connection speed, where a 1 gbps connection costs
$250,000 a year.
A related, but more fundamental, change is the ability to specify a desired lev-
el-of-service quality. Currently, if traffc increases on the Internet, all communica-
tions slow down. This situation is annoying but not troublesome for simple tasks
like sending e-mail or browsing a Web site. On the other hand, full video transfer
requires a constant minimum level of transmission capacity. So participants need
a mechanism to tell all components that a specifc set of messages should take
priority to receive a certain level of service. Some people have suggested that
the system should enable participants to pay a fee to gain their desired levels of
service, for each type of message. For example, basic e-mail messages would be
free if there is no rush in delivering them. But to reserve a time slot for video-
conferencing, participants would pay an additional fee. Then all of the Internet 2
components would give the video packets a higher priority. So far, there has been
Reality Bytes: Faster Speeds but Data Caps
In April 2011, AT&T announced that it would limit broadband customers to no more
than 250 gigabytes of Internet data transfer per month. DSL users would be limited
to 150 GB. The 250 GB value matches that imposed by Comcast. AT&T claimed
that only 2 percent of its users would be affected by the caps, but that people in that
group use up 20 percent of the network bandwidth. Data transfers above the caps will
be priced at $10 for each 50 GB. Mark Siegel, spokesman for AT&T noted that the
companys customers averaged 18 GB per month. He also claimed that the reason for
the change was because customers said that the people who use the most should pay
more. The 250 GB cap could be reached by watching 109 hours of high-defnition
video from Netfixor about 3.6 hours every day for a month. The comparison to
video is important because customers are moving to online video and dropping TV
services from AT&T and Comcast. Vince Vittore, a broadband analyst with the Yan-
kee Group, noted that the data cap isnt absolutely necessary. Its mostly a move to
prevent customers from cutting off video services. Note that Cisco forecasted that
video on-demand usage will double every 2.5 years. Verizon Wireless followed with
data caps and substantially higher prices for heavy users in July 2011 (Web site).
Adapted from David Goldman, AT&T Starts Capping Broadband, CNN Online,
May 3, 2011.
no agreement on whether additional fees should be charged, or on how the qual-
ity-of-service issues can be resolved. Although the system is being designed for
academic and government users, the industry participants (e.g., Cisco) ultimately
intend to transfer any useful technologies to the commercial Internet. Businesses
could fnd many uses for high-speed connections and service-quality guarantees.
For starters, better video transfer may fnally open the way for desktop videocon-
ferencing to replace travel to meetings.
Wireless Networks and Mobile Commerce
Beginning with cell phones in the 1990s people and businesses have become fas-
cinated with wireless communication. Wireless Internet and mobile commerce
have the potential to revolutionize the Internet, business, and society. Technologi-
cally, wireless is different from the traditional Internet in only two ways: (1) the
transmission medium is microwave radio, and (2) the client devices are smaller
with smaller screens and less computing power. Yet wireless connections open
thousands of new possibilities. The client devices can consist of anything from en-
hanced cell phones, such as the device shown in Figure 3.30, to laptops and digi-
tal tablets. Many of these devices are more powerful than some older computers.
Most cell phones and tablets have built-in cameras, Web browsers, and their own
programming or application capabilities. Touch screens have made it possible to
enter basic text and notes on these portable devices.
To date, the wireless communication presents the greatest challenges. As the
portable devices become more sophisticated, people want to use them for every-
thing from simple text to voice to video. And video includes video calls as well as
downloaded high-defnition movies. The catch is that 3G cellular networks were
designed to handle perhaps 2 mbps of data transmission. Even the early 4G sys-
Figure 3.30
Mobile commerce. Wireless connections offer new capabilities, such as e-mail and
Internet browsers on cell phones.
tems provide download speeds of only 3 10 mbps. Upload speeds are substan-
tially lower, often less than 1 mbps. Transmitting video on these wireless con-
nections can be slow for individual users. Worse, the cell-phone space is shared
by everyone in the area. Throw a few thousand people into the same location all
trying to send and receive large images or video and the network slows to a crawl
for everyoneas AT&T learned when it was the frst to sell the Apple iPhone. De-
spite the increased money pumped into expanding the network, the growth of cell
phones and data applications has expanded even faster. At some point, the cell-
phone companies might be able to install enough capacity to handle the increasing
traffc, but it is likely that data caps and usage charges will remain for heavy users.
Some of the more interesting possibilities of m-commerce come from the abil-
ity to use the phone to fnd and pay for virtually anything you might need. Rather
than rely on credit cards or cash, your cell phone could record all of your transac-
tions and give you complete instantaneous access to your data. But an incredibly
reliable and secure system must exist before consumers will trust it. In 2011, sev-
eral phone vendors began rolling out near-feldcommunication(NFC) payment
systems. NFC is a very-short range wireless communication method where either
the cell phone itself or a tag placed within the cell phone can carry basic payment
data. When waved next to a payment terminal, the ID is used to transfer money
from your account to a vendor. Some credit cardsparticularly those used in Eu-
rope use similar technology. The actual source of the funds (your bank account,
credit card, cell phone bill, or PayPal) depends on the system being used. Several
companies have created competing versions and it is not clear which one might
win the largest market share.
Another interesting possibility for m-commerce comes from the ability to use
location to identify potential customers. Figure 3.31 shows that it is relatively
easy to identify a persons location based on the way the cellular system works
Figure 3.31
Location knowledge in m-commerce. Businesses could contact potential customers
on learning their location through the wireless system.
using signal strength and triangulation. Most cell phones also have built-in GPS
receivers and can locate your position down to a few feet. Currently, people can
turn off most of the location broadcast data on their cell phones, but it is available
for 911 emergency calls.
Think about the business opportunities that this system could provide, even
from the perspective of a consumer. As you walk into a mall, you could enter your
shopping list into a browser on your phone, which would contact all of the local
stores for prices and availability, and then provide a local map and directions to
each item. You could enter notes and make your selection through the browser,
then press a couple of keys and purchase the items. Most cell phones and search
engines provide a method to fnd companies based on your current locationsuch
as the closest gas stations or restaurants that carry specifc types of food. Similar-
ly, some companies use a system that automatically tracks your location and noti-
fes your friends (www.loopt.com). A few companies such as www.foursquare.
com have built a system to encourage people to check-in to report their specifc
location in commercial stores.
Businesses have even more opportunities to collect data on customers, such as
sending advertising to your phone and tracking the number of browsers versus
buyers. But marketing systems have greater knowledge of exactly what products
and features each customer wants. For instance, the systems could offer a mini-
survey to each customer who chooses to buy a competitors product to fnd ways
to improve products. Of course, the privacy aspects are interesting. They are ex-
amined in Chapter 14.
Cloud Computing
Are personal computers necessary anymore? As network speeds in-
crease, new computing methods are being created and rolled out to businesses and
individuals. The generic term cloud computing is often used to describe applica-
tions where the server, and most of the software, are run on computers stored on
Internet serversusually leased from a third party. Figure 3.32 shows the basic
concept. The main data and applications are running on servers located on the
Internet. The results, whether data, images, or even video, are transmitted and dis-
played on client computers, laptops, or even smart phones and tablets. The key is
that the main applications and data are centralized. Ideally, the user can access the
data and applications using any device from almost any location. For these ap-
plications, the user needs only a simple client display device, often only running a
Web browser. The data and applications remain on the servers.
Scalability and reliability are important aspects to cloud computing. The ap-
plications and data are run on distributed servers connected by high-speed net-
works. If one server fails, the load is automatically picked up by the other servers.
Similarly, new servers can be added at any time to extend the performance of the
collection. Data is stored on storage area networksalso connected with high-
speed networks. Most service providers also have multiple server clusters run-
ning in different physical locations around the world. Data is replicated across the
systems so it is closer to any user on the Internet. Each local cluster has its own
high-capacity Internet connections. Most sites have dual connections from dif-
ferent providers. If one network provider goes down or has a line cut, the system
automatically routes data through the backup connection.
Cloud service providers also use virtual machines to lease space on their hard-
ware to multiple customers. Multiple VMs can run on one physical server. Be-
tween the hardware, networks, and personnel, the cloud providers make money
through economies of scale. Consider the network costs alone. A T1 line provid-
ing 1.544 mbps might cost $400 a month to a single company. But, a giant pro-
vider can get an OC3 line for perhaps $20,000 a month that provides 155 mbps;
or about half the cost per megabit per second. Plus, the higher-capacity line can
be allocated across even more customers because most of them will have minimal
traffc. In a large data center, the monthly network and power costs constitute a
large portion of the expenses. A handful of employees with the appropriate soft-
ware can monitor hundreds or thousands of servers, so personnel costs are rela-
tively low. Selling capacity to multiple businesses helps cover the monthly costs.
The more services they can offer, the more money they can attract.
But, it is diffcult to answer the question of whether everything can be shifted to
the cloud. Even if you completely trust the servers and the network, is there some
data that you might wish to keep on your own computers? Or, is it always going to
be easier to pay someone else to handle the software applications and data? These
are the main questions that need to be answered.
Global Telecommunications
What problems are you likely to encounter if you need to con-
nect to a supplier in a different country? Business frms are becom-
ing more dependent on international markets. This internationalization increases
the demands on the telecommunications system. The international transmission
of data is becoming part of the daily business routine. A manufacturing company
may have factories in several different countries, with the headquarters located in
yet another country. Supplies have to be sent to factories. Finished and intermedi-
ate products have to be tracked. Customer demands have to be followed in each
country. Quality control and warranty repair information have to be followed from
the supplier through the factory and out to the customers. Financial investments
have to be tracked on stock markets in many countries. Different accounting and
payroll rules have to be provided for each country. Basic accounting and fnancial
data have to be available to management at any time of day, in any part of the
organization.
Creating networks across international boundaries creates many problems.
Some of the complications are technical, some are political or legal, and others
are cultural.
Technical Problems
The biggest technical complication is that each country may have its own tele-
communications standards. Today, most people rely on cell phonesparticularly
in smaller, developing countries. Just as in the U.S., several different standards
exist for cell phones. Consequently, a cell phone from one country might or might
not work in a different area. Many of the European nations use a version of the
GSM standard, so phones on those systems will generally work in multiple areas.
However, some American travelers have learned the hard way that between roam-
ing and international charges, using a phone in a different nation can be extremely
expensive. Many European travelers purchase new SIM cards (subscriber identity
module) to work in each country. A new SIM card essentially enables you to make
local phone calls without roaming charges, but international calls can still be ex-
pensive. To handle those, frequent travelers can use VoIP systems such as Skype
and fnd a WiFi connection to make low-cost international calls over the Internet.
One possible way to avoid the public telecommunications hassles is to use sat-
ellite phones. Some of these phones (notably Iridium) will work almost anywhere
on the planet. Unfortunately, they do not work well inside large buildings because
the signal gets blocked. Still, they can provide international reach even in remote
areas and support small data transfers as well as voice. The traveler needs to pur-
chase a special phone and sign up for a package of minutes. The per-minute cost
(in 2010) averages about $0.90 1.30, which might seem high to people used to
unlimited calls. However, even in the U.S., some cell phone plans charge $0.45 a
minute; and international roaming charges in Europe can often exceed $1 a minute
and some people have been charged over $3 a minute.
Fortunately, the Internet has made it easier to share data, images, and video
regardless of location. So, although the U.S., Europe, and Asia use different tele-
vision standards, the computers all handle the standard types of images and video
used on the Internet. You might have to ensure that the recipient has an up-to-date
computer to handle some newer types of video, but as time goes by, most comput-
ers should be acceptable. However, keep in mind that it might not be possible to
plug your cell phone or laptop into a monitor in a different nation if you want a
larger display. Still, you can usually transfer the data and use a local system.
Legal and Political Complications
Some important problems can be created when a frm wants to transmit informa-
tion across national boundaries. These transfers are called transborderdatafows
(TBDFs). The problem arises because the information has value to the sender.
Some of the biggest complications today are political. Some nations exert strong
control over the Internet. For example, in China the government physically owns
and runs all of the Internet routers. That gives them complete control over traffc.
The government commonly flters traffc for certain keywords and blocks sites
that it deems to be against public policy. As another example, during the Egyp-
tian turnover in January 2011, the existing president (Mubarak) appears to have
ordered the shutdown all Internet routers in an attempt to prevent protestors from
communicating. He also shut down all cell phone service for a couple of days.
[Vijayan 2011] Ultimately, all he managed to do was demonstrate to all citizens
that he had no respect for them and he was forced to step down.
Several other nations have expressed concerns over cell phonesparticularly
the Research-in-Motion Blackberry which promises strong encryption. In 2010,
some Middle Eastern countries along with India threatened to ban Blackberry un-
less the company provided a way for government authorities to intercept civilian
communications. They are not alone. The Obama administration also worked on
a plan to obtain access to decrypted messages using wiretap authority [Rashid
2010]. The issues of tradeoffs between privacy and government monitoring are
explored in more detail in Chapter 14. For now, it is important realize that tech-
nologies commonly used in one country might be banned in another country. It is
important for travelers and businesspeople to keep up with the laws of the nations
in which they operate.
Another important issue revolves around typical marketing data about custom-
ers. It is common for marketing departments to maintain huge databases. These
databases contain customer names, addresses, phone numbers, estimated income
levels, purchases, and other marketing information. Problems have arisen because
the western European nations have much stricter laws concerning privacy than the
United States. In most European nations, it is illegal to sell or trade customer data
to other companies. It must also be stored in protected fles that cannot be seen by
unauthorized employees or outsiders. In most cases, it is the responsibility of the
company to prove it is meeting the requirements of the law. In many cases, this re-
quirement means that customer data must be maintained on computers within the
original nation. Also, this data cannot then be transmitted to computers in other
countries. As a result, the multinational company may be forced to maintain com-
puter facilities in each of the nations in which it does business. It also needs to
impose security conditions that prevent the raw data from being transmitted from
these computers.
There is one more important political issue involving international computer
centers. Many nations, especially the developing nations, change governments
quite often, as well as abruptly. There are many nations where terrorist activities
are prevalent. Oftentimes, large multinational companies present tempting targets.
Because computer centers tend to be expensive, special security precautions need
to be established in these countries. Probably the most important step is to keep
the computer center away from public access. Several U.S. security specialists
publish risk factors and suggested precautions for each country. They also provide
security analysis and protectionfor a fee.
A host of other political complications affect any multinational operation. For
example, each nation has different employment laws, accounting rules, investment
constraints, and local partnership requirements. Most of these can be surmounted,
but they usually require the services of a local attorney.
Cultural Issues
All of the typical cultural issues can play a role in running multinational computer
networks. The work habits of employees can vary in different nations. It may be
diffcult to obtain qualifed service personnel at some times of day or night. These
issues can be critical for computer networks that need to remain in operation 24
hours a day. In many nations, it is still considered inappropriate to have female
managers when there are male subordinates. Collecting information may be ex-
ceedingly diffcult or even culturally forbidden. In some countries, you will lose a
customer if you try to obtain marketing data such as age and income.
In some nations, the connections between suppliers and customers are estab-
lished by culture. For instance, in Japan, distribution of products is handled by
only a few large frms. These companies have established relationships with the
suppliers and retail outlets. In any country, it can be diffcult for an outside frm to
create a relationship between suppliers and customers. Trying to build computer
networks with the various companies could cause severe repercussions. The es-
tablished frms may think you are trying to steal their knowledge or information.
Summary
One of the most important concepts in MIS is the necessity of sharing data. But
networks have gone beyond simple data to becoming a major part of the use of
computers and accessing information. Many applications and business tools rely
on almost-constant connection to the Internet and other devices. As a manager,
you need to work in this world where you can search and connect to data and
people from many different parts of a company. It also means that you need to be
accessible and make your knowledge and skills available to other workers on the
team.
Communication and networks require standards, and they need to evolve over
time. One of the greatest shifts taking place now is the move to high-speed wire-
less communication. Physically, computer components have shrunk far enough so
that even relatively small cell phones carry huge amounts of computer power. As
people carry these devices everywhere, they want access to everything from any-
where in the world. The current challenge lies in building and paying for a wire-
less infrastructure that can handle the increased load created by everyone demand-
ing high-speed access at the same time. The second challenge for business lies in
determining how to use these connections and applications to improve manage-
ment and increase profts.
Building networks requires specialized workers and skills, but managers need
to keep up with some of the changing technology to make benefcial choices. As
technologies and prices change, managers often need to determine when to imple-
ment new technologies and when to wait. Making these decisions often requires
understanding the underlying technologies and trends. Managers also need to be
aware of the fundamental economics, particularly the economies of scale involved
in networks. Decisions need to be made about leasing space and network capacity
versus building and managing your own connections. And these decisions need to
be reviewed as prices change and new frms enter the industry.
Globally, telecommunications reach around the world. But capabilities and
technologies vary by nation. And the U.S. is not always at the top of the world.
Still, if a company wants to connect to customers and suppliers in other nations,
managers need to be aware of potential conficts. Technology differences remain,
pricing issues can be huge barriers, and political restrictions can cause complica-
tions with many types of communications and applications. Most of the problems
can be avoided or bypassed but it helps if you search for them and make plans
ahead of time.
A Managers View
The goal of telecommunications is to remove location as a factor in manage-
ment and decisions. Technologies exist to enable you to connect to almost
anyone, anywhere in the world to share data, voice, and video. However,
bandwidth or the speed of transferring data varies considerably, and transfer-
ring large amounts of data can be expensive. Still, as speed and availability
increase, new businesses and new ways of operating businesses are being
developed. Opportunities exist to create expand and proft as the world be-
comes more interconnected.
Key Words
Web Site References
Financial News and Quotes
Big Charts bigcharts.marketwatch.com
Bloomberg www.bloomberg.com
Dun & Bradstreet www.dnb.com
Global Financial Data www.fndata.com
Reuters www.reuters.com
SEC Edgar www.sec.gov
Motley Fool www.fool.com
Wall Street Journal wsj.com
Yahoo Finance fnance.yahoo.com
Discount Online Trading
TDAmeritrade www.tdameritrade.com
E*Trade www.etrade.com
Fidelity www.fdelity.com
Charles Schwab www.schwab.com
ScotTrade www.scottrade.com
bandwidth hogs
blog
Bluetooth
Broadcasts
coaxial cable
co-locate (colo)
direct sequence spread
spectrum (DSSS)
domain name system (DNS)
dynamic host control pro-
tocol (DHCP)
e-discovery
e-mail
extranets
fle transfer protocol (FTP)
frequency division mul-
tiplexing (FDM)
grid computing
Groupware
hypertext markup language (HTML)
instant messaging (IM)
Internet service provider (ISP)
Internet Assigned Num-
bers Authority (IANA)
Internet Corporation for Assigned
Names and Numbers (ICANN)
Internet Engineering
Task Force (IETF)
Intranet
IPv6
last mile
latency
local area network (LAN)
media access control (MAC)
near-feld communication (NFC)
network address translation (NAT)
network interface card (NIC)
packets
peer-to-peer networks.
podcasts
point of sale (POS)
portable document format (pdf)
really simple syndication (RSS)
scalability
server farm
social networking
standards
storage area network (SAN)
time division multiplexing (TDM)
transborder data fows (TBDFs)
transmission medium
twisted-pair
ultra-wideband (UWB)
voice over IP (VoIP)
whois
WiFi
Wikis
WiMax
Review Questions
1. Does it matter where data is stored anymore? Are networks reliable and
available enough to assume data and processing can be handled anywhere?
2. Why do businesses use networks to share hardware?
3. How is cloud computing like renting software?
4. List the main components of a network.
5. List the types of transmission media that are available. How do they compare
in transmission rates and cost?
6. What are the main advantages and drawbacks to wireless networks?
7. How do networks handle multiple users at the same time?
8. Why are standards so important in networks?
9. What organization is responsible for creating (or not) new top-level domains
on the Internet? Should this group be adopted and controlled by the United
Nations?
10. What type of network service do you need to run a Web server? Newer (LTE)
phone speeds are reported to be over 10 mbps; would it be possible to use
those to run a Web server?
11. Why are some cell phone networks so slow?
12. Why is the domain name system so important and what is the role of domain
registrars?
13. What is the Internet2 and how will it affect businesses?
14. What problems arise with global telecommunications?
15. Describe the process by which a computer is assigned an IP address.
16. Some people and a few members of the Internet committees have suggested a
substantial increase in the number of top-level Internet domains. Why is this
a bad idea for businesses?
Exercises
1. Some cell phones have the ability to function as a Wi-Fi to connect other
devices to the Internet using the cell phone network. Most tablets have an
optional cell phone connection. Choose a tablet and a cell phone from the
same service. Compare the costs of paying to connect the tablet directly
to the cell network versus paying for the Wi-Fi connector and connecting
through a cell phone.
2. Find at least one method to connect a printer to your home (or school)
network so that it can be shared by all users on the network. How would you
limit access to specifc people, or track the number of pages printed by each
person?
3. Using the Internet, fnd at least two software packages that will back up
data across a LAN. Briefy explain how the software functions and what
components need to be installed. Estimate the price of the software for a
network of four servers and 100 clients.
4. Research to fnd the service that generates the most amount of traffc on the
Internet today.
5. Design your perfect wireless communication device. Possibly draw it, but
identify the features that you would want. Rank the features by importance
(descending order) and write a brief reason for your top three selections.
6. Assume you want to start an e-business. What steps do you have to take to
obtain and establish a domain name for the business? How much will it cost?
7. Choose two existing Web site names. To make it more interesting, one of
them should be from a spam site. Use the Internet resources to obtain as
much information about the company and owner as you can. Note, you
should also check the offce of the appropriate secretary of state. [Answer
depends on companies, but at least provide the output from www.arin.net (or
www.apnic.net) and www.networksolutions.com (or www.internic.org)]
8. Check vendor advertising and identify the costs and speed of the following
services:
a. DSL from the phone company.
b. Cable modem.
c. Cell phone Internet access.
d. Dial-up Internet access.
e. ClearWire.
f. Wireless access at a local hot spot.
9. Identify the cost of running a server on Amazons cloud system using an
entry-level Web server running Microsoft Windows Server. Assume the site
you create will contain about 5 GB of data and handle about 500 gigabytes of
data transfer (outbound) in a month.
10. Estimate how long it would take to transmit the following information.
a. A 5 megabyte image fle over a 3G cell phone (70 kbps).
b. An e-mail message with a 120 megabyte video clip attachment over a 5
mbps cable-modem connection.
c. A 1 gigabyte data folder over a 100 mbps local area network connection.
d. A 1 gigabyte data folder onto an SSD drive that can write 270 MB/second.
e. If possible, test the above examples with real data.
f. Why would the actual transfer times be slower than the estimated
numbers?
11. Identify at least three major features provided by HTML 5.
Technology Toolbox
12. Create three Web pages that are linked. Include at least one image.
13. Create a style sheet for at least two Web pages and demonstrate how the look
of the site can be changed by altering the style sheet.
14. Create a simple Web page using Microsoft Word and save it. Create a similar
page using either straight HTML or an HTML editor such as FrontPage.
Compare the two fles and comment on the differences.
15. Find three Web hosting companies and identify the fle transfer methods
supported by each company. Does the company charge extra for some
methods?
16. Briefy describe the purpose of server programming for Web pages and how
it differs from client-side scripting.
Technology Toolbox
17. How are fles uploaded to YouTube?
18. How are fles uploaded to the Amazon S3 service?
19. Find a commercial Web hosting company and list the methods that it supports
to upload fles.
Teamwork
20. For each member in the team identify (a) how many e-mail messages they
send a day, (b) the number of text messages sent per day, (c) the number of
phone calls made per day. Combine the results from all team members and
comment on any patterns. Why would people send more cell phone text
messages than e-mails?
21. As a group answer the following two questions. Is it possible to live with
only one wireless device? Can it be standardized within the group?
22. You have an American cell phone and will be traveling to Europe for
two months. You want to have cell phone service so that your friends and
coworkers can reach you in an emergency. You also want to be able to call
people and places in Europe while you are there. Each team member should
fnd a different way to provide this service. Share your results, compare the
costs and benefts and choose a plan.
23. Each person should fnd a software package that is used to create Web
pages. Summarize the basic features, ease of use, target market, and its
price. Combine the data and recommend a package that could be used by a
marketing department to design basic Web pages.
24. If a network is available with Microsoft Outlook (or similar package) on each
client, add each team member to the Contacts list. Each person should enter a
few items in a personal calendar to block out some times for one week. Then
share the calendars and use the system to schedule a meeting.
25. Interview managers, friends, or family members who have jobs and identify
where their primary fles are stored. Are they stored on local computers, on a
department fle server, or at a more distant location? Who is responsible for
backup? What is the process for recovering lost fles? Combine each result
into a document. Create a table and a chart to summarize the results. Write a
short paragraph suggesting improvements.
26. If you have access to computers running Windows 7 or above, have each
team member determine the IPv6 address of a machine and compare the
results. Hint: You can always use the command line program ipconfg.
27. Design a network for the Rolling Thunder Bicycle Company. Identify who
will need access to the network; how many workstations you will need (and
where to place them); the data, input forms, and reports users will need.
Using the existing data, estimate the storage requirements and transmission
needs. Specify how changes and growth will affect the type of network
needed.
28. Describe how the Internet could be used to increase sales at Rolling Thunder
Bicycles.
29. Rolling Thunder Bicycles wants to expand international sales. What changes
would need to be made to the application? What problems would you expect
to encounter, and how would you overcome these potential problems?
30. The manager of Rolling Thunder Bicycles wants to access the database from
home. Describe some options of how this could be done. What potential
problems might arise?
31. If you have access to laptops and a wireless network, test the database
running over a wireless connection. If fve or six people want to use the
system with wireless devices, what problems might arise? Do some research:
Is there a system that can help?

Additional Reading
The capacity of optical fbre, The Economist, June 28, 2001. [Theoretical
capacity of fber-optic cables.]
Federal Communications Commission, Broadband Performance, OBI Technical
Paper No. 4, 2010. http://www.fcc.gov/Daily_Releases/Daily_Business/2010/
db0813/DOC-300902A1.pdf. [FCC evaluation of U.S. broadband usage and
speeds, suggesting that actual consumer speeds are about half the advertised
rates.]
Jeff Hecht, Ultrafast Fibre Optics Set New Speed Record, New Scientist, April
29, 2011. [Two researchers show demonstration speeds of 100 terabits per
second in fber.]
The Internet, untethered, The Economist, October 13, 2001. [Special section on
wireless issues with examples from Japan.]
Kahn, J. M. and K.P. Ho. Communications technology: A bottleneck for optical
fbres. Nature, June 28, 2001, pp. 1007-1009. [Theoretical capacity of fber-
optic cables.]
Kahn, Robert E. The Role of Government in the Evolution of the Internet,
Communications of the ACM, August 1994, p. 15-19. [Early days of the
Internet.]
Kurose, James F and Keith W. Ross, Computer Networking: A top-down
approach featuring the Internet, 2001, Addison Wesley Longman: Boston.
[Technical details on networking and TCP/IP.]
Mehta, Stephanie, Bend it Like Corning, Fortune, August 6, 2007. [New fber
optic cable from Corning that can handle tight bends.]
Fahmida Y. Rashid, Obama Wants to Wiretap the Internet, eWeek, September
27, 2010. [Several countries and the U.S. lean on Blackberry to provide
decryption keys so police can listen in on calls and messages.]
Jaikumar Vijayan, Egypts Net Blockage an Armageddon Approach,
Computerworld, January 28, 2011. [Initial report on Egypts shutdown of the
Internet and cell phones.]
Wiggins, R. How the Internet Works, Internet World, 8(10), 1996. [Basic
explanation of terms and connection points.]
Waltner, C. Meet your connection, Fortune Technology Review, Summer 2001,
59-66. [Discussion of business Internet connection costs and options.]
Cases: Wholesale Suppliers
Te Industry
0
1
2
3
4
5
6
7
8
9
10
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
B
i
l
l
i
o
n

$
Annual Revenue
W.W. Grainger
Univar
Owens & Minor
Wholesale suppliers are the original business-to-business providers. In many
ways, they are like any other manufacturer or retailer. The difference is that their
primary customers are other businesses. Selling to other businesses often means
that the company supplies bulk items, but some wholesalers specialize in small,
hard-to-fnd items. Some wholesalers provide commodities, such as raw chemi-
cals. That raises one of the big challenges: differentiating your company from the
competitors. What makes one companys products different from the others? In
most cases, it is not the product that matters, but the customer service. Some-
times location is a key aspect to providing service, so small frms can survive by
-0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
R
a
t
i
o
W.W. Grainger
Univar
Owens & Minor
carving a local niche. But the big frms often have an advantage in economies of
scale, driving costs down by serving large geographical regions through distribu-
tion centers. Some of the large commodity providers are global in scope. But size
raises a new set of problems: managing a global company that has grown through
multiple mergers. Communicating within the company is diffcult enough. Com-
municating a consistent message to customers can be exceedingly hard.
Payments are another challenging issue for wholesalers. Manufacturers often
squeeze vendors by stretching payments to the future. Small frms often pay the
price in these cases, because it puts pressure on their cash fow when they do not
receive payment for 90 days. Technology cannot always solve this problem be-
cause it is a power balance between vendor and customer. However, it can at least
track the payment history and make the entire process transparent to both sides.
B2B Ecommerce
Communication is the entire purpose of electronic networks. Internal networks
are useful for collaboration and coordinating employees. For customers, the big
leap in the last decades has been e-commerceparticularly business-to-business
(B2B) sites. B2B sites can be set up by one company, focused on providing a
single interface for its customers. Alternatively, it might be a site shared across
the industry by multiple suppliers and many customers. Several auction-based
Web sites were created around 2000 to increase competition and make it easier
for customers to fnd and purchase various items. Many of these sites later failed;
however, a few remain in some key industries. Both sellers and purchasers have to
decide what type of B2B site works best for each type of product.
A shared site can be run by a neutral party, with costs shared by everyone.
Hence, it can be easier for smaller frms to participate. But auction-based sites
tend to drive down the prices of commodities, so sellers might not want to use
them. From the perspective of the buyers, they might prefer a tighter relationship
with one or two suppliers. If you compete only on the basis of price, suppliers
might not be responsive when crises arise. Ultimately, one of the big questions for
buyers is how much of the logistics they want to handle themselves, versus how
much they want to outsource to the vendors.
Logistics
Distribution is a key factor in the wholesale industry. Suppliers need to get prod-
ucts to the customers. Sometimes the products are bulky. In the case of chemicals,
they might be volatile and diffcult to transport. But that is the point of customer
service by the supplier. On the other hand, logistics involves more than just trans-
portation. It includes tracking orders and payments, timing the deliveries, fnding
rare products, tracking shipments, and recording everything in a format that can
be shared. Elemica is a multi-vendor Web site for the chemical industry. Its Web
site notes that an average international shipment requires 6 participants, 20 docu-
ments, and between 4 and 12 weeks of time. One of Elemicas selling points is a
software tool to track the entire process and integrate the data between the sup-
plier and purchaser enterprise resource planning (ERP) systems.
Wholesalers also have to make diffcult decisions about inventory levels. In-
creasingly, through just-in-time ordering systems, manufacturers have pushed in-
ventories back to the suppliers. Consequently, suppliers have to continually evalu-
ate and forecast customer demands, compare them to world supply levels, and
determine how much to stock of each item in every location. Linking to customer
databases can be a useful step, because it enables suppliers to quickly spot trends
and make more accurate forecasts (Babcock 2003).
Additional Reading
Babcock, Charles, Distribution: Distributors Look To Crystal Balls,
Information Week, September 22, 2003.
Elemica Web site: http://www.elemica.com
Case: Owens & Minor
Founded in 1882, Owens & Minor (ticker: OMI) is a leading distributor of medi-
cal and surgical supplies to hospitals in the United States. It is also a health care
supply chain management company that can run the entire logistics side for hos-
pitals. In 2006, the company distributed almost 120,000 products from over 1,000
suppliers. It had 41 distribution centers and a total of 3,200 employees distributing
products to about 4,000 customers (annual report 2006). By 2010, the number of
customers had increased to 4,400 supported by 52 distribution centers and 4,800
employees (annual report 2010).
The company purchases medical items in bulk from suppliers and stores them
in its warehouses, which are close to customers. Most warehouses deliver within
a 200-mile radius. Customers can order through a variety of systems, including a
stockless, automated system where Owens & Minor employees deliver individ-
ual items as needed. Although the company purchases items from almost all of
the medical manufacturers, about 16 percent of the net sales in 2003 consisted of
products from Johnson & Johnson and 14 percent from subsidiaries of Tyco Inter-
national, which includes the Kendall Company (annual report). Its major nation-
wide competitors are Cardinal Health (whose roots lie with American Hospital
Supply Corporation), and McKesson. In 2006, sales revenue reached $5.5 billion
with a net income of $49 million, which represent a 16 percent increase in revenue
but a 25 percent decline in proft from 2005. But the move into new offces ac-
counts for some of the increased costs (2006 Annual Report).
Delivering Supplies
Until 1998, O&M was simply delivering supplies to hospitals. In that year, the
company was hit with the loss of substantial revenue when HCA Inc., the biggest
health care provider in the United States, decided to take its business to another
companyimmediately reducing O&Ms revenue by 11 percent. The company
decided that the answer was to go beyond simply selling and delivering products.
It needed to provide entire supply chain management tools to hospitals. The com-
pany estimates that hospitals spend an additional 40 percent of their supply costs
just managing the logistics, tracking orders, and restocking. David Guzman, hired
as CIO in 2000, notes that theres only so much blood you can squeeze from a
turnip in terms of the product price. But theres a lot you can do with the logistics
costs (Kontzer 2003). Consequently, O&M decided to extend itself deeper into
the supply chain on both the hospital and manufacturer ends. CEO Bilmer Minor
III notes that our business has been built on supply-chain services and customer
service. We realized our difference would be based on providing information
(Kolbasuk 2001).
Logistics Services
As a distributor, O&M understands the importance of networks for connecting
to customers and suppliers. The company survives on price differentialsbuying
products at discounts from manufacturers and distributing them to hospitals. With
ferce competition, it has to use technology to hold down costs to make a proft
about 1 percent on sales (Stahl 2004).
One challenge that O&M faces is that its customers (hospitals, nurses, and phy-
sicians) are several years behind in technology. Craig Smith, president of O&M,
estimates they are as much as 20 years behind. Smith believes that radio frequen-
cy identifcation (RFID) chips can make a substantial difference by helping hospi-
tals track everything they own from beds to drugs. Were headstrong into RFID.
It will have a signifcant impact on our business (Stahl 2004).
Beyond the simple warehouse and transportation services, O&M has created
several tools to help hospitals reduce their costs. The OMSolutions division is a
professional services unit that handles consulting and outsourcing to help hospi-
tals reduce costs. The team handles everything from stockroom and process re-
design within a hospital to outsourced management of supplies (annual report).
The CostTrack system is another tool to help customers analyze their costs. Using
activity-based costing, the system shows customers exactly how much each step
of the purchasing process is costing. It enables hospitals to then choose exactly
which parts of the process they want to outsource to O&M. In 2003, 32 percent of
O&M sales were generated through the CostTrack system (annual report).
One big problem that hospitals face is that they buy supplies from many sourc-
es. As much as 50 percent of the purchases are direct from manufacturers. Other
items might be purchased from multiple suppliers and distributors. So O&M de-
veloped the Wisdom2 analytical tool. It collects all purchasing data from a hospi-
tal and makes it available as a single dataset. It tracks purchases from competitor
suppliers as well as directly from the manufacturers. Judy Springfeld, director of
corporate standardization and contracting at Baylor Health Care System (a nine-
hospital network), says that weve just been dying for data (Kontzer 2003). The
Wisdom2 system provides the detailed data she wants, across all of the purchases,
and even puts it into a spreadsheet. Wisdom and Wisdom2 are also useful to O&M
directly.
Ultimately, Owens & Minor would like to simplify the ordering process even
moreto the point of automated replenishment. If the system can forecast usage
rates, O&M employees could deliver new items to the hospital and track the entire
process, without needing a separate order from the hospital. Guzman thinks the
process is feasible, Youd be surprised how remarkably predictable demand is
in the healthcare system. And I dont mean in general. I mean system by system.
Massachusetts General is different from Stanford Hospital, but Stanford in its own
right is predictable, as is Massachusetts General (Kontzer 2003).
O&M also uses technology and networks to improve its operations. Remember
that margins are extremely tight, and saving costs is critical. A new warehouse-
management system in 2001 that uses wireless technology to guide workers in-
creased productivity by 20 percent (Kolbasuk 2001). Notice that O&M manages
to record over $4 billion of sales a year with only about 3,200 total employees.
New Systems
To provide new services, Owens & Minor needed new systems. But before build-
ing new systems, the company frst needed to consolidate its information technol-
ogy team. In 2002, the company canceled an outsourcing contract with IBM and
expanded an arrangement it had with Perot Systems. The goal was to consolidate
data centers to save costs, but also to begin creating new technology systems (Vi-
jayan 2002). The companys main goal is to redesign the legacy systems so they
use Web services to integrate all of the components. Ultimately, every informa-
tion system the company has will need to connect to the OMDirect Web portal.
Integrating the systems is still diffcult and requires a new system architecture that
uses XML to transfer the data. Guzman notes that you cant go out and buy Web
services. Its clear you have to be the one to build [them] (Murphy and Bacheldor
2003), so the process is scheduled to take three years.
In 2006, Owens & Minor extended its outsourcing agreement with Perot Sys-
tems, committing to migrating their mainframe systems through 2014. Perot Sys-
tems also helped transfer and integrate data from the purchase of McKesson. The
most important step was linking the O&M systems with those of the new cus-
tomers (2006 Annual Report). The purchase of McKesson reduces the number of
national providers to two: O&M and Cardinal Health. With the purchase of Ac-
cess Diabetic Supply, LLC in 2005, O&M has become the third-largest mail-order
supplier of diabetes supplies; but faced problems with not getting paid by many
end-customers. The company prides itself on being an early adopter of computer
technologylargely used to reduce costs and improve productivity. The company
distributes 180,000 products from 1,200 suppliers, so tracking inventory and pur-
chases are important. They also provide software management tools for operating
rooms across the nation, including QSight, SurgiTrack, Wisdom, and PANDAC,
an operating room inventory management program. O&M uses these tools to con-
trol and reduce inventory costs. Wisdom is a business intelligence tool that helps
customers analyze their supply usage. Sales to the top 10 customers constituted 20
percent of revenue, so providing support and linkages are critical to the success of
O&M.
In 2010, O&M needed to move its custom ERP system onto new Windows
servers. The original system was a custom COBOL application running on main-
frame hardware, consisting of about 10 million lines of code. The company de-
cided that the complex business logic built into the code was too valuable to lose
and too expensive to recreate. So it used Micro Focus COBOL and tweaked the
code to run on Windows-based hardware (Thibodeau 2010). In addition to saving
money by not buying and customizing a new ERP system, the company cut its
hardware costs in half.
The company had been looking at the project for several years and struggled to
fnd a way to transfer the system without bringing the business to a grinding halt
for six to nine months, as CEO Craig Smith put it (Nash 2010). Transferring to
a new ERP system ultimately saved $100 million. The new system simplifes us-
ability. When a customer calls in, sales representatives can see all of the relevant
data on one screen instead of searching 11 or 12 screens to get an answer.
Security
With any network system, and particularly with Internet-based systems, security
is a critical factorparticularly in the health care industry. At one level, security
today is relatively straightforward: identify each resource and user, and then as-
sign the appropriate permissions to users. Except that with thousands of users and
hundreds of applications, it is expensive to manage the tasks. Owens & Minor has
more than 12 administrators who are dedicated to managing, adding, and deleting
employee and customer access rights (Hulme 2003). And, the company has to
be extremely careful to monitor access rights to ensure they are correct and that
they are updated as employees change jobs or leave the company. To simplify the
process, O&M is consolidating the identity databases onto a single centralized
Microsoft Active Directory repository. Active Directory (AD) runs on Microsoft
servers and holds user credentials. It can be accessed by a variety of applications
and services. Users essentially login to AD, and the system authenticates the user
to other applications. O&M is also using IdentityMinder from Netegrity Inc. to
transfer the user rights into the 20 or 30 applications accessed by each user. Paul
Higday, chief architect for O&M, notes that what this will allow us to do is set up
a user based on what theyre allowed to access with a single click instead of hav-
ing to manually add each account (Hulme 2003).
Questions
1. How does Owens & Minor use networks and information systems to reduce
costs?
2. Given the innovations by American Hospital Supply in the 1980s, why are
hospitals not even more integrated into the supply chain?
3. How are logistics services different from simply delivering supplies?
4. How will RFID affect the use and purchasing of hospital supplies? Will the
technology be widely accepted?
Additional Reading
Ante, Spencer E., Owens & Minor, BusinessWeek, November 24, 2003, p. 92.
Hulme, George V., Identity Checkpoint, Information Week, January 20, 2003.
Kolbasuk McGee, Marianne, Supply Chains Missing Link, Information Week,
September 17, 2001.
Kontzer, Tony, Owens & Minor Takes Supply Chain Deeper, Information
Week, September 22, 2003.
Murphy, Chris and Beth Bacheldor, Tying It All Together, Information Week,
March 17, 2003.
Nash, Kim S., What Getting IT Right Means to a CEO, CIO, May 18, 2010.
Stahl, Stephanie, Prescription For Change, Information Week, February 2,
2004.
Thibodeau, Patrick, One Firms Story: The Mainframe Goes, but Cobol Stays
Behind, CIO, April 20, 2010.
Vijayan, Jaikumar, Owens & Minor Axes IBM For Perot, Computerworld, July
29, 2002.
Case: W.W. Grainger
Founded in 1927 in Chicago, W.W. Grainger (ticker: GWW) is in an interesting
industry. The company supplies products to other businesses for maintenance, re-
pair, and operations (MRO). The products include janitorial supplies, light bulbs,
and bolts. But the repair side can include any number of unique parts or tools
used to repair equipment, from gaskets to pumps. The company estimates the U.S.
market for MRO to be $100 billion a year, giving Grainger about a 5 percent share
of the marketthe largest of any competitor. About 90 percent of the market is
flled by local and regional suppliers, and Graingers strategy is to capture market
share from these fragmented rivals. In 2003, the company had 575 sales outlets
backed by 17 distribution centers and carried more than 500,000 different prod-
ucts (2003 Annual Report). In 2006, Grainger had sales of $5.9 billion with net
earnings of $383 million. In 2004, Grainger began installing more outlet stores in
major metropolitan areasin a goal to make it more convenient for people to pur-
chase products. The company expected to spend $50 to $80 million on the project,
as well as $10 to $15 million on information technology. (2006 Annual Report). In
2010, Grainger had sales of $7.2 billion with 607 branches, including new outlets
in Asia in Latin America. By 2011, Grainger carried 354,000 different products
and James Ryan, the CEO, has a goal of reaching the sweet spot of 450,000
500,000 different products (2010 Annual Report).
As a distributor and supplier, Grainger makes money on the price differential
by purchasing items in bulk and selling them at a markup to customers. But the
markup is small. However, Grainger is expanding into the logistics market. The
company estimates that about 40 percent of the total procurement cost comes
from trying to locate and purchase MRO products. One of Graingers competitive
strengths is the number of products it carries and makes available almost imme-
diately at any of its stores. Customers need to know that Grainger will carry the
parts and supplies they need, regardless of the manufacturer. Parts that are not
stocked in the local branches can usually be shipped for next-day delivery out of
the distribution centers. To strengthen its local presence, Grainger is adding more
showrooms and warehouse capacity across the United States. The goal is to have
an outlet within 20 minutes of most businesses and institutions (annual report).
Of course, it does not make sense to carry products that rarely sell, so Grainger
has to carefully choose the products that it holds in inventory. The company also
works closely with manufacturers to reduce cycle times so that they can quick-
ly refll warehouses or custom-order products if needed. Grainger manages the
inventory problem by using multiple channels. Its 3,700-page CD-ROM cata-
log describes 82,000 products. Its Web site offers more than 200,000 products.
When customers need more specialized repair parts, Graingers Parts division can
obtain more than 2.5 million parts from its suppliers. The FindMRO division is
even more specialized and can track down over 5 million facilities maintenance
products (annual report). George Rimnac, vice president and chief technologist at
Grainger, observes that many of the products we sell are products our customers
didnt know they needed until today (Pratt 2001).
The FindMRO division demonstrates a key element of Graingers success: its
knowledge base. Between its systems and workers, the company is extremely suc-
cessful at fnding parts, including chemicals to fuel Universal Studios fog ma-
chines. This knowledge has a huge value to Grainger. A typical online order is
worth $250 with small proft margins. Orders on FindMRO average $1,200 and 80
percent of the items are shipped directly from the manufacturergiving Grainger
an immediate proft (Sviokla 2001).
Grainger has been adversely affected by the transfer of manufacturing from the
United States to other nations. The company has countered this shift somewhat by
focusing on health care and government agencieswhich cannot easily move off-
shore. At the same time, the company is expanding its purchases from lower-cost
offshore suppliersparticularly for house brands (Top Distributor June 2003).
Logistics Services
Selling to businesses and government agencies is a key component to growth for
Grainger. But Grainger knows that simply offering them products is not good
enough. Grainger has developed sophisticated information tools to help customers
analyze and replace their purchasing systems. The companys Integrated Supply
division is a professional services group that will reengineer a customers stock-
room and provide a just-in-time inventory system.
Like other distributors, Grainger was hurt by the economic downturn of 2001
and 2002. Although sales revenue declined, the company used its supply chain
software to boost its proft margins, resulting in an increase in net income (Konicki
2002). The company also expanded its investments in warehouses and information
technology. Rimnac observes that if you have fnancial resources, a recession is
a good time to invest so when the rebound comes you can excel (Konicki 2002).
After another decline in sales in 2002, yet a continued improvement in profts,
CEO Richard L. Keyser commented that we remain committed to improving ser-
vice to our customers as they continue to look for ways to reduce costs. Enhance-
ments to our logistics network and local availability of the right products will
provide higher levels of service. Our initiatives should accelerate sales growth as
more customers experience this improved service (Lower Sales August 2003).
To combat these changes, Grainger is expanding into additional services. In
2001, the company opened an on-site branch at Florida State University. It fol-
lowed by opening a second on-site branch at Langley Air Force Base in 2002.
In 1996, Grainger took on a more in-depth role at the American Airlines facility
at the Dallas/Fort Worth airport. The company essentially took over all janitorial
and MRO services for the facility. With its success, two years later, the contract
was extended to cover the nearby American Airlines headquarters building. In
2003, the company ran integrated supply programs for more than 40 customers.
Large customers with over $2 million a year in MRO purchases from Grainger are
eligible for the program. Grainger customizes the service for each company, but
essentially, Grainger identifes the inventory needs and handles everything from
ordering to stocking and might even include an on-site center distributing products
to employees. At American Airlines, Grainger was able to substantially reduce the
amount of MRO inventory sitting around. At the same time, facility worker com-
plaints about not having the necessary tools available disappeared (Fraza 2003).
Grainger experimented with several Web sites and e-business approaches be-
fore reaching its current confguration. In 1999, the company had multiple Web
divisions, including OrderZone (a marketplace), FindMRO (a search site for hard-
to-fnd parts), and MROverstocks (an auction site). It also had relationships with
logistics sites such as Ariba and Commerceone as well as several other e-market-
places. Carol Rozewell, vice president research director at research frm Gartner
Inc., noted that Grainger customers were confused. They offered such a wide va-
riety of products, customers needed guidance to navigate [the Web sites] (Mad-
dox 2002). Ultimately, Grainger killed off all of the sites except the companys
main site (grainger.com) and the FindMRO site that is accessed only through the
main site. But Grainger spent more than $180 million on Internet technologies
and took a $23.2 million charge, followed by a $13.4 million write-off of digital
properties (Maddox 2002). The main Web site was redesigned to make it easier
and faster to use.
Internal Systems
Despite the growing importance of Web-based sales and in-house ties to custom-
ers, Grainger is also emphasizing increased sales through its local branches. Push-
ing products more effciently through the local stores is the main reason for the
$200 million redesign of Graingers distribution system (Buss 2002).
One step Grainger took to improve effciency was to install an SAP R/3 ERP
system in 1999. Unfortunately, Grainger had several problems installing and
confguring the system. For example, the ERP software miscounted the items on
handpartly because of problems in the transaction-processing subsystem. In
the frst year alone, the system cost Grainger $19 million in lost sales (Stedman
2000). Ultimately, Grainger got the system fxed and consolidated its fnancial
data onto a single system. One of the problems Grainger had was trying to connect
too many outside products to the SAP system. In 2003, the company began phas-
ing out the old systems and using standard SAP components instead. Jarnail Lail,
vice president of business systems, observed that the move enabled the company
to reduce the number of outside consulting frms from 100 down to 10 (Colkin
Cuneo 2003).
Despite the integrated enterprise resource planning system, Vice President of
Finance Laura Brown wants more. In particular, she wants to be able to match
sales fgures with expense data in real time (Colkin and Whiting 2002). The prin-
ciples of the Sarbanes-Oxley Act are scaring fnancial managers and executives
who want to ensure that the numbers the systems are spitting out are true repre-
sentations of the business fnances.
In January 2006, Grainger implemented an Enterprise Resource Planning Sys-
tem from SAP. The goal was to integrate the information across the channels,
products, and logistics network. CEO Richard L. Keyser in a September 2004
press release stated that for us, technology is not a mere indulgence or a nice to
have. It is at the very heart of our business and helps us drive our multi-channel
strategy. Providing customers access to our broad product line through several
channels has been a competitive weapon for us because it spells speed and conve-
nience for our customers. He also observed that 60 percent of the companys or-
ders came in via telephone, so the company also upgraded its telephone system to
make it simpler to reach people and place orders. The early-phase implementation
of SAP resulted in a $115 million reduction in inventory and a 25 percent increase
in productivity, so Grainger was fnally happy with the system. According to SAP,
the system supports 2 million customer contacts, 800,000 products and 115,000
orders per day (SAP Web site).
Grainger has repeatedly been high in the rankings of Best Places to Work in
IT, compiled by Computerworld. In 2010, the company had 586 IT employees
(and 13,000 total employees), with a turnover rate of only two percent. It probably
helps that James Ryan the CEO was once the CIO at Grainger.
Questions
1. How is Graingers business different from that faced by Owens & Minor?
2. Why did Graingers initial Web approach using multiple sites fail?
3. Graingers problems implementing its ERP system are often cited for being
unusually severe. Why did Grainger experience so many problems?
4. How does Grainger support local stores as well as online sales? Why is that
approach successful for Grainger but not for Dell?
Additional Reading
Buss, Dale, The New Deal, Sales and Marketing Management, June 2002, Vol
154(6), pp. 24-29.
Colkin, Eileen and Rick Whiting, Inadequate IT? Information Week, September
9, 2002.
Colkin Cuneo, Eileen, Less Is More In Software, Information Week, January
13, 2003.
Fraza, Victoria, Making Integrated Supply Work, Industrial Distribution, April
2003, Vol 92(4) p. 69.
Industrial Distribution, Top Distributor Talks Business, June 2003, Vol 92(6),
p. 46.
Industrial Distribution, Lower Sales, Higher Earnings, August 2003, Vol 92(8),
p. 34.
Konicki, Steve, and Eileen Colkin, Optimisms Back! Information Week,
March 25, 2002.
Maddox, Kate, Growing Wiser, B to B, September 9, 2002, Vol 87(9), pp. 1,2.
Pratt, Mary K., Maintenance, Repair And Operations, Computerworld,
February 26, 2001.
Stedman, Craig, ERP Woes Cut Grainger Profts, Computerworld, January 7,
2000.
Sviokla, John, Knowledge Pays, Information Week, February 15, 2001.
Case: Univar: Van Waters & Rogers
Van Waters & Rogers is a respected name in the chemical supply industry. Estab-
lished in Seattle in 1924 by George Van Waters and Nat S. Rogers, the company
grew through acquisitions to become the largest wholesale provider of chemicals
in the United States and Canada. Through international mergers, the company
is also a major (but not the largest) supplier in Europe. In 1973, the company
changed its offcial name to Univar, but kept the esteemed Van Waters & Rog-
ers name on its local units. In 1986, the company acquired McKesson Chemi-
cal to extend its reach across the United States. Starting in 1991, the company
began a series of mergers and acquisitions in Europe, forming Univar Europe.
The rest of the chemical supply industry was consolidating at the same time. In
1996, Royal Pakhoed, a major worldwide player in liquid chemicals, purchased
the shares of Univar it did not already own. In 1999, Royal Pakhoed merged with
its major competitor (Royal Van Ommeren) and created Royal Vopak. The growth
continued with the purchase of Ellis & Everard in 2001, giving Vopak a substan-
tial market share in the United Kingdom and Ireland. However, with the global
economic downturn after 2001, all chemical suppliers were suffering. In 2002,
Vopak spun off the chemical and distribution assets and reestablished Univar as an
independent company. Univar is a Dutch company, listed on the Euronext Amster-
dam stock exchange as UNIVR. Its headquarters is in Rotterdam, but major ad-
ministrative offces remain in Bellevue, Washington (company history Web site).
Largely because of the economic recession of 2003 and the high prices for oil (a
major ingredient for many chemicals as well as a factor in transportation costs),
all of the major chemical suppliers have suffered. Analysts estimate that Vopak
has invested about $1 billion in Univar. Shortly after stock was issued in the new
Univar company, its shares fell to 5.6, which gave the company a valuation of
only about 150 million, less than one-ffth of that billion dollars (Tilton 2002).
Univar has progressed through considerable turmoil with the various mergers
and the fat economies. In the early 1990s, the company tried to centralize manage-
ment and run its distribution centers using a hub-and-spoke system. Analysts have
pointed out that the system worked poorly and sales suffered. With the purchase
by Vopak, the Van Waters & Rogers division was moved back to a decentralized
approach, with more responsibility pushed to regional directors and a focus on
specifc product segments. In the late 1990s, the company moved more operations
out to multiple local warehouses (Morris 1996). The multiple acquisitions during
the late 1990s and early 2000s did not really help sales in the short run. Managers
and investors faced huge uncertainties about what was going to happen next.
Initially, splitting the company back from Vopak appears to have helped. Sales
for 2003 increased 9 percent to 1.62 billion eruos. However, much of that gain
was due to currency effects from the 40 percent Euro appreciation. (In the chart at
the top of this section, Univars sales have been converted to dollars using a sin-
gle exchange rate to eliminate the exchange-rate effects.) Sales in North America
were fat and hampered by the manufacturing slowdown and the rise in oil prices
(Van Arnum 2003). A key step in the mergers was to consolidate the European
operations under a single brand. Managing the company globally instead of re-
gionally is a second imperative. As part of Vopak, the strategy had been to balance
regional storage with distribution. John Phillpotts, president of Univar Europe,
notes that we never operated commercially as a global business before. For the
last 18 months, weve had no distractions about where the company is going or
whos going to own it. Chemical distribution is all we talk about now (Young
2004).
In 2007, Univar began the process of selling to CVC Capital Partners, a Lon-
don-based private equity frm (Berman 2007). In 2007, Univar acquired CHEM-
CENTRAL Corporation and expanded into new industries including coatings,
personal care, and cleaning. Univar servers more than 250,000 client customers
worldwide and leads the industry in chemical distribution (2006 Annual Report).
Much of Univars proft margins come from processing, packaging, and distribu-
tion. Most of the inventory turns over every 40 days. Most of the costs are fxed,
particularly tied up in processing and storage facilities such as the 160 distribution
centers in the U.S., Canada, and Europe. Net sales in 2006 were $6.6 billion with
a net income of $134 million. For four years, sales grew at an 11 percent com-
pound growth rate and earnings grew at 21 percent due to declining costs (2006
Annual Report). Net sales in 2009 were $7.2 billion and Univar had 179 distribu-
tion facilities and 6,900 employees around the world (Web site/annual report).
The company continued its acquisition of smaller distributors in Europe.
E-Business
Univar has followed two approaches to e-business. The company is part of the
consortium that drives Elemica, the chemical industry Web site. Univar also runs
its own B2B Web portal, ChemPoint. The ChemPoint site is also designed to sup-
port smaller chemical manufacturers and highlight new products. In particular,
the site is aimed at providing distribution services to chemical producers. Cus-
tomers place orders through the site, and Univar handles the distributionlinking
electronically to Yellow Transportation, its main less-than-truckload (LTL) car-
rier. Chad Steigers, managing director of ChemPoint.com, notes that there are still
some diffculties getting companies to adopt electronic transactions. The goal of
many frms is to increase productivity. His frm is looking to more technology to
encourage the use of e-business. We think, for example, RFID will be a major
factor especially in the chemical arena (Cottrill 2003). Ultimately, buyers should
be able to place an order, and then track it all the way through shipping, storing,
and delivery.
ChemPoint focused on ensuring a broad market by signing up at least one sup-
plier for each type of chemical, with the goal of providing buyers with a depend-
able source. Reliability and reduction of search costs are important factors for
many of the hard-to-fnd chemicals (Cuny 2001).
When the ChemPoint site was started, it faced intense competition. As Chad
Steigers, managing director, pointed out in 2000, There are 25 trading companies
that offer auctions of chemicals. We think that will drop down to two or three. At
the time, the Web site had 127 customers and sold 1.3 million products provided
by 2,200 suppliers, amounting to $24 million in transactions a year (Seideman,
2000). The emphasis on specialty chemicals and distribution are key elements that
have kept the site active today. Steigers observes that e-distribution is a new seg-
ment for the industry. Were the only ones focused on the less-than-truckload seg-
ment of the specialty and fne chemicals market, which is about a $70-billion/year
niche (Fuller 2000).
Univar also participates in Elemica, an electronic marketplace for the huge
chemical producers and distributors. The site was set up at the end of 2000, pri-
marily to facilitate long-term contracts. It was also designed to assist frms with
fnancing and transportation issues. The decision to avoid spot market auctions
was probably due to the competing systems already in development (most of them
have since failed) (Rosencrance 2000). Because of the emphasis on large sales,
Elemica is emphasizing the ability to connect ERP systems between buyers and
sellers. Larger purchasers can transmit orders, receive confrmations, and transfer
the data directly into their accounting ERP systems (www.elemica.com).
Univar also runs specialty Web sites for certain uses. For instance, the pestweb.
com site is used to provide information and sell products to pest and weed con-
trol companies. To provide wider access to data and ordering, the site created an
iPhone app to provide information on products and the industry. It enables large
and small frms to obtain data and order chemicals directly from their nearest Uni-
var provider (pestweb.com).
The company also wants to build in more relationship-based features to its
Web sites. The goal is to help customers fnd related information and products
that might be useful to them. As CIO Cummings points out, it is challenging for
customers to understand the full breadth of products provided by Univar, so If
we knew a bit more about what the customer does, we could direct them towards
more of our capabilities (Beacham 2010).
Internal Network Technologies
The global nature of Univar makes it more diffcult to manage the company. Even
across the United States, the battles between centralization and decentralization
led to sales problems. Dealing with multiple offces around the world, in a com-
pany that now sees itself as a single global entity, can be diffcult and expensive.
To improve communications without increasing travel costs, Van Waters & Rog-
ers turned to PlaceWares hosted collaboration system (now owned by Microsoft
and renamed LiveMeeting). The system provides voice and video to any desktop
that is connected to the Internet. It also has shared drawing boards, chat facilities,
and the ability to show slides. It can also store a presentation for later playback
if some people cannot make the initial meeting. In one example, Van Waters &
Rogers used it for a strategy meeting between a team in North America and one in
Switzerland to discuss a product rollout (Agnew 2000). The system is also used
for training and general meetings. Ron Miazga, the human resources training di-
rector, notes that the system has been well-received and cost effective. However,
building training sessions takes some practice, since you have to make the pre-
sentations very media intensive. You should ask a polling question every three to
fve minutes. You have to change the screen every minute or so to hold their atten-
tion (Agnew 2000).
In 2006, Univar signed with Odyssey Logistics to handle in-bound deliveries
including transportation and management. Odyssey claims that one reason for the
strategic alliance was the power and capabilities of its Global Logistics Platform
software (OdysseyLogistics.com). The software integrates with commercial and
private carriers of rail and trucks to provide scheduling and tracking. It also deter-
mines rates, availability, and scheduling over the Web. Outsourcing these opera-
tions simplifes management and provides capabilities that Univar would not be
able to create by itself.
Around 2007, Univar replaced its IBM computers with an IBM System z9 en-
terprise system because of the high growth rates and increased server loads. The
companys custom ERP system was designed and built to run on IBM large com-
puters and the company needed a new system that could expand and grow with
future needs. The system can also run virtual Linux servers which could be used
to reduce the number of physical servers Univar runs.
In 2010, Univar hired Dave Cummings as a new CIO. His main experience was
with ConocoPhillips as a program manager with SAP applications (press release).
Univar has an online ordering system that supports standard ordering, and track-
ing, as well as material safety data sheets (MSDS).
Questions
1. Why does decentralization at Univar require more use of networks?
2. Why does Univar pursue a multiple-site e-commerce strategy, and why does
it work for Univar and not for Grainger?
3. Why are some customers reluctant to adopt electronic ordering for
chemicals?
Additional Reading
Agnew, Marion, Collaboration On The Desktop, Information Week, July 10,
2000.
Beacham, Will, Interview with CIO Dave Cummings, March 10, 2010, http://
www.icis.com/Articles/2010/03/10/9341649/e-commerce-technologies-
could-revolutionise-chemicals.html.
Berman, Dennis K., CVC Capital Is Set to Purchase Univar, The Wall Street
Corporate History Web site: http://www.univarcorp.com/ourhistory.htm
Cottrill, Ken, Overcoming Inertia, Traffc World, January 27, 2003, p. 1.
Cuny, Tim and Marvin Richardson, Redefning Business: Proving Grounds,
Information Week, February 12, 2001.
Fuller, John, Vopak Launches ChemPoint.com, Chemical Week, March 15,
2000, Vol 162(11), p. 56.
IBM Case Study, Univar Extends Computing Capabilities with IBM System
z9, http://www.mysiriuszone.com/component/option,com_docman/
task,doc_download/gid,2828/.
Morris, Gregory DL, Pakhoeds VW&R Buy Accelerates Expansion, Chemical
Week, October 16, 1996, Vol 158 (39), p. 60.
Rosencrance, Linda, Twelve Chemical Companies Are Setting Up Online
Exchange, Computerworld, May 17, 2000.
Seideman, Tony, E-Marketplace Vision Collides With Reality,
InformationWeek, June 6, 2000.
Tilton, Helga, Analysts Positive on More Focused Univar, Chemical Market
Reporter, November 18, 2002, Vol 262(18), p. 19.
Van Arnum, Patricia, North American Distributors, Wait and See, Chemical
Market Reporter, November 24, 2003, Vol 264(18) p. FR3.
Young, Ian, European Unifcation Takes Shape at Univar, Chemical Week, May
19, 2004, Vol 166(17), p. 22.
adopted by rivals?
sectors?
210
How do you store and retrieve the vast amount of data collected in a modern company?
How do you ensure that the transaction data is accurate?
Why is the database management approach so important to business?
How do you write questions for the DBMS to obtain data?
How do you create a new database?
How do you create business applications using a DBMS?
What tasks need to be performed to keep a database running?
Why are databases so important in e-business?
How are databases used in cloud computing?
Chapter Outline
Database Management
4
Chapter
Introduction , 211
Relational Databases, 213
Tables, Rows, Columns, Data Types,
214
Data Quality, 215
Data Integrity, 216
Multitasking, Concurrency, and Integrity,
217
Data Volume, 218
The Database Management Approach, 218
Focus on Data, 219
Data Independence, 219
Data Integrity, 219
Speed of Development, 220
Control over Output, 221
Queries, 221
Single-Table Queries, 222
Different Types of Conditions, 230
Computations, 233
Computing Subtotals, 238
Joining Multiple Tables, 241
Examples, 241
Multiple Tables, GROUP BY, and
WHERE, 243
Views, 247
Converting Business Questions to
Queries, 248
Designing a Database (optional), 250
Notation, 252
First Normal Form, 252
Second Normal Form, 254
Third Normal Form, 254
Database Applications, 257
Data Input Forms , 257
Reports, 259
Putting It Together with Menus, 261
Database Administration, 262
Standards and Documentation, 263
Testing, Backup, and Recovery, 264
Access Controls, 264
Databases and e-Business , 264
Summary, 268
Key Words, 269
Exercises, 270
Cases: Pharmaceuticals, 276
211 Chapter 4: Database Management
Eli Lilly and Company
Introduction
How do you store and retrieve the vast amount of data collect-
ed in a modern company? Airlines have a diffcult problem: they need to
track every seat and passenger on every fight. Thousands of people might be try-
ing to book the same fight at the same time, but the system can never sell the
How do you store and retrieve huge amounts of data? Eli Lilly is a giant pharma-
ceutical company. Creating new drugs requires enormous efforts in research. Get-
ting drugs approved takes years and dozens of lawyers. Except for the occasion-
ally unique blockbuster, selling a drug requires delicate, but expensive, marketing.
All three areas beneft from the use of information technology, particularly database
systems. Evaluating chemicals, searching the huge database of existing results, and
tracking progress generate huge amounts of data and require sophisticated analytical
systems. Tracking clinical trials, the paperwork, and the progress of drugs through
various agencies requires nontraditional databases and support. Marketing is a rela-
tively new area for pharmaceutical companies, and they are still looking for ways to
sell such complex products. Sometimes problems arise, such as when Lilly acciden-
tally included the e-mail addresses of 669 subscribers to its prozac.com service in a
message to 700 users.
Reports and
ad hoc queries
Database
DBMS
Applications
Sales and transaction data
Figure 4.1
Without a DBMS, data can be scattered throughout the company, making it more
diffcult to share information. Inconsistent data, duplication and errors are common.
A DBMS collects data from many sources and maintains data through a common
interface. Data defnition, access, consistency and security are maintained by the
DBMS.
same seat to different people. Many other business problems have similar char-
acteristics. All of the accounting and payroll data, sales, and purchases have to
be saved for any company. Database management systems were specifcally de-
signed to solve these problems. As shown in Figure 4.1, the primary elements of a
database system are to collect and store data, produce reports, and provide data to
answer business queries. Today, database systems form the foundation of almost
every business application. What database features do you need as a manager?
How can you retrieve the data stored by the system? How can you create reports?
A database management system (DBMS) is one of the most important tools
in business and MIS. The systems have changed the way that computer applica-
tions are developed, and they are changing the way that companies are managed.
The database approach begins with the premise that the most important aspect of
the computer system is the data that it stores. The purposes of a database manage-
ment system are to provide shared access to the data, answer questions, and create
reports from the data.
A crucial factor with databases is that they can become massive. Several com-
panies such as American Express and UPS have indicated that their databases
contain several terabytes (trillions of bytes) of data. Even small companies deal
with databases with megabytes (millions of bytes) of data. The size of the data-
base greatly affects its performance and the ability of users to fnd the data they
need. Large databases need to be designed and maintained carefully to ensure that
they run properly.
Trends
In the 1960s and 1970s, companies typically built their own transaction-pro-
cessing systems by writing programs in COBOL. These programs consisted
of millions of lines of code. Each program created and used its own set of
fles. As companies expanded, more programs were createdeach with its
own format for fles. Whenever a manager wanted a new report or additional
information, a programmer had to modify the old code or create a complete-
ly new program.
A database management system (DBMS) presents a different approach
to data, reports, and programming. The most important task is to defne and
store the data so authorized users can fnd everything they need. Report writ-
ers and input screens make it easy to enter data and create reports without
relying on programmers. Data is stored in a special format so that it can be
shared with multiple users.
In the early 1970s, E. F. Codd created a fexible approach to storing data,
known as the relational model, that avoided these problems. Today, relation-
al databases are the dominant method used to store and access data. Rela-
tional databases have a query system that enables managers to get answers
to questions without relying on programmers.
Early databases were designed to handle business types of data, such as
customer names and account data. Some modern database systems can store
entire books, pictures, graphs, or even sound clips as types of data. A few
companies are working on object-oriented DBMSs that enable users to cre-
ate their own data types and continue to manipulate and search the data.
Another important characteristic of databases is that they are designed to help
users examine the data from a variety of perspectives. Instead of simply printing
one type of report, they enable users to ask questions and create their own reports.
Figure 4.2 illustrates how a DBMS is used in an organization. It collects data for
transaction processing, creates reports, and processes ad hoc queries for manag-
ers. Figure 4.2 also indicates that databases usually require programmers to defne
the initial database and maintain programs to perform basic operations. The over-
all design is controlled by the database administrator.
A DBMS is complex software that can be purchased separately or it can be em-
bedded into an application. The primary vendors are Oracle (Oracle DBMS), IBM
(DB2), and Microsoft (SQL Server). A few open-source systems also exist, led
by MySQL and PostgreSQL. Today, MySQL is largely controlled by Oracle. The
open source tools generally can be acquired at low or no cost; however, technical
support or maintenance contracts can be purchased. Without the support contract,
you would need skilled developers and administrators to handle all of the sup-
port tasks. Microsoft Access is commonly available with the Offce suite, but it is
limited to relatively small projects. Also, Microsoft is encouraging developers to
switch to SQL Server.
Relational Databases
The goal of a relational DBMS is to make it easy to store and retrieve the data
you need. All data is stored in tables, which consist of columns with rows of
data. Each table has a name and represents objects or relationships in the data. For
instance, most businesses will have tables for customers, employees, orders, and
inventory.
Figure 4.2
MIS employees and databases. The database administrator is responsible for defning
and maintaining the overall structure and data. Programmers and analysts create
applications and programs that collect data and produce reports. Business operations
generate data that flls the database. Managers use the application programs and ask
ad hoc questions of the data.
Program Program
Database Administrator
Programmer
Analyst
Managers
Data
Programs
& Revisions
Database
Management
System
Business Operations
Business Needs
Ad Hoc Queries
and Reports
Data Collection
and Transaction
Processing
Standards, Design, and Control
Besides storing data, a modern DBMS has several useful tools. Input screens
are built to help users enter data. Reports can be created by laying out data, text,
and graphics on the screenoften with no programming. You can get answers
to questions with a query language or even by pointing to tables and data on the
screen. You can establish security conditions by granting or denying access to por-
tions of the data. Most systems include an application generator that can tie input
screens, queries, and reports together with a menu system. A complex application
can be created by typing a few titles on the screen, without writing a single line of
traditional program code.
Tables, Rows, Columns, Data Types
If you understand how spreadsheets work, it is easy to comprehend relational da-
tabases. A single spreadsheet consists of rows and columns of data. Each column
has a unique name, and a row contains data about one individual object. A data-
base consists of many of these tables that are linked by the data they contain.
In a database, each table contains data for a specifc entity or object. For ex-
ample, most companies will have a table to hold customer data. There are certain
attributes or characteristics of the customers that you want to store. In Figure 4.3,
each customer is assigned a unique CustomerID and has a name, address, and city.
In practice, there will be more columns.
Figure 4.3 also illustrates one of the most important features of a database sys-
tem: Relational databases are specifcally designed to allow many tables to be cre-
ated and then combined in interesting ways. If you had only one table, you could
use a spreadsheet or virtually any fling system, assuming it could handle the num-
Figure 4.3
Creating table defnitions. Tables are defned so that they can be linked by common
columns. For a given row in the Orders table, you can fnd the corresponding
customer data by locating the row with the matching phone number. In practice, this
matching is handled by the DBMS query system.
Customer Table
CustomerID Name Address City
12345 Jones 125 Elm Chicago
28764 Adamz 938 Main Phoenix
29587 Smitz 523 Oak Seattle
33352 Sanchez 999 Pine Denver
44453 Kolke 909 West Denver
87535 James 374 Main Miami
Sales Table
SaleID CustomerID Date Salesperson
117 12345 3/3/09 887
125 87535 4/4/09 663
157 12345 4/9/09 554
169 29587 5/6/09 255
ber of rows you needed. However, most business problems involve data stored in
different tables. In the example, customers can place many different orders. Each
order is stored in a separate line in the Orders table.
Notice that the tables are joined or linked by the CustomerID column. The Cus-
tomerID is the primary key column in the customer table. Each row in a table
must be different from the rest; otherwise, it is a waste of space. Consequently,
each table must have a primary key. A primary key is a set of one or more columns
that uniquely identifes each row. If someone gives you a key value (e.g., 12345),
you can immediately locate the appropriate row and fnd the rest of the data for
that entity (name, address, city).
Each primary key value must be unique, but that is hard to guarantee with most
common identifers. In some cases, you might use phone numbers, but what hap-
pens if a customer moves and gets a new phone number? In most cases, it is safest
to have the computer generate identifer values that are guaranteed to be unique.
However, do not expect customers (or salespeople) to memorize these numbers.
The identifers are simply used in the database to ensure there is a way to separate
customers. You can always use names or phone numbers to look up other data for
customers.
More complex tables require multiple columns to identify a row. These keys
are called composite keys. They represent many-to-many relationships. For ex-
ample, a typical order system requires an OrderItem table that contains two col-
umns as keys: OrderID + ItemID. Both columns are keyed because many items
can be ordered at one time (so ItemID is keyed), and an item can be ordered at
many different times (so OrderID is keyed).
Unlike a spreadsheet, each database column can contain only one type of data
at a time. For example, in the Date column you can store only dates. You would
not be allowed to put names or totals in this column. Most relational databases
were designed to hold business types of data. The basic choices are text, dates
(and times), numeric, and objects (graphics, video, and sound). Some systems
enable you to be more specifc about how to store numeric data. For instance,
you might want to store data with only two decimal places for monetary values.
Whenever possible, dates should be stored in a date format instead of text. That
way you can perform arithmetic on the values. For example, a formula like (today
+ 30) could be used to fnd a due date that is 30 days from today.
Data Quality
How do you ensure that the transaction data is accurate? Trans-
action-processing systems can become quite complex. Problems are going to oc-
Reality Bytes: Database Terminology
When E. F. Codd created the relational database model, he deliberately introduced
new terms to describe the way that databases should store information. His terms
are attribute, tuple, and relation. Although Codds terms are precisely defned math-
ematically, they can be confusing. As a result, many people use the slightly easier
words: column, row, and table. Before relational databases, several different terms
were used to refer to the various parts of a database. The problem is that many of
the terms had several defnitions. Common terms include feld, record, and fle. You
should avoid these terms.
cur in any systemespecially because all business organizations change over
time. That means the computer system has to change to match the business. It is
virtually impossible to change the system at exactly the same time as the business,
so problems will occur. Other problems arise simply because the systems are so
complex. Many processes involve humans who make mistakes. If you understand
what types of problems might arise, they will be easier to solve.
Data Integrity
One of the most important concepts in information processing is the issue of data
integrity. Data integrity means keeping data accurate and correct as it is gathered
and stored in the computer system. There is little value in an information system
that contains out-of-date or inaccurate data. A common complaint among shop-
pers today is that stores using bar-code scanners might have a different price in the
computer than the amount displayed on the shelf. It is easy to change prices in the
computer; it is more diffcult to change the signs in the store. Shoppers will feel
cheated if the computer tries to charge them a higher price than the amount listed
on the shelf. Some states, such as Michigan, have passed laws requiring that the
scanned price cannot be higher than the amount listed on the package or display.
Similar errors cause problems when the computer shows more items in stock than
actually exist.
The frst step to ensure data integrity lies in its capture. Each item must be cor-
rectly entered and the complete information recorded. It is sometimes possible to
check the data as it is entered. Item code numbers usually include a check number
that is based on the other digits. In the item code 548737, the frst fve digits add
up to 27, so the number 7 is included as the last digit. If the person or machine
makes a mistake entering one of the digits, they will probably not add up to 7,
so the computer can immediately determine that there is an error. Sophisticated
methods exist to catch mistakes involving more than one digit.
Even with machine entry of data, validity problems can arise. What happens
when a shipment arrives but the receiving department forgets to record it? The
same problem occurs when a clerk holds an item for a customer and does not re-
Reality Bytes: The Government Pays Dead People
It can be hard to track the almost 2.5 million people who die each year in the United
States. Over three years, the federal government sent $180 million worth of beneft
checks to 20,000 people who had died. Although the Social Security Administration
maintains a database of people who have died, it was not routinely shared with other
agencies. The White House Offce of Management and Budget also noted that over
three years, checks valued at $230 million were sent to 14,000 convicted felons
some in jail, some still running. Although the General Services Administration main-
tains an Excluded Parties List System, few agencies actually check it for ineligible
contractors. Of course, these numbers are dwarfed by the $65 billion in erroneous
payments made in a single year through the Medicare and Medicaid programs. Still,
the government is working on building a comprehensive database that will make it
easier for agencies to verify status before checks are sent out.
Adapted from The Wall Street Journal, Database Aims to Prevent Uncle Sam From
Paying Dead, June 18, 2010.
cord it in the computer. Data integrity can be destroyed by indiscriminately allow-
ing people to change the numbers in the computer. It is one of the main reasons
for creating secure computers and controlling access to each piece of information.
Multitasking, Concurrency, and Integrity
A useful feature offered by more sophisticated operating systems is the ability
to perform more than one task at a time. In many situations it is useful to have
several jobs running at the same time. What happens if you are searching a huge
database and your boss calls and asks you for a sales fgure? With a multitasking
computer operating system, you could switch to a new program, look up the num-
ber, and allow the database to continue searching in the background.
If you use a multitasking operating system, it is important that your application
software understand that other applications might be running at the same time.
Each application needs to protect its data fles from concurrency problems. Con-
currency arises when applications attempt to modify the same piece of data at the
same time. If two people are allowed to make changes to the same piece of data,
the computer system must control the order in which it processes the two requests.
Mixing the two tasks will result in the wrong data being stored in the computer.
These problems can be avoided by only using software that was specifcally writ-
ten for multiuser (or multitasking) computers.
Consider the case of a mail-order frm shown in Figure 4.4. On the left side,
customer Sanchez sent a payment on his account. At the same time the clerk be-
gins to process the payment, Sanchez calls a second clerk and places a new order.
The fgure shows what happens if both transactions continue and interfere with
each other. What should the fnal balance be? Does the computer have the correct
number?
To solve this problem, the application program must know that several people
might try to access the same piece of data at the same time. The software locks out
all users except one. When the frst process is fnished, the other users can try to
gain access again. To keep data accurate, applications used by many people at the
same time must be written to handle these concurrency problems. Early personal
Figure 4.4
Concurrency and data integrity. Multiuser and multitasking systems can cause
problems with concurrent changes to data. Two processes cannot be allowed to
change the same data at the same time. A key strength of a DBMS is that it is built to
handle these problems with minimal effort.
Transaction A
Customer Accounts
Sanchez: Balance Transaction B
1.Receive300payment
2.Readbalance(500)
Sanchez: 500
3.Newpurchase(350)
4.Readbalance(500)
5.Subtractpayment
6.Storenewresults(200)
Sanchez: 200
Sanchez: 850 7.Addpurchase
8.Storenewresult(850
computers were designed for only one user, so much of the software did not pre-
vent concurrency problems. Software designed for computer networks generally
handles this issue. When you use this software, you will occasionally receive a
message that says a piece of data you desire is currently being used by anoth-
er person. If you get this message, simply wait for a few minutes and try again.
When the frst person is fnished, you should be able to proceed.
Data Volume
A common problem experienced by a growing business is the increase in the
amount of data or data volume. Consider the huge databases handled by Infor-
mation Resources, which processes data from supermarket checkouts, or United
Parcel Service, which tracks every package every day.
As the business grows, there will be an increase in the number of transactions.
As the price of a computer drops, more applications are placed on the computer.
Additional transactions become computerized. Several problems can be created
from this increase: (1) processing overload or system slowdowns, (2) greater dif-
fculty in making sure the data is accurate, (3) insuffcient storage within the com-
puter system, and (4) data not captured fast enough.
Visa International processes more than 6 billion electronic transactions a
year. By the year 2000, the company was handling 15 billion annual transactions.
There are 18,000 banks offering Visa cards, used by 10 million customers. So
much data is generated on a daily basis that Visa cannot keep transaction data
online beyond six months. All older records are moved to backup storage, making
them inaccessible for additional research or decisions.
Sloppy practices and huge datasets can lead to inaccurate data. As the system
slows down or the computer runs out of storage space, people avoid using it, so
data is no longer up to date. With the increase in volume and the computerization
of new types of data, it is more diffcult for programmers and managers to check
the data. If parts of the computer system are too slow, data may not be captured
fast enough. As a result, some data might be lost. A tremendous amount of infor-
mation is stored in raw data. The raw data could be analyzed to offer new services
or improve the quality of existing services. However, the huge volumes require
too much storage space and too much processing time.
Careful planning is required to avoid these problems. At best, new computers
and storage usually take a month or two to purchase. It could take a year or more
to evaluate and purchase a large, expensive central computer. The MIS depart-
ment would like to forecast the demands that will be placed on the computers at
least a year in advance.
Te Database Management Approach
Why is the database management approach so important to
business? In many ways, the database approach has revolutionized the way in-
formation systems function and altered the way businesses operate. Originally,
all programs handled their own data in separate fles. It took enormous coordina-
tion and documentation to try and make the multiple programs work together. The
DBMS changes everything by focusing on the data instead of the programs. Its
primary purpose is to store and share the data.
Focus on Data
The database management approach is fundamentally different from the older
programming methods. Whenever someone needs a computer application, the frst
step is to identify the data that will be needed. Then a database management sys-
tem is used to store the data. It takes care of storing the raw data, as well as infor-
mation about that data. Everything you want to know is stored within the DBMS,
not in an application program. This situation is illustrated in Figure 4.5. The goal
of the DBMS approach is to collect accurate data and make it available to users.
The system should also minimize unnecessary duplication of data.
Data Independence
Defning the data separately from the programs is called data independence. The
main advantage is that it is possible to change the data without having to change
the programs. For instance, you might want to add a second phone number to the
customer data. With a DBMS, you can make this change, and it will not affect any
of the existing programs. Similarly, the reports can be modifed without having to
change the data. That means when the programmer is called in at 3 A.M., she has
to change only one program. All the other programs will be unaffected. Besides
making the programmers life easier, the database is more likely to be accurate
and there will be less duplication of data.
Data independence means that the data and programs are separate, which
makes it possible to alter the database tables as needed, without destroying the
programs. As the business grows, you might decide to collect additional data, such
as keeping track of sales by salesperson or by sales route. As the company ex-
pands and changes, the underlying tables can be expanded and new tables can
be addedwithout interfering with the existing tables or current programs. Just
be careful to avoid deleting tables or removing entire columns. Of course, as the
business changes, managers will undoubtedly want to modify the reports to add
the new information.
Data Integrity
Data integrity is an important consideration in designing and maintaining databas-
es. Data integrity means that the database holds accurate, up-to-date data. In the
airline case, it means not selling the same seat to two different people. If there are
All data les
Database Management
System
Billing
Program
Invoice
Program
Figure 4.5
DBMS approach. The database management system controls all access to the data.
Programs and queries are controlled by the DBMS. Security, consistency, and
integrity rules are always maintained by the DBMS.
business limits on certain values, the database should force the data entry to abide
by those rules. For example, prices will always be positive numbers. Another in-
tegrity concept is the importance of identifying missing (null) data. Computations
should be able to skip missing data. From a managers viewpoint, an important
integrity design element is naming columns carefully. The columns should have
names and descriptions so that the users understand what is stored in the database.
If a column is simply labeled Revenue, users might wonder if that means revenue
from all products, all divisions, the current year, or perhaps just monthly totals.
All of this information is stored with the database in the data dictionary.
An important component of database integrity is that the data needs to be con-
sistent. For example, consider a table that describes products for sale. Perhaps the
products are grouped into categories (cleaning supplies, paper goods, clothing,
etc.). Each item belongs to only one category. What happens if the categories are
not entered consistently? The Cleaning Supplies category might be entered as
just Cleaning, or maybe as Clean Supplies, or even Cl Sup. These variations
in the data make it diffcult to search the table based on that category because the
user would have to know all of the variations. A good DBMS supports rules that
can be used to minimize these problems. However, when dealing with databases,
it is good practice to be careful when you enter data to ensure that your entries are
consistent.
Speed of Development
It is possible to create an entire database application without having to write a
single line of traditional programming code. As a result, an application can be
built in a fraction of the time it would take to create it by writing COBOL pro-
grams. Studies indicate that most systems can be created 10 times faster using a
DBMSif the data already exists in the database. As the commercial database
productssuch as Oracle, SQL Server, and DB2continue to add features, they
can be used to solve even more complex problems.
Keep in mind that it is possible to use traditional programming tools such as
COBOL in conjunction with the DBMS. If complex reports or complicated cal-
culations are involved, it is sometimes easier to use a traditional programming
Reality Bytes: Government Databases can be Really Big
Take 130 million people, multiply by every healthcare visit, and every diagnoses
and treatment; and you end up with a giant database. As the Federal government in-
creases its involvement in healthcare, it needs to build a giant database to track doz-
ens of aspects of healthcare for individuals and groups. Of course, a giant database
raises giant privacy questions. The government says that it can anonymize the data.
The goal is to track price and quality patterns to focus on groups and not individu-
als. Initially, the database will focus on Federal employees, and include the ability
to evaluate longitudinal effectsessentially tracking treatments over time. Analysts
will have access only to de-identifed data. The inspector generals offce also plans
to use the database to monitor for fraud and waste, which means it needs access to
provider information.
Adapted from Jaikumar Vijayan, Feds Move Toward Health Claims Database De-
spite Privacy Fears, Computerworld, June 16, 2011.
language. These programs retrieve the base data from the DBMS and print their
own reports or store computed values in the database for later use.
One of the most important steps of developing a solution is to break the prob-
lem into smaller pieces. One major piece of any problem is the data. A DBMS
makes this portion of the problem easier to solve. By putting the DBMS in charge
of maintaining the data, keeping track of security, automatically supporting mul-
tiple users, and printing reports, the developer can concentrate on solving the spe-
cifc business problems. By starting from scratch with COBOL, each of these fea-
tures would have to be rewritten for every application that was designed, which
would be expensive.
Control over Output
Another strong advantage of database management systems is their ability to pro-
vide many different views of the output. In fact, a primary objective of the rela-
tional database approach is to store the data so that users can retrieve it any way
they need. The other feature of databases is that they are designed to make it easy
to combine related data. An older programming/fle approach generally limits the
user to using data in only one way.
With a DBMS, output can be created from report writers, which make it easy
to format the data; some systems even draw graphs. The other common method of
retrieving data is to use a query language such as query by example (QBE) or SQL
discussed in the next section. Queries enable managers to search for answers to
questions without using a programmer to write special programs.
Queries
How do you write questions for the DBMS to obtain data? Most
of the time, managers work with databases that have been created by someone
else. You will need to learn how to retrieve data to answer questions. It might be
nice to be able to ask questions in a natural language (such as English), but it turns
out to be hard to make computers understand these questions and you might not
always be certain that the answer is what you asked for. A DBMS provides at least
one method of asking questions and retrieving data. Two common methods are
QBE and SQL. SQL is an international standard method for retrieving data from
database management systems. It is supported by most of the major commercial
relational database management systems. By the way, according to the standard,
the name SQL is just three letters and not an acronym. QBE stands for query by
example and is a visual method of examining data stored in a relational database.
Figure 4.6
Four questions to create a query. You will always have to answer these four
questions. In many cases, there will be only one table (or view), so the second and
last questions are easy.
Four Questions to Create a Query
What output do you want to see?
What do you already know (or what constraints are you give)?
What tables are involved?
How are the tables joined?
You ask questions and examine the data by pointing to tables on the screen and
flling in templates. Queries can only answer questions for which you have col-
lected the appropriate data.
Regardless of the method used to look up information in a database, there are
four basic questions you will answer, as listed in Figure 4.6. It does not matter in
which order you think of the questions. With some methods (such as QBE), it is
easier to choose the tables frst. With other methods (such as SQL), it is some-
times easier to choose the output frst. In many cases, you will switch back and
forth among the four questions until you have all of the components you need. As
you learn more about databases, keep these four questions handy and write down
your answers before you attempt to create the query on the DBMS.
Single-Table Queries
Consider a simple customer table that contains columns for CustomerID, Name,
Phone, Address, City, State, and AccountBalance. Each customer is assigned a
unique number that will be used as a primary key. The AccountBalance is the
amount of money the customer currently owes to our company. The table with
some sample data is shown in Figure 4.7. The tables in these examples are inten-
tionally small to make it easier to understand the concepts. A real table of custom-
ers would contain more columns and thousands of rows of data. But the basic
concepts will remain the same. Each column represents some attribute or charac-
teristic. Each row holds data for a single customer. The CustomerID is the primary
key for this table and it must always be a unique number. In this example, the ID
was randomly assigned by the marketing department when a customer was added
to the database. It is more common to have the DBMS generate unique ID num-
bers, but you have to remember that the values are often random.
Query by Example
Query-by-example systems that were designed for graphical user interfaces
(GUIs) are especially easy to use. Microsofts Access illustrates a common ap-
proach. The basic mechanism is to make selections on the screentypically by
pointing to them with a mouse. You then fll out a template like the one shown in
Figure 4.8. Note that Access has several ways to create queries. The Design (or
Customers
CustomerID Name Phone City AccountBalance
12345 Jones 312-555-1234 Chicago 197.54
28764 Adamz 602-999-2539 Phoenix 526.76
29587 Smitz 206-676-7763 Seattle 353.76
33352 Sanchez 303-444-1352 Denver 153.00
44453 Kolke 303-888-8876 Denver 863.39
87535 James 305-777-2235 Miami 255.93
Figure 4.7
A sample table for customer data. CustomerID is the primary key and is used to
uniquely identify each customer.
QBE) approach is probably the easiest to learn. The Wizard is not very useful. You
should also be able to read the text SQL statements, which are shown later in this
chapter. In the end, the query systems are basically the same. You must always an-
swer the four basic questions: (1) What output do you want to see? (2) What con-
straints are you given? (3) What tables are involved? (4) How are the tables joined
together? This section focuses on a single table and saves the complications of the
last two questions for later.
Figure 4.8
Design: List all of the customers. First select a table, then drag columns you want to
see onto the grid. Checking the Show box ensures that the column will be displayed
when the query is run.
Figure 4.9
Results for the query that selects all rows of the Customers table.
The Design view or QBE approach begins by asking you a hard question frst:
Which tables are involved. Fortunately, you can add tables later, so it is usually
best to focus on one table at a time. For now, you can start a new query and choose
the Customers table. Your next step is to decide what columns you want to see in
the results. In this case, drag the CustomerID, Name, Phone, City, and Account-
Balance columns onto the grid. You can drag multiple columns at a time, double-
click a column name, or select it from the drop-down list on the top of the grid. It
does not matter how you get the column name in place, as long as it is in the list
and the Show checkbox is set, data will be displayed for that column.
Figure 4.10
Design query. List the customers with an AccountBalance of more than $200. Results
of the query can be sorted in Ascending or Descending order. Multiple levels of sorts
are created by selecting additional columns. You will use multiple column sorts when
the frst column contains several identical values. For example, sort by City, Name.
Figure 4.11
Rows that match the conditions. It is easy to check that the AccountBalance is larger
than 200 in each row. Even if your query has more rows, you should check random
rows to ensure you entered the conditions correctly.
You can run the query by selecting the DataSheet view or by clicking the Run
button. For most queries, the DataSheet view is more convenient because you of-
ten want to select the Design view to return and edit the query. Figure 4.9 shows
the results of running the simple query that selects all of the rows in the Custom-
ers table. If you ever want to copy the rows into a spreadsheet or some other docu-
ment, you can click the small gray square at the top-left of the results to select all
of the rows and columns. Selecting the Copy option will put the results onto the
clipboard.
Most of the time, you will want to see only some of the rows of data. For in-
stance, you want a list of customers who owe you the most money. You decide
to restrict the listing to customers who have account balances greater than $200.
With QBE, you enter the appropriate restriction in the criteria row under the de-
sired column as shown in Figure 4.10. Look at the fgure carefully and check the
greater-than sign. You read the condition as: AccountBalance > 200. You can think
of this condition as a flter on the rows that will be displayed. Only rows that meet
the specifed conditions will be displayed.
Figure 4.11 shows the results of running the query. Double-check the data to
ensure that all of the rows meet the condition you specifed. Checking your results
is a good habit to learn. One of the trickiest aspects of queries is that the DBMS
will almost always return somethingbut you need to verify that the rows it re-
turns are actually the rows that you want to see.
You can specify other conditions for the other columns. Placing them on the
same row means they will be interpreted as AND conditions. If conditions are
placed on separate rows, results will be computed for rows that match at least one
of the criteria (OR condition). Figure 4.12 shows the QBE screen, which tells the
DBMS to display the ID, City, and AccountBalance for customers who live in
Denver and have account balances of more than $200.
Figure 4.12
List the customers from Denver with an AccountBalance of more than $200.
Conditions entered on the same row are joined with an And. Conditions on different
rows are connected with a logical Or.
As shown in Figure 4.13, the AND condition creates a flter so that rows are
displayed only if they match all of the conditions on that row. In this example,
the customer must be from Denver and owe more than 200. Because of the small
number of customers, only one person meets both conditions. But, you begin to
see the value of using conditions to limit the rows you want to see. Many busi-
nesses have huge databases, and it is impossible to search through the data by
hand. You can use a simple query to quickly fnd the rows that match almost any
set of conditions.
AND conditions are used to narrow a search. Each time you add an AND con-
dition, you are creating one more hurdle that a row must meet to be displayed.
Sometimes you need to widen your search. In these cases, you use OR condi-
tions to impose an alternate set of criteria. Figure 4.14 shows the query for search-
ing for customers who live in either Denver or Chicago. Notice that there is no
AccountBalance condition. If you wanted to add the AccountBalance condition,
Figure 4.13
Rows that match the conditions. To be displayed, all of the conditions on the row
must be true. In this case, only one customer is from Denver and owes more than
$200.
Figure 4.14
List the customers from Denver or Chicago. Conditions entered on different row are
joined with an Or, which widens the search because rows need to match only one of
the conditions.
you would have to enter it onto both of the criteria rows. With multiple condition
rows, the query processor tests to see if a row meets all of the conditions on one
row to decide whether to display it. If you were to put the AccountBalance>200
condition only on the Denver rowit will apply only to customers who live in
Denver. All of the Chicago customers would be displayedregardless of their
AccountBalance.
Figure 4.15 shows the list of customers who live in Denver or Chicago. Again,
it is important to check the results to ensure that you entered the conditions cor-
rectly. If you accidentally enter the conditions with an AND clause by putting
them on the same row, you would not get any matches. In this database, no cus-
tomer can live in both Denver and Chicago at the same time.
It is also easy to sort the results. For example, you probably want to which
customer owes the most money. Instead of trying to fnd just one customer, busi-
nesspeople often want to see the list sorted by the AccountBalance. That way, if
two people owe close to the same amount, they will both be displayed. Figure
4.16 shows that you can sort the results simply by selecting the sort-order in the
Sort row. In this case, you want to sort the data in descending order (high-to-low)
so that the customers who owe the most money are displayed at the top of the list.
Create and run the query to see which person ends up on top.
You will also encounter situations when you need to sort by multiple columns.
Names are the most common example. In an English-speaking nation, your data-
base might contain many people with the last name Smith. If you sort only by last
name, the rows will still be diffcult to read. Instead, you should sort by LastName
and FirstName. The one catch in QBE systems is that you need to display the
LastName column to the left of the FirstName column so that the query sorts frst
by LastName, followed by FirstName.
SQL
Another method of retrieving data from a DBMS is with the query language SQL.
Although some people fnd SQL more diffcult to learn, it has two advantages.
First, it is a standard language that is supported by many different database sys-
tems, so the commands will work the same in many situations. Second, it is easier
to read than QBE, so it is easier to understand your queries. It is also simpler to
write down on paper because it relies on words instead of pictures or layout. Keep
in mind that SQL requires the same answers as QBEthe main difference is that
you have to type all of the column names instead of dragging them into place.
SQL is a moderately complex language. Fortunately, this chapter uses only a
few simple SQL statements. You can take a database class to learn more SQL
Figure 4.15
Rows that match the Or condition. Customers are displayed who either liven in
Denver or Chicago.
details and understand the true power of the language. If SQL seems complex,
remember that it is considerably easier than trying to write programming code to
retrieve the data. It is even much simpler than trying to build a spreadsheet to fnd
the answers.
The standard command for retrieving data in SQL is SELECT. To be clear,
SQL command words in this book will be written in capital letters with tabs to
separate the words, but you can type them into the computer as lowercase. The
simple form of the command is shown in Figure 4.17. The fve parts are written
on separate lines to make the command easier to read. One of the benefts of SQL
is that you can write down the keywords and then fll in the blanks in any order.
Typically, you can start with the list of columns to display and then fll in the con-
ditions. From these two lines you can fgure out exactly which tables are needed.
In most cases, the best approach is to decide which columns you want to see.
When you read a business question, think about the values that you want to see
that would answer the question. For example, if you want to see a list of custom-
ers that meet some condition, you want the fnal result to include at least their
FirstName, LastName, and possibly the phone number or address. These columns
can be listed in whatever order you want. The column names should be separated
by commas. If you want to see all the columns, you can use the key word ALL or
an asterisk (*). This approach (SELECT *) reduces the typing, but often results in
a huge number of columns. But it is also useful if you cannot remember the names
of the columns or how they are spelled.
To illustrate the power of SQL, try creating a simple query to list all of the col-
umns and all of the rows from the Customers table. Figure 4.18 shows the query
you want to enter. Microsoft Access supports SQL. In fact, all of the QBE queries
you build are actually converted to SQL. Start a new query in Access and choose
Design view. Do not add any tables. Click the SQL button to switch to SQL text
mode. Type the SELECT command shown in the fgure and run the query. You
Figure 4.16
Sorting the results. Select Ascending or Descending on the Sort row for the column
that needs to be sorted.
will receive an error message if you type one of the words incorrectly, such as
forgetting the s in Customers. The results are similar to those in Figure 4.10, but
also include the Street column.
Next, you need to know the name of the table. The SQL command to retrieve
all of the customer data is SELECT * FROM Customers. The result can be sorted
by adding the ORDER BY clause. For example, SELECT * FROM customers
ORDER BY City.
To get a list of customers who live in Denver with account balances greater
than $200, you need to add a WHERE clause. The command becomes SELECT
* FROM Customers WHERE (AccountBalance > 200) and (City = Denver).
Notice the similarity to the QBE command. Of course, with SQL, you need to
remember (and type in) the names of the tables and columns. NULL values and
BETWEEN commands are also available in SQL.
SQL is also useful when you have complex conditions with several AND and
OR connectors. It is easy to enter conditions in the QBE grid, but they can be dif-
fcult to read. Remember that you get completely different results if a condition is
entered on a different line. But, with many conditions spread across several col-
umns, it can be diffcult to see if conditions are on the same line. With SQL, you
simply read the text and verify the AND or OR connectors. Figure 4.19 shows the
query that lists the customers who live in Denver and owe more than $200. Notice
that it is easy to read the AND connector. However, you should always use paren-
theses to isolate the separate conditions. Also, you have to type in the quotation
marks for any text comparisons (Denver).
One nice feature of Microsoft Access is that you can switch back and forth be-
tween SQL and QBE. You can even edit the query in either mode and get the same
results. You can use drag-and-drop to create the query, and switch to SQL to veri-
fy the AND and OR connectors. To see the effect, use QBE to rebuild the query in
Figure 4.16 that lists customers from either Denver or Chicago. Figure 4.20 shows
the text when you switch to SQL view. First, note that Access always uses the full
Figure 4.17
The SQL SELECT command. This command is used to retrieve and display data. It is
the foundation of many of the other SQL commands.
SELECT columns
FROM tables
JOIN matching columns
WHERE conditions
ORDER BY column {ASC | DESC}
SELECT *
FROM Customers
Figure 4.18
Display all columns and rows from the Customers table. You begin to see the power
of SQL when you realize you can instantly get all of this data with only three words
(and a symbol).
name of columns, which includes the table name, such as Customers.Name. This
approach is reasonable because different tables might use the same column name
(such as Salespeople.Name and Customer.Name), and Access needs to keep track
of exactly which column you mean. Second, Access uses too many parentheses in
the conditions, which makes it hard to edit; but the OR connector is easy to see
and you can always return to the QBE grid to change the query conditions.
Diferent Types of Conditions
It is relatively easy to enter conditions on the criteria rows of the QBE grid or to
add them to the SQL WHERE clause. However, to handle more complex cases,
you need to understand all of the comparison operators. Some special operators
also exist that can save you effort when creating queries. As a manager, you need
to be able to convert business questions into database queries. Identifying and
specifying conditions is a key step in this process. Once you understand the ba-
sic elements of the conditions, the real challenge lies in correctly interpreting the
business question.
Figure 4.21 shows the primary comparison operators used in business queries.
SQL does support some advanced techniques, such as the IN ( list ), but you need
to understand the basic operators before trying to tackle the complex ideas. The
standard comparison operators (=, <, >) work exactly as you expect them to work
from basic math. The not-equal operator seems a little tricky (< >), but you get
used to it after a while.
The BETWEEN operator is a nice way to enter a range of values. It is com-
monly used for testing a date because you often need to see if a business date falls
between two specifed days. The condition will match any row where the value
is between or equal to the start and end value. For example, to retrieve rows of
sales data that occurred in January 2009, you would use: WHERE SaleDate BE-
Figure 4.19
SQL query to list customers who live in Denver and owe more than $200. It is
relatively easy to verify the AND connector. Be sure to use parentheses to separate
the conditions.
SELECT Name, Phone, City, AccountBalance
FROM Customers
WHERE (AccountBalance>200) AND (City=Denver)
SELECTCustomers.CustomerID,Customers.Name,Customers.Phone,
Customers.City,Customers.AccountBalance
FROM Customers
WHERE(((Customers.City)=Denver))
OR(((Customers.City)=Chicago));
Figure 4.20
Displaying a query built in Design view. Microsoft Access always uses the full name
of the column which includes the table name. It also puts too many parentheses in the
conditions. But, the OR connector is easy to read.
TWEEN 01-Jan-2009 AND 31-Jan-2009. However, note that date values are
sometimes tricky to enter into a DBMS. The format used here will work on most
systems, but Microsoft Access uses pound signs (#) instead of quotation marks
to signify dates: BETWEEN #01-Jan-2009# AND #31-Jan-2009#. Also, when
you use the BETWEEN command remember that you must always have two val-
ues separate by the AND keyword. Figure 4.22 shows the use of the BETWEEN
clause to obtain a list of the sales that took place in June. Figure 4.23 shows the re-
sults. Always look through some of the rows of data to verify that you entered the
conditions correctly. The DBMS will almost always return results, but you have to
inspect them to ensure it is selecting the rows you want to include.
The Is Null and Is Not Null operators are useful but have an important trick.
Any value that was not entered is automatically assigned a missing value and can
only be found by searching for Is Null. For example, if someone failed to enter the
name of the city the customer lives in, you would search: WHERE City Is Null.
The important trick is that you cannot use City=Null. Most systems will actually
accept the equals sign, but will never return any matching rowseven if the City
value is missing. You can stare at that last command a long time trying to fgure
out what went wrong, until you remember that you must use the IS operator in-
stead of equality. Along the same lines, the condition City= (quoted text with
no value), will also fail. There is a difference between a zero-length string and a
value that was never entered. This difference presents problems if someone hap-
pens to enter blank spaces for a text value. The DBMS will often display missing
values as blank spaces, but they are stored differently internally.
The LIKE command is a useful tool but it can be diffcult to understand at frst.
Essentially, you are specifying a search pattern or flter. Only rows that match
that condition will be displayed. The LIKE command uses two special charac-
Figure 4.21
Query conditions. The standard comparisons are easy to understand. BETWEEN is
most useful for dates but can be used in other cases. Pattern matching with LIKE is
used for simple searches of text data. Be careful to use Is Null when searching for
missing data.
Operator Meaning Examples
= Equals City=Denver
Salary=60000
<
>
Less than
Greater than
Salary < 60000
Sales > 15000
< > Notequal City < > Denver
BETWEEN Between x and y SaleDate BETWEEN 01-Jan-2012 AND 28-Feb-2012
Sales BETWEEN 10000 AND 20000
LIKE Simple pattern
matching % or
* matches any
characters _ or ?
matches one
LastName LIKE J%
ProductID LIKE BL_ _DR_ _ _
Null Missing data City Is Null
NOT Negation Not City = Denver
ters to form patterns. The SQL standard states that a percent sign (%) will match
any other characters, including no characters; and an underscore ( _ ) will match
any single character. Microsoft Access uses an asterisk (*) and question mark (?)
to create the same patterns. You can see the power of pattern matching even for
simple business questions. For example, to list the customers with a last name that
starts with B, you would use: WHERE LastName LIKE B% (or WHERE Last-
Name LIKE B* in Access). The B in the pattern requires the name to have that
letter in the starting position, followed by any other characters (including none).
As another example, the condition WHERE Description LIKE %smart% will
Figure 4.22
The BETWEEN clause. This query uses BETWEEN to retrieve sales that occurred in
June.
Figure 4.23
Date results. Verify that the rows returned match the condition you desire. The
DBMS will rarely tell you if you make a mistake. It will almost always return results,
so you need to check to ensure you entered the conditions correctly.
search for any product that contains the word smart anywhere within the de-
scription. The word smart must exist, but can be preceded or followed by any
other letters. The single-character search is used less often, but can be useful if
you have specially-formatted codes. For instance, many companies create product
codes that describe the product. Perhaps the frst two letters are a color code, fol-
lowed by a size code, a category, and a three-digit identifer, such as: BL09DR293
to represent a blue (BL) dress (DR) of size 9 (09) with an product identifer of
293. A salesclerk could search the inventory for all blue dresses using the pat-
tern: WHERE ProductID LIKE BL_ _DR _ _ _. You need to know the details
of how the product codes are created to build these conditions. In other cases, it
is probably easier to search for items by using other columns of data, such as:
(ProductType=Dress) AND (Color=Blue).
It is generally easy to create any individual condition. The process becomes
more diffcult, with more chances for error, when you are given many conditions
at the same time. The key is to be patient and write one condition at a time and test
the values one at a time. As you will see in the next section, it is important to test
the conditions with a simple SELECT statement so you can ensure that the query
is retrieving only the rows you need.
Computations
Many business questions involve totals or other calculations. All database systems
have some mechanism to perform simple calculations. The calculations are usu-
ally simpler than those found in spreadsheets, but they are easier to use and can
operate on millions of rows of data.
Query processors can perform two types of computations: (1) Compute totals
across multiple rows of data, or (2) Use arithmetic and functions to operate on data
one row at a time. You can also combine these two methods, but it is best to treat
them separately for now. You often need to compute totals or average of business
data. Figure 4.24 shows the aggregation functions commonly provided by SQL
and QBE. SUM, AVG, and COUNT are the most-used functions in business.
SUM computes the total value of a numeric column. Essentially, it looks at
each row of data that matches the WHERE clause and adds up the values stored
in each row. Similarly, AVG works only on numeric columns, but it computes the
simple average. COUNT is slightly trickier. It simply counts the number of rows
of data that would be displayed, so it works with any column or even without a
Figure 4.24
The standard aggregation functions available in SQL and QBE. These functions
operate on multiple rows of data at a time. SUM, AVG, and COUNT are the most-
used functions in business.
SUM total value of items
AVG average of values
MIN minimum value
MAX maximum value
COUNT number of rows
STDEV standard deviation
VAR variance of items
Technology Toolbox: Building Forms in Access
Problem: Every month you have to create a report that uses several new pieces of
data. You have an assistant to help enter data, but need to create a system that is
simple to use.
Tools: Most DBMSs have a forms tool to quickly generate forms. In Microsoft Ac-
cess, the Form Wizard is easy to use, as long as the tables are defned correctly.
It is relatively easy to make
forms to enter and edit data
for a single row at a time. You
can quickly build a form to
edit data for a single customer.
Start the Form Wizard, select
the Customer table and all of
its columns by moving them
to the right-side column. Work
through the remaining selec-
tions by choosing design op-
tions. If you do not like the re-
sult, you can always delete the
form and start over. When the
form is complete, use it to enter
data for a couple of customers.
The Form Wizard can build
more complex forms, such as the Sale form that includes data from the tables: Sale,
Customer, SaleItem, and Item. You need to be cautious in choosing columns. Start
the Form Wizard and choose all columns from the Sale table. Then choose most of
the columns from the SaleItem table, but do not include the SaleID. The SaleItem ta-
ble represents the repeating section. If you include the SaleID, it will waste space by
displaying the same value on every row. Next, choose columns from the Customer
table, but do not include the CustomerID. You should not include the CustomerID
column from the Customer table, because you will not use this form to add new cus-
tomers. Keep your forms relatively simple, so each form has a clear purpose. Finally,
select any desired columns from the Item table (Description, Color, Price, and so
on). Once again, do not include the ItemID column from the Item table, because you
will not use this form to add items. Once the columns have been selected, follow the
prompts to fnish building the form.
To make the form easier to use, you will want to switch to Design View and rear-
range the items on the form so they are grouped in a layout that is easier to read.
Look at some of the problems with the form. To enter a new sale, you will have
to enter the CustomerID. Did you memorize all of the CustomerID values? Instead,
delete the CustomerID text box. Select the Combo box in the Toolbox and click on
the form where the CustomerID box used to sit. Follow the Wizard prompts to select
the Customer table; choose at least the CustomerID and Name columns. On the last
screen, make sure you mark the option to store the result in the CustomerID column.
Test the result. This procedure displays all of the customers in the drop-down box.
When you select a customer, the corresponding CustomerID is transferred to the
CustomerID column in the Customer table.
Quick Quiz:
1. Create a simple customer form and enter data to test it.
2. Create a basic order form and add a combo box to select customers.
column name: Count(*). Although the functions are similar, you must be cautious
when choosing them. It can be diffcult to determine when you should use SUM
instead of COUNT. Figure 4.25 shows the AccountBalance column for the rows
of data in the Customer table. Due to the small amount of data, you can easily
verify that there are 6 rows of data.
Figure 4.26 shows the QBE query and the values returned by the COUNT,
AVG, and SUM functions. A few seconds with a calculator will convince you that
the sum is correct. Always remember that Avg and Sum work only on numeric
data because they compute the total of the values stored within each row. You
have to carefully evaluate any business question to decide if it requires adding
numbers or simply counting rows. The VAR and StDev functions are for simple
statistical calculations and compute the variance and standard deviation of the
data in the chosen column, using n-1 weighting. The older versions of the SQL
query language did not specify standard names for these functions so they are
spelled differently depending on the DBMS you are using. The more advanced
Figure 4.25
AccountBalance column from the Customer table.
Amount
$197.54
$526.76
$353.76
$153.00
$863.39
$255.93
Count Avg Sum
6 $391.73 $2,350.38
Figure 4.26
Count versus Sum. Count returns the number of rows that would be returned with a
SELECT statement. Sum adds up the values within each row.
(more expensive) DBMSs support additional aggregation functions, and you can
read about them in the vendors documentation or help system.
Query by Example
Although most database management systems provide a means to compute totals
and averages, there is no standard method for entering the commands. Typically,
the commands are displayed on a menu. Access uses an extended grid, which is
shown in Figure 4.27. The total row is enabled by clicking the summation (sigma)
button. You then select the desired function on the Total row using the drop-down
list in the appropriate column. The example shows how to get the number of cus-
tomers and the average account balance.
Computations within a row are a little more complicated because you have to
type everything instead of picking from a list. Consider the Items table that lists
all of the products for sale. It contains a Category column and a Price column. You
are told that all items in the Electronics category have a cost that is 70 percent of
the listed price. You want to display the list of items in theElectronics category,
their price, and their estimated cost. Figure 4.28 shows the query. Start a new que-
ry, choose the Items table, and display the Category and Price columns. Enter the
Clothes condition to restrict the rows. Create a new column by typing the formula:
0.7*Price. When you move the cursor to another location, Access will add the
brackets and enter Expr1 as the title of the new column. Change Expr1 to EstCost,
but be careful to leave the colon separator.
Figure 4.27
QBE for aggregation. In Access, you select the table and columns. Clicking the
summation button (sigma) adds the Total row to the grid, where you can pick the
Count and Avg functions.
Figure 4.29 shows the results of running the query. If desired, you can use the
Round or Format functions to alter the calculation or the way it is displayed. The
main thing to remember is that these computations are performed using only the
data on one row. You can use data from multiple columns, but not from different
rows.
SQL
Because QBE actually uses SQL, it should be clear that these same queries can
be written directly in SQL. In many cases SQL is so easy that it is faster to type
the simple SQL commands than it is to work through the QBE grid. Figure 4.30
shows the SQL command to count the number of customers and return the aver-
age amount owed. The functions are easy to use, but notice the use of the AS
Figure 4.28
Row-by-row calculations. You can use basic arithmetic (+, -, *, /) and common math
functions to create a new, computed column in a query. Just type in the formula and
then change the text before the colon EstCost) to be displayed as the new column
heading.
Category Price EstCost
Electronics $1,000.00 700
Electronics $400.00 280
Figure 4.29
Row-by-row calculations. You can use basic arithmetic (+, -, *, /) and common math
functions to create a new, computed column in a query. Just type in the formula and
then change the text before the colon EstCost) to be displayed as the new column
heading.
command to provide a new name for the column. If you leave out this name, the
system will pick something that is unique but meaningless.
Row-by-row calculations are also straightforward in SQL. In fact, because you
have to type the formula even within QBE, it is often easier to type the entire
query within SQL. But, if you have complex WHERE clauses (or many tables),
you might want to build the base query in QBE and switch to SQL to enter the
formula. Figure 4.31 shows the SQL statement to list and estimate the cost of the
items in the Electronics category.
Computing Subtotals
One of the most powerful features of SQL is its ability to create subtotals by add-
ing one line to the SELECT statement. Many business questions require the com-
putation of subtotals. In fact, it is diffcult to fnd business questions that do not
require subtotals. A subtotal is an aggregate calculation (usually SUM) that you
want to perform for all members within a group. Common business examples in-
clude: Find the customer with the most sales, or the employee who sold the most
last month, or identify the product or category that is the best or worst seller. Sub-
totals are also used whenever you want to compute totals by time period, such
as total costs per month or sales this year compared to last year. Each of these
questions requires adding up the value for each element specifed (customer, em-
ployee, product category, month, year, and so on). It is diffcult to compute these
subtotals by hand. It is time-consuming even if you write your own program code.
First you have to sort the data by the selected attribute to group the values togeth-
er, then you fnd the break points where the value switches (such as from James
to Jones). SQL (and QBE) solve this problem with one simple line: GROUP BY.
SQL
It is easier to see how subtotals are created by using SQL. Once you understand
the process, you can use QBE to defne the calculations. The frst point to remem-
ber is that you want to see subtotalsso you have to use the aggregation functions
in the output (SELECT) line. Figure 4.32 shows how to defne a basic subtotal
Figure 4.30
Aggregation functions in SQL. The use of the function is straightforward, but note
the use of the AS command to create a new name for the output column.
SELECT Count(CustomerID) AS NCustomers,
Avg(AccountBalance) AS AverageOwed
FROM Customers
SELECTCategory,Price,0.7*PriceASEstCost
FROM Items
WHERE (Category="Electronics")
Figure 4.31
Row-by-row computations in SQL. Simply type the formula as a new column and
use the AS keyword to give the column a meaningful name.
using SQL. The frst two lines are similar to what you have already done. You are
asking to see the name of the City and the total of the AccountBalance column.
The third line tells the query processor that you want the subtotal computed for
each value of city. The nice part is that you do not need to know which cities ex-
ist in the database. It automatically sorts and groups the data and then reports the
total for each of those cities.
Figure 4.33 shows the total amount due by each city. The query processor au-
tomatically sorts the data by the City column because that is how it computes the
subtotals. More commonly, you would add an ORDER BY line to sort by the sub-
totalso you can fnd the city where customers owe you the most money.
Take a couple of minutes and study the SQL statement. You will need to use it
many times. The query itself is relatively simplejust remember to keep it that
way. A common mistake is to add too many columns or extra tables. For GROUP
BY queries, you must the smallest number of tables and columns possible. The
easiest way to create the query is to think about the output. For any business ques-
tion, decide what you want to see. In this example, you want a list of cities and the
total of the AmountDue for each city, so the SELECT statement contains only the
City and Sum(AmountDue). Both columns are in the Customers table, so you list
that on the FROM statement. The phrases for each or group in any business
questions are clues that you need to add the GROUP BY line.
An important catch with basic SQL is that any query using GROUP BY can
display only subtotals (or other aggregate functions). It cannot display the detail
rows and then the subtotals. If you want both details and totals, you must write
a second query. A newer version of the SQL standard does have the ability to
display both detail rows and subtotals, but it is relatively complex and it is not
supported by Microsoft Access. More importantly, if you truly want to see that
much information, better tools exist to examine the data interactively, and these
are covered in Chapter 9.
SELECT City, Sum(AccountBalance) AS SumOfAccountBalance
FROM Customers
GROUP BY City
Figure 4.32
Subtotals by City. The output line contains a SUM function and the attribute you
want to group. Then add the GROUP BY City line to tell the query parser what you
want to see.
City SumOfAccountBalance
Chicago $197.54
Denver $1,016.39
Miami $255.93
Phoenix $526.76
Seattle $353.76
Figure 4.33
Subtotals by City. Each city is listed followed by the total amount owed by customers
in that city. It is important that you include only these columns and nothing extra.
Customers
CID LastName Phone City AccountBalance
12345 Jones 312-555-1234 Chicago $197.54
28764 Adamz 602-999-2539 Phoenix $526.76
29587 Smitz 206-656-7763 Seattle $353.76
33352 Sanchez 303-444-1352 Denver $153.00
44453 Kolke 303-888-8876 Denver $863.39
87535 James 305-777-2235 Miami $255.98
Salespeople
SPID LastName DateHired Phone Commission
255 West 5/23/75 213-333-2345 5
452 Tomas 8/15/94 213-343-5553 3
554 Jabbar 7/15/91 213-534-8876 4
663 Bird 9/12/93 213-225-3335 4
887 Johnson 2/2/92 213-887-6635 4
Items
ItemID Category Color Description Price
1154 Shoes Red Red Boots $100.00
2254 Clothes Blue Blue Jeans $12.00
3342 Electronics Black LCD-40 inch $1,000.00
7653 Shoes Blue Blue Suede $50.00
8763 Clothes Black Mens Work Boots $45.00
9987 Electronics Silver Blu-Ray Player $400.00
ItemsSold
SaleID ItemID Quantity
117 1154 2
117 3342 1
117 7653 4
125 1154 4
125 8763 3
157 7653 2
169 3342 1
169 9987 5
178 2254 1
Sales
SaleID CID SPID SaleDate Amount
117 12345 887 3/3/2012 $1400.00
125 87535 663 4/4/2012 $535.00
157 12345 554 4/9/2012 $100.00
169 29587 255 5/5/2012 $1800.00
178 44453 663 5/1/2012 $12.00
188 29587 554 5/8/2012 $1180.00
201 12345 887 5/28/2012 $100.00
211 44453 255 6/9/2012 $2800.00
213 44453 255 6/9/2012 $12.00
215 87535 887 6/9/2012 $62.00
280 28764 663 5/27/2012 $1036.00
285 44453 887 6/15/2012 $100.00
Figure 4.34
Multiple tables. The true power of a database lies in the ability to combine data from
multiple tables. Actual databases can have hundreds or thousands of related tables.
Notice that each table is related to another table through matching columns. You
should be able to draw lines between column labels that will connect each of the
tables.
Joining Multiple Tables
The true strength of a database management system lies in its ability to combine
data from several tables. Part of the Customer table is shown in Figure 4.34, with
additional tables that show a list of sales to those customers and the salespeople
involved. Notice that the tables were designed so they can be connected. For ex-
ample, the Sales table can be connected to the Customers table by matching the
CustomerID (abbreviated as CID to save space). The Sales table can be matched
to the Salespeople table through the SalespersonID (SPID). Once you have joined
the tables together, the database system retrieves and displays the data as if it were
stored in one table.
The chief advantage to using multiple tables is that data is stored one time
even for one-to-many relationships. For example, each salesperson may be associ-
ated with many different sales. Instead of repeating the salesperson information
on every order, you only need to include the salespersons ID (SPID) number.
Joining the tables together tells the DBMS to automatically look up the corre-
sponding data from the Salespeople table.
Query by Example
Most people fnd that database systems that use graphical QBE commands to
join tables together are much easier to use than straight SQL commands. With a
DBMS like Access you join the tables together by pointing to the column name in
one table and dragging it to the matching column in the other table. The DBMS
displays the connection between the two columns. Whenever you want to retrieve
data from more than one table, you must frst join them together.
SQL
In SQL, connections between tables are typically made with the INNER JOIN
clause in the FROM statement. For example, to join the Customers and Orders
tables by equal customer numbers and get the combined data, use the command
SELECT * FROM Customers INNER JOIN Orders ON Customers.CustomerID =
Orders.CustomerID
Notice that both tables must be listed in the FROM statement. Always remem-
ber that if you list more than one table, the tables must be joined. The dot before
the column (CustomerID) separates the table name from the column name (table.
column). You can use this form any time you type in a column name, but it is only
required when there might be confusion about which table the column is in. In the
example, both tables have a column called CustomerID. To keep them straight,
you have to specify which table you want to use.
Examples
You now have the basics to begin asking questions of the data. Start with an easy
one. Which customers (CustomerID) have placed orders since June 1, 2012? The
query and result are shown in Figure 4.35. Notice that customer number 44453
has placed two orders. Some systems will show you the order number twice; oth-
ers will automatically delete the duplicates.
It can be diffcult to remember each customers number, so it is better to use the
customer name and have the DBMS automatically look up the customer number.
This second query is shown in Figure 4.36. Note that the Customer table is joined
to the Orders table by the matching values of CustomerID. Be cautious when add-
ing tables to a query and make sure that all of the tables are connected. If you
place two tables in the query with no connection, the query will still runvery
slowly. The query processor will match every row in the frst table with every row
in the second table. For example, if each table has 1,000 rows, the resulting query
would contain 1 million rowsand the results would be meaningless.
SELECT DISTINCT
CustomerID,
Name, SaleDate
FROM Sales
INNER JOIN Customers
ONSales.CustomerID
=Customers.CustomerID
WHERE SaleDate >= #6/1/2012#
CID Name OrderDate
28764 Adamz 6/27/2012
44453 Kolke 6/9/2012
44453 Kolke 6/15/2012
87535 James 6/9/2012
Figure 4.36
Multitable queries. Queries that use more than one table are slightly more complex. Because
columns can have any name, you must tell the database system how the tables are connected.
What are the names of the customers who placed orders since June 1?
SELECT CustomerID, SaleDate
FROM Sales
WHERE SaleDate >= #6/1/2012#
CustomerID SaleDate
44453 6/9/2012
44453 6/9/2012
87535 6/9/2012
28764 6/27/2012
Figure 4.35
QBE and SQL. Which customers have placed orders since June 1, 2006? QBE and SQL are
based on the same fundamental concepts. You build each query by asking the same four basic
questions.
Now, try a more complicated query: List the salespeople (sorted alphabetically)
with the names of the customers who placed orders with that salesperson. This
question sounds diffcult, but the command is easy when you join all three tables
together. The query and the result are shown in Figure 4.37. Notice there is no
entry for the salesperson (Zeke) who has no sales at this point.
Multiple Tables, GROUP BY, and WHERE
The basic concept of multiple tables is straightforwardonce you have chosen
the tables and joined them correctly, you can select columns from any of the ta-
bles. However, be cautious about using extra tables. Each query should use the
minimum number of tables needed to answer the question. Extra tables can create
problems and incorrect answersparticularly when computing totals.
Figure 4.37
Multitable queries with several joins. More complicated queries follow the same basic rules.
Note that some database management systems can automatically switch displays between
QBE and SQL. This feature is useful so that you can check the joins and the criteria to be sure
they are being interpreted correctly. salespeople (sorted alphabetically) along with the names
of customers who placed orders with that salesperson.
SELECT DISTINCTSalespeople.LastNameName,
Customers.LastName
FROM Salespeople
INNER JOIN (Customers INNER JOIN Orders ON
Customers.CustomerID=Sales.CustomerID)
ONSalespeople.SalespersonID
=Sales.SalespersonID
ORDERBY Salespeople.LastName
SalesName Cust.Name
Bird Adamz
Bird James
Bird Kolke
Jabbar Jones
Jabbar Smitz
Johnson James
Johnson Jones
Johnson Kilke
West Kolke
West Smitz
With the ability to use multiple tables, it is now time to look at more realistic
questions. Consider the business question: Who are the top salespeople in June?
The key to understanding this question is to recognize that the answer requires
computing the total sales by employee in the month of June. Begin by creating a
query that lists the sales in June along with the name of the salesperson.
As shown in Figure 4.38, you can add a WHERE condition to control the rows
that will be used in the computations. The other important step is to create the
Value column which multiplies price by quantity. Notice that this query requires
four tables: One for the name of the salesperson, two for the Sales data, and one
for the price of the items.
Figure 4.39 shows the rows from the initial query. Each row displays the sale of
one item, although there can be multiple quantities of that item. Notice that Price
and Quantity are displayed so you can verify that the Value column is computed
correctly. When building queries, it is important to display intermediate results to
Figure 4.38
Subtotals with a WHERE condition. The WHERE condition is applied frst and limits the rows
used to compute the totals.
Figure 4.39
Initial query results to compute the value column. Each row represents the sale of one item.
Price and quantity are temporarily displayed to verify the Value computation.
SalespersonID Name SaleDate Quantity Price Value
255 West 6/9/2012 2 $1,000.00 $2,000.00
255 West 6/9/2012 5 $50.00 $250.00
255 West 6/9/2012 1 $12.00 $12.00
887 Johnson 6/9/2012 1 $12.00 $12.00
887 Johnson 6/9/2012 1 $50.00 $50.00
ensure that computations are specifed correctly. Because a WHERE condition is
also used, the SaleDate values should also be checked to ensure the correct data is
being displayed. If you make a mistake when specifying a WHERE condition, the
only way to catch it is to look at the rows that are returned.
The next step is to remove the Quantity and Price columns and then compute
the total of the Value column. Figure 4.40 shows the SQL statement that answers
the question about total sales by salesperson in June. Note the Sum function in the
SELECT clause which specifes the total computation. More importantly, observe
that the WHERE clause holds the date condition (June). This condition flters the
rows that will be used in the computations. Finally, the GROUP BY clause speci-
Figure 4.40
SQL query to compute subtotals with WHERE condition. Notice that the date
condition is specifed in a WHERE clause because that flters the rows that will be
used in the computation.
SELECT Salespeople.SalespersonID, Salespeople.Name,
Sum([Quantity]*[Price]) AS [Value]
FROM Items INNER JOIN ((Salespeople
INNER JOIN Sales
ON Salespeople.SalespersonID = Sales.SalespersonID)
INNER JOIN ItemsSold
ON Sales.SalesID = ItemsSold.SaleID)
ON Items.ItemID = ItemsSold.ItemID
WHERE (Sales.SaleDate Between #6/1/2012# And #6/30/2012#)
GROUP BY Salespeople.SalespersonID, Salespeople.Name
ORDER BY Sum([Quantity]*[Price]) DESC;
Figure 4.41
Initial attempt to compute subtotals. Click the Totals button and set the Sum option
for the Value column. Notice that the other columns are automatically set to GROUP
BY.
Technology Toolbox: Creating Database Reports
Problem: Your company has a large database that contains the information you
need. However, you need to produce a relatively complex report every month based
on data in that database.
Tools: A database Report Wizard can help you build reports with a visually oriented
tool that makes it easy to create the layout and subtotals that you want.
If you need in-line computa-
tions, such as price * quantity,
it is generally easiest to create a
query that performs these sim-
ple calculations and selects the
columns you will need on the
report. Just create a query, test
the data, and save it as a view.
Then open the Report Wizard
and select the saved view. Oth-
erwise, you can select individ-
ual tables and choose the col-
umns you want from each table.
The key to understanding
reports is to use the groupings
and levels correctly. First, identify the data that you want to see at the most detailed
level. For example, you might want to see individual sale items. In creating the un-
derlying query, make sure you select the tables and columns necessary to display the
detail level you need. Sometimes, you will want to use subtotals in the query, but
many times you will simply let the report writer create the totalsit all depends on
what you want to see at the detail level. The next decision is to identify what group-
ings and levels you need. For instance, you might want to see a list of items ordered
by each customer, so you set a grouping by customer. Alternatively, you might want
to see a list of orders by each customer, so you set groupings for customer and then
for order.
The Form Wizard contains an option to compute totals for the numeric columns.
The button to select this option is easy to miss, so look carefully. Once the Wizard
creates the design, you will still have to set properties to improve the format and
layout of the report. You can also add new text boxes to compute other values. For
example, to create a total, just place a text box in the desired footer and enter the
formula =Sum(Value), using whatever column you want to total.
Creating reports is detail work. It takes time to get the layout and the format right.
It takes a sense of design to make the report look good. Remember that the goal is to
present the important information so that it can be found quickly.
Quick Quiz:
1. Create a report that prints all of the items ordered by each customer.
2. Create a report that prints each customer, followed by the orders for that cus-
tomer.
3. Create a report that displays a chart of total sales by customer.
fes that the totals are to be computed for each sales person. The ID and the name
are both included because there is a chance that two employees have the same
name. Including the ID ensures that employees with the same name will be treated
separately.
Look at the SQL statement again and be sure that you understand the main
points; then you can use the Access design grid to fnish the query. Figure 4.41
shows the initial attempt. Remove the Quantity and Price columns. Use the totals
button to add the Total row and select the Sum option for the Value column. Leave
the other columns at the default GROUP BY option.
Figure 4.42 shows what happens when this initial query is executed. Look at
the ID and Name columns. The question calls for computing totals for each sales-
person, so the query should have only one row for each person. These results con-
tain two rows for ID 255 (West). Why? Look back at the query grid, or better
yet, look at the SQL statement. The query lists three columns as GROUP BY: ID,
Name, and SaleDate. So the DBMS computed exactly what you asked for: The
total value for each salesperson for each day. The second row for West arose be-
cause the sale took place on a second day. If the database contained a complete set
of monthly data, the query would return different totals for every day in the month
for each employee.
The solution to the problem is to remove the SaleDate from the GROUP BY
condition. In the Access grid, this task is accomplished by changing the GROUP
BY entry on the Total row to WHERE. Figure 4.43 shows the fnal query in the
grid. Set the SaleDate row to Where and sort the Value in descending order to
display the highest value at the top of the list. When you change the SaleDate to
the WHERE entry, notice that the display checkbox is unset. Do not attempt to
restore the check markthat would tell the query to display both the individual
detail rows and the totals. But Access, and the standard GROUP BY SQL query,
cannot display details and totals at the same time. Run the query now and you will
see that it returns exactly one row for each employee who participated in a sale in
June. Remember this lesson. When computing subtotals that involve date condi-
tions, you almost always want to move the date condition to the WHERE clause.
The only exception is when you actually want to compute sales by day.
Views
There is one important feature of queries that you will fnd useful. Any query can
be saved as a view. For example, if you have a complex query that you have to run
Figure 4.42
Unfnished query. GROUP BY with SaleDate results in two rows for Salesperson
255one for each day.
SalespersonID Name SaleDate Value
255 West 6/9/2012 $2,250.00
255 West 6/10/2012 $12.00
887 Johnson 6/9/2012 $62.00
every week, you (or a database specialist) could create the query and save it as a
view with its own name. The important point is that the view can now be treated
like any other table. In the example, you might defne a view that combines the
tables for customers, orders, and salespeople, and call it SalesOrders. Then, to get
the total sales from customers in Miami, you run the query on the SalesOrders
view and you no longer have to worry about joining tables because the query has
already performed the step.
The advantage of views is that you can look at the data in different ways with-
out having to duplicate the data. As a manager, you can create complex views so
your employees can look up information using much simpler commands. Think of
a view as a mirror. If you stand in front of a three-way mirror in a clothing store,
you get different views of yourself although there is still only one person.
Converting Business Questions to Queries
The query grid and SQL are relatively easy to use. The real challenge lies in un-
derstanding the business question and correctly converting it into a query. In fact,
this task demonstrates the true role of MIS. One of the most diffcult tasks in
applying information technology and creating systems is to identify the business
needs and translate them into the computer realm. Students who become good at
this translation, can solve problems quicker and become more successful.
The key to translating business questions into queries is to focus on the four
questions posed at the start of this section: (1) What do you want to see? (2) What
constraints are you given? (3) What tables are involved? And (4) How are the
tables joined? And, it is usually easiest if you answer the questions in that order.
In business, the second key is to understand exactly what data is stored in the
database tables. Ultimately, you need to know what all of the tables mean and ex-
actly what type of data is stored in each column. The good news is that once you
have worked for a company for a couple of years, the terms and data will become
Figure 4.43
Correct version of the query grid. The WHERE identifer under the SaleDate column
removes the SaleDate from the GROUP BY clause and places it in the WHERE
clause.
familiar to you. The bad news is that company databases can hold hundreds of
tables and thousands of columns.
For longer queries, you should write down the basic SQL clauses as shown in
Figure 4.44. As you read the business question, you fll in the blanks. Of course,
you will need the relationship diagram or a list of tables in the database so you
know the names of the tables and columns. Consider an example from the Roll-
ing Thunder Bicycle company database. The business question: Which customers
from California spent the most money on race bikes in 2010? The relationship
diagram is shown in the Exercises at the end of this chapter, or you can open the
database and examine the diagram or the tables directly.
The frst step is to decide what you want to see. The business question calls
for information about customers, so you should include at least the customer last
name and perhaps their phone numbers. You have some fexibility to add other
data but keep it simple at the beginning. You also want to see the total amount of
money spent by each customer. The best way to represent that value is to look at
the SalePrice of each bicycle that was purchased. This number ignores taxes and
shipping costs which are not revenue to the frm, but managers might want these
additional values in some cases. You can now write the SELECT clause:
SELECT LastName, Phone, Sum(SalePrice) As TotalSales
Now look at the constraints you are given. Reread the business question and
you should see three constraints: California, 2010, and race bikes. Each of these
becomes a condition in the WHERE clause. Since all three conditions must be met
at the same time, they are all connected with AND clauses. Look through the list
of columns in the Bicycle table and you can fnd the appropriate columns to use
(SaleState, OrderDate, and ModelType). If you are unsure about the data held in
each column, you can display a few rows of data from the table to check the val-
ues. You can now write the WHERE clause.
WHERE (SaleState=CA)
AND (OrderDate BETWEEN #1/1/2010# AND #12/32/2010#)
AND (ModelType=Race)
Because you need to compute the total value for each customer, you must add a
GROUP BY clause. This clause must list each column in the SELECT clause that
is not part of an aggregation function.
GROUP BY LastName, Phone
Now that you know all of the columns needed, you can see which tables were
used. In this case, the Bicycle and Customer tables, so you can write the FROM
clause.
SELECT
FROM
INNER JOIN
WHERE
GROUP BY
ORDER BY
Figure 4.44
Primary SQL clauses. Write these down and fll in the blanks to start a new query.
You can simplify the FROM and INNER JOIN parts by just listing the tables and
drawing connector lines.
FROM Bicycle, Customer
To avoid memorizing the INNER JOIN syntax, you will eventually use the
query grid tool to enter the tables and build the joins graphically. But, you should
still verify that the tables will be connected. On the INNER JOIN clause, draw
lines between the tables to indicate how they are connected.
INNER JOIN Bicycle.CustomerID = Customer.CustomerID
In some cases, you will initially list tables that are not directly connected. For
instance, you might use the Bicycle table and the Components table in a query.
But, remember that every table must be connected when you build a query. In
these situations, go back to the relationship diagram and look for an intermediary
table that connects the ones you need. In this example, the BikeParts table con-
nects to both the Bicycle and Component table, so you must add it to your FROM
clause.
Finally, you can sort the data to organize it to answer the question. In this case,
you want to know which customers spent the most money, so sort the results in
descending order of the total.
ORDER BY Sum(SalePrice) DESC
Notice that this query will give you multiple rows. The best customers will
be listed at the top, but does that really answer the question? Technically, you
could cut off the list to display just the values at the top (SELECT TOP 5 Last-
Name, Phone, ). However, as a businessperson, you probably want to see more
than just the frst row. You generally want to know if there are ties, how close the
second-place person is to the frst, and so on. So, in reality, when a businessperson
asks for the best or worst he or she generally means to sort the list and show
at least the frst few top items in the list.
The other big hint in building queries is that sometimes they get very complex
and it is diffcult to determine how to approach the problem. In these situations,
always start with the parts you know how to do. For example, you generally know
how to enter conditions, so write those down. Then fgure out roughly what you
want to see for output and write those columns down on the SELECT line. Build
the query in pieces and test it to ensure that you are obtaining exactly the rows you
need. If necessary, work with one table at a time and add them one at a time, veri-
fying each step by running the query. If you need computations, verify that you
know how to do them, and add them to the SELECT line. Hold off on the totals
until the end because you cannot tell if the totals are correct when looking at just
one number.
Designing a Database (optional)
How do you create a new database? Database management systems are
powerful tools with the ability to present data in many ways. They are used by
managers to answer many different types of questions. However, this fexibility is
not automatic. Databases need to be carefully designed; otherwise, managers will
not be able to get the information they need. Poor design also leads to unneces-
sary duplication of data. Duplication wastes space and requires workers to enter
the same data several times. Normalization is an important technique to design
databases.
To understand the process of normalization, consider the example of retail
sales. You begin by thinking about who will be using the database and identi-
fying what data they will need. Consider the situation of the salespeople. They
frst identify the customer then record each item being purchased. The computer
should then calculate the amount of money due along with any taxes. Figure 4.45
shows a sample input screen that might be used.
The key design point is that you will need multiple tables to store the data. If
you try to cram all of it into a single table, you will end up with unnecessary du-
plication of data and plenty of problems when you try to delete or insert new data.
Each entity or object on the form will be represented by a separate table. For this
example, there are fve objects on the form: Customers, Salespeople, Items, Sale,
and ItemsSold..
Before explaining how to derive the fve tables from the form, you need to
understand some basic concepts. First, remember that every table must have a
primary key. A primary key is one or more columns that uniquely identify each
row. For example, you anticipate problems with identifying customers, so each
customer will be assigned a unique ID number. Similarly, each item is given a
unique ID number. There is one drawback to assigning numbers to customers: you
cannot expect customers to remember their number, so you will need a method to
look it up. One possibility is to give everyone an ID card imprinted with the num-
berperhaps printed with a bar code that can be scanned. However, you still need
a method to deal with customers who forget their cards. It is usually better to build
a method to lookup customers by name.
The second aspect to understand when designing databases is the relationships
between various entities. First, observe that there are two sections to the form: (1)
the main sale that identifes the transaction, the customer, the salesperson, and the
date, and (2) a repeating section that lists the items being purchased. Each cus-
tomer can buy several different items at one time. There is a one-to-many rela-
tionship between the Sale and the ItemsSold sections. As you will see, identifying
one-to-many relationships is crucial to proper database design.
In some respects, designing databases is straightforward: There are only three
basic rules. However, database design is often interrelated with systems analysis.
Figure 4.45
The order form is used in almost any frm. We need to determine the best way to
store the data that is collected by this form.
In most cases, you are attempting to understand the business at the same time the
database is being designed. One common problem that arises is that it is not al-
ways easy to see which relationships are one-to-many and which are one-to-one
or many-to-many.
Notation
It would be cumbersome to draw pictures of every table that you use, so you
usually write table defnitions in a standard notation. The base customer table is
shown in Figure 4.46, both in notational form and with sample data.
Figure 4.46 illustrates another feature of the notation. You denote one-to-many
or repeating relationships by placing parentheses around them. Figure 4.47 repre-
sents all the data shown in the input screen from Figure 4.45. The description is
created by starting at the top of the form and writing down each element that you
encounter. If a section contains repeating data, place parentheses around it. Pre-
liminary keys are identifed at this step by underlining them. However, you might
have to add or change them at later steps. You can already see some problems with
trying to store data in this format. Notice that the same customer name, phone, and
address would have to be entered several times.
Remember that some repeating sections are diffcult to spot and might con-
sist of only one column. For example, how many phone numbers can a customer
have? Should the Phone column be repeating? In the case of the retail store, prob-
ably not, because you most likely want to keep only one number per customer. In
other businesses, you might want to keep several phone numbers for each client.
Data normalization is directly related to the business processes. The tables you
design depend on the way the business is organized.
First Normal Form
Now that you have a way of writing down the assumptions, it is relatively straight-
forward to separate the data into tables. The frst step is to split out all repeating
sections. Think about the problems that might arise if you try to store the repeat-
ing data within individual cells. You will have to decide how many rows to set
aside for storage, and you will have to write a separate search routine to evaluate
data within each cell, and complications will arise when inserting and deleting
data. Figure 4.48 illustrates the problem.
Reality Bytes: Internationalization: ZIP Codes
Databases often contain addresses (of customers, suppliers, employees, etc.), that
typically use zip codes. In the United States, zip codes typically consist of fve digits,
so it is tempting to set up a ZipCode column that restricts input to fve integers. How-
ever, bulk mail is often cheaper if it uses nine-digit zip codes (zip + 4).
Even more important, if your addresses might someday include international data,
you have to be more careful in column restrictions. For instance, Canadian and Brit-
ish postal codes include alphabetic characters in the middle of the code. Some areas
(such as Hong Kong) do not use any postal codes.
Similarly, when you set up databases that include phone numbers, be sure to allo-
cate enough space for area codes. If international phone numbers will be listed, you
need to add three extra digits on the front for the international country code.
The answer to this problem is to pull out the repeating section and form a new
table. Then, each item purchased by a customer will fll a new row. Figure 4.48
uses the notation to show how the table will split. Notice that whenever you split
Figure 4.46
Notation for tables. Table defnitions can often be written in one or two lines. Each
table has a name and a list of columns. The column (or columns) that makes up the
primary key is underlined.
Figure 4.47
Converting to notation. The basic rental form can be written in notational form.
Notice that repeating sections are indicated by the inner parentheses. If we actually
try to store the data this way, notice the problem created by the repeating section:
Each time a customer checks out a video we have to reenter the phone and address.
CID Last Name Phone Street City Balance
12345 Jones (312) 555-1234 125 Elm Street Chicago $197.54
28764 Adamz (602) 999-2539 938 Main Street Phoenix $526.76
29587 Smitz (206) 676-7763 523 Oak Street Seattle $353.76
33352 Sanchez (303) 444-1352 999 Pine Street Denver $153.00
44453 Kolke (303) 888-8876 909 West Avenue Denver $863.39
87535 James (305) 777-2235 374 Main Street Miami $255.93
SaleID SaleDate CID Name Phone Street ItemID Qty Description Price
117 3/3/2012 12345 Jones (312) 555-1234 125 Elm Street 1154 2 Red Boots $100.00
117 3/3/2012 12345 Jones (312) 555-1234 125 Elm Street 3342 1 LCD-40 inch $1,000.00
117 3/3/2012 12345 Jones (312) 555-1234 125 Elm Street 7653 4 Blue Suede $50.00
125 4/4/2012 87535 James (305) 777-2235 374 Main Street 1154 4 Red Boots $100.00
125 4/4/2012 87535 James (305) 777-2235 374 Main Street 8763 3 Men's Work Boots $45.00
157 4/9/2012 12345 Jones (312) 555-1234 125 Elm Street 7653 2 Blue Suede $50.00
169 5/6/2012 29587 Smitz (206) 676-7763 523 Oak Street 3342 1 LCD-40 inch $1,000.00
169 5/6/2012 29587 Smitz (206) 676-7763 523 Oak Street 9987 2 Blu-Ray Player $400.00
178 5/1/2012 44453 Kolke (303) 888-8876 909 West Avenue 2254 1 Blue Jeans $12.00
188 5/8/2012 29587 Smitz (206) 676-7763 523 Oak Street 3342 1 LCD-40 inch $1,000.00
188 5/8/2012 29587 Smitz (206) 676-7763 523 Oak Street 8763 4 Men's Work Boots $45.00
201 5/23/2012 12345 Jones (312) 555-1234 125 Elm Street 1154 1 Red Boots $100.00
SaleForm(SaleID, SaleDate, CustomerID, Phone, Name, Street, (ItemID, Quantity, Description, Price ) )
Repeating Section
Causes duplication
Table name Table columns

Customer(CustomerID, LastName, Phone, Street, City, AccountBalance)
a table this way, you have to bring along the key from the prior section. Hence,
the new table will include the SaleID key as well as the ItemID key. When a table
contains no repeating sections, you say that it is in frst normal form.
Second Normal Form
Even if a table is in frst normal form, there can be additional problems. Consider
the SaleLine table in Figure 4.49. Notice the two components to the key: SaleID
and ItemID. The nonkey items consist of the Quantity, Description, and Price of
the item. If you leave the table in this form, consider the situation of selling a new
item. Every time an item is sold it will be necessary to reenter the Description and
list Price. It means that you will be storing the description every time an item is
sold. Popular items might be sold thousands of times. Do you really want to store
the description (and other data) each time?
The reason you have this problem is that when the SaleID changes, the item de-
scription stays the same. The description depends only on the ItemID. If the Price
represents the list price of the item, the same dependency holds. However, what
if the store offers discounts on certain days or to specifc customers? If the price
can vary with each transaction, the price would have to be stored with the SaleID.
The fnal choice depends on the business rules and assumptions. Most companies
resolve the problem by creating a list price that is stored with the item and a sale
price that is stored with the transaction. However, to simplify the problem, stick
with just the list price for now.
When the nonkey items depend on only part of the key, you need to split them
into their own table. Figure 4.50 shows the new tables. When each nonkey column
in a table depends on the entire key, the table is in second normal form.
Tird Normal Form
Examine the SaleForm2 table in Figure 4.49. Notice that because the primary key
consists of only one column (SaleID), the table must already be in second normal
form. However, a different problem arises here. Again, consider what happens
when you begin to collect data. Each time a customer comes to the store and buys
something there will be a new transaction. In each case, you would have to record
SaleID SaleDate CID Name Phone Street ItemID, Quantity, Description, Price
117 3/3/2012 12345 Jones 312-555-1234 125 Elm Street 1154,2,RedBoots,$100.00
3342,1,LCD-40inch,$1,000.00
7653,4,BlueSuede,$50.00
125 4/4/2012 87535 James 305-777-2235 374 Main Street 1154,4,RedBoots,$100.00
8763,3,Men'sWorkBoots,$45.00
157 4/9/2012 12345 Jones 312-555-1235 125 Elm Street 7653,2,BlueSuede,$50.00
169 5/6/2012 29587 Smitz 206-676-7763 523 Oak Street 3342,1,LCD-40inch,$1,000.00
9987,2,Blu-RayPlayer,$400.00
178 5/1/2012 44453 Kolke 303-888-8876 909WestAve. 2254,1,BlueJeans,$12.00
188 5/8/2012 29587 Smitz 206-676-7763 523 Oak Street 3342,1,LCD-40inch,$1,000.00
8763,1,Men'sWorkBoots,$45.00
201 5/23/2012 12345 Jones 312-555-1234 125 Elm Street 1154,1,RedBoots,$100.00
Figure 4.48
A table that contains repeating sections is not in frst normal form. Each table cell
can contain only basic data. Storing it in repeating form makes it diffcult to search,
insert, and delete data. This version is not in frst normal form.
the customer name, address, phone, city, and so on. Each entry in the transaction
table for a customer would duplicate this data. In addition to the wasted space,
imagine the problems that arise when a customer changes a phone number. You
might have to update it in hundreds of rows.
The problem in this case is that the customer data does not depend on the pri-
mary key (SalesID) at all. Instead, it depends only on the CustomerID column.
Again, the solution is to place this data into its own table. Figure 4.51 shows the
split. Splitting the table solves the problem. Customer data is now stored only one
time for each customer. It is referenced back to the Rentals table through the Cus-
tomerID. The same rule applies to Salespeople, resulting in the ffth table.
The fve tables you created are listed in Figure 4.52. Each table is now in third
normal form. It is easy to remember the conditions required for third normal form.
First: There are no repeating groups in the tables. Second and third: Each nonkey
column depends on the whole key and nothing but the key.
Note in that if the Customers table contains complete address data, including
ZIP Code, you could technically split the Customers table one more time. Because
ZIP codes are uniquely assigned by the post offce, the city and state could be de-
termined directly from the ZIP code (they do not depend on the CustomerID). In
fact, most mail order companies today keep a separate ZipCode table for that very
reason. For our small retail frm, it might be more of a nuisance to split the table.
Although you can purchase a complete ZIP code directory in computer form, it is
SaleForm(SaleID, SaleDate, CID, Phone, Name, Street, (ItemID, Quantity, Description, Price) )
SaleForm2(SaleID, SaleDate, CustomerID, Phone, Name, Street) Note replication
SaleID SaleDate CID Name Phone Street
117 3/3/2012 12345 Jones (312) 555-1234 125 Elm Street
117 3/3/2012 12345 Jones (312) 555-1234 125 Elm Street
117 3/3/2012 12345 Jones (312) 555-1234 125 Elm Street
125 4/4/2012 87535 James (305) 777-2235 374 Main Street
125 4/4/2012 87535 James (305) 777-2235 374 Main Street
SaleLine(SaleID, ItemID, Quantity, Description, Price) Note replication
SaleID ItemID Quantity Description Price
117 1154 2 Red Boots $100.00
117 3342 1 LCD-40 inch $1,000.00
117 7653 4 Blue Suede $50.00
125 1154 4 Red Boots $100.00
125 8763 3 Men's Work Boots $45.00
157 7653 2 Blue Suede $50.00
169 3342 1 LCD-40 inch $1,000.00
169 9987 2 Blu-Ray Player $400.00
178 2254 1 Blue Jeans $12.00
188 3342 1 LCD-40 inch $1,000.00
188 8763 4 Men's Work Boots $45.00
201 1154 1 Red Boots $100.00
Figure 4.49
Splitting a table to solve problems. Problems with repeating sections are resolved by
moving the repeating section into a new table. Be sure to include the old key in the
new table so that you can connect the tables back together.
ItemsSold(SaleID, ItemID, Quantity
SaleID ItemID Quantity
117 1154 2
117 3342 1
117 7653 4
125 1154 4
125 8763 3
157 7653 2
169 3342 1
169 9987 2
178 2254 1
188 3342 1
Items(ItemID, Description, Price)
ItemID Description Price
1154 Red Boots $100.00
2254 Blue Jeans $12.00
3342 LCD-40 inch $1,000.00
7653 Blue Suede $50.00
8763 Men's Work Boots $45.00
9987 Blu-Ray Player $400.00
SaleLine(SaleID, ItemID, Quantity, Description, Price)
Figure 4.50
Second normal form. Even though the repeating sections are gone, we have another problem.
Every time the ItemID is entered, the description has to be reentered, which wastes a lot of
space. There is a more serious problem: if no one has purchased a specifc item yet, there is no
way to fnd its description or price since it is not yet stored in the database. Again, the solution
is to split the table. In second normal form, all nonkey columns depend on the whole key (not
just part of it).
Sales(SaleID, SaleDate, CustomerID, SalespersonID)
SaleID SaleDate CID SPID
117 3/3/2012 12345 887
125 4/4/2012 87535 663
157 4/9/2012 12345 554
169 5/6/2012 29587 255
178 5/1/2012 44453 663
188 5/8/2012 29587 554
Customers(CustomerID, Phone, Name, Address, City, State, ZIPCode, AccountBalance)
CID Name Phone Street City Balance
12345 Jones (312) 555-1234 125 Elm Street Chicago $197.54
28764 Adamz (602) 999-2539 938 Main Street Phoenix $526.76
29587 Smitz (206) 676-7763 523 Oak Street Seattle $353.76
33352 Sanchez (303) 444-1352 999 Pine Street Denver $153.00
44453 Kolke (303) 888-8876 909 West Avenue Denver $863.39
87535 James (305) 777-2235 374 Main Street Miami $255.93
SaleForm2(SaleID, SaleDate, CustomerID, Phone, Name, Address)
Figure 4.51
Third normal form. There is another problem with this defnition. The customer name does
not depend on the key (SalesID) at all. Instead, it depends on the CustomerID. Because the
name and address do not change for each different SalesID, the customer data must be in a
separate table. The Sales table now contains only the CustomerID, which is used to link to the
Customers table and collect the rest of the data. The same rule applies to Salespeople.
a large database table and must be updated annually. For small cases, it is often
easier to leave the three items in the Customer table.
Database Applications
How do you create business applications using a DBMS? Data-
base systems typically have tools to help build applications. The tools make it
relatively easy to create input forms and reports. Even if you never learn how to
design a database, or become a programmer, you can learn to use these tools to
build small applications and customized reports.
Data Input Forms
Rarely is data entered directly into the databases tables. Instead, input forms are
used to enter some data automatically and to present a screen that is easier for us-
ers to understand. It is common to use colors and boxes to make the screen easier
to read. Input screens can be used to perform calculations (such as taxes). Longer
descriptions and help screens can be included to make it easier for the user to
remember what goes in each column. A sample form from the Rolling Thunder
Bicycles case is shown in Figure 4.53.
Many times, input screens look like older paper forms. Consider a typical or-
der form, which frst collects customer information such as name and address. It
also contains lines for items ordered, descriptions, and prices. These are usually
followed by subtotals and totals. If these forms exist on paper, it is easy to create
them as a DBMS input screen. If you are creating a completely new form, it helps
to draw it on paper frst to get a feel for what you want it to look like.
Most input forms begin as a screen that is empty except for a menu line or
some other help message. Three types of information can be placed on an input
screen: (1) simple text, (2) input blanks, or (3) data retrieved from the database. A
Windows-based DBMS can also include pictures, graphs, sound, and video.
Figure 4.52
Third normal form tables. There are no repeating sections and each nonkey column
depends on the whole key and nothing but the key. This fgure also shows the
relationships between the tables that will be enforced by the DBMS. When referential
integrity is properly defned, the DBMS will ensure that rentals can be made only to
customers who are defned in the Customers table.
Paper forms have labels to tell the user what is supposed to be entered into
each blank. For instance, many paper forms ask for a name: NAME ________.
The label (NAME) tells you what you are supposed to enter on the blank line. A
DBMS input form works much the same way. The frst step is to type in the vari-
ous labels. Move the cursor to a spot on the screen and type in a label or sentence
that will tell the user what needs to be entered.
Most database systems automatically enter some types of data, such as the cur-
rent date. If necessary, users can change the date, but it saves time by enabling
them to press ENTER to accept the displayed value. The same situation holds for
sequential items like order numbers, where the DBMS can automatically generate
each unique order number.
After you have typed in the basic labels, the next step is to add the data-en-
try boxes. Just as you would type a blank line on a paper form, you need to tell
the DBMS exactly what data will be entered by the user. For instance, move the
screen cursor to a position next to the Date label, and then tell the DBMS to enter
data at that point. You will specify the name of the column where the data will be
stored. You can also specify default values. A default value is a value that is auto-
matically displayed by the computer. For the case of the date, the DBMS will let
you enter a name like Date() that will display the current date.
Text/Labels Data Variables
Scrolling Region/Subform Record Selectors
* Subform
* Main
Command Buttons
Figure 4.53
DBMS input forms. Input forms are used to collect data from the user and perform
basic computations. Subforms or scrolling regions are used when there is a one-to-
many relationship.
When a DBMS prints out data, it can be formatted in different ways. You can
control the way the data is displayed by using a format command. A date might be
displayed as 10/24/2012 by entering the format MM/DD/YYYY. There are sev-
eral common date formats; most frms tend to use one standard format. Note that
many European frms use a format that is different from the common ones used in
the United States. Most database software automatically picks up the regional set-
tings from the Windows environment and displays dates in the local format.
The next section of the order form contains basic customer information. This
data is stored in the Customer table, not the Orders table. When you select the
Orders table, you might have to indicate that the Orders and Customer tables are
connected to each other by the phone number. Now, place the text labels on the
screen (customer name, address, etc.). Then place a data entry box after each label.
Next, you can add the Sales table; it is connected to the Orders table by the
order number. Type in the column names for Item#, Description, Price, and Quan-
tity. The DBMS input form will defne this part of the table as a scrolling region
or subform. To users, this subform will behave somewhat like a spreadsheet. They
can see several rows at a time, and keys (or the mouse) will move the screen cur-
sor up and down as users enter data into any row.
The only items entered in the Sales table are the Item# and the Quantity or-
dered. The Description and Price can be found by creating a look-up in the Items
table. If the clerk using this screen types in the item number, the description and
price will appear. With a good DBMS, it is possible to defne a pop-up form or
combo box in case the clerk does not know the number. This way, by pressing a
certain key, a table listing each Item# and Description will be displayed in a win-
dow on the screen. The clerk can then scroll through the list to fnd the item.
Reports
Most of the time, the data listed in one table is not complete enough to help man-
agers make decisions. For example, a listing of a Sales table might provide only
phone numbers, item numbers, and the quantity ordered. A more useful report
would print sales grouped by customer. It would also compute total sales for each
customer. Because this report relies on data from several tables, it is best to base
the report on a view.
Reality Bytes: JPMorgan Chase Needs Better Database Design
In late 2010, JPMorgan Chase bank had a problem with its online site for three days.
For at least 24 hours, customers could not conduct any transactions. The problem
was eventually traced to a third-party authentication system that relied on an Oracle
database. Apparently, some fles in the Oracle database were corrupted and through
mirroring the corruption was spread to the replicated copies. These copies then af-
fected several other systems including Automated Clearing House (ACH) transac-
tions and some loan applications. It also took developers several hours to clean up
the problems because a large amount of extra data was being stored in the authen-
tication database that should have been located elsewhere, including customer Web
usage data.
Adapted from Jaikumar Vijayan, Oracle Database Design Slowed Chase Online
Banking Fix, Computerworld, September 24, 2010.
The view for the sales report example needs four tables. An OrderReport view
is created that joins the Customer table to Orders by CustomerID, Orders to Item-
Sold by OrderID, and ItemsSold to Items by ItemID. The DBMS will have a cre-
ate report option to create the sales report. The report will be based on the Order-
Report view. The report writer consists of a blank screen. You can put simple text
statements anywhere on the page. You also can place data values on the page, and
you can compute totals and make other calculations.
Most reports can be broken into categories. For example, there might be report
titles that appear only at the front of the report (such as cover pages). Other in-
formationsuch as the report title, date, and column labelswill be repeated at
the top of each page. All of these items are called page headers. Similarly, there
can be page footers at the bottom of each page. Reports may also contain group
breaks. For instance, the sales report needs subtotals for each customer, so you
need to break the report into subsections for each customer. Generally, you can
specify several levels of breaks. For instance, you might break each customer or-
der into totals by date. Each break can have a break header, a detail section, and
a break footer. In the example, the customer name is printed on the break header.
There is a detail line that lists the item information. The subtotals are displayed
on the break footers. The report design or layout is illustrated in Figure 4.54. The
report with sample data is printed in Figure 4.55.
To create this report, you frst tell the DBMS that the report will contain one
break based on customer phone number. You also defne the variable Extended,
Figure 4.54
DBMS report writers. Reports are created in sections. The report header is printed
one time at the top of the report. Data in the page header section is printed at the top
of every page. There are corresponding page footers and report footer. Primary data
is printed in the detail section. Data can be organized as groups by creating breaks.
Titles are often printed in the break header with subtotals in the break footer.
which is price multiplied by
quantity. Now you move the cur-
sor to the top of the screen and
type in the titles for the top of
the page. Then place each col-
umn and variable on the report.
You can format each item to
make it look better. For exam-
ple, you might want to format
dates as MM/DD/YYYY so that
all four digits of the year are dis-
played. Similarly, you can add
dollar signs to the subtotals and
totals.
When you have fnished cre-
ating the report, you can print
it. When you print this report,
it should be sorted by customer
name. The DBMS will also en-
able you to print the report so
that it contains data just for one
month. Notice that only fve or
six lines are needed to create
a complex report. Without the
DBMS report writer, it would
take a programmer several hours
to create this report, and it would
be much harder to make changes
to it in the future.
Putting It Together with
Menus
If you are creating a database for
yourself with just a couple of in-
put screens and reports, you can probably quit at this point. On the other hand, for
more complex databases or for projects other people will use, it would be wise to
make the system easier to use. Application generators are tools that enable you
to combine the various features into a single application. The resulting application
can be used by selecting choices from a menu, much like users do with commer-
cial software. The important design feature is that you can create the entire appli-
cation without writing any programming commands.
Consider a simple example. As a manager, you need a sales report printed ev-
ery day that shows the best-selling items. Every week you want a list of total sales
for each employee to bring to your sales meetings. You also send letters to your
best customers every month offering them additional discounts. You want to put
your secretary in charge of printing these reports, but you do not have time to ex-
plain all the details about how to use the database program. Instead, you create a
simple menu that lists each report. The secretary chooses the desired report from
the list. Some reports might ask questions, such as which week to use. The secre-
tary enters the answers and the report is printed.
Figure 4.55
Sample report. Reports are often printed by groups or
breaks with subtotals for each group. With a report writer,
the layout, typefaces, and computations are easy to
change.
The frst step in creating an application is to think about the people who will
use it. How do they do their jobs? How do the database inputs and reports ft into
their job? The goal is to devise a menu system that refects the way they work.
Two examples of a frst menu are shown in Figure 4.56. Which menu is easier for
a clerk to understand? The one that best relates to the job. Once you understand
the basic tasks, write down a set of related menus. Some menu options will call
up other menus. Some will print reports; others will activate the input screens you
created.
Once you know how you want the menu structure to appear, you fll in the
menu templates in the application generator. To create a menu, you type in a title
and fll in the choices. Then you assign an action to each choice. Usually you just
pick from a list of actions and type in specifc data such as the name of the report
and how you want it sorted. When you are fnished, the application generator cre-
ates the application.
Database Administration
What tasks need to be performed to keep a database running?
Managing a database can be a complex job. Often hundreds of choices need to be
made when the database is designed. Someone needs to be in charge of defning
the data, making sure that all useful facts are captured, and managing security for
this valuable asset. Databases have to be evaluated and fne-tuned on a regular
basis. Someone has to keep track of these maintenance changes and decide when
major updates should be installed. A database administrator (DBA) is usually
appointed to manage the databases for the frm. The DBA needs to know the tech-
nical details of the DBMS and the computer system. The DBA also needs to un-
derstand the business operations of the frm.
The database administrator is responsible for all operations involving the da-
tabase. These duties include coordinating users and designers, establishing stan-
dards, and defning the data characteristics. When new programs are created, the
DBA makes sure they are tested and documented. The DBA also schedules back-
ups and recovery, and establishes security controls.
In a few large companies, an additional person known as the data administrator
(DA) is charged with overseeing all of the data defnitions and data standards for
the company. In this case, typically several DBAs are used to monitor and con-
Main Menu
1.SetUpChoices
2.DataInput
3.PrintReports
4.Utilities
5.Backups
Customer Information
Daily Sales Reports
Friday Sales Meeting
Monthly Customer Letters
Quit
Figure 4.56
Designing menus for users. Which menu is easier for a secretary to understand?
When designing applications, you should organize the application to match the
processes users perform.
trol various databases. The DA is responsible for making sure data can be shared
throughout the company.
Standards and Documentation
In any company of moderate size, many different databases will be used by hun-
dreds of workers. These databases were created at different points in time by
teams of employees. If there are no standards, each piece will be unique, making
it diffcult to combine information from multiple databases or tables. The market-
ing department may refer to customers, but management calls them clients. The
DBMS needs to know that both terms refer to the same set of data. Also, someone
has to determine the key values for each table. Consider the Customer table. One
department might assign identifcation numbers to each customer; another depart-
ment might use customers phone numbers, and a third department might use the
customer names. To prevent confusion and to be able to combine information, it is
best for all users to use only one of these methods to identify the customers.
Many standards are related to the database process. It is easier to use a database
if all input screens have similar characteristics. For instance, the base screen might
use a blue background with white characters. Data that is entered by the user will
be displayed in yellow. Similarly, certain function keys may be predefned. ESC
might be used to cancel or escape from choices. F1 might be used for help and F3
to display a list of choices. If each application uses keys differently, the user will
have a hard time remembering which keys do what with which database.
Likewise, it is helpful to standardize certain aspects of reports. It might be nec-
essary to choose specifc typefaces and fonts. Titles could be in an 18-point Hel-
vetica font, whereas the body of reports could be printed in 11-point Palatino. To
provide emphasis, subtotals and totals could be printed in boldface, with single
and double underlining, respectively.
One of the crucial steps in creating a database is the defnition of the data. Many
important decisions have to be made at this point. Besides the issues of what to
call each item, the DBMS has to be told how to store every item. For instance, are
phone numbers stored as 7 digits, or should they be stored as 10 digits, or perhaps
stored with the 3-digit international calling code? Postal zip codes pose similar
problems. The United States uses either a fve-digit or nine-digit zip code, but is
considering adding two more digits. Other countries include alphabetic characters
in their codes. Someone has to determine how to store this information in the
manner that is best for the company.
There are many other aspects of database design that need standards to make
life easier for the users. However, whenever there are standards, there should be a
mechanism to change these standards. Technology always changes, so standards
that were established fve years ago are probably not relevant today. The DBA
constantly reviews and updates the standards, and makes sure that employees fol-
low them.
Even though databases are easy to use, they would be confusing if the design-
ers did not document their work. Picture a situation where you want to fnd infor-
mation about customers but the designers named the table Patrons. You might
never fnd the information without documentation.
Documentation can assume many forms. Most DBMSs allow the designers to
add comments to each table and column. This internal documentation can often
be searched by the users. Many times it can be printed in different formats so that
it can be distributed to users in manuals. Because it is maintained in the database
along with the data, it is easy to fnd. It is also easy for the designers to add these
comments as they create or change the database, so the documentation is more
likely to be current. It is up to the DBA to ensure that all designers document their
work.
Testing, Backup, and Recovery
One advantage of the DBMS approach is that it provides tools such as report writ-
ers and application generators that end users can employ to create their own sys-
tems. Although it is easier for users to create these programs than to start from
scratch, the programs still need to be tested. Corporate databases are extremely
valuable, but only if the information they contain is accurate. It is the responsibil-
ity of the DBA to keep the information accurate, which means that all software
that changes data must be tested.
Most companies would not survive long if a disaster destroyed their databases.
For this reason, all databases need to be backed up on a regular basis. How often
this backup occurs depends on the importance and value of the data. It is pos-
sible to back up data continuously. With two identical computer systems, a change
made to one can be automatically written to the other. If a fre destroys one sys-
tem, the other one can be used to continue with no loss of information. Obviously,
it is expensive to maintain duplicate facilities. Many organizations choose to back
up their data less frequently.
The main point of backup and recovery is that someone has to be placed in
charge. Especially in small businesses, there is a tendency to assume that someone
else is responsible for making backups. Also, remember that at least one current
copy of the database must be stored in a different location. A major disaster could
easily wipe out everything stored in the same building. There are some private
companies that for a fee will hold your backup data in a secure, freproof building
where you can access your data any time of the day.
Access Controls
Another important task in database administration is the establishment of security
safeguards. The DBA has to determine which data needs to be protected. Once
basic security conditions are established, the DBA is responsible for monitoring
database activity. The DBA tracks security violation attempts and monitors who
is using the database. Because there are always changes in employees, removing
access for departed employees and entering new access levels and passwords can
be a full-time job.
Databases and e-Business
Why are databases so important in e-business? Many people still
think of Web sites as simple pages of text with a few images. But e-business re-
quires interaction with customers and the company data. Consequently, most e-
business Web sites are connected to databases. In e-commerce, customers want
to know if a product is in stockthis information is in the database. Similarly,
customer, order, and shipping data have to be maintained and shared throughout
the company. Other e-business sites use databases to provide services, store trans-
action data, and provide search and matching capabilities.
Designing an e-business database is no different than traditional business ap-
plications. However, the technologies for building Web-based applications are
still evolving. Currently, two leading systems are being developed: Oracle/Sun
is championing a Java-based approach known as J2EE, and Microsoft is building
the .NET platform. Databases are the heart of both systems, and both are server-
based technologies designed to build interactive Web sites. Unfortunately, the two
systems are completely independent and incompatible. If you build an application
to run with one approach, you would have to completely rewrite it to use the other
method.
Because the two approaches are so different, if you are building an e-business
system, one of your frst actions is to select one of these systems. In most cases,
the J2EE approach runs on UNIX-based computers, but some versions exist for
Microsoft-based systems. The Microsoft approach will probably only run on Mi-
crosoft servers, but some companies are experimenting with building versions that
run on UNIX systems. Because the two systems are both new, it is diffcult to
evaluate them on technical grounds. The Microsoft approach offers a little more
fexibility with its support for multiple languages. Microsoft .NET also provides
a complete development environment with several easy-to-use tools that make it
relatively easy for beginners to create database-oriented Web sites.
The basic approach is shown in Figure 4.57. Web developers create script pag-
es that interact with the database. When customers request a page, the server ex-
ecutes the associated program script. The script sends queries to the database and
retrieves the desired data. For example, the script might retrieve product descrip-
tions, prices, and in-stock status. The data and images are added to the Web page
and sent to the customer, who sees only the simple results.
Customer
Web Server
Database Server
Results
SQL
Order Form
Description
Prices
Web program script
<HTML>
Text
Data
</HTML>
Figure 4.57
E-business database. When a customer requests a page, the server runs a script
program that interacts with the database by sending queries and formatting the data to
build a new Web page.
Generally, the database runs on a separate serverwhich reduces the load on
the Web server and makes it easier to handles backups and other database mainte-
nance chores. However, with increasingly powerful servers, one server can handle
smaller applications.
Cloud Computing
How are databases used in cloud computing? Several of the pioneers
in cloud computing created their own special-purpose database systems. These
are not relational database systems. For example, Google needs a way to store
and retrieve huge amounts of data quickly. The company built huge sets of paral-
lel servers, and then created a database system called BigTable that is designed
to work over this distributed network. The system works well for storing docu-
ments and has built-in support for versioning. But it works nothing like SQL. For
instance, there are no JOIN commands. Instead, you specify key values for the
row, column, and time and store a single object at that location. BigTable can be
accessed through the Google App Engine application. It is useful for storing com-
plex document-type data. A more generic version of the database known as Ha-
doop is available from several open-source providers, and Facebook has created
(and distributed) the Cassandra database system that works in a similar manner.
Amazon is another early Web company that provides cloud computing ser-
vices. It currently offers three ways to store data: S3, SimpleDB, and Relational
Database Service (RDS). S3 is primarily designed to store and deliver Web pages
and other common Web objects. It is easiest to think of it as a fle service. You
upload fles to the S3 servers, and build links to them on your Web pages. When
users view your pages, the content linked to the S3 service is delivered from the
Amazon network. Because Amazon has distributed servers and high-speed In-
ternet connections, this content is delivered relatively quickly even under heavy
loads. It is best used for storing large media fles (music, video, and so on) that
will be needed by many people. Pricing is based on the amount of data stored
and transferred each month with no fxed costs. Amazon SimpleDB is similar to
Googles BigTable. For starters, you need a programmer to use either one. You
basically store and retrieve data by specifying an item name, an attribute name
and the value. Differences exist between the Google and Amazon systems, but the
Reality Bytes: Salesforce and Database.com
Salesforce.com is often listed as a leader in cloud computingmoving customer
information to the Web. In late 2010, the company released Database.com, a cloud-
based system largely based on Oracles tools to support applications running on any
platform. The system uses application programming interfaces (APIs) to deliver data
to other systems. Essentially, the site becomes a giant database store to hold any
information programmers need for Web-based services. The basic free version sup-
ports three users, 100,000 records and 50,000 transactions per month. Each set of
100,000 rows costs $10 per month, and each set of 150,000 transactions carries a
charge of $10. They system does not really use a relational data model, which reduc-
es its fexibility, but simplifes data storage and retrieval for some types of problems.
Adapted from Chris Kanaracus, Salesforce.com Unveils Database.com, Comput-
erworld, December 7, 2010.
underlying concepts are similar. More recently, Amazon introduced a cloud-based
relational database service. It is largely based on the MySQL DBMS and uses the
same syntax. Amazon has also indicated that it will implement an Oracle 11g da-
tabase as well.
Microsoft sells the online service Azure which is a complete implementation
of SQL Server as cloud computing. The system is built on multiple servers run by
Microsoft administrators and your database can scale up to relatively large sizes.
It is relatively easy to convert applications built using standard Microsoft devel-
opment tools to run on the cloud servers.
The main beneft to using cloud computing for databases is that you do not
have to run the hardware or software. That means you do not need to hire people
to maintain and update everything, and you do not have the initial fxed costs. You
are free to concentrate on the data and the applications. With most cloud database
services, pricing is based on usage rates. If your company is starting up and you
have small amounts of data and few users, monthly prices will be low. As the
company grows and data increases, you will pay higher fees. But, you should be
generating more revenue to cover those fees. And, most services include econo-
mies of scale so the rates decrease when you hit certain levels.
Cloud based databases are ideal for building Web-based applications. The re-
liability and scalability provided by the big service providers ensures that your
applications will be available even as demand increases. Yet, the fexible pricing
keeps the prices low during the start-up or low-use periods. Figure 4.58 shows an
estimate of costs using Amazon RDS for a mid-size database. The transfer values
are estimated to be about equal to a full T1 utilization rate. The point is that the an-
nual cost is less than 10 percent of the cost of hiring a single IT worker to run the
systemand one worker would not be able to provide 24-hour support. Another
approach is to look at the cost of the T1 lineprobably $300-$400 a month or
half of the lease cost alone. Server and network hardware costs (3-year life) would
easily make up the other half of the costs. So even if you ignore personnel costs, it
is unlikely that a similar system could be confgured in-house for this price.
On the other hand, the costs continue to increase even as your company be-
comes large. At some point, it becomes cheaper to pay the fxed costs of buying
and running your own hardware, software, and networks. It is up to you as a man-
ager to monitor the operating costs and compare them to the costs of running your
Figure 4.58
Amazon RDS (MySQL) estimated cost. Calculation is from Amazons Web site
calculator. Values are estimated and might not include all costs. But the example
provides a mid-size database with T1-level data transfer for less than 10 percent of
the cost of a single DBA.
Amazon RDS (MySQL), U.S. East
1 Extra large instance
20 hours/day
20 GB/month at 50 million I/O per month
10 GB/month data transfer in
500 GB/month data transfer out
20 GB/month regional transfer
Estimated cost: $616 per month or $7400/year.
own facilities. The cloud computing providers have expertise and specialize in
providing their support, so it will be expensive to acquire that same level of exper-
tise. Also, as the number of cloud providers increases, the competition should help
to hold down prices. Documentation for the various services along with pricing
models are available online, so it is straightforward to estimate the costs based on
the amount of data and the monthly data transfer rates.
Summary
Everyone needs to search for information. Computers make this job easier, but
someone must set up and maintain the databases to ensure their integrity. Rela-
tional database systems are increasingly used as the foundation of the information
system. They make it easy to share data among users while maintaining access
controls. Equally important, the databases are easy to alter as the organization
changes. Sophisticated databases can handle all the data types in use today, not
just simple numbers and text.
It is relatively easy for users to obtain data using SQL or query-by-example
tools. Because SQL is a recognized standard query language, it is worth remem-
bering the basic elements of the SELECT command. The syntax is easy (SELECT
columns, FROM tables, WHERE conditions, ORDER BY columns). Just remem-
ber that whenever you use more than one table, they must be joined by related
columns.
An important step in databases is to design them correctly. The trick is to split
the data into tables that refer to exactly one concept. Most organizations have a
database administrator to help users create the initial database tables, defne stan-
dards, establish access rights, and perform backups and testing. Once the tables
have been defned, users can create input screens, reports, and views by using
graphical tools to draw the desired items on the screen.
It is important to choose the right tool for each job. Databases excel at handling
huge amounts of data and sharing it with other users. On the other hand, spread-
sheets are designed to perform calculations and create graphs. One indication that
a problem should be solved using a DBMS instead of a spreadsheet is that several
tables of data are involved.
Several database systems, using different technologies, are available as cloud
computing services. These services provide scalability in the amount of data
stored and transferred across the Web, with almost no fxed costs. But they can
be expensive for some applications so managers must compare the in-house fxed
costs to the total variable costs of using a cloud service.
A Managers View
Every business has to store data. Every manager needs to do research. Some-
times you will have to summarize and evaluate transaction data. Sometimes
you will use external databases to evaluate the industry and your competi-
tors. Database management systems provide important capabilities to man-
agers. One of the most useful is a query language, such as QBE or SQL, that
enables you to answer questions without the need for hiring an MIS expert.
A DBMS also speeds the development of new systems and provides basic
features such as report writers and input forms.
Key Words
Web Site References
Database References
Computerworld data Management www.computerworld.com/s/
topic/173/Databases
eWeek Database www.eweek.com/c/s/Database/
Car Shopping Services
America's Automall www.automall.com
Autobytel www.autobytel.com
AutoNation www.autonation.com
AutoTrader www.autotrader.com
Carmax www.carmax.com
Dealer Net www.dealernet.com
Edumnds www.edmunds.com
Kelley Blue Book www.kbb.com
Vehix www.vehix.com
Review Questions
1. How is data stored in a relational database?
2. What are the main benefts of using databases to build business applications?
3. What four questions do you need to answer in order to create a database
query?
4. How do you join tables with QBE? With SQL?
5. How do you enter restrictions or constraints with QBE? With SQL?
6. How do you perform computations with QBE? With SQL?
7. Would you prefer to use QBE or SQL to access a database? Why?
8. What tasks are performed by a database administrator?
9. Why are forms (input screens) important in a relational database?
application generator
BETWEEN
break footer
break header
break
column
composite key
concurrency
data independence
data integrity
database administrator (DBA)
database management sys-
tem (DBMS)
default value
detail section
LIKE
normalization
one-to-many
page footer
page header
primary key
QBE
query by example
row
scrolling region
SQL
table
view
10. Do you think business users can create their own reports with a DBMS report
writer? What limitations does it have?
11. Why are standards important in a database environment?
12. Why is a DBMS important for e-business?
13. How are non-SQL databases different from relational databases?
Exercises
C04Ex15.mdb
It is best to answer the frst 15 exercise questions using a DBMS, but if one is
not available, you can use the tables in the text and write your queries by hand. If
you have a DBMS that handles both QBE and SQL, you should do the exercise
with both methods. The number at the end of the question indicates the degree of
diffculty.
1. List the customers who live in Chicago. (1)
2. List the items where the color is Blue. (1)
3. List the salespeople with a commission rate greater than 4%. (1)
4. List the customers with a last name that begins with the letter S. (1)
5. How many customers have an account balance of more than $400? (2)
6. Which customers have purchased a Blu-Ray player? (3)
7. List all of the customers of salesperson West. (3)
8. List all of the items purchased by Customer Kolke. (3)
9. Which salesperson made the most sales by count? (3)
10. What was the total value of sales in May? (4)
11. Compute the commission owed to each salesperson in June based on the
Sales Amount. (4)
12. What is the total amount of sales value in June? (4)
13. Which item category had the highest sales value in May? (4)
14. Write queries to test if the Amount value in the Sales table is correct for all
sales. (5) Hint: Save a separate query to perform the computations.
15. Which items had no sales in June? (5) Hint: This query requires a LEFT
JOIN or a subquery, which are not covered in this book.
16. You have been hired by a small county government offce to track properties
for tax purposes. The annual property tax rate is set by the local governments,
but the assessors offce is responsible for tracking the value of each property.
Whenever a house is sold, the sale price is reported to the offce. Since this
price is a concrete measure of the price, it has to be recorded permanently. In
other cases, the assessor estimates the value of a property from its primary
features, such as the lot size (measured in fractions of an acre), house size
(square feet), age (year built), neighborhood, and general condition. New
values are estimated every year if a house has not been sold. If a property has
not sold for more than fve years, an assessor performs a simple inspection of
the property and takes pictures from the road. Property owners can appeal an
assessment to a board. The appeal date and any comments are recorded. The
data is currently recorded on a form organized by each property. Recently,
the offce has been assigning GPS coordinates (latitude, longitude, altitude)
to each property. Many do not have data yet but will be added when the
properties are inspected. Create a list of normalized tables need to build this
database.
Property ID
Year Built Construction Price
Year Remodel
Address
Tax Code Area
Zone Code
Subdivision Township/Range/
Section
Latitude
Longitude
Altitude
Lot size
House size
Sewer
Water
Utilities
Bedrooms
Bathrooms
Fireplaces
Garage
Basement
Property Description
Year Valuation Method Comments
Appeals
Date Claimed Value Resulting Value Comments
17. A friend has asked for your help in training for a triathlon. No, you do not
have to run with her at 5 AM. She wants a personal log to track her progress
and well-being. She has been writing entries in a journal, but she wants to be
able to plot the data and compute totals. The basic layout of her existing log
is shown in the form. For each day, she tracks the distance and time of each
run, bike ride, and swim. In all three cases, distance is recorded in miles; but
often includes fractions. Eventually, when she plots the data, she wants the
ability to exclude some days. For instance, one day her bike broke and she
had to run home, so the time would not match with the others. She usually
records her basic meals and estimates the total calories for the day. For the
health category, she estimates the number of hours of sleep she gets at night,
but the other entries are subjective comments and notes about the day.
Date Run Bike Swim Food Health
Sunday Dist.
Time
Comment
Dist.
Time
Comment
Dist.
Time
Comment
Breakfast
Lunch
Dinner
Snacks
Calories
Sleep
Illness
Comments
18. Using your experience in a job, school, or any organization; identify a
problem that could be supported with a database management systemthat
requires at least three separate tables. Sketch a sample form.
19. Think about business or Web-based applications and identify a task that you
believe could best be supported by a non-SQL DBMS. Explain why.
Technology Toolbox
20. For the Sales database, create input screens for Items, Customers, and
Salespeople. Add an ItemCategory table that holds a list of the categories
so you can add a combo box to the Items form to select from the list of
categories.
21. Using a different design, rebuild the Sales form.
22. Add a button to the Sales form that enables users to open the Customer form
to edit the data or enter a new customer.
23. Create an inventory report that lists all of the products; group them by
category; and, within each category, sort them by price.
24. Create the customer sales report that is described in the chapter.
25. Create a start-up form that can be used as a menu. Begin in Design View and
add buttons that open the other forms and reports. Use colors or images to
enhance the appearance of the form.
Teamwork
26. As individuals, each person should investigate a cloud-based DBMS in terms
of capabilities and pricing. Combine the results as a team and select one tool
that you would choose for a start-up Web-based company that anticipates
starting with 1,000 customers a month and increasing to 10,000 customers a
month within two years.
27. Assign the tables in the Sales database to each person on the team. As a
group, identify the primary users of the system. Each person should then
specify the security access rights for each user on each table: SELECT,
INSERT, UPDATE, or DELETE.
28. Assume that you need to buy a DBMS to create a Web-based company.
Research the components needed and have each person fnd information
and evaluate a DBMS package. Try to identify costs as well as strengths and
weaknesses of the package. Share the individual results and create a report
that makes a recommendation.
29. With the cooperation of a local small business, create a database for that
company. Note that you should verify the initial layout of the tables with
your instructor or someone who has studied database design. Assign specifc
forms and reports to individual team members and combine the pieces.
30. Each team member should write up three business questions related to either
the C04Ex15.mdb or Rolling Thunder database. Exchange the questions with
the other team members, and then create the queries to answer each question.
Share your answers.
CustomerID
Phone
FirstName
LastName
Address
ZipCode
CityID
BalanceDue
Customer
CustomerID
TransDate
EmployeeID
Amount
Description
Reference
CustomerTrans
StoreID
StoreName
Phone
ContacFirstName
ContactLastName
Address
Zipcode
CityID
RetailStore
State
TaxRate
StateTaxRate
SerialNumber
CustomerID
ModelType
PaintID
FrameSize
OrderDate
StartDate
ShipDate
ShipEmployee
FrameAssembler
Painter
Construction
WaterBottle
CustomName
LetterStyleID
StoreID
EmployeeID
TopTube
ChainStay
HeadTubeAngle
SeatTueAngle
ListPrice
SalePrice
SalesTax
SaleState
ShipPrice
FramePrice
ComponentList
Bicycle
CityID
ZipCode
City
State
AreaCode
Population1990
Population1980
Country
Latitude
Longitude
City
ModelType
Description
ComponentID
ModelType
Paint
EmployeeID
TaxpayerID
LastName
FirstName
HomePhone
Address
ZipCode
CityID
DateHired
DateReleased
CurrentManager
SalaryGrade
Salary
Title
WorkArea
Employee
SerialNumber
TubeID
Quantity
BicycleTube
ModelType
MSize
TopTube
ChainStay
TotalLength
GroundClearance
HeadTubeAngle
SeatTubeAngle
ModelSize
LetterStyle
Description
LetterStyle
PurchaseID
EmployeeID
ManufacturerID
TotalList
ShippingCost
Discount
OrderDate
ReceiveDate
AmountDue
PurchaseOrder
SerialNumber
TubeName
TubeID
Length
BikeTubes
SerialNumber
ComponentID
SubstituteID
Location
Quantity
DateInstalled
EmployeeID
BikeParts
PurchaseID
ComponentID
PricePaid
Quantity
QuantityReceived
PurchaseItem
ManufacturerID
ManufacturerName
ContactName
Phone
Address
ZipCode
CityID
BalanceDue
Manufacturer
CompGroup
GroupName
BikeType
Year
EndYear
Weight
Groupo
ComponentID
ManufacturerID
ProductNumber
Road
Category
Length
Height
Width
Weight
Year
EndYear
Description
ListPrice
EstimatedCost
QuantityOnHand
Component
ManufacturerID
TransactionDate
EmployeeID
Amount
Description
Reference
ManufacturerTrans
TubeID
Material
Description
Diameter
Thickness
Roundness
Weight
Stiffness
ListPrice
Construction
TubeMaterial
GroupID
ComponentID
GroupCompon
ComponentName
AssemblyOrder
Description
ComponentName
PaintID
ColorName
ColorStyle
ColorList
DateIntroduced
DateDiscontinued
Create queries to answer the following questions. (Diffculty level: 1 = easiest.)
31. List all race bikes that sold in 2009 with a sale price greater than $5000. (1)
32. List all the purchase orders in 2008 with a total value greater than 200,000.
(1)
33. List the employees hired after January 1, 1995 with a salary grade higher
than 10. (1)
34. List the road bikes purchased in December 2009 with a FrameSize greater
than 50 cm. (1)
35. List the paint colors introduced after 1995. (1)
36. List the retail stores in Delaware. (2)
37. List the customers (name and phone) who bought full suspension mountain
bikes in December 2008. (2)
38. List all the race bikes ordered in 2010 that were painted with red in the color
list sold to the state of New York. (2)
39. List the manufacturers who supplied mountain bike (MTB) forks in 2010. (2)
40. List the employees who painted race bicycles in January 2010. Hint: change
the relationship from EmployeeID to Painter. (2)
41. Compute the number of bicycles sold by state for 2009. (3)
42. Which letter style was the most popular for Road bikes in 2010? (3)
43. What was the average discount (ListPrice SalePrice) given on mountain
bicycles in 2009? (3)
44. Which manufacturer did RT buy the most components from by value in
2010? (3)
45. Which employee sold the most race bikes in March 2010 by count? (3)
46. By count, what was the most popular crank installed on full suspension
mountain bikes in 2010? (4)
47. How many bikes were sold each month in 2010. Hint use the function
Format(OrderDate, yyyy-mm). (4)
48. For 2009, was the average price of mountain bikes lower than the average
price of full suspension bikes? (4)
49. What was the average effective sales tax rate (tax/SalePrice) by state in
2010? (4)
50. In 2009, did race bikes have more Cranks made by Campy or Shimano? (4)
51. Which customers who have purchased mountain bikes have also bought
Road bikes at any time? (5) Hint: Create two separate queries and save them.
52. Which customers purchased two or more bikes in at least two different years?
(5)
Additional Reading
Disabatino, Jennifer, NHL Scores With Database On Draft Day,
Computerworld, July 9, 2001. [NHL uses a database to provide more
information and cut hours off the time to draft players.]
Fox, Prim, Tax Filing Gets Connected to the Web, Computerworld, August
8, 2001. [The importance of connecting databases to the Web even for
governments.]
Mayer, Merry, New DNA Database Extends The Long Arm Of Law
Enforcement, Government Computer News, October 19, 1998, p. 47.
[Government database to identify criminals.]
Post, Gerald, Database Management Systems: Designing and Building
Applications, Irwin/McGraw-Hill, 2002. [How to design databases and use
them to build business applications.]
Sliwa, Carol, Lee Copeland, and Don Tennant, Dead voters in Florida?
Computerworld, November 13, 2000. [Only 10 states have an online central
voter registration database, and 16 have no state database at all.]
Trombly, Maria, Schwab Database Glitch Cuts Users Off From Some
Information, Computerworld, February 13, 2001. [The importance of
database availability and reliability.]
Tiboni, Frank, FEMA Automates Property Inspection Scheduling, Government
Computer News, November 9, 1998, p. 14. [Emergency agency uses
telephone registration and database to help disaster victims in less time.]
Cases: Pharmaceuticals
Te Industry
The pharmaceutical industry has received considerable attention in the past few
years. Several organizations, including the AARP, have criticized the industry for
price increases on name-brand drugs that exceed the infation rate. The AARP
reported that prices have increased by 6 to 7 percent a year in 2002 and 2003
(Kaufman and Brubaker 2004). Prescription drugs often form a signifcant per-
centage of household budgets for elderly retirees. Yet the pharmacy industry notes
that identifying and testing a new drug is expensive. On average, it takes 10 to
15 years and costs $800 million to bring one new drug to the market. And only
one in fve potential drugs makes it through the entire process. The industry spent
an estimated $33.2 billion on R&D in 2003 (PhRMA 2004). On the fip side, the
pharmaceutical companies spend about as much on marketing as they do on R&D
(fnancial statements). In the same report, the industry observes that 10.5 percent
of total health care dollars in the United States are spent on prescription drugs.
0
10
20
30
40
50
60
70
80
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
B
i
l
l
i
o
n

$
Annual Revenue
Pfizer
GlaxoSmithKline
Bristol-Myers Squibb
Eli Lilly
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
R
a
t
i
o
Pfizer
GlaxoSmithKline
Bristol-Myers Squibb
Eli Lilly
Drug Prices
The combination of high research costs, high risk, and the need to make a proft
makes it diffcult to set public policy. Profts encourage additional research and
development to create new drugs. High prices keep the drugs out of the reach of
some people who need them. Some governments have imposed price controls in
an attempt to keep prices lower for patients. Canada presents an interesting ex-
ample. Because trade with Canada is relatively unrestricted, U.S. patients have
turned to Canadian pharmacies to fll their prescriptions. Even some state and lo-
cal governments have tried to save costs by purchasing drugs from Canada, where
the prices are held down by government rules (Schiavone 2003). Drug companies
have staged an all-out battle to prevent the purchase of drugs from other countries.
GlaxoSmithKline went as far as stopping sales to Canadian pharmacies (CNN
2003).
In 2004, the U.S. government introduced a Medicare insurance policy for pur-
chasing name-brand prescription drugs. Participants have to sign up for the plan
and pay a specifed amount. The policies are sold by private companies, and the
terms vary. Essentially, the policy covers a certain level of drug costs. However,
with the copayments, limits on total value, and rising prices, it is not clear how
many people will be able to beneft from the plan.
Research and Development
Research and development is a diffcult process. Many times, luck and accidents
play a role in creating new drugs. Yet companies need a huge infrastructure to take
advantage of that luck. Research requires people and labs. Increasingly, it also
needs information technologyin the form of databases, analytical tools, and col-
laboration software.
Finding a potential drug is only the frst step (research). Development requires
extensive testing, fnding ways to deliver the drug with minimal side effects, and
manufacturing it cheaply and consistently to narrow tolerances.
A pharmaceutical frm has to recover all costs within the life of a patented drug.
Once the patent expires, the drug can be manufactured generically, and the invent-
ing frm no long makes any money. Since 2000, several of the leading frms have
suffered substantial declines in revenue as their patents expired on several block-
buster drugs.
An interesting twist on pharmacy research arose in 2004. A couple of frms,
particularly GlaxoSmithKline, were charged with withholding research data. All
potential drugs have to go through clinical human trials to identify the specifc
gains and evaluate potential side effects. The drug companies provide the results
of the studies to the Federal Drug Administration (FDA), which has to give fnal
approval to the drug. In at least some cases, negative research was withheld. Many
of the trials are somewhat secretiveneither patients nor physicians can be al-
lowed to know which patients have the specifc drugs. Consequently, it is relative-
ly easy to bury negative results and only publish the positive ones. Of course, easy
is not the same thing as ethical. As a result of substantial publicity in 2004, several
agencies, companies, and people are pushing for a national registry of clinical tri-
alsbefore they are started. This way, everyone can check on the progressboth
good and bad (Martinez 2004).
Genetics is playing an increasing role in drug discovery and testing. Several
specialist frms (e.g., Genentech) have been struggling for several years to use ge-
netic modifcations techniques to create completely new drugs. A few of the drugs
have been successful, but the sub-sector has not yet matched its early promises.
Marketing
Pharmaceutical frms have substantially pushed the boundaries of drug marketing
in the last decade. Originally, the frms relied on publications in medical journals
to highlight the success of their drugs. Of course, sales representatives made di-
rect calls to physicians to point out the results, offer comparisons with existing
drugs, and answer basic questions. These activities are still important to the indus-
try. However, the industry changed enormously when it began marketing drugs di-
rectly to consumers. The FDA has relatively strict rules about the contents of these
ads, and it routinely monitors the ads and requires companies to alter the ads.
Nonetheless, physicians are now bombarded with questions from patients about
a new drug they saw on television. Of course, patients rarely read the fne print
telling about the side effectsand few can make comparisons among the vari-
ous drugs. So, physicians have to spend time explaining their choices. The phar-
maceutical industry points out that increased communication between doctor and
patient is a good thing. But, there remain many unresolved questions about the
effectiveness of the campaignsother than to boost pharmacy company profts.
Increasingly, pharmaceutical frms are being asked to track drugs and ship-
ments more closely. California and Florida both passed laws that will require the
industry to track the entire supply chain of drugs from the manufacturer to the
wholesaler and retailer. Californias law takes effect in January 2009. The security
company VeriSign has established a database that can be used to track ownership
of drugs, and sales will not be allowed until the seller can verify its ownership
of the drug. AmerisourceBergen, a drug wholesaler, is working with the pharma-
ceutical companies to develop the system. In the early stages, the companies are
considering the use of RFID tags instead of simple bar codes to make it easier to
read the data (Weier 2007). Many Schedule 1 drugs, such as the addictive pain-
killer OxyContin, are already tracked with RFID tags. Pfzer also announced that
to reduce thefts and diversions to drug dealers, it would begin tracking bottles
of Viagra (Carr and Barrett 2005).
By 2011, pharmaceutical frms began changing their approach to marketing.
The frms were facing declining profts, loss of patents, and the fact that physi-
cians do not like being interrupted. By 2010, about 25 percent of physicians put
themselves on do-not-see lists and about 75 percent of sales rep visits did not
result in face-to-face meetings. So tens of thousands of sales reps were eliminated,
and largely replaced with Web sites, iPad apps, and other digital tools. Older-style
Web sites were discontinued in favor of interactive sites that enable physicians to
ask questions, or get information from online reps (Whalen 2011).
Information Technology Needs
The pharmaceutical industry runs on data, information, and knowledge. The re-
search departments generate and need to scan enormous amounts of data. They
experiment and produce papers that disseminate knowledge. Researchers need to
collaborate and share this data and knowledge. But the size of the companies, the
variety of products, and the geographical dispersion all add complications. Ad-
ditional data comes from physicians and even customers. On top of the research
data, the marketers need to track salespeople, physician contacts, and marketing
plans. The legal staff needs to monitor progress, evaluate patent documents, and
shepherd thousands of documents through the FDA approval process. On top of
all that, the CFO and CEO have to run the businesswatching the fnancial data
and creating the standard business reports.
Additional Reading
Carr, David F. and Larry Barrett, Philip Morris International: Smoke Screen,
Baseline Magazine, February 1, 2005.
CNN Online, GlaxoSmithKline Stops Selling Drugs To Canadian Pharmacies,
January 21, 2003.
Kaufman, Marc and Bill Brubaker, Study: Drug Prices Outpace Infation, The
Washington Post, May 25, 2004.
Martinez, Barbara, Spitzer Charges Glaxo Concealed Paxil Data, The Wall
Street Journal, June 3, 2004.
Pharmaceutical Research and Manufacturers of America (PhRMA),
Pharmaceutical Industry Profle 2004, Washington, DC: PhRMA (http://
www.phrma.org/publications/publications//2004-03-31.937.pdf).
Schiavone, Louise, Governments Eye Canada For Cheap Drugs, CNN Online,
October 16, 2003.
Simons, John, When Big Pharma Gets Too Big, Fortune, December 4, 2006.
Weier, Mary Hayes, Pharma Industry Debating Bar Codes Vs. RFID For Drug
Tracking, Information Week, March 28, 2007.
Whalen, Jeanne, Drug Makers Replace Reps with Digital Tools, The Wall
Street Journal, May 9, 2011.
Case: Eli Lilly and Company
Eli Lilly (ticker: LLY) is one of the leading pharmaceutical companies. Founded
in 1876, the frm now has more than 46,000 employees (www.lilly.com). Like
other pharmaceutical frms, Lilly is dependent on revenue from blockbuster drugs.
When patent protection expires for a blockbuster drug, as it did for Lillys Prozac
in the early 2000s, revenue and profts drop. The challenge for companies is to fll
the pipeline with new drugs that will provide a new revenue stream. Research and
development are key elements in this process. But fnding new blockbuster drugs
that solve a major medical problem is not easy. Lilly, like the other frms, is turn-
ing to biotechnology to help create and test new drugs. Lillys genetically engi-
neered insulin has been one of the frst successes in this new line of research. Lilly
is also hoping to capitalize on the erectile dysfunction sales by heavily advertising
its Cialis drug directly to consumers.
R&D is critical to generating profts. But R&D has become considerably more
complicated. Research is no longer performed by one or two scientists working
alone in a lab. Instead, Lilly has thousands of researchers working on various proj-
ects. Each of the compounds being created needs to be shared across the company.
Collaboration has become a critical success factor. Throw in the fact that most of
the large companies have expanded through mergers. Now, how do you combine
the systems and the people to provide effective collaboration? Charles Cooney,
codirector of the Program on the Pharmaceutical Industry at MIT, observes that
R&D is useless unless it can be integrated into a bigger picture that helps you
convert leads to drug candidates. The diffculty for the CIO is getting the team to
work between the various disciplines needed to integrate the information in order
to create knowledge (Overby 2002).
Genetics is an increasingly important factor in pharmacy, and it adds many
complications. Lilly, as well as the other frms, is looking at a future where drug
selection may be tailored to a patients DNA. Some drugs are more effective when
specifc genes are present. This level of knowledge will require a tremendous
amount of information fowboth in R&D and in marketing.
In 2006, Lilly took an interesting step by outsourcing some data management
and data analysis to Tata Consultancy Services (TCS) in India. TCS is well known
for providing off-shore consulting for programming and other IT services. But
this contract calls for scientifc analysis using a facility dedicated to working on
data analysis for Lilly. Steve Ruberg, Lillys group director for global medical
information sciences said that the agreement includes gaining access to a global
talent pool, increasing fexibility and scalability of our resources, and maintaining
a global workfow that is operational 24 hours a day (McDougall 2006).
Information technology with its connections to customers provides a new way
to evaluate drug effects. All drugs go through scientifc trials where physicians
carefully monitor patients and evaluate side effects. But testing on a few hundred
or even a thousand subjects does not always provide complete information. His-
torically, drug frms have relied on physicians to evaluate and report on effective-
ness and side effects. Lillys drug Strattera was the frst nonstimulant treatment for
children with attention-defcit hyperactivity disorder. On the basis of early trials,
the company knew that sleepiness could be a side effect. To evaluate the drug
when it was released to customers, Lilly designed a customer relationship man-
agement system to track customer contacts through its call center. With the feed-
back from actual customer use, the company was able to identify that the sleepi-
ness was minimized if the drug was taken at a meal and at night. Sidney Taurel,
Lillys CEO, notes that weve seen that IT makes you more effective. Its a big
shift. Weve changed to seeing IT as an enabler of effectiveness (Murphy 2004).
Big shift is right. When Lilly hired Tom Trainer as its frst CIO in 1994, he
found a messwith 17 different IT organizations and minimal use of comput-
ers for business. By the time Trainer left in 1999 to head global IT for Citigroup,
the technology had improved considerably. But IT was still largely used to re-
duce costs. CFO Charlie Golden says, We had gone through quite a change in
our perspective and strategy in IT. It was important to try to sort through what it
would take to continue to change, but also to make IT a more integral part of our
strategy (Ewalt 2003). Roy Dunbar was chosen as the new CIO. He began as a
pharmacist, had some business experience, but little experience with information
technology. With some intense education, and a lot of humility, Dunbar learned
the foundations of information systems. His ultimate charge was to fnd ways to
use technology to support the strategy of the frm. For that, he needed everyone to
start with a business perspective. One of the key tools his team created in 2000 was
the Molecule Library. It is a knowledge management system that makes it easy for
researchers to fnd information about chemical compounds being worked with in
the companys pipeline. The system reduces research time from days to minutes.
Another new tool, the Sample Identifcation Database, functions as a registry of
all compounds being developed. The Gene Anatomy Made Easy (Game) project
helps scientists identify DNA analysis without requiring expert help. Shaving a
few hours a week off the research time of every scientist can help get projects to
market earlier, saving billions of dollars in costs (Ewalt 2003).
In 2001, the IT department began expanding its use of SAP (its enterprise re-
source planning system) to improve connections to suppliers and customers.
Combining the companys shipping, billing, and sales data makes it easier to track
sales and keep on top of billing problems. The company also implemented a sales
management system to provide better information to salespeople.
Coming from the pharmacy side of the business, Dunbar knew that it was criti-
cal for the scientists to not only understand the value of the technology but also
defne and shape it to make it useful. He notes that even as the projects move
into production, its senior-level scientists who are championing the new tools and
evangelizing about their benefts (Overby 2002). To improve communication be-
tween the IT staff and scientists, the IT development teams hold meetings at least
once every two weeks between the two groups. Moreover, Dunbar doubled his
IT department staff to 2,700 people, with an emphasis on hiring people with dual
computer science and biology or chemistry degrees. In 2003, Mike Heim was ap-
pointed as the new CIO. Mike began his career at Lilly as a systems analyst in
1979.
In 2008, Lilly signed an agreement with TopCoder Inc. to help build some of
its large applications for global drug discovery operations. TopCoder sponsors
worldwide competitions trying to attract creative programmers. Code written by
the programmers is stored in libraries that can be used by sponsoring companies
such as Lilly (Havenstein 2008).
From various reports, Lilly uses Amazon Web Services (AWS) to handle some
computing tasks. Using the cloud services is faster than trying to buy and set up
a new physical server. But, rumors also exist about issues surrounding the basic
contract and potential liability issues (Vaughan-Nichols 2010).
In 2010, Lilly began reducing costs by eliminating 5,500 workers, including
340 information technology jobs in addition to 140 cuts due to layoff, retirement,
and resignations; out of 1,250 IT workers (Sears 2010). By 2013, drugs generating
half of Lillys revenue will lose patent status and face competition from generics.
Marketing
Developing a new drug is only the frst step to making a proft. Companies also
have to convince physicians to prescribe the drugs. With competition, and a huge
array of new drugs, physicians need to evaluate the benefts and side effects. In
the past, drug companies had sales representatives call on physicians. But Brian
Weed, global marketing and sales IT director for Lilly, points out that doctors
offces are closing their doors to pharmaceuticals sales representatives and citing
HIPAA and privacy concerns as reasons (Greenemeier 2003).
Sales force management and customer relationship management software are
becoming more important at the pharmaceutical frms. Salespeople need to tell a
consistent story to physicians. In terms of patients, Lilly is encouraging customers
to contact the drug frm directly, instead of frst going to their physician (Greene-
meier 2003). The direct contacts can be benefcial in building brand awareness. As
Lilly learned with Strattera, the call centers can provide more immediate feedback
on actual usage and side effects. Ultimately, the process might require tighter links
between the patients, physicians, and pharmaceutical frms, but those are not yet
on the drawing boards.
Over the years, Lilly, and many other pharmaceutical frms, have faced bad
publicity with various lawsuits. A common complaint is that the drug frms with-
hold or bury important information from physicians, usually regarding side ef-
fects. Thousands of plaintiffs claimed that Lillys Zyprexas anti-psychotic drug
led to obesity and diabetes. In an interesting twist, a doctor and a lawyer involved
in the case obtained many of the documents from plaintiffs and posted them on a
Web site. Lilly attempted to obtain an injunction to (1) get the documents returned
and (2) have all copies removed from all Web sites. Jack B. Weinstein, a federal
judge in Brooklyn, ruled that the documents had to be returned, but it would be
impossible to enforce any court order to prevent Web sites from displaying copies
(Thomas 2007).
Questions
1. How does Eli Lilly use databases to improve effciency?
2. What types of data does Lilly store in its databases?
3. How could Eli Lilly use additional information technology in marketing?
Additional Reading
Claburn, Thomas, Eli Lilly Recovers Confdential Documents But Loses Secrets
To The Web, Information Week, February 15, 2007.
Ewalt, David M., Roy Dunbar, Information Week, December 15, 2003.
Greenemeier, Larry, Pharmaceutical Makers Turn To CRM, Information Week,
October 29, 2003.
Havenstein, Heather, Eli Lilly Aims to Tap Rock Star Programmers,
McDougall, Paul, Drug Company Eli Lilly Outsources Clinical Data to India,
Information Week, November 20, 2006.
Murphy, Chris, Answers That Matter, Information Week, February 2, 2004.
Sears, Don E., Eli Lilly Eliminates 340 IT Jobs, eWeek, July 16, 2010.
Overby, Stephanie, They Want a New Drug, CIO, October 15, 2002.
Vaughan-Nichols, Steven J., Falling Through the Clouds, Computerworld,
August 3, 2010.
Case: GlaxoSmithKline
GlaxoSmithKline (ticker: GSK) is one of the leading global pharmaceutical frms.
Created through the merger of several frms over the past decade, it is headquar-
tered in England and has over 100,000 employees worldwide. Like the other large
pharmaceutical frms, GlaxoSmithKline (GSK) is suffering from the transition of
its blockbuster drugs to generic status. It could easily lose $3 billion to $4 bil-
lion in revenue from its antidepressants Wellbutrin and Paxil and the antibiotic
Augmentin (Business: Glaxo 2004. Because the company is headquartered in
England, but almost half of its sales are in the United States, profts have suffered
from the 35 percent depreciation of the dollar in 2003-2004. Nonetheless, CEO
Jean-Pierre Garnier is optimistic. Through the mergers, he was able to trim $6
billion in costs. He has rearranged the research teams into divisions focused on
specifc areas (cancer, AIDs, and so on). More important, the company has 80
new products in the pipeline and half have reached advanced stage clinical trials
(Business: Glaxo 2004).
GSK faces other issues. In 2004, the New York attorney general, Eliot Spitzer,
fled suit against GSK for concealing negative data about Paxil. Spitzer contends
that of all the studies Glaxo performed, only one generated the results they wanted
to promote the drug. Their effort to suppress the other studies was harmful and
improper to the doctors who were making prescribing decisions and it violated the
law (Martinez 2004). The primary concern is that Paxil might lead to suicidal be-
havior if used in children under 18 years of age. Glaxo notes that Paxil was never
approved for use by children in the United States and was not promoted that way.
Glaxo spokesperson Mary Anne Rhyne notes that there are many, many stud-
ies each year. Its impractical to believe that every company in the industry will
be able to publish from every study (Martinez 2004). But Glaxos problems are
compounded by a couple of internal memos. One noted that the company would
have to effectively manage the dissemination of these data in order to minimize
any potential negative commercial impact. Another stated that it would be com-
mercially unacceptable to include a statement that effcacy had not been demon-
strated (Martinez 2004).
In response to these problems, several organizations have called for a clinical
trial registry to track the progress of all trials and outcomes. The leading pharma-
ceutical companies are also backing the registry. However, the process is not that
simple. Many of the Phase I trials are designed to test for side effects and do not
focus on effcacy. So, some drugs might not appear to be useful. Some companies
have noted that the National Institute of Health already maintains an online regis-
try (www.clinicaltrials.gov), and it could be expanded to include results from all
Phase III trials (Hovey 2004).
Relying on blockbuster drugs for profts is a risky strategyparticularly when
bad news erupts concerning the heavily-used drugs. In 2007, some medical infor-
mation was reported suggesting that GSKs second-biggest selling drug Avandia,
a diabetes treatment, was linked in some way to a risk of heart attack. The data
on both sides of the argument appeared weak and additional studies were being
conducted. Nonetheless, physicians had basically stopped prescribing the drug for
new patients. In managing the crisis, Jean-Pierre Garnier, CEO of GSK suggested
fve tips (Whalen 2007):
(1) Fight data with data.
(2) Communicate to employees. Daily phone calls are essential.
(3) Study doctor opinion to catch any changes.
(4) Put data on company Web site so everyone can see.
(5) Keep working on long-term business goals.
In 2007, it was estimated that GSK spent about $560 million on U.S. consumer
ads; which is only a fraction of the $5.2 billion spent annually on U.S. pharmaceu-
tical ads. But, in part to head off potential new regulations, Glaxos CEO Andrew
Witty said the company would cut back on television advertising. (Whalen 2009).
Information Technology in Research
In one sense, pharmaceutical R&D is similar to a university. Sharing knowledge
within the organization is a key factor. The company generates and acquires infor-
mation in the form of documentssome electronic, some in paper. GSK built a
centralized library system (Lynx) to give researchers immediate access to the doc-
uments. The system provides access to over 2,000 electronic journals. Requests
for paper books or articles are handled by librarians, who frst try to locate the
appropriate title within one of the companys nine libraries. If it is not available,
they check with their outside suppliers. Of course, the company would prefer to
digitize all paper documents, making it simpler to deliver the information, plus
make it available to other users later. However, copyright limits the number of
documents that can be digitized and prevents the company from disseminating
them (Delaney 2003).
Genetics, particularly individual differences among patients, is one of the big
target areas for pharmaceutical frms. Glaxo has forged a relationship with First
Genetic Trust to study how variations in individual DNA can affect the effcacy
and side effects of various drugs. Glaxo even built a proprietary high-speed net-
work to communicate with First Genetic Trust. The company uses data mining
and huge databases to compare drug effects across individuals. The main system
runs on several HP 16L-Series 9000 UNIX servers. The company uses commer-
cial applications to enable patients to provide informed consent and specify how
their genetic data can be used. In general, only the genetic data without personal
patient identifers is given to Glaxo researchers (Greenemeier 2002).
In 2009, GSK was looking at many of the same cost issues facing other phar-
maceutical frmsnotably the number of proftable drugs losing patent status. VP
of IT Ingo Elfering said that he was using this pressure to push for changes in the
IT operations. The company signed a deal with Microsoft to shift standard e-mail,
calendars, and collaboration services to Microsofts hosted system (Lai 2009).
The deal will transfer 100,000 employees from the existing IBM Lotus Notes to
Microsoft Exchange, SharePoint, and Offce Live Meeting. The company expects
to reduce costs by 30 percent.
Questions
1. How does information technology help Glaxo in its research?
2. Would a national database of clinical trials prevent the problems that Glaxo
had with approval of its Paxil drug?
3. Assuming that new drugs are developed that are tailored to specifc DNA
markers, how would you build an information system to take advantage of
this data? How will you handle the privacy issues?
Additional Reading
Delaney, Emma L., Glaxosmithkline Pharmaceuticals Research And
Development, Interlending & Document Supply, Vol 31(1), 2003.
The Economist, Business: Glaxos Big Challenge, May 15, 2004.
Greenemeier, Larry, Genetic Research Drives High-End Computing,
Information Week, February 18, 2002
Hovey, Hollister, H., J&J Adds Its Support For Idea Of Clinical Trial Registry,
The Wall Street Journal, June 18, 2004.
Lai, Eric, GlaxoSmithKline Deal Highlights Microsofts Overseas Launch of
Hosted Collaboration Software, Computerworld, March 2, 2009.
Martinez, Barbara, Spitzer Charges Glaxo Concealed Paxil Data, The Wall
Whalen, Jeanne, Glaxos Garnier Is Taking the Heat, The Wall Street Journal,
July 9, 2007.
Whalen, Jeanne, Glaxo to Pare Ads on U.S. Television, The Wall Street
Case: Bristol-Myers Squibb
Bristol-Myers Squibb (ticker: BMY) has headquarters in New York City and em-
ploys about 44,000 people worldwide. One of the companys recent innovations
was the development of TAXOL to treat breast cancer. The company was able to
create a synthetic compound that duplicated the original drug derived from the
bark of the Pacifc Yew (an endangered tree). In 2004, Bristol-Myers Squibb was
approved to market the genetically created drug Erbitux in conjunction with Im-
Clone Systems. You might remember the drug as the one that led Martha Stewart
to be found guilty of lying to SEC investigators. Bristol-Myers Squibb also sells
many consumer-level products (www.bristolmyers.com). In 2006, the company
earned $1.6 billion on sales of $17.9 billion. However, 18 percent of sales revenue
came from one drug: Plavix. The patent for Plavix expires in 2008 in Europe and
2011 in the United States (2006 Annual Report).
Research Data
Laboratory equipment today is generally automated. Often, it can be controlled by
computers and most machines generate data that can be collected by computers.
The catch is that machines can generate enormous amounts of data. BMY was
having trouble keeping up with the amount of data generated. The BioAnalytical
Sciences (BAS) group turned to the Scientifc Data Management System (SDMS)
developed by NuGenesis Technologies Corporation. The system collects all in-
coming data and reports and archives them to a central repositorywhich frees
up the hard drives on the local workstations. The data and reports are then avail-
able to other researchers from a secure Web site. The system also provides access
to the binary data, which is needed for compliance with federal laws. The system
enables researchers to tag the raw data by researcher, project, analytical method,
and so on. It collects data from diverse systems and makes the data accessible via
a common interface. The BAS group connects to over 100 instruments in two sites
in New York and New Jersey. The central server also backs up two main servers.
Because the system consolidates historical data, it makes it easy for researchers
to compare current results to older analyses. Data is stored in a secure Oracle
database and marked with time-stamp and audit trail data so everything can be
tracked. In a six-month pilot test, the system generated 5,000 reports. The system
automates most data management tasks, freeing up time by researchers and IT
administrators. More important, it enables greater collaboration because the data
and reports are accessible throughout the company, regardless of the geographic
location or software installed on a specifc machine (DeVincentis 2002).
Knowledge is critical to pharmaceutical researchers. And knowledge is vastly
more complex than simple data. How can a researcher in one location quickly fnd
an expert in another division of the company? How does the research in one area
compare to that by another department? Bristol-Myers Squibb purchased tools
from LexiQuest to help solve these types of problems. LexiQuest is a linguistic
search specialist, providing the ability to search based on information instead of
just key words. LexiQuest Guide uses language-recognition technology to search
documents based on everyday language. It semantically analyzes each document
to evaluate its content. The two companies are also building a custom dictionary
and search system to handle the specifc terms used in pharmaceutical research
(Jezzard 2001). In essence, the system reads the documents, enabling it to pro-
vide a more precise match to researcher requests. By providing more accurate and
faster responses, researchers can save considerable time. And time is critical when
you need to bring a new drug to the market.
Business Operations
Pharmaceutical frms are pressured by expiring patents, government attempts to
reduce drug costs, and increased competition. Like any business, they are turning
to information technology to reduce costs and improve the business operations.
Bristol-Myers Squibb is turning to the Web for its purchasing needs. It acquired
Aribas Buyer e-procurement software in 2000, saving over $90 million a year
within the frst year. The system standardizes purchasing methods. Employees
who need to buy something, such as a PC, check the suppliers and options on the
Web site. Then a request for quote (RFQ) is posted on the Ariba network and sup-
pliers bid for the contract. Employees can also use the system to purchase smaller
items directly. The system aggregates the purchases where possible and orders in
bulk. In 2001, as much as 16 percent of sales orders were conducted over the sys-
tem (Yasin 2001).
In 2007, Bristol-Myers Squibb lost an important initial lawsuit. U.S. District
Court Judge Patti B. Saris in Boston ruled that Bristol, AstraZeneca, and Scher-
ing-Plough were guilty of infating average wholesale prices (AWP) for certain
drugs. Some insurers use the AWP to establish reimbursement rates. On the other
hand, Medicare stopped using the list in 2003, largely because of the inaccuracies
that it perceived. Judge Saris wrote that unscrupulously taking advantage of the
fawed AWP systemby establishing secret mega-spreads far beyond the stan-
dard industry markup was unethical and oppressive (Tesoriero and Korn 2007).
In 2008, Bristol faced the many of the same issues as the other pharmaceutical
frmsand is also responded by cutting jobs. It cut 10 percent of its workforce
in 2008 and again in 2009. The goal was to reduce costs by $2.5 billion (Wang
2008).
Questions
1. Can machines replace lab workers and scientists in conducting experiments
and evaluating data? What problems might arise and how would you
minimize the risks?
2. How does Bristol-Myers use information technology to reduce costs
particularly in purchasing?
3. Why is a central research database system so important to Bristol-Myers?
Additional Reading
DeVincentis, John, Performance, Speed Improve With Scaleable SDMS, R&D,
October 2002.
Jezzard, Helen, LexiQuest Aids Drug Co., Information World Review,
December 2001
Tesoriero, Heather Won and Melissa Korn, Ruling on Drug Pricing Faults Three
Companies, The Wall Street Journal, June 22, 2007.
Wang, Shirley S., Bristol-Myers Sets More Job Cuts, The Wall Street Journal,
December 16, 2008.
Yasin, Rutrell, E-Procurement is Drug for Bristol-Myers Squibb, InternetWeek,
June 11, 2001.
Case: Pfzer
With almost $68 billion in sales in 2010, Pfzer (ticker: PFE) is the largest of the
pharmaceutical frms (www.pfzer.com). As a research organization, the company
has produced some well-known brands, including prescription drugs such as Ce-
lebrex, Difucan, Lipitor, and Viagra. It also has a big presence in the consumer
market with Benadryl, e.p.t., Listerine, Neosporin, and Rolaids, among others. In
2010, the company had 15 brands generating over $1 billion a year in sales (2010
Annual Report). It has 130 potential drugs in the pipeline (Overby 2002). But, $12
billion of revenue came from Lipitor, the biggest-selling medicine ever produced
by any pharmaceutical company. But Lipitor faced competition from generics in
2010, and the company has had no success in creating another blockbuster drug.
The company had been betting on a new drug torcetripib to boost good cho-
lesterolbut a 15,000-patient study was cancelled when patients developed high
blood pressure and started dying (Simons 2006). Pfzer offcials tried to impress
investors by revealing it had 242 research programs in progress. But it is diffcult
to say that any of the drugs will make it to market or produce the blockbuster rev-
enue Pfzer depends on. Hank McKinnell, the CEO who engineered many of the
mergers that made Pfzer a giant, stated in 2002 that size helps. Chemical sup-
pliers return our calls faster, we can run very large-scale global clinical studies,
and we can try out the newest technologies and implement them rapidly in order
to do things others cant. So, there are real advantages to scale. But the board of
directors ousted McKinnell in 2005 when the promised gains did not appear. Ad-
ditionally, some people are concerned that the big companies are creating fewer
new drugs. Submissions to the FDA fell from an average of 41 per year to 27 per
year, and many of those are created by smaller companies but licensed by the big
four (Simons 2006). The company also created a Web site to enable investors and
potential customers to track drugs in the pipeline (www.pfzer.com/pipeline). The
concept of making this information openly available is a radical change for any
pharmaceutical company.
Pfzer continued its expansion by purchasing Pharmacia Corporation in 2003,
Esperion Therapeutics in 2004, and Vicuron Pharmaceuticals in 2005. Closing off
a co-development agreement with sanovi-aventis, Pfzer purchased the worldwide
rights to Exubera an inhaled insulation therapy. However, Pfzer did sell off its
consumer products division to Johnson & Johnson for $16.6 billion in 2006 (2006
Annual Report). In 2010, Pfzer acquired Wyeth, another large pharmaceutical
frm (2010 Annual Report).
The company faces many of the same problems as the other pharmaceutical
companies. One of the biggest issues is how to integrate data and knowledge
across the companyparticularly when growth came through mergers.
Walter Hauck, vice president of worldwide informatics at Pfzer, knew he faced
a huge problem in trying to integrate the many research departments. He needed
to create tools that would support collaboration not only among the scientists but
also among the lawyers and marketers who bring the drugs to market. He began
with an ambitious task: integrate all of the changes at once. At La Jolla Labs in
California, his IT team introduced one new application a week for six months.
The scientists almost rebelled, saying that the sheer number of changes was tak-
ing away time from their research. Hauck heard similar complaints from many
other divisions: It was a lot of change to drop on people while expecting them to
continue to deliver (Overby 2002). With little control over introducing new tools,
some employees were asked to sit through training sessions on the same prod-
uct multiple times. Hauck knew it was time to implement a change management
strategy. One of the keys was to standardize the process by creating a checklist,
making sure that the business units needed the changes and understood the value.
The lists also ensured that the developers captured the feedback and evaluated the
pilot tests accurately.
Pfzer also tried to help employees by creating a new application to enter ex-
pense reports. But, the new system was harder to use than paper forms, help calls
increased, and many people stopped using it. Joseph C. Schmadel Jr., senior di-
rector of business technology, observed that that was an indication to us that
something was wrong. To solve the problem, Pfzer brought in netNumina, Inc.,
a consulting frm, to rethink the user interface. By rebuilding the user screens and
linking them to the back-end reporting system, the system became easier to use
and still collected the necessary data. The system also creates a digital dashboard
that provides summary statistics for executives (Weiss 2004).
Integrating data is critical to a company as large as Pfzer. The company had 14
different fnancial systems and wanted to combine the data into a data warehouse.
The IT group built a data warehouse as a central repository, and then built links
to each of the individual systems. But the IT department quickly encountered a
problem. Each of the systems had different defnitions for the data. It took months
to clean up the data so that it could be combined and managed as a single source.
Danny Siegel, senior manager of business technology, says, We saw that we had
to put in place some rigorous data standards. This kicked off a six-month, totally
non-technical effort to devise a set of standards that allow users to slice and dice
data in whatever context they need it (King 2003).
Vita Cassese, VP of global business technology at Pfzer summarized the chal-
lenges the company faces in trying to reduce costs but still support research. She
said there are two opposing challenges. We need to drive a level of consistency
and standardization to operate effectively while still fostering innovation. She
also noted that the size of the business makes it more challenging to manage IT,
noting in many ways, it demands different approaches to data demand and infor-
mation management. That makes for very big challenges, because we also need
to drive down the costs of managing this information (King 2006). Pfzer has
tackled these challenges by centralizing the primary IT services, including ERP,
the data center, help desk, and networking. But local support is provided by plac-
ing IT staff members into the various divisions around the world. Ms. Cassese
noted that IT still needs to be close to the business and have a deep, deep un-
derstanding of the business, but we also need a deeper level of discipline around
how we deploy technology to control the costs. To reduce costs, the company is
trying to centralize much of the data handling. In 2005, the company initiated a
project to consolidate 30 document management systems. The company is also
trying to standardize desktop hardware. The document consolidation program is
also driven by new federal rules that require greater sharing of data on drug trials
and submission of labeling data in XML. Consequently, the system will use XML
to store all of the data, integrating information across divisions in 26 nations.
Because Pfzer has acquired or partnered with many companies since the in-
troduction of computerized database management, the company has a staggering
amount of legacy data. The diffculty is to get the data into a shape where it can
be easily accessed by researchers that span the globe and be nimble enough to be
able to accept new data and new systems as more companies come in via mergers
and acquisitions. In addition, they need to convince scientists that what they have
is not just an archive of past experiments but also a rich resource of information
for future drugs (Derra 2004).
Pfzer, like many other pharmaceutical companies, is pushing to become more
effcient in drug discovery, and is looking to its legacy data as a source of use-
ful information and insights into new drug development. But the sheer number
of legacy systems within even one research facility is overwhelming. Bhooshan
Kelkar, PhD, advisory software engineer at IBM Life Sciences, Dallas, says that
the comments of a colleague speak to the enormity of the issue. He was working
with a big pharmaceutical company last year and reported that they had 500 lega-
cy applications just in their discovery and clinical areaIt was costing them 50%
of their IT budget every year, just in maintenance and reconciliation of different
legacy systems. That is huge. The key to unlock the power of mining legacy sys-
tems will be a combination of robust systems, standard methods, consistent data,
and cultural change from within the drug discovery environment (Derra 2004).
While the task is daunting, it is essential to new drug development to effectively
catalog all the known compounds within Pfzers extensive library in order to rap-
idly develop new pharmaceuticals or delivery methods. Transferring all of the cor-
porations legacy data to a common format is the best possible way for researchers
to have quick access to important older data.
This same need for a universal data standard for collecting, sorting, and pro-
cessing clinical trial data for FDA submission is currently occupying the entire
pharmaceuticals industry. But the data-collection standards have been slow to
catch on. However, in June 2004, a consortium of American pharmaceutical com-
panies debuted a data-interchange standard for electronically submitting drug-ap-
proval applications to the U.S. Food and Drug Administration. This could speed
the entire drug-review process, reducing the time between submission and FDA
approval. The major pharmaceutical companiesincluding Eli Lilly, Merck, Aven-
tis, and Pfzerare the sponsors of the move, under the heading of the Clinical
Data Interchange Standards Consortium (CDISC). Among the standards are a uni-
versal fle structure and submission form, both within XML programming. CDISC
previously developed standards for collecting lab data and clinical-trial data such
as patient information. Those are designed to simplify the collection of clinical-
trial data by pharmaceutical companies and the contract research organizations
such as Quintiles Inc.that drug makers hire to manage clinical trials (Whiting
2004). The FDA will not require CDISC standards for drug-approval applications,
but it has endorsed CDISCs efforts.
In 2007, a Pfzer employee used a company laptop to work at home. She in-
stalled a peer-to-peer fle sharing application on the computer, and inadvertently
made the entire machine a sharing device on the P2P network. The action exposed
the personal data of 17,000 Pfzer employees, including Social Security numbers.
Pfzer reported the breach and noted that various outsiders had downloaded data
on about 15,700 employees (Vijayan 2007).
With 86,000 employees, and several thousand scientists in R&D, it can be chal-
lenging to coordinate and share information. So Pfzer customized Imaginatiks
Idea Center tool to enable employees to submit ideas for new products or process
improvements. Rob Spencer, a senior research fellow noted that the tool saved
the company at least $20 million and helped solve hundreds of business problems
(Rosencrance 2010). The goal is to use collective intelligence of the employees
to attack diffcult problems. The ideas or comments from one person can trigger
ideas or solutions by other people.
Pfzer is facing the same cost-cutting challenges as other companies. The ac-
quisition of Wyeth and King help increase total sales, but the company closed
major research facilities in 2010 to achieve a 24 percent reduction in R&D costs.
Pfzers CEO Ian Read noted that This is a fundamental change in culture. We
have to fx this innovative core (Loftus 2011).
Questions
1. Why did Hauck introduce so many new applications at one time? Why was it
such a failure? Could you have predicted that outcome?
2. Why does Pfzer have so many different research systems and why is it so
diffcult to integrate them? What information technology tools might help?
3. How did Pfzer manage to have so many different fnancial systems?
Additional Reading
Derra, Skip Legacy Systems: More than Old Data in the Attic Drug Discovery
and Development, May 1, 2004.
Havenstein, Heather, Rules Prompt Pfzer to Consolidate Content Management
Systems, Computerworld, May 23, 2005.
King, Julia, Business Intelligence: One Version of the Truth, Computerworld,
December 22, 2003.
King, Julia, Pfzer: Striking a Balance, Computerworld, October 30, 2006.
Loftus, Peter, Pfzer to Cut Research, Shift Funding to Buybacks, The Wall
Street Journal, February 1, 2011.
Overby, Stephanie, They Want a New Drug, CIO, October 15, 2002.
Rosencrance, Linda, Tap the Wisdom of Employees -- and Boost the Bottom
Line, Computerworld, February 24, 2010.
Vijayan, Jaikumar, Personal Data on 17,000 Pfzer Employees Exposed,
Computerworld, June 12, 2007.
Weiss, Todd, Pfzer Case Study: Hiding Complex Apps Behind User-Friendly
Interfaces, Computerworld, March 9, 2004.
Whiting, Rick Standards May Speed Approval of New Drugs Information
Week, June 14, 2004.
adopted by rivals?
sectors?
2
Part
Business Integration
How do you protect data? How do you process transactions and
what problems will arise? How do you integrate data across the
organization and with suppliers and customers, including across
the Internet? What tools are used to support teamwork?
Protecting computers and data today is no longer the job of
just the security director. Because of the network-based threats, it
is everyones responsibility to be aware of potential problems and
maintain good computing habits.
A hundred years ago, most businesses were tiny. As frms be-
came larger, owners needed a way to manage and control the
huge number of employees. Managers needed assistance with
hundreds of daily operational decisions. Information technol-
ogy provides the means to both share and protect data. It is relied
on today to collect transaction data. Firms, even large ones, are
increasingly creating enterprise systems that store fnancial and
transaction data in a common central formatmaking it available
to managers throughout the company.
The Internet offers new ways to extend companies into rela-
tionships with suppliers and customers. As a manager, you need
to use information technology to reduce costs, create new sales
opportunities, and organize the company.
Teamwork is increasingly important to managersparticular-
ly in service organizations. Many tools have evolved to support
groups of people working on documents and projects.
Chapter 5: Computer Security
Chapter 6: Transactions and Entreprise Resource Planning
Chapter 7: Electronic Business
Chapter 8: Teamwork
How are information systems used to integrate business data
and support teams?
293
How do you protect your information resources?
What are the primary threats to an information system?
What are the primary security threats faced by individuals?
What primary options are used to provide computer security?
How do you protect data when unknown people might be able to fnd it or intercept it?
What additional benefts can be provided by encryption?
What non-computer-based tools can be used to provide additional security?
How do you prove the allegations in a computer crime?
What special security problems arise in e-commerce?
If you have to track everyones computer use to improve security, what happens to
privacy?
Chapter Outline
Computer Security
5
Chapter
Introduction, 294
Threats to Information, 295
Disasters, 297
Employees and Consultants, 299
Business Partnerships, 301
Outsiders, 302
USB Drives, 303
Threats to Users, 303
Virus/Trojan Horse , 304
Phishing: Fake Web Sites, 306
Updates and Patches, 307
Intercepted Data, 308
Computer Security Controls, 310
User Identifcation, 311
Access Control, 313
Administrator Access Rights, 314
Single Sign-On and Lack of Standards,
314
Encryption, 315
Single Key, 316
Public Key Infrastructure, 318
Additional Security Measures, 322
Audits, 322
Physical Access, 323
Monitoring, 323
Hiring and Employee Evaluation, 325
Computer Forensics, 325
E-Commerce and Cloud Computing
Security Issues, 326
Theft of Data from Servers, 327
Firewalls, 328
Intrusion Detection and Prevention, 329
Denial of Service, 330
Privacy, 332
Summary, 333
Key Words, 334
Exercises, 335
Cases: Professional Sports, 339
294 Chapter 5: Computer Security
National Football League
Introduction
How do you protect your information resources? What are the ma-
jor threats? Figure 5.1 presents some of the issues: outside hackers, people inter-
cepting data, attacks on the server, physical threats to the equipment (including
natural disasters such as foods and fre), and internal threats from employees, as
well as privacy issues such as abuse of personal data. Think about the problem
for a second: Who stole more money in 2010 (or almost any other year): teenage
hackers or CEOs? Computer security has to prepare for both of these threats and
others.
Security is a challenging problem for any business. It has become even more
diffcult with computing devices and data used everywhere in the company, on the
Web, and around the world. Many companies rely on part-time and contract work-
ers and need to give them access to data to complete their tasks. Monitoring tools
are needed to watch for security and privacy issues. Advances in encryption and
biometrics have provided powerful tools, but security ultimately comes down to
people. If a worker is careless with passwords, or a programmer makes a mistake,
or a network engineer falls behind on updates, or an auditor fails to test an ac-
count, holes are created that can be exploited by a thief or miscreant.
On the other hand, organizations can go overboard with security rules. Com-
puter security is a balanceyou must protect the data but still enable workers
to do their jobs. The challenge is that IT security workers often see their jobs as
identifying potential problems and preventing them from occurring. If an attacker
steals data, the security staff is likely to be blamed, but if security is so tight that
workers need to create workarounds to do their jobs, management might never see
the problem, much less blame the security staff. So there is an incentive to try and
security as tight as possibleeven if it interferes with jobs.
Encryption plays an important role in protecting systems. It can also be used to
authenticate the sender of a message. A key aspect of security and encryption is
How do you keep data secure? Any professional football team has dozens of coaches
and playerswhether it is American football or European soccer, the characteristics
are similar. Teams win through cooperationthat means everyone needs to share in-
formation. Coaches create playbooks, players provide feedback, and scouts identify
weaknesses in opposing teams. All of this information has to be sharedso teams
increasingly put the data onto computers. But securing this data is critical. Imagine
the problems that arise if an opposing team gets a copy of the latest player health
reports and new plays. Yet there are dozens of players and coaches, most of whom
are not computer experts (to put it politely, since you really do not want to insult a
300-pound linebacker). Plus, coaches and players are sometimes replaced or switch
teams.
Teams increasingly have to offer more data to fans. Some stadiums are now of-
fering wireless networks. But security again becomes critical. How do you keep the
fans from accidentally interfering with the coaching networks? What other threats
can you think of? Hundreds of things can go wrong with technology, and someone
has to be responsible for protecting the systems and creating contingency plans to
handle problems.
the need to identify users. Consequently, the fip side of many security policies is
the loss of privacy.
The Internet, e-commerce, and cloud computing add some challenging aspects
to security. E-commerce requires that portions of the computer systems be avail-
able to consumers and other businesses. Greater business benefts are generated
when Web sites are integrated with corporate datasuch as inventory levels so
customers can determine if an item is in stock. Yet, allowing public access to these
systems creates greater security risks. Furthermore, since the Internet is a shared
public network, data needs to be protected in transmissionto ensure it is not in-
tercepted or altered. Wireless networks are even more open to eavesdropping and
interception. Because of the public nature of the Internet, even a well-protected
system can be brought down with denial-of-service attacks.
Tightening security can easily lead to a loss of privacy. One way to improve
security is to completely identify every person and every activity performed. But
even completely honest people are not willing to give up that much privacy. So,
security faces another trade-off. These trade-offs are important, but they make the
job harder for the corporate security expert.
Treats to Information
What are the primary threats to an information system? Infor-
mation threats are often described in two categories: Physical and logical. Some
people add a third category: Behavioral. Physical threats are things that go wrong
with the hardware and buildingsparticularly fres, foods, earthquakes, and
hurricanes. Logical security refers to the ability to defne and control access to
Trends
Security has been an issue for thousands of years, from the simple substitu-
tion ciphers of Caesar to the importance of codes and code breaking in World
War II. As more data was moved to computers, several complications arose.
One of the biggest obstacles has been the need to identify people. Passwords
have been the most common method, but they cause many problems. Newer
technologies are available, but they require standards and people will have
to agree to use them. Since security requires identifying people, increased
emphasis on security can result in a reduction in privacy. Firms have col-
lected data on consumers for years, but only recently have technologies ad-
vanced to the point where it is relatively inexpensive and easy to collect and
analyze data on millions of consumers. Despite Hollywoods portrayals, the
greatest security threats come from insiders. On the other hand, it used to be
diffcult to attack servers, and required programmers with a deep knowledge
of the system. Today, with millions of computers connected to the Internet,
it is relatively easy for beginners to download code from a site and run au-
tomated attacks against known bugs in operating systems. This technique
is commonly used for creating denial-of-service attacks on Web sites. As
e-commerce expands in importance, it becomes increasingly critical to de-
velop a more robust Internet protocol that can identify and stop denial-of-
service attacks. Many security tools exist to protect servers and to encrypt
data transmissions, but it is diffcult to stop denial-of-service attacks that
rely on fooding the server.
data. Behavioral security refers to specifc problems related to people. You might
have strong physical and logical security, but employees who write their pass-
words on notes stuck to their computer, or give out private information over the
phone can quickly defeat any security system.
Many potential threats exist to information systems and the data they hold. The
complicated aspect is that the biggest information threat is from legitimate us-
ers and developers. Purely by accident, a user might enter incorrect data or de-
lete important information. A designer might misunderstand an important function
and the system will produce erroneous results. An innocent programming mistake
could result in incorrect or destroyed data. Minor changes to a frail system could
result in a cascading failure of the entire system.
You can detect and prevent some of these problems through careful design,
testing, training, and backup provisions. However, modern information systems
are extremely complex. You cannot guarantee they will work correctly all of the
time. Plus, the world poses physical threats that cannot be avoided: hurricanes,
earthquakes, fres, and so on. Often, the best you can do is build contingency plans
that enable the company to recover as quickly as possible. The most important
aspect of any disaster plan is to maintain adequate backup copies. With careful
planning, organization, and enough money, frms are able to provide virtually con-
tinuous information system support.
A second set of problems arises from the fact that as technology changes, so do
criminals. Today, only a desperate person would rob a bank with a gun. The prob-
ability of being caught is high, and the amount of money stolen is low. Do not take
it as an invitation to become a thief, but the computer offers much easier ways to
steal larger amounts of money.
Figure 5.1
Security and privacy are important issues in the Internet era. People can intercept
data, attack servers directly and indirectly, and companies routinely monitor all
transactions.
Server Attacks
+ Physical Dangers
Data interception
+ external attackers
The Internet
Monitoring/
Spyware
Internal + Privacy
The threats mentioned in Figure 5.2 provide a useful way to organize some of
the concepts of computer security. In particular, the physical threats need to be
separated from the logical or data attacks.
Disasters
Fortunately, fres, foods, hurricanes, and other physical disasters do not hap-
pen too often. But when a disaster does hit a companys data center, it could de-
stroy the company. Without advance preparations, the loss of a data center could
shut down the operations. How long can a company survive without transaction
processing?
Today, there are many ways to plan for and recover from disasters. Figure 5.3
shows a traditional method of using scheduled backups and a disaster recovery
services provider. When computer servers are expensive, it makes sense to con-
tract with a company to provide a facility and possibly spare computers. One lev-
el of support, called a hot site, consists of a fully confgured computer center.
Specifc computer equipment is already installed and ready for immediate use.
When the MIS staff declares a disaster, they install the backup tapes on the hot-
site computers and use telecommunication lines to run the day-to-day operations.
Another alternative is to contract for a cold site, which provides fully functional
computer room space, without the computer equipment. If a disaster occurs, either
the company or the disaster recovery services provider can arrange for the neces-
Figure 5.2
Threats to information. By far, the most serious threats are from insiders: employees,
mistakes, consultants, and partnerships. Businesses have to trust insiders to stay in
operation, but you need to put limits on the access to data. It is possible for outsiders
to break into most computer systems, but it is fairly diffcult. Viruses are a serious
threat both to data and security in general.
Employees & Consultants
Links to
business
partners
Outside
hackers
Virus hiding
in e-mail or
Web site.
sary equipment to be shipped to the cold site. However, there might be a delay of
several days before the new data center will be operational, so a cold site is often
used in conjunction with a hot-site contract.
The problem with the traditional approach is that companies can no longer af-
ford to run without computer supporteven for a few hours. Fortunately, comput-
er prices have also declined while network transfer speeds have increased. Con-
sequently, it is possible for many companies to provide continuous backupboth
in terms of data and processing capabilities. Figure 5.4 shows the basic concepts.
Server clusters are created in separate locations and connected by a network. Each
server contains redundancy in terms of multiple processors, disk drives, network
connections, and even power supplies. If one component in the server dies, the
others pick up the load instantly. Data is often stored on a network-attached stor-
age area network, with its own duplication of drives and RAID confguration.
Finally, in case a fre, food, or network outage knocks out the entire facility, a
second facility hosts another server cluster. Data from one site is continuously
replicated to the second site. User applications are connected by a network so if
one site goes down, everything shifts automatically to the backup site.
Large frms might build or lease their own remote sites with multiple server
clusters. But even small companies can gain many of the same advantages are
relatively low cost by using cloud computing. Large service providers on the
cloud, such as Amazons EC3, automatically provide servers in multiple locations.
Companies can lease almost any level of computing which is handled off site by
Amazon.
Figure 5.3
Disaster planning. 1. You have to back up your data. 2. You can contract with a
disaster facility to have a site ready. 3. Create a detailed plan to move your MIS
people and restore the system. 4. If a disaster strikes, shift the business network to the
temporary facility. 5. Rebuild the original data center.
Business/Operations
Network
Backup/Safe storage Recovery Facility
MIS Employees
Employees and Consultants
Employees are the heart of any company. Companies function and succeed by
trusting their employees. Although almost all employees are honest and diligent,
there is always the chance that one employee will use the companys knowledge,
experience, and trust to misappropriate resources.
It can be diffcult to identify people who might cause damage to the frm.
Many companies today use psychological tests, background checks, and random
drug tests to indicate potential problems. Most companies are wary of employees
whose employment has been terminated. Businesses follow specifc steps when
employees leave, being particularly careful to remove the employees access to
company computers.
A more complicated problem arises with MIS employees. Programmers and
analysts have to be trusted. Without them, there would be no software. However,
it is generally best if the programmers are not the users of the program. Com-
panies enforce a separation of duties among staff programmers and users. Think
about what might happen if a bank teller was also responsible for writing the com-
puter program used by tellers. It would be easy to use the computer to steal money
from different accounts. Auditing transaction-processing systems is an important
task for auditors.
Unscrupulous programmers have also been known to include time bombs
in their software. Whenever the software runs, it checks a hidden fle for a se-
cret word. If the programmer leaves the company, the secret word does not get
changed. When the program does not fnd the correct word, it starts deleting fles.
On large projects, these bombs can be impossible to spot (until they go off). Keep-
Figure 5.4
Continuous backup. Duplicate servers are installed at different locations and
connected across a networkoften the Internet. The servers themselves also have
built-in redundancy. Data is written to multiple drives and multiple locations. If one
site fails completely, the other site can instantly pick up the load.
Server cluster
with built-in
redundancy
Storage area
network with
redundancy
and RAID
Off-site or cloud
computing
processing and data
Users connect
to the servers
Use both sites
continuously or switch
DNS entries to transfer
users in a disaster.
Secure Internet
connection
ing good backups can usually minimize the damage. As a related side note, the
software industry is pushing states to adopt a new set of laws (UCITA) that makes
it legal to include a shutdown time bomb if a software company has a dispute with
a business that uses its software.
Another danger area is that programmers might include a trap door or secret
password that allows them to gain access to the software even if they leave the
company. Sometimes these trap doors are installed innocently, to enable program-
mers to make corrections faster. The important point is to make sure they are re-
moved when the system is permanently installed.
An interesting twist on passwords arose in 2008 when Terry Childs, the system
administrator for the San Francisco network refused to give up the administrative
username and password. Apparently the department suffered from management
and personality differences and he argued that no one else was qualifed to be
an administrator on the network. After sitting in a jail cell for 12 days he gave
the login credentials to the mayor. Later, he was convicted of interfering with a
network and sentenced to 4 years in prison. Ignoring the personality issues, the
case does point out the necessity of providing backup access to administrative ac-
counts. What would happen in a large organization if a network administrator died
in an accident?
An interesting class of threats to securing your data arises from negligence in-
stead of deliberate actions by the users. For instance, employees might acciden-
tally delete data. Or carrying disks, tapes, or even laptop computers past magnetic
felds can sometimes damage the fles. In these cases, the best bet is to have back-
ups readily available. More complicated problems arise when laptop computers
are lost or even stolen. In addition to the data stored on the machines, the fles
often hold passwords for corporate computers. Many laptops provide passwords
and encrypt the data to minimize these problems. One other problem that falls into
this category is a warning to be careful about how you dispose of old tapes, disks,
and computer equipment. Businesses run similar risks when they send computer
equipment out for repairs.
In general, the best way to minimize problems from employees stems from typ-
ical management techniques. Hire workers carefully, treat employees fairly, have
separation of jobs, use teamwork, and maintain constant checks on their work.
Consultants present the same potential problems as employees. However, consul-
Reality Bytes: Hacking is Easy When You Tell Everyone the Answers
In a well-publicized case, attacker David Kernell broke into Sarah Palins Yahoo
e-mail accountlargely by scouring the data that she publicly listed on various ser-
vices. In 2010, George Bronk was arrested for using a similar technique to break
into more than 3,200 e-mail accounts of women. He searched the victims Facebook
accounts for answers to security questions typically used to recover forgotten pass-
words at online sites. Once he was able to obtain the password, he changed the pass-
word, searched for nude photos and posted them on the victims Facebook pages. He
was charged in Sacramento Superior Court and faced six years in prison.
Adapted from Robert McMillan, Nude Photos Stolen from Womens e-Mail Ac-
counts, Computerworld, January 13, 2011.
tants tend to be hired for a short time, so the frm knows even less about them than
about regular employees. Consultants are generally used for specialized jobs, so
there may not be any internal employees who can adequately monitor their work.
Business Partnerships
As computers spread throughout every aspect of business, many companies share
their data. For example, General Motors asks its suppliers to provide all infor-
mation electronically. This electronic data interchange (EDI) means that business
information is processed faster and with fewer errors. The problem is that in many
cases, it means GM gives other companies considerable access to GMs comput-
ers and vice versa. For instance, if GM is thinking about increasing production,
the managers might want to check supplier production schedules to make sure the
suppliers could provide enough parts. To do it electronically, GM needs access
to the suppliers computers. To participate in business today, you must trust your
partners. However, you have limited ability to evaluate all of their employees.
The issue of partnerships becomes more important in an Internet world of soft-
ware and cloud computing. Application service providers (ASPs) on the cloud,
such as NetSuite or Salesforce.com, run software on their Web servers and you
store data on their sitesusing only browsers to access the data and applications.
Cloud computing has advantages such as: (1) experts set up and run the site so
you do not have to hire specialists, (2) storing the data on the Web means it is ac-
cessible to your employees wherever they have Web access, and (3) you can start
small and scale up to a reasonable size without hassles. From a security perspec-
tive, a potential drawback is that all of your fnancial data is stored on a site run
by someone else. Of course, the reputation of the ASP depends on protecting your
data and maintaining security, so it is probably safer than what a small business
could handle independently; however, you should still investigate the ASP secu-
rity procedures.
Attachment
01 23 05 06 77 03
3A 7F 3C 5D 83 94
19 2C 2E A2 87 62
02 8E FA EA 12 79
54 29 3F 4F 73 9F
1
2 3
From: afriend
To: victim
Message: Open
the attachment
for some
excitement.
Virus code
1. User opens an attached program that
contains hidden virus.
2. Virus copies itself into other programs on the
computer.
3. Virus spreads until a certain date, then it
deletes les.
Figure 5.5
Virus activity. Once a virus is loaded on your machine, you will need an antivirus
software package to remove the virus. Several versions are available at low cost. A
virus can come from any software that supports executable code. Attachments sent
through e-mail are currently the most common method of being infected.
Outsiders
There is some threat from outsiders who might dial up your computer and guess
a password. Using some common sense can minimize most of these threats. For
example, in the 1980s, some groups gained access to computers because the op-
erators never changed the default password that was shipped with the computer!
The Internet causes additional problems because it was designed to give other
people access to your machines. The key lies in providing people with just the
level of access they need. The biggest problems today arise from a group labeled
script kiddies, who download system scanning/attack software from the Internet
and randomly search computers for holes. Another major problem with passwords
is a technique hackers call social engineering. A hacker calls up a victim (you),
tells some story, and convinces you to reveal your password. Never give your
password to anyone.
In theory, modern computer security systems are effective and can prevent
most outside attacks. The problem is that operating systems are complex software
programs that have errors. Experts search these systems to fnd the errors, and
ultimately, the vendor fxes the errors. However, this process can result in dozens
of patches a year. Some businesses do not keep up with the patches, and some
patches interfere with other programs and are not applied. Consequently, there
can be thousands of systems connected to the Internet that suffer from published
faws. Software downloaded from the Internet can automatically search for these
faws and provide access even to inexperienced hackers. One key to protecting
your servers is to make sure they have all the current operating system patches.
Figure 5.6
Spyware. Software loaded into your computer captures keystrokes, websites you
visit, or other personal information. It transfers the data to a website run by a
company or hacker who steals your money. The problem is particularly dangerous on
public computers that are used by several people.
Capture
keystrokes
hacker
Password
Credit card
Accounts
USB Drives
USB or fash drives are a useful technology for transferring fles. But, that also
makes them a prime target for carrying viruses and Trojans. Reports indicate that
USB drives were a main vector for spreading the Stuxnet virus that appeared to be
targeted to specifc industrial machines, such as those used by Irans nuclear pro-
cessing. But, even simpler viruses, Trojans, and key-stroke loggers can be spread
with USB drives. USB drives are physically small but can contain huge amounts
of data or complex malware. Some organizations (including at times the U.S. mil-
itary) have banned them, but they are now everywhereincluding cameras and
cell phonesso they are diffcult to stop. Some software tools exist that can turn
off USB ports in computers, or monitor them to prevent anyone from inserting
USB drives.
On the fip side, USB drives can be used to improve computer security in some
situations. In particular, if you need to use a shared or public computer, it is likely
that the computer is already infected with any number of viruses or key-stroke
loggers. If you only need to use the system to browse a few open Web sites, the
risks are minimal. But public computers are risky when you need to connect to
fnancial accounts. (Student labs might be a problem, but many schools are care-
ful to monitor and remove common viruses.) USB drives have the ability to solve
problem of infected public computers. You can install an entire operating system
onto a USB drive along with software and a Web browser. Then if you need to use
a public computer, you can reboot from the USB drive and run a system that you
know is clean. You are only using the computers processor and RAM, not its in-
fected software. However, you should double-check the keyboard cable to ensure
no one installed a physical device on the line to capture keystrokes. Of course, you
need the operating system and licenses to install the software onto the USB drive.
Treats to Users
What are the primary security threats faced by individuals?
Business face threats to data and servers, but they also face threats to user comput-
ers. Individuals face similar threats to their personal computers. Overall, the main
Reality Bytes: You Would Expect TSA to Understand Security
James Duchak was employed as a contractor at the Colorado Springs Operations
Center of the U.S. Transportation Security Administration. He was a data analyst
for about fve years and updated the TSA servers with data pulled from the terror-
ist screening database and the U.S. Marshals Service Warrant Information Network.
He learned in 2010 that he was being let go, and was training his replacement. His
replacement noticed that Duchak deleted code that was used to format the birth dates
for people entered into the system. In October 2010, Duchak, 47, pleaded guilty to
the charges and was sentenced to two years in prison, $60,000 in restitution, and
three years of supervised release. The restitution claim is interesting because the
code should have been backed up and date formatting code is usually fairly easy to
write.
Adapted from Robert McMillan, Former TSA Contractor Gets Two Years for Dam-
aging Data, Computerworld, January 12, 2011.
threat is that an attacker can take over the computergaining administrator ac-
cess to see all of the data and use the computer for any task. Several bad outcomes
can arise from this threat. Your data and passwords can be stolenresulting in
someone impersonating you and stealing your money or trashing your reputation.
By monitoring your use of the computer, the hijacker can steal credit card and
bank account data. The attacker could also turn your machine into a zombie that is
used to attack other computers. Or, it could be used to commit crimesthat would
then be traced back and blamed on you.
Several methods are used to attack individual users. Some of them are diffcult
to detect and hard to prevent. All of them require users to be cautious and pay at-
tention to their computers. Can you see the problem? Many people treat comput-
ers as simple machines that handle basic tasks and never need attention. Software
and operating system vendors (notably Microsoft) have attempted to automate
many of the maintenance tasks to ensure they get handled properly to protect user
computers. But, from the perspective of society, it is not possible to protect all
computers.
Virus/Trojan Horse
Everyone should know the ancient story of the Trojan Horse, where Greek sol-
diers hid inside a giant wooden horse which was then brought into the walls of
the city of Troyleading to the destruction and downfall of the city. In comput-
ing terms, a program or complex data fle might contain a special section of code
designed to take over your computer. When you run the program, the hidden code
executes with your permissions and can take over your computer. A computer vi-
rus is a special Trojan Horse that frst copies itself into other programs, and then
tries to spread every time those programs are run. One complication is that many
applications support macro programming languages, such as Microsofts Visual
Basic for Applications, where code is stored and executed within data fles or Web
pages. These programming languages provide powerful features to integrate ap-
plications but they provide even more opportunities for nasty code to sneak onto
your computer.
As shown in Figure 5.5, the virus code can be diffcult to spot. A virus can be
picked up from many sources, but e-mail attachments are the prevalent method
today. The obvious solution would seem to be the same as for the original Trojan
Horse: Look inside the program to see if soldiers are hiding there. Of course, a
person would never be able to look through the hundreds of thousands of fles
and programs on a computer by hand. Instead, antivirus software is used to scan
your computer by looking through every single fle. But the second complication
is that it is diffcult to recognize a virus. The most common approach is to keep a
signature list of all known viruses and check every line in every fle against every
possible known virus.
Viruses and Trojan Horses remain a threat despite the use of antivirus software.
The reasons are clear when you understand how the antivirus software works.
For the most part, the software works on existing, known viruses. Anything new
or anything that modifes itself can sneak through. Additionally, with the huge
number of fles and the growing number of known viruses, the antivirus software
can take a substantial amount of processing timewhich can slow down your
computer. The software can also make it diffcult to install, update, or create new
software. Many commercial software packages even recommend disabling the
antivirus software before attempting to install anything. Consequently, antivirus
software can help clean fles once a virus is identifed, but it has not proven very
useful in stopping attacksparticularly since new viruses are created every day.
Today, it is easy to create a virussimply fnd a virus software kit on the Web,
make a few changes, and send it to someone. You would need only minimal tech-
nical skills. Of course, it is illegal to create and release viruses and other destruc-
tive software (in most nations).
Instead, the best way to stop a virus is to avoid running software acquired from
the Internet and to never open script attachments sent to you by e-maileven if
they appear to come from a friend. Be cautious, because some attachments that
appear to be pictures are actually virus scripts. Most e-mail services now have
flters that can block script attachments, but they tend to be heavy-handed and also
block useful fles.
Ultimately, the most important step with viruses is to make certain that you al-
ways have current backup fles. Then, if a virus deletes your fles, you can recover
the data, run an antivirus software package, and remove the virus. It will cost you
time, but at least you will save the data.
The problem with viruses and Trojan Horses is that they often install spy-
wareor software that sits on your computer and monitors all of your activity. It
can capture your keystrokes and record Web sites visited, passwords entered, and
credit card numbers. As shown in Figure 5.6, periodically, the spyware software
sends the information to a Web site where it is collected by an attacker.
Viruses, worms (essentially viruses that do not destroy data), Trojan horses,
and spyware are often called malware, because they are designed to do bad (mal)
things to your computer system. Spyware tools are particularly dangerous on pub-
lic computerssuch as those at Internet cafs, print shops, or libraries. Someone
could install the software and capture all of your keystrokes and passwords. You
should avoid entering passwords and credit card data on computers that are shared
with the public. Microsoft Windows includes software that scans your machine
Reality Bytes: Brute Force Password Attacks Using Amazon
Brute force attacks on passwords have always been a possibilitywhere a computer
tests every possible combination of letters and numbers to fnd a password. But, with
reasonable passwords, brute force attacks historically would take many years to fnd.
The obvious solution is to split the task across multiple computers. Still, an attacker
would have to fnd a way to get access to thousands of computers. With cloud com-
puting, anyone can buy cloud computing time, such as that from Amazons Elas-
tic Compute Cloud (EC2) service. In 2010, Thomas Roth, a security consultant in
Cologne, Germany, used a cluster of Nvidia graphics processors through Amazon
to test 400,000 possible passwords per second. He used the system to break into a
wireless network in 20 minutes at a cost of 28 cents per minute. With some improve-
ments, he says he can break wireless passwords in 6 minutes. Roth noted that "The
speed of computers is increasing incredibly fast, and so brute forcing will get faster
and faster, and the new cloud offerings make parallelization of such use tasks easy
and affordable,"
Adapted from Stuart J. Johnston, Researcher Breaks Wi-Fi Passwords Using Cloud
Computing Power, eSecurity Planet, January 12, 2011.
for some known, dangerous spyware. You can also use the Windows Task Man-
ager to show you what processes are currently running on your computer. If you
recognize a malware tool, you can stop the process to shut it down. But you have
to be careful, because a standard computer runs many processes. You need to de-
cipher the cryptic names of the processes to understand which ones are good and
which ones are bad. It is usually best to use a spyware toolwhich contains a list
of the known malware tools.
Phishing: Fake Web Sites
Attackers often want to gain access to your computer or at least your credit card
and bank account numbers. One attack method is phishing for passwords. As a
side note, news writers have been converting words beginning with f to ph
ever since the 1970s episodes of phone phreakingwhere hackers used various
methods to make free phone calls. Figure 5.7 shows one basic method of phish-
ing. An attacker sends mass e-mails pretending to be from a big-name bank. A
certain percentage of users will click the link in the e-mail which takes them to a
fake Web site that looks like the bank site but is actually run by the attackers. If
you do not recognize the fake and enter your username and password, the attack-
ers can quickly turn around and log into the real bank site using your credentials.
From there, they can transfer all of your money to their own accounts. Hundreds
of variations exist.
If you are always alert and moderately paranoid, you can avoid becoming prey
to a phishing attack. Avoid clicking links sent by e-mail. Always double-check the
URL of a site that asks for a login. Verify that the security certifcate exists and is
assigned to the bank. But, some fsh are not as smart as others, or you might be
tired one night and not pay attention and take the bait.
Figure 5.7
Phishing. A fake e-mail contains a link to a fake bank Web site. You inadvertently
click the link and enter your username and password. The attackers running the site
can now log into the real bank and steal your money.
Bank account
is overdrawn.
Please click
here to log in.
E-mail
Really good fake of
your banks Web site.
Username
Password
Sent to hacker
who steals your
money.
Banks have implemented some security features to make it easier for customers
to recognize the real bank site versus fake sites. Figure 5.8 shows the basic pro-
cess of a two-step login. Initially, customers select an image or enter a phrase that
is associated with the username. These values are stored on the banks server and
would not be accessible to an attacker. During subsequent logins, the customer
frst enters the username or ID value for the account. The real bank site retrieves
the matching image or phrase. When the customer recognizes the proper display,
it is safe to enter the password. Of course, you should never use the same phrase
or image for multiple sites. So, customers have to remember one more thing to
log in. It is relatively easy for any business to implement a similar scheme, but it
ultimately makes it more diffcult for people to use the systems. And it raises the
question of how far a company should go to protect customers from their own ac-
tions. Browsers have tried to implement warnings for known phishing sites, but
sites can still get past the warnings. Ultimately, it comes down to people paying
enough attention to avoid fake sites.
Updates and Patches
In the past couple of years, one of the favored methods of attacking your com-
puter has been to exploit faws in the operating system and software running on
your computer. Software that was created with sloppy programming practices can
create holes for outsiders to exploit. Many of the problems are caused by buffer
overruns, where the program allocates a limited amount of memory for data and
an attacker enters a carefully crafted entry that takes up more space than allocated.
This extra data then overwrites the program code and your computer executes the
attackers code. Microsoft, Apple, and other software vendors have been work-
ing hard to fnd and fx these errors. Periodically, these vendors release patches to
their software, which you need to load and install on your computer. The good
news is that the companies are actively fxing the problems. The bad news is that
every time they announce a problem, the attackers also learn of the issues. You
need to immediately patch your computer before the attackers take advantage of
the faw.
Figure 5.8
Two-step login. Customers enter an image or phrase at the real bank site. In future
logins, the username is entered frst, which retrieves the matching image or phrase.
If the customers recognize the image, it is safe to enter the password. Fake/phishing
sites would not have the correct image or phrase.
Username
Real bank site
URL
Security indicators
Image or phrase you
created earlier
Password:
Password
Figure 5.9 shows a basic time line. A vendor or security researcher fnds a prob-
lem and reports it to the software developer. The developer fxes the problem and
releases a patch. Patches are released with varying levels of importance. Critical
faws that could allow an attacker to take over your machine are given the highest
priorityparticularly if the attack has already been seen in the wild or outside
of a lab setting. You need to install these updates as soon as possible, and most
computers should be set to automatically download and install critical updates
when they are released.
This threat is magnifed because attackers put scripts on the Internet that auto-
matically scan for unprotected machines. These scripts make it possible for people
with almost no computer skills to attack your system. Of course, there is still a
risk that attackers can fnd faws before software vendors fnd and patch them.
These zero-day attacks could happen at any time, but they require considerably
more work by the attackers and are likely to be aimed at specifc targets. Still,
you truly need to keep your system up to date to stop the known attacks. Many of
these attacks are based on browsers and add-in software such as Java and Flash.
Intercepted Data
The Web and online content and services continue to be an increasing part of in-
formation systems and even daily life. Mobility and wireless connections are also
gaining importance. Both of these trends lead to a potentially serious problem:
Wireless data is relatively easy to intercept. Any device connected to the same
network has the ability to monitor any packet sent on the network. The device is
supposed to ignore packets not addressed to it, but it is relatively easy to tell it to
capture all packets on the network.
Figure 5.9
Zero-day attack. Attackers exploit a known bug or fnd their own and launch e-mails
or even Web sites that can attack your computer. Because the bug has not yet been
patched, it is easy for many computers to become infected.
time
Researchers
find bug
Vendor
announces
patch
Hacker attacks your
computer when you go
to a Web site
You should
update
immediately
Zero-day attack.
Hacker finds bug/hole first.
Everyone is vulnerable.
In late October 2010, Eric Butler released an add-on for the Firefox Web
browser called Firesheep that makes it incredible easy for anyone to monitor
WiFi traffc and hijack sessions from other users. Note, it is illegal to use an-
others computer accountno matter how easy it is to do. Figure 5.10 shows the
basic process. A user signs into a Web service and the username and password are
usually encrypted. The server generates a cookie and sends it to the userso the
user identity is maintained for the session without requiring a new login for every
page. This cookie and subsequent activities are often unencrypted. The eavesdrop-
per can grab the session cookie and hijack the sessionmaking any changes to
the users account or Web contents.
Wireless transmissions are relatively open and broadcast over an area. The only
way to protect the traffc is to encrypt everything. Web servers and browsers are
designed to encrypt traffc, but most Web site designers tended to limit encryption
to just the username and password pages. The feeling was that encryption slowed
down the server, the browser, and the data transmissions. In the early days of the
Web, these delays were measurable; so designers got in the habit of encrypting
only what they felt were essential items. With faster computers and networks,
sites will now be forced to encrypt all transmissions.
But, as a user of Web services, can you do anything while waiting for sites
to turn on full encryption? You could simply avoid using any important services
while connected to public wireless networks. Browsing Web sites to read the news
is certainly not going to matter to anyone. Logging into a social network or e-mail
site runs the risk of giving access to an eavesdropperand with Firesheep, that
eavesdropper could be anyone. The only other option is to sign up for a service
that provides a virtual private network (VPN). A VPN establishes an encrypted
Figure 5.10
Intercept WiFi connections. Often, only the username and password are encrypted for
many sites. But the session cookie that maintains the login is often unencrypted and it
can be capturedallowing an eavesdropped to hijack the session and make changes
as the user.
Initial page, encryption keys
Username/password
(encrypted)
Cookie/identifier
(Not encrypted)
Session and additional pages
not encrypted. With
unencrypted cookie/identifier.
User
Server
Intercepted
Eavesdropper
hacker
Hijacked
session
network from your computer (laptop) to the VPN server. No one can intercept the
traffc between those two points. Companies often set up VPN services so workers
can securely connect from home into the company network and operate as if they
were on site. Several commercial companies provide VPN services to individuals
for a monthly fee. If possible, you should test the VPN services before signing
onto a long-term contract. You want to ensure that the company has fast and reli-
able connections and servers that are available whenever you need them. It takes
several steps to install a VPN and it might interfere with some applications.
Computer Security Controls
What primary options are used to provide computer security?
Transaction and accounting data is clearly valuable to a company and needs to
be protected. Computer security systems need to protect against three general
problems: (1) unauthorized disclosure of information, (2) unauthorized modifca-
tion, and (3) unauthorized withholding of information. These three problems are
sometimes referred to as: confdentiality, integrity, and accessibility, leading to the
too-cute acronym CIA. As an example of the frst problem, you would not want
Reality Bytes: Rogue IT Employees
Fortunately, most IT employees are extremely honest. They need to be honest be-
cause they have access to most of the company operations. But, that does not mean
you should just ignore employees. Ronald Reagans famous saying when negotiat-
ing a nuclear-arms reduction treaty with the Soviet Union seems to apply to many
situations: Trust but verify. In a different situation involving a Fortune 500 com-
pany, a rogue IT employee had lost 11 laptops over three years. A fag that should
have raised questions earlier. Anyway, in 2008, a retailer in Pennsylvania had to hire
a security consultant to investigate a problem after the Business Software Alliance
(BSA) reported that the company might be using pirated software. The investigation
found that the software was illegal, and that it was sold to the retailer by a com-
pany secretly owned and operated by one of its own IT employees. More digging
found that the seven-year employee had been running a for-pay pornography site
on the companys servers and that he had stolen 400 customer credit-card numbers
from the retailers Web site. And he was the only person who had the administrative
passwords. The U.S. Secret Service and CERT provide a whitepaper that lists basic
steps you can take to reduce threats from insiders. (http://www.cert.org/archive/pdf/
CSG-V3.pdf) To recover the systems, the company sent the rogue employee on a
long overnight fight to California, then used the fve hours to reset every possible
password and lock him out of the systems. When the fight landed, the COO met
him at the airport and fred him on the spot. Although a background check might not
have prevented the hiring of the rogue employee, it probably would have spotted is
lie about having an MBA degree. Other people have suggested that his personality
should have indicated some problems. One of the investigators said that He was
extremely confdent, cocky and very dismissive of other people. Most good IT em-
ployees realize it is impossible to know everything, are willing to consult with other
experts, and are helpful to employees.
Adapted from Tam Harbert, Security Fail: When Trusted IT People Go Bad, Com-
puterworld, January 18, 2011.
hackers to get access to your customers credit card data. An example of the sec-
ond problem would be employees modifying their payroll records to change their
pay rates. The third problem is less obvious, but just as important. Imagine what
would happen if you needed to look at the latest inventory to decide how much
to reorder, but the computer refused to give you access. This problem is often re-
ferred to as denial of service (DoS) and is a diffcult problem faced by Web sites.
User Identifcation
One of the primary diffculties with providing computer security lies in identify-
ing the user. For years, the most common means of identifying computer users
is through usernames and passwordsbecause this method is easy to program.
The programmers simply set up a database of users that contain a username and
password. No standards are needed and every system can be independent. More
recently, interest in biometrics has been increasing; particularly the fngerprint
readers installed on laptops; but the lack of standards has prevented widespread
adoption of these technologies.
Passwords
Each user is given an account name and a password that are known only to the
computer and the user. If someone correctly enters both the name and the pass-
word, the computer assumes it must be the user. This method is cheap, fast, and
does not require too much effort by the user. However, there are problems. The
biggest diffculty is that users are afraid of forgetting their password, so they
choose words that are easy to remember. Unfortunately, passwords that are easy
to remember tend to be obvious to other people. For instance, never use the words
password or secret as a password. Similarly, do not use the names of relatives,
pets, or celebrities. Most of these can be obtained by looking in a phone book,
talking to someone you know, or browsing your Facebook page.. In fact, you
should not use any actual words. Most people use only a few thousand words in
typical conversation. The goal is to make it hard for someone to guess the pass-
word. You need to choose passwords from the largest possible set of characters
and numbers. Two other rules about passwords: Change them often and do not use
the same password for everything. Most systems have a method to enable you to
change passwords. Some systems force users to change passwords on a regular
basis, such as every 30 or 60 days.
One drawback to passwords is that you need too many of them. Everything
from ATM cards to phone calls to computer accounts uses passwords or personal
identifcation numbers (PINs). It is too risky to use the same password for every
account, but it is diffcult to remember several different passwords, especially if
you choose random letters and numbers and change them often. With so many
passwords, it is tempting to write them down, which defeats their purpose. This
confict is the main reason a new system is needed to identify users.
Passwords are not a perfect solution to identifying users. No matter how well
they are chosen or how often they are changed, there is always a chance that
someone could guess the password. Companies are moving to two-factor au-
thentication, where users need a password and a second method of identifcation.
Password generators described in the next section are sometimes used as the sec-
ond method. .
Password Generators
Password generators are small electronic cards that generate new passwords every
minute. The same system is embedded on the main computer. When you want to
log in, you simply enter the number on the card that you are carrying. Since the
password is changed every minute, you do not have to worry about anyone guess-
ing it or intercepting it. On the other hand, you have to carry around the card. You
also have to enter a password when you log in to safeguard against loss or theft of
the card. However, the password does not have to be changed constantly because
the risks are lower. Banks and government agencies have been leading adopters of
this technology for providing secure access by employees. The cards are relatively
inexpensive, easy to confgure, and can be revoked if they are lost or an employee
leaves. The technology could become even more useful if it could be integrated
into your cell phone. You cannot ask people to carry around dozens of security
cards, so it makes more sense to integrate the security into a device they already
carry. At this point in time, no one has built software to perform this task, but it
would be relatively straightforward.
Biometrics
Biometrics is a feld of study that attempts to identify people based on biological
characteristics. The most promising devices are fngerprint and handprint readers.
As shown in Figure 5.11, there are even devices that recognize the pattern of your
iris (the colored ring surrounding the pupil of your eye). These systems work with
a simple camera that can be installed cheaply. They are being tested now for iden-
tifcation at airports and in ATMs. The Canadian government is building a large-
scale system to handle customs check-in for returning Canadian citizens.
As costs decline, the biggest drawback to biometric security devices is a lack
of standards. You can use a fngerprint scanner to log into Windows, and the Win-
dows ID can be used to get access to in-house systems. However, the biometric
Figure 5.11
Biometric devices. Several methods exist to identify a person based on biological
characteristics. Common techniques include fngerprint, handprint readers, and
retinal scanners. The iris scanner is a relatively useful technology since it requires
only a camera and is noninvasive.

data is not used beyond the single computer, and there are no standards for trans-
ferring the data securely to other servers.
Biometric security devices have some important advantages. The user does not
have to remember anything or carry keys around. They are reasonably accurate
and diffcult to fool by an unauthorized person. But the industry still needs stan-
dards so that the security information can be transferred securely and validated by
the fnal server.
Some of the worrisome issues of biometrics have held back its adoption. Many
people are concerned about privacythe perceived ability of governments to
track people if biometrics become widely adopted. Another concern is that to use
biometrics, you must frst register your information (fngerprint or iris scan) and
have that data stored within the system. People worry that if someone steals this
data, the thief could use it to impersonate anyone. The assumption is that it is easy
to change passwords if they are stolen, but diffcult to change your fngerprints.
The fallacy to this argument is that the biometric data is never stored in raw form.
Instead, a one-way hash is used that converts the raw data into a new, encrypted
set of digital data. It is impossible to retrieve the raw data from the stored set. By
using unique encoders each time the data is scanned, such as time stamps and
fuzzy adjusters, it is possible to prevent these replay attacks, where an attacker
captures data as it is entered and uses it on a different system or at a later time.
Access Control
As long as the computer can identify each user, you can control access to any
piece of data. As manager of the marketing department, you could allow other
managers to read the sales data but not change it. Similarly, as shown in Figure
5.12, the accounting department could allow managers to see the accounts pay-
Figure 5.12
Access control. In Windows, right-click the folder or fle to set its properties. Under
the Security tab, you can set permissions to any person or group.
able data, but only the accounting department would be able to modify the data
and write new checks. With a good security system, it is possible for the human
resources manager to allow employees to look up work phone numbers in the cor-
porate database but not to see salaries or other confdential information.
The common access controls available are read, write or modify, execute, and
delete. With these security features, the owner of the information can give other
users exactly the type of access they need. Windows and other operating systems
support additional permissions for folders, so you can control exactly which fles
people can see, whether they can change permissions, or even deny specifc tasks
to individuals.
As a creator of data, it is your responsibility to set the appropriate access per-
missions. Today, most of your fles will be shared through a Web site. You can set
aside different directories for each group of users and assign permissions to each
directory. To avoid accidents, you generally do not give anyone delete permis-
sions. Your main choice is which users should be able to read the data, and which
ones need to be able to change it. Of course, if multiple people have permission to
change a document, you should set the document to track changes so you can see
who made each change.
Administrator Access Rights
Virus, Spyware, and Phishing attacks present diffcult problems in terms of secu-
rity. These attack methods take over the permissions of the user. If you (the user)
have high permissions, the attacking software can gain total control over your
machine. One of the most common tricks is to install software on your computer
that provides back-door access to an attacker. Because the attacking software has
your permissions, you would generally not even know the software was installed.
In early versions of personal computer operating systems, users were often made
Administrators on the computergiving you complete control over all features
of the computer. This permission was generally needed to install new software
and new hardware devices. However, it also gives attacking software total control
over the computer. In most cases, users should not be included in the Administra-
tors group. Microsoft Vista, and Windows 7 go a step further and ask you to verify
the installation of new software. This approach was frst proposed in 1987 (Post
1987), and is designed to notify you if an attacker is trying to install software or
alter your machine. It can be a nuisance when you frst confgure a computer, but
in daily use, the warning notices rarely appear. If they do, you should think care-
fully about whether you want to allow the software to alter your computer.
Single Sign-On and Lack of Standards
Figure 5.13 shows one partial solution to the large number of passwords is to use
a single sign-on method. At least within a company, a central server handles all of
the login tasks. For example, Microsofts Active Directory uses a server based on
Kerberos to authenticate users. This information is then provided to other serv-
ers throughout the company. Users log in once and the security server provides
authentication to all of the authorized servers. At this point in time, it is not used
for access to Internet sites.
It some ways, it would be nice if this feature could extend to Web sites across
the Internet. Once you have logged into your machine, it could authenticate you to
other computers on the Internet. This way, you would need only one password
or perhaps even a biometric scanner attached to your laptop. At this point in time,
the world is not even close to this solution. The problem: lack of standards. Few
standards exist for collecting, storing, or sharing authentication data. Even if you
buy a fngerprint reader for your laptop or cell phone, it cannot be used to verify
your identity to external Web sites. So far, almost no work has been done on de-
veloping standards to support this level of authentication. The lack of standards is
less important within a single company. Within a company, the MIS department
has the authority to create and defne its own standards, which gives it the ability
to purchase and create software so that it can all be integrated. But, it will likely
be several years before everyone sits down and agrees to the standards needed to
share authentication information across the Internet.
Several vendors provide short-term solutions to manage passwords. Several
store your Internet passwords in a fle and then provide them to the server when
you log in. Most browsers have a similar type of password caching mechanism.
A couple of vendors provide USB drive-based solutions that work the same way.
Passwords are cached in an encrypted fle on the thumb drive and software deliv-
ers it to the Web site when you want to log in. Of course, if you lose the USB
drive, you will lose your login information, so you need to be careful.
Encryption
How do you protect data when unknown people might be able
to fnd it or intercept it? What additional benefts can be pro-
vided by encryption? Encryption is the foundation of many aspects of secu-
rity. For example, encryption protects messages sent across the Internet and pro-
tects fles stored on servers. Cryptography has been around for thousands of years,
but computers have radically altered the types of codes available. One important
Figure 5.13
Single sign-on. A security server handles the initial login authentication. Any other
servers or devices verify the user identifcation through a token that was issued by the
security server.
User
login
Security Server
Kerberos
RADIUS
Request
access
Web server Database
Request
access
validate
validate
feature to remember in terms of cryptography and computers is the concept of
brute force attacks. If a hacker knows the algorithm method used to encrypt a
message, it might be conceivable to have a computer test every possible key to
decode the message. The essence of stopping a brute force attack is to have a
key that is so long that it would take millions of years to try every combination.
The problem is that computers get faster every year. So encryption technologies
that were secure 20 years ago can be broken in hours today. Also recognize that a
process that takes a million years could be completed in one year using a million
computers. And, today, it is possible to fnd a million computers to use.
Encryption should be seriously considered for any communications that are
sent between computers. Without encryption, it is relatively easy for unauthorized
people to deliberately or accidentally read or change the messages. Encryption
is available with many personal computer software packages. Almost all spread-
sheets and word processors permit you to encrypt your fle when you save it. To
read it back, you have to enter the correct password. You also can fnd encryption
packages on the Internet that will protect your e-mail messages.
Single Key
For many years, single-key encryption codes were the only systems available.
Figure 5.14 shows the basic steps required to encrypt and decrypt a message with
a single-key system. Both the sender and receiver have the software that handles
the encryption and decryption. Both people also need to have the same key, which
is the diffcult part. How do you deliver a secret key to someone? And if you can
deliver a secret key, you might as well send the message the same way.
On the other hand, single-key systems are fast. They can encrypt or decrypt a
message with almost no delay. Since the late 1970s, most of the business world
standardized on the Data Encryption Standard (DES). However, this system only
supported keys of 56 bits, and by 2000, messages encrypted with DES were bro-
ken in under 24 hours by brute force attacks in various contests. Triple DES was
popular for a whileessentially encrypting the message three times. But in 2001,
the U.S. government chose a new method known as the Advanced Encryption
Standard (AES) because it is fast and users have a choice of a key length of 128,
192, or 256 bits. Keep in mind that longer keys make the message more secure
(harder to break by brute force) but increase the time it takes to encrypt and de-
crypt the message.
Figure 5.14
Single key encryption. Both the person who encrypts the message and the person
who decrypts it use the same key. The systems are fast, but it is diffcult to safely
distribute the key.
Key: 9837362 Key: 9837362
AES AES
Single key
Plain text
message
Plain text
message
Encrypted
text
Encrypted
text
Technology Toolbox: Encryption
Problem: You need to send a fle to someone without it being read by anyone else.
Tools: or single key encryption you can use Offce and protect the fle with a pass-
word. Dual-key encryption requires buying and installing certifcates. You can also
use BitLocker or other commercial programs to encrypt an entire drive or USB drive.
The easiest way to secure a data fle is to encrypt it by adding a password. With Of-
fce 2010, once the document has been created, use the menu: File/Info tab/Protect
Document /Encrypt with Password (in the Permissions section). You will be prompt-
ed to enter a password and then enter it again to ensure you typed it correctly. The
fle is encrypted with AES when it is saved and the password is required to open and
decrypt it. You can then e-mail the fle to someone and not worry about it being in-
tercepted. However, you must still fnd a way to tell the recipient the password. Ob-
viously, you cannot send it by e-mail. A phone call would be better but not perfect.
That is the heart of the problem with single-key encryption.
It is possible to purchase and install digital security certifcates to automatically
encrypt e-mail. After the certifcates are installed, and the e-mail client (Outlook) is
confgured, e-mail and attachments can be encrypted by checking a single box. The
problem is that everyone participating has to obtain and install security certifcates
and exchange public keys with everyone else. And the certifcates usually expire
after a year, so everyone has to repeat the process each year. The detailed steps are
explained on Web sites that sell the personal security certifcates.
A related issue is the need to encrypt an entire driveeither the main disk drive
on a computer or a USB drive. Laptops and USB drives are particularly danger-
ousevery week, thousands of people lose both of these items. Laptops might seem
secureyou need a password to log inbut it is relatively easy to remove the hard
drive, put it in a case and read the entire drive by plugging the case into another com-
puter. Several encryption programs can be purchased to encrypt the entire disk drive
(or volume). Windows 7 (Ultimate and Enterprise) comes with the BitLocker Drive
Encryption program. Other systems require you to purchase a commercial program.
BitLocker is installed by opening the Control Panel and selecting the BitLocker
Drive Encryption option. It works best if the computer has a Trusted Platform Mod-
ule (TPM) chip, which is often installed in business-level laptops (especially Leno-
vo). The software requires additional confguration if the computer does not have a
TPMcheck the Internet for details. Basically, the encryption keys can be stored on
a USB drive which must be inserted to startup the computer. The TPM is easier and
safer because it stores the key on a hardware chip in the computer. BitLocker is also
useful for encrypting USB drives, where a password is required to unlock the drive.
In a corporate environment that uses Active Directory and Group Policies, the
BitLocker encryption keys can be stored centrally. The policies can even require that
USB drives use BitLocker encryption or the system will be blocked from writing to
them.

Quick Quiz:
1. 1. Why would a business want to use encryption?
2. When would it be useful to set up dual-key encryption for e-mail?
3. In a typical company, which drives should use drive-level encryption?
Encryption is useful for almost any type of data that you need to protect. For
example, you can save a Word document or Excel with a password, which en-
crypts the document with AES. If you want anyone else to open the document, you
will have to provide the password. However, be sure to remember the password.
Without it, you cannot retrieve any of the data. A few companies sell software that
attempts to crack the Microsoft Offce passwords. The process takes a long time
if you use long, complex passwords, but keep in mind that these passwords might
not be foolproof.
A second common use of encryption is useful for laptops. Windows (and other
programs) include the ability to encrypt the entire hard drive. Without encryption,
it is possible to remove the disk drive and connect it to another computerwhich
will be able to read all of the data on the drive. With encryption, the drive is tied
to the specifc computer, and the data cannot be read without logging into the
specifc drive. Since laptops are commonly lost or stolen, this level of encryption
provides a useful method to protect the data from theft. The integration with Win-
dows makes it relatively easy to install. On the other hand, encryption makes it
more complicated to upgrade hard drives.
Public Key Infrastructure
Public key infrastructure (PKI) is a substantial leap in encryption technology. The
method arose from a military-political question. Given a U.S. embassy in the mid-
dle of a foreign nation that can intercept all communications, how can a secret
message be transmitted out of the embassy when there is no way to exchange a
secret key? The answer was found by two mathematicians (Diffe and Hellman),
and later refned into a system (and company) named after three other mathemati-
cians (RSA: Rivest, Shamir, and Adleman). The solution is to create an encryption
system that uses two keys: a public key and a private key.
Reality Bytes: Don't be an Idiot
Despite the repeated press reports (and fake actions in Hollywood movies), com-
puter security systems have improved considerably over the past few years. But, a
weak link exists in every system: the component between the chair and the keyboard
(you). It is easy to break into a system or steal money if you give away your pass-
wordor download and run software that gives control to hackers. In a 2010 scam,
criminals called thousands of people and ran a basic scam. They claimed to provide
free security checks and used a variety of social engineering scams to get pass-
words and account information. Microsoft paid for a survey of 7,000 computer users
in the U.K., Ireland, the U.S., and Canada. In total, 15 percent of the respondents
had received a phone call from the scammers. In Ireland, the rate was as high as 26
percent. Of those who had received the call, a whopping 22 percent followed the
instructions and gave the scammers access to their computers. The victims lost $875
on average (but only $82 in Ireland). Perhaps they will consider it an educational fee.
When a random person calls and asks for passwords, accounts, to run software or go
to a specifc Web site; just hang up the phone.
Adapted from Nathan Eddy, Microsoft Survey Reveals Extent of Emerging Internet
Phone Scam, eWeek, June 17, 2011.
Dual Key Encryption
The essence of a dual-key system is that it takes both keys to encrypt and decrypt
a message. Whichever key is used to encrypt the message, the other key must be
used to decrypt it. Figure 5.15 illustrates the process. The beauty of the system is
that anyone can be given your public keyin fact, this key can be published in
a directory. Then, whenever someone wants to send you a secure message, he or
she simply uses the RSA algorithm and your public key. At that point, the mes-
sage is gibberish and can only be decrypted using your super-secret private key.
No one can read or alter the message. However, someone could destroy it before
it reaches you.
Todays Web browsers use this method to encrypt credit card transmissions.
The Web server sends your browser a public key. The browser encrypts the con-
tent and sends it across the Internet. Only the Web server can decrypt the con-
tentsusing the private key. A similar system called Pretty Good Privacy (PGP)
is available on the Internet to encrypt e-mail messages.
The one drawback to dual-key encryption systems is that they tend to be
slowparticularly for long messages. One common solution is to use dual-key
encryption to establish the identity of the parties and to exchange a one-time se-
cret key to be used for the rest of the transmissions. The single-key system is fast
and protects the transmitted data, and the initial dual-key system makes it possible
to distribute the secret key without anyone stealing it.
Authentication
A second aspect of dual keys has even more uses. PKI can be used for authenti-
cation. Consider the case where Bob works for a bank, and she receives a mes-
sage that claims to be from Alice, and it says to pay you $1 million. How does
Figure 5.15
Dual-key encryption. Alice sends a message that only Bob can read. With a dual-key
system, one key encrypts the message, the other decrypts it. Once Bobs public key is
applied to the message, only Alices private key will retrieve the message. Keys are
usually very large prime numbers.
Alice
Bob
Message
Public Keys
Alice 29
Bob 17
Message
Encrypted
Private Key
13
Private Key
37
Use
Bobs
Public key
Use
Bobs
Private key
Bob know that the message is authentic and not forged by you? Consider the case
where Alice want to pay you some money (but only $100).
Figure 5.16 shows the answer. To make sure that only Bob can read the mes-
sage (and that no one else can modify it), Alice encrypts it with her public key. To
ensure that the message is authentic, Alice also encrypts it with her private key.
Remember that the keys always work in pairs. When Bob receives the message,
she uses Alices public key and her private key to decrypt it. If the message is de-
crypted correctly, then it was not modifed in transit, and it could only have come
from Alice. This process is used to create digital signatures. In 2000, the federal
government passed a law declaring digital signatures to carry the same legal au-
thority as a traditional signature for legal purposes.
Certifcate Authorities
The proper name for dual-key encryption is public key infrastructure (PKI). Why
is the word infrastructure so important? Think about how a hacker might attack
the system in Figure 5.16. What if Alice did not know much about technology and
encryption? So, posing as Alice, you create a private key and publish the public
key in a directory under Alices name. Then you send your e-mail message to the
bank pretending to be Alice, using her public key and asking the bank to pay
you $1 million. The message decrypts fne, and Bob believes the message is legiti-
mate. Similar problems can arise by impersonating the bank.
To make the PKI system work, it is critical that the directory of public keys
accurately represent the people indicated. So, some organization needs to be in
charge of the public directory, and people who wish to use it need to verify their
identity before registering a public key. At the moment, several commercial com-
panies operate as certifcateauthorities and sell digital encryption certifcates to
anyone. Almost no oversight or regulation exists in the industry. At one level, the
browser developers (Microsoft) control which companies get listed as trusted root
providers in the browser, but there is no governmental regulation. The U.S. mili-
Figure 5.16
Dual-key encryption for message authentication. Alice sends a message to Bob at
the bank. Using her private key ensures that the message must have come from her.
Using Bobs public key prevents anyone else from reading or changing the message.
Alice
Bob
Public Keys
Alice 29
Bob 17
Private Key
13
Private Key
37
Message
Message
Encrypt+B Encrypt+A
Transmission
Use
Alices
Private key Use
Alices
Public key
Use
Bobs
Public key
Use
Bobs
Private key
Encrypt+B+A
tary (and probably others) does run its own certifcate servers but those are used
only within the military networks.
These public companies, including Verisign and GoDaddy sell digitalcertif-
cates that verify the identity of the person and generate the public/private key
pairs. Companies and individuals can purchase these certifcates, and you are sup-
posed to verify your identity before receiving the certifcate. However, in 2001,
Verisign announced that it accidentally issued a digital certifcate to an imposter
who claimed to be from Microsoft. Eventually Verisign caught the mistake and
invalidated the certifcate, but the incident points out that the process is far from
foolproof. The troubling point is that for the PKI system to work, the certifcates
and keys must be controlled by a trusted organization.
A handful of public CA frms exist todayleading to some competition in
price, but there is still no oversight. In fact, you can easily install a server and
generate your own security certifcates. These certifcates are useful for your own
employees. Companies can install them on employee computers and use the cer-
tifcate to identify a user instead of relying on passwords. Generating your own
certifcates reduces your costs. However, they are not useful for e-commerce, be-
cause your company will not be recognized by the customers browser. You can
look at your Web browser options to see the list of Trusted Root Certifcation
Authorities. This list represents CAs that are automatically recognized by your
browser. A certifcate issued by anyone else creates a warning message.
Encryption is used in many areas in computer security, but it is not the only
factor in security. It is important to protect data during transmission, and it is use-
ful to encrypt sensitive data on disk drives. But, decryption keys need to be stored
someplace, so attackers often target those storage locations. So, the security threat
simply shifts to a different level. Any security system needs to incorporate mul-
tiple levels of protection. Encryption is one level of security that is designed to
solve specifc problems.
Encryption also presents issues to societywhich are covered in more detail
in Chapter 14. For now, note that encryption is available to everyone: individu-
als, businesses, criminals, spies, and terrorists. Consequently, many government
Reality Bytes: Microsoft gets Binged
Just to show that insider theft can happen in any company, Microsoft sued a former
Bing toolbar promotions director for allegedly stealing $460,000 from the company
and trying to walk away with another $1.5 million. Microsoft claims that Robert
D. Curry set up a fake company (Blu Games) and submitted fake invoices for ser-
vices it never performed. Microsoft claimed that Curry created false documents and
even forged his managers signature. As promotions director, his job was to convince
companies to encourage people to download the Bing toolbar to increase the use of
the Bing search engine. Microsoft spotted the problem only after Curry convinced
the fnance department to increase the amount of funds for the project to $3.7 mil-
lion. He was then going to transfer the $1.5 million to his company through a legiti-
mate vendor (Pentad). But, the size of the transaction apparently caught the eye of
internal auditors.
Adapted from Gregg Keizer, Microsoft Claims Employee Stole $460,000 from the
Company, Computerworld, January 27, 2011.
agencies want methods to decrypt messages. The U.S. government has proposed
several encryption methods that contain back-door keys that would allow agents
to decrypt messages easily.
Additional Security Measures
What non-computer-based tools can be used to provide addi-
tional security? A fundamental issue in computer security is that logical
controls are never enough to protect the computer. For example, anyone who has
physical access to the computer can either circumvent the security controls or de-
stroy the data. Besides, many employees have extended access to the data and
applications. To be safe, you need to implement some standard business policies.
You also need to train users periodically to warn them about new attacks and re-
mind them to never give their passwords to anyone else.
Audits
Accountants have long known that to maintain security over data, it is necessary
to perform audits. There are too many ways for unscrupulous people to make
changes to stored information. Audits are used to locate mistakes and to prevent
fraud. Existing criminology practice states that in many cases, the threat of getting
caught (by an audit) will convince most people to be more careful and to avoid
fraudulent behavior. The same principles extend to security audits. By monitor-
ing computer activity, auditing fnancial records, and periodically checking to see
whether everyone is obeying security regulations, users are encouraged to follow
the security guidelines of the company.
Of course, audits cost money and they interfere with the daily operations of the
frm. As a result, it is important to schedule audits carefully and to keep an eye
on the costs as well as the potential benefts. Several professional organizations
(such as the EDP Auditors Association) can help security employees learn
more about the latest technologies and teach them what to look for in audits. The
Reality Bytes: Hacker Arrests
The hacker group Anonymous obtained huge publicity during the WikiLeaks
discussions of late 2010. The group encouraged people to download its software so
their computers could be used to attack business and governmental sites. In mid-
2011, the Spanish police arrested three individuals claimed to be senior members
of the Anonymous group within Spain. A computer found at the home of one of the
individuals was alleged to have been used to launch attacks against many Web sites.
Anonymous responded by initiating a denial of service attack against Spanish police
computers, and denying that the three arrested were part of the group; or that Anony-
mous was even a group. Earlier, Dutch police arrested two teenagers in December
2010; and the UK police arrest fve males in early 2011. In the U.S. the FBI searched
homes of several alleged members of Anonymous, but no arrests were made. Turkish
police arrested 32 people in early June 2011. In early June 2011, NATO declared that
Anonymous and similar groups were a threat to international security.
Adapted from Cassell Bryan-Low, Spain Arrests Three in Hacker Crackdown, The
Wall Street Journal, June 10, 2011; and Ben Rooney, Turkey Arrests 32 in Hacker
Swoop, The Wall Street Journal, June 13, 2011.
American Institute of Certifed Public Accountants (AICPA) also pro-
vides standards and audit guidelines that are useful at combating fraud.
Physical Access
Because it is so diffcult to provide logical security to a computer, other mecha-
nisms have been developed. Many of them rely on controlling physical access to
the computer. For instance, computers and terminals should be kept in controlled
areas. They must certainly be kept away from visitors and delivery people. Many
types of locks and keys can be used to protect terminals and personal computers.
Similarly, all documents should be controlled. Paper copies of important reports
should be shredded. All fnancial data is routinely audited by both internal and
external auditors. Sometimes hidden control data is used to make sure procedures
are followed.
Monitoring
Another effective security provision is to monitor access to all of the data. Most
computers can track every change to every fle. They can keep a log of who ac-
cesses each fle. They track every time someone incorrectly enters a password. An
audit trail of every fle change can be provided to management. That means it is
possible to fnd out who changed the salary data, what time it was changed, and
from which terminal.
Remember that every device connected to the Internet is given a unique number
as an address. Every Web site that you visit can track this number. In some cases,
you can only be identifed down to the company you work for, but in many situ-
ations, companies can monitor exactly what each machine is doing at any time.
Additional software can be installed on computers to provide even more detail
including storing all keystrokes.
Reality Bytes: Find the Weakest Point
A common problem in security is that you can strongly protect one point, only to
lose because the system is vulnerable at a different location. Security certifcates are
strong protection for transmitting data on the Internet. The encrypted traffc is ex-
tremely diffcult to break. But, in 2011, someone, possibly in Iran, was able to obtain
server security certifcates from Comodo in the names of Google, Microsoft, Skype,
and Yahoo. Melih Abdulhayoglu, the CEO and founder of security company Co-
modo noted that One of the origins of the attack that we experienced is from Iran.
What is being obtained would enable the perpetrator to intercept Web-based email/
communication and the only way this could be done is if the perpetrator had access
to the countrys DNS infrastructure (and we believe it might be the case here). For-
tunately, Comodo identifed the attack and revoked the bogus certifcates. Browser
vendors (Google, Microsoft, and Mozilla) quickly followed by tagging the certif-
cates as invalid within their browsers. The ability to withdraw and fag certifcates
is an important Web security feature. But it still requires that certifcate authorities
identify problems and react quickly.
Adapted from Gregg Keizer, Firm Points Finger at Iran for SSL Certifcate Theft,
Computerworld, March 23, 2011.
Technology Toolbox: Assigning Security Permissions
Problem: You need to share some fles with members of the marketing department.
Tools: Windows has the ability to set detailed security permissions on folders and
fles.
In a student lab, you might not
have permissions to create fold-
ers and set fle permissions. Some
versions of Vista might not support
detailed security options. Browse
to a location where you want to
put a shared folder. Create the new
folder.
In a business setting, the users
and groups have probably already
been created by the network ad-
ministrator. If you need to cre-
ate groups and users, use the Start
menu/All Programs/Administrative
Tools/Computer Management. If
you do not see the administrative
tools, you need to enable it by set-
ting the properties of the main task-
bar, or use Start: compmgmt.msc /s. Right-click the Users icon and select New User.
Make up a username and password for a user. Repeat this until you have three sam-
ple users. (If the users are already defned on the network, you can skip this step.)
Right-click the Groups icon and select New Group. Name it Marketing and provide
a description. Click the Add button and enter the usernames of the three users you
just created.
Return to your folder and right-click the folder icon and select the Sharing and
Security option. Click the option to share the folder and name it Marketing. Click the
Permissions button and Remove the Everyone group. Add the new Marketing group
and assign Read permission. Click the Apply button, then click the Security tab.
Click the Add button and enter the Marketing group so they have read access. Click
the Add button and enter the name of one of the users. Give this person Modify (and
Write) permissions. This user will be able to read and change fles stored in this
folderthe others will only be able to read them. Log on as one of the new users and
testing the fle permissions. Note, if you are familiar with command-line commands
(DOS), you can use the runas command without logging off. When you are fnished,
remove the Marketing group and users using the Computer Management tool.
Quick Quiz:
1. Why is it important to defne groups of users?
2. Why is it important to delete this test group and users when you are fnished?
Hiring and Employee Evaluation
Because insiders raise the greatest security issues, it makes sense to be careful
when you hire employees. Employers should always check candidates referenc-
es. In more extreme situations, employers can check employee backgrounds for
criminal records. There are several instances of disgruntled employees causing se-
curity problems. In many cases, the best security solution is to establish close re-
lationships with your employees and encourage teamwork. Employees who work
closely together can defuse potential problems and informally monitor the work
of other employees. Figure 5.17 notes that several Web sites will search public re-
cords to perform basic background checks for small businesses. Validating social
security numbers is an important step for many U.S. businesses.
Computer Forensics
How do you prove the allegations in a computer crime? Some-
times, stopping computer attacks is not good enough. You want to catch the crooks
or attackers and have them charged with the appropriate crime. The problem is
that you have to be extremely careful when you collect evidence that will be used
in a criminal case. In particular, the investigator has to be able to guarantee the au-
thenticity of the evidence from the moment it is collected to when it is presented
in court. This process can be tricky with digital evidence. It is not a task for ama-
teurs. You have to bring in a professional investigator to handle the evidence cor-
rectly. Of course, few police departments have people trained in computer crime,
and they are probably busy chasing murderers and drug dealers.
Figure 5.17
Employee background checks are important. For a fee, several websites help small
businesses perform basic background checks to verify SSNs and check public
criminal records.
Audits
Monitoring
Background checks:
http://www.lexisnexis.com/risk
(bought ChoicePoint)
http://www.knowx.com/
(also lexis nexis)
http://www.casebreakers.com/
http://www.publicdata.com/
Some private companies help with investigations, but make sure you contact
your lawyers and the prosecuting attorneys early in the process. Several techni-
cal companies exist to help examine computer evidence, such as recovering data
from hard drives and decrypting fles. Throughout the entire process, you have to
keep good records. Most actions have to be logged; and be sure to record the date,
time, and people involved. Also, remember that computer logs and backup tapes
often get recycled after a certain time, so be sure to maintain secure copies. Figure
5.18 shows some of the capabilities of commercially available forensic hardware
and software. Typically, a drive is cloned with a special device to prevent data
from being altered on the original drive. Copies of the cloned drive are then cre-
ated and used for searching. Special software recovers deleted fles and performs
keyword searches. Photos are often important and are organized by the software.
Some software can highlight time sequences based on fle dates, computer logs,
and the browser history. Increasingly, the challenge lies in the vast amount of data
stored on computers. Computerized searches are required to cover the amount of
data, but a human investigator is needed to identify keywords and look through
potential matches. Even if the computer can fnd photo and movie fles, the com-
puter is not very good at fnding objects or classifying images in photos or video.
The task of examining computer fles is time consuming and highly detailed. But,
several job openings are available for specialists trained in computer forensics.
E-Commerce and Cloud Computing Security Issues
What special security problems arise in e-commerce? E-commerce
and cloud computing use the same security technologies available to any busi-
ness. However, some aspects of online businesses are more sensitive and require
more careful security planning. These issues are highlighted in this section, with
Figure 5.18
Computer forensics. Careful steps must be followed to protect the integrity of the
evidence and the chain back to the alleged criminal. Once the data is copied, special
software is used to recover deleted fles and search for keywords, images, and
activity sequences. A human investigator is still needed to interpret the results and
direct the search.
Original
drive
Exact
copy
Write blocker:
Physically prevent
data from being
altered on the
original drive.
Software:
Verify copy.
Tag/identify files.
Scan for key words.
Recover deleted files.
Identify photos.
Attempt to decrypt files.
Time sequence
Browser history
File activity
Logs
a discussion of the common solutions. The earlier sections already covered the
issues of data transmission and the importance of encryption. Most Web sites use
dual-key encryption through a system known as secure sockets layer (SSL). On-
line Web services should use SSL for all communications. Servers also need to be
protected against direct attacks. One layer of protection is provided by segmenting
the network holding the servers using frewalls. Additionally, intrusion detection
and prevention methods can be used to monitor for various attacks. One other
serious problem is denial of service attackswhich are diffcult to solve for an
individual company.
Teft of Data from Servers
Because of the powerful encryption systems available, interception of transmis-
sions is a relatively minor problemas long as you use the encryption techniques.
Instead, the servers connected to the Internet have become tempting targets. Sev-
eral incidents have been reported of hackers stealing millions of records of cus-
tomer credit card data from e-commerce frms. While credit laws protect consum-
ers, the loss of a credit card is still painful and time consuming. In addition, the
e-commerce frm faces liability issues if it did not adequately secure the data.
Securing servers uses the same technologies as any computer system: (1) make
sure the latest operating system patches have been applied, (2) set access control
rights to the smallest number of people possible, (3) encrypt the sensitive data, (4)
hire trusted employees, and (5) monitor access to the sensitive data. A sixth step
(frewalls) is explained in a later section. Figure 5.19 shows the primary categories
that are required by the major credit card companies. Merchant banks and the card
companies require vendors to agree to these conditions or they will not allow you
to handle credit card data. They also require you to pay for periodic tests by an
outside approved company. No support or negotiation is provided for any of the
terms. The card companies have effectively become a monopoly in their attempts
to push security (and blame) onto the merchants. Consequently, it is increasingly
diffcult for small merchants to handle their own credit-card processing. Instead,
anything less than huge frms should use payment mechanisms through a third-
Figure 5.19
Payment card industry security rules. Any merchant or business that handles
credit card data must agree to abide by these rules. Source: https://www.
pcisecuritystandards.org
1. Install and maintain a frewall confguration to protect cardholder data.
2. Do not use vendor-supplied defaults for passwords.
3. Protect stored cardholder data.
4. Encrypt transmission of cardholder data across open, public networks.
5. Use and regularly update anti-virus software.
6. Develop and maintain secure systems and applications.
7. Restrict access to cardholder data by business need to know.
8. Assign a unique id to each person with computer access.
9. Restrict physical access to cardholder data.
10. Track and monitor all access to network resources and cardholder data.
11. Regularly test security systems and processes.
12. Maintain a policy that addresses information security.
party, such as PayPal, Google, FirstData, or even Amazon. These frms all charge
fees for handling the credit cards, but you avoid most of the risk and the costs of
maintaining detailed security systems.
Firewalls
The Internet and e-commerce add some complications to protecting company
data. You need to give customers access to some important company data to pro-
vide the best e-commerce site. For example, customers like to know if an item is
in stock before they place an order. To offer this service, you need to connect your
Web server to your company inventory system. But any time you open a connec-
tion from the Internet to your company data, you have to be extremely careful to
control that interaction. Security access controls and database security controls are
two important provisions.
Beyond access control, simply connecting your company computers to the In-
ternet could cause problems within the network itself. You do not want company
network traffc being sent to the Internet, and you do not want outsiders to be able
to see your internal computersgiving them the chance to try and break into your
servers. Figure 5.20 shows how frewalls are used to isolate sections of the net-
work. Firewalls are essentially routers that examine each packet of network data
passing through them and block certain types to limit the interaction of the compa-
ny network with the Internet. The Internet protocols were designed as an open net-
work to transport many types of data and to enable computers to connect in many
different ways. For example, servers have logical ports on which they listen for re-
Figure 5.20
Firewalls. Firewalls are packet analyzers that are used to keep local data from
moving to the Internet and to limit the actions allowed in from the Internet. Firewalls
examine each packet of data on the network and block certain types to limit the
interaction of the company network with the Internet.
Company PCs
Internal company data servers
Internet
Firewall router
Firewall router
Keeps local data
from going to
Web servers
Examines each
packet and
discards some
types of requests.
quests. Since only a few of these ports are used for common Internet activities, the
frewall is confgured to block all of the other ports to prevent outsiders from fnd-
ing a back way into one of your servers.
Firewalls are also used to segment your network. Internet traffc can be con-
trolled based on a set of rules. These rules can include the IP source and des-
tination address, the incoming and outgoing ports, and the protocol or primary
purpose of the packet. By adjusting the rules, security experts can force Internet
traffc to specifc servers and prevent that traffc from going anywhere else on the
network. It is common practice to put Web servers in a special section of the net-
work that only allows certain traffc into and from the servers.
Intrusion Detection and Prevention
Monitoring the networks and servers is an important step in providing a secure
system. If nothing else, you need to know if someone has broken into your serv-
erspreferably before the police knock on the door and tell you. As shown in Fig-
ure 5.21, an intrusion detection system (IDS) is a combination of hardware and
software that continuously monitors the network traffc. The hardware is similar
to that used in a frewall, but instead of blocking the packets, it performs a more
complex analysis. The systems use a set of rules to monitor all Internet traffc and
search for patterns. For instance, a common attack often begins with a sweep of a
targets network to look for open ports. The IDS observes this repeated scanning,
blocks the requests, identifes the source, and sends a warning to the security man-
ager. An IDS is an effective monitoring tool, but the cheaper ones tend to generate
too many false warnings.
Figure 5.21
Intrusion detection and prevention. The IDS monitors packets and logs from routers,
servers, and special network devices. It has rules to recognize certain attacks and can
monitor for changes in traffc patterns. An IPS can shut off a network connection if it
detects a threat.
IDS/IPS
Company PCs
Collect packet
info from
everywhere
Analyze packet data in real time.
Rules to evaluate potential threats.
IPS: Reconfigure firewalls to block IP
addresses evaluated as threats.
An intrusion prevention system (IPS) begins with an IDS but adds the abil-
ity to shut down a network connection that has been identifed as a threat. For
instance, if the system detects a repeated scan from a specifc IP address, it will
set a rule in the main Internet router and tell it to discard any incoming packets
from that IP address. Typically these blocking rules are set in place for a specifc
amount of time and then released after the immediate danger has passed.
The main problem with and IDS and an IPS is that servers today are constantly
under attack. An IDS will identify almost continuous scans of various types from
around the world. An IDS that triggers alerts for every one of these attacks would
constantly be sending out warnings. Humans would be overloaded and ignore the
warnings. Using an IPS to block out some of the obvious scans is a partial answer.
But, remember that many organizations use network address translationwhich
means IP addresses are shared and reused. So, if a crazed lunatic from a large uni-
versity uses a PC to scan outside servers, the IPS on that server could respond by
blocking the server for a large chunk of users at that university. The problem with
an IDS and IPS is that it is diffcult to draw this line between identifying attacks
and allowing people to use the servers.
Denial of Service
Denial-of-service attacks have gained importance in the last few years. The es-
sence of an e-business site is the ability to reach customers 24 hours a day. If
someone foods the site with meaningless traffc, then no one can use the service
and the company may go out of business. Several variations of DoS have been
used in the past couple of years, sometimes dragging down large chunks of the
Internet at one time. Most of the techniques take advantage of faws in the Internet
protocols or in their implementation on a particular type of server. Figure 5.22 il-
lustrates the process. A hacker breaks into some weakly protected computers and
loads a special program that is hidden. On a signal, the machines all send requests
to the targeted server. With some known Internet design faws, these messages can
Reality Bytes: RSA Hack from the Top
RSA Security, a subsidiary of EMC, is one of the big players in computer security.
The company was one of the frst to develop and sell dual-key encryption and SSL
certifcates. When the original patents expired and competitors undercut its prices,
the company turned to other technologies and services. In particular, the company is
known for its SecurID, which is a small device that generates a new password key
every minute. The keys are synchronized to the server, so the passwords are only
good for one-time use and change constantly. The process makes it relatively dif-
fcult for anyone to attack a login to a protected server. So, instead of attacking the
protected computers, in 2011, hackers went after RSA itself. Apparently, someone
was able to breach the security and obtain access to the underlying keys. These at-
tackers then used the data to break into computers at defense contractor Lockheed
Martin. Not long after, RSA issued millions of replacement security tokens for virtu-
ally every customer.
Adapted from Siobhan Gorman and Shara Tibken, Security Tokens Take Hit, The
Wall Street Journal, June 7, 2011.
be multiplied so that a few thousand messages become millions and bog down
the server. This type of attack is hard to trace to the original source unless inves-
tigators fnd monitor logs on the zombie machines. Several Internet sites provide
simplifed instructions on how to perform these attacks, so even weak hackers or
script kiddies can create havoc.
In many cases, a hacker has to break into thousands of computers to turn them
into zombies that can be used for DoS attacks. Some attackers use scanning tools
to search for PCs that lack critical updates and then attack individual comput-
ers. These attackers then install their software and link the computers to a central
server that sends control messages to the zombies. The attackers can then lease
these botnets to other hackers. An interesting twist to this process arose in 2010
and 2011 during the release of war data and an investigation of Julian Assange,
the founder of Wikileaks. A hacking group (Anonymous) decided to use their bot-
net to attack Web sites of people and organizations that were critical of Wikileaks.
They also decided to expand the size of their botnet by letting people volunteer to
have their computers become zombies. Of course, they did not use those terms,
but apparently, some people did volunteer. Voluntarily adding your computer to a
botnet conducting illegal activity is a really bad idea.
Stopping a DoS attack on your servers can be exceedingly diffcult. Most com-
panies will not have the resources to deal with the attack. The goal of DoS is to
food the network and servers with activity. One answer would be to add servers
or network capacity to handle the load. One possible solution is to pay for cloud
services such as Amazons EC2 which spreads the processing across multiple
servers in many locations and has multiple Internet connections around the globe.
But, servers and bandwidth cost money, and the food attacks could dramatically
increase these costs.
Figure 5.22
Denial of service attack. A bored hacker frst breaks into weakly protected home and
school systems and loads a hidden program on thousands of machines. At a signal, all
of the zombie machines send messages to the server, preventing everyone else from
using it.
Zombie PCs at homes,
schools, and businesses.
Weak security.
Break in.
Flood program.
Coordinated flood attack.
Targeted server.
Occasionally, researchers and Internet-based companies will call on people to
fnd a way to reduce the threat from DoS attacks. For a while, it was suggested that
ISPs should be responsible for their clients. In a way, the imposition of data caps
by ISPs is a response to this threat. If a client computer is spewing huge amounts
of datait is likely that this data is bad, either as spam or DoS attacks. The ISP
should be able to recognize these offenders and shut them down. Of course, ISPs
are reluctant to shut down paying customers; however, they are fnding it easier to
bill them more money for excessive use. In effect, these actions push the responsi-
bility back to the owner of the computers. Microsoft helps users a little by embed-
ding tools into Windows to monitor from some known highly-active spyware and
botnet tools. In the end, it makes sense to use economics to encourage people to
monitor their own computers.
Privacy
If you have to track everyones computer use to improve secu-
rity, what happens to privacy? Computer security is a complex topic.
One of its most challenging aspects is that many of the tools available to improve
security can also reduce individual privacy. For example, a single sign-on system
used across the country could be used to track people through their purchases,
when and where they used a computer, when they passed through toll booths or
boarded planes, or almost anything else. Even without single sign-on, the security
systems enable companies to track detailed usage by employees and customers.
These issues are explored in detail in Chapter 14. For now, you need to remem-
ber that many proposed security controls scare even honest peoplebecause of
the loss of privacy. As a businessperson, you must be aware of these problems and
establish policies to minimize the effects and keep customers happy.
Reality Bytes: Anything on a Network has Security Risks
In April 2011, Sony realized that someone had hacked its game network and taken
complete control over the server. The network was used to run the PlayStation video-
game service for more than 75 million customers. Sony had to take the entire system
offine and rebuild many of the components. It took almost two months to restart the
network. Later, Sony found attacks on a second network. Nintendo also found at-
tacks against its game network. Many early networks were built without security as
a primary component. Taking security on as an afterthought often causes problems.
In June, Sega confrmed that its gaming network had also been attacked and that al-
most 1.3 million people lost their personal information. However, the Sega personal
information was mostly limited to name, birthday, and e-mail addresses. The com-
pany uses external credit card processors to handle payments and does not store any
fnancial data on its own servers.
Adapted from Ian Sherr, Sony Shuts Down PlayStation Network Indefnitely, The
Wall Street Journal, April 25, 2011; and Daisuke Wakabayashi, Sega Confrms Cy-
berattack, The Wall Street Journal, June 19, 2011.
Summary
Companies have to trust employees, consultants, and business partners, but this
group presents the greatest security threats. To maintain secure systems, compa-
nies often perform background checks on employees, separate tasks so multiple
people can observe the work of others, and monitor fnancial transactions. Natu-
ral disasters are a threat to the physical assets, but their business damage can be
minimized by having continuous backups in different locations. Cloud computing
makes these facilities available even to small companies. The Internet provides
more avenues of attack for outsidersparticularly through phishing schemes and
attacks on unpatched systems. The best defenses are to install all current operat-
ing system patches, to assign access rights carefully, and to monitor the computer
usage with an intrusion detection system. However, denial-of-service attacks are
particularly hard to prevent.
Encryption protects data during transmission. It is particularly useful for send-
ing credit card data over the Internet. It can also be used to provide digital sig-
natures that authenticate users to validate the source of messages. If computer
crimes are detected, the law requires careful collection and analysis of the data us-
ing validated computer forensic tools. Experienced experts are generally required
to conduct these investigations.
A Managers View
Computer security is a critical issue for any company. For Web-based busi-
nesses, careful controls and continual vigilance are mandatory. Information
systems have many potential weaknesses and threats. But overall, electronic
security can be stronger than any other formif it is maintained by experts.
Encryption is a key component in securing systems, communications, and
protecting privacy. The drawback to security is that it imposes limits on em-
ployees and customers. Some tasks become more diffcult and some loss of
privacy occurs. To protect corporate and personal data, we have to be able to
trust the people who collect the data.
Key Words
Web Site References
Computer Crime
CERT www.cert.org
Computer Crime Research Center www.crime-research.org
Computer Security Institute www.gocsi.com
FBI: Internet Crime Complaint www.ic3.gov
Interpol www.interpol.int/Public/
TechnologyCrime
US CERT www.us-cert.gov
National Security Agency www.nsa.gov
Department of Justice/cyber crime www.cybercrime.gov
Review Questions
1. What are the primary threats to information processed by computers?
2. How do viruses spread over the Internet? How do you stop them?
3. Why is data backup the most important aspect of security?
4. How do you protect yourself from false Web sites?
5. Why is it important to ensure updates and patches are installed soon after
they are released?
6. What methods are available to identify computer users?
7. What are access controls and how are they used to protect information?
8. What are the advantages and disadvantages of dual-key encryption compared
to single-key encryption?
Advanced Encryption Standard (AES)
Application service providers (ASPs)
Authentication
Biometrics
brute force
certifcate authorities
cold site
denial of service (DoS)
digital certifcates
digital signatures
Firewalls
hot site
intrusion detection system (IDS)
intrusion prevention system (IPS)
Kerberos
malware
phishing
Pretty Good Privacy (PGP)
private key
public key
replay attacks
script kiddie
secure sockets layer (SSL)
single sign-on
social engineering
spyware
two-factor authentication
virtual private network (VPN)
virus
zero-day attacks
9. How can dual-key encryption be used to authenticate a message?
10. Why are certifcate authorities so important in a public key infrastructure?
11. Why do we not have a better system to identify users of Web sites?
12. What threat are audits trying to protect against?
13. If you wish to investigate a computer attack, what is the most important rule
you need to follow?
14. What are the main issues in protecting e-commerce Web sites?
15. Why are wireless transmissions more of a security problem than wired
systems?
16. What is a denial-of-service attack? Why is it so important in e-commerce?
Why is it so diffcult to prevent?
17. What is a frewall and how does it protect a company from Internet attacks?
18. How does an intrusion detection system work?
Exercises
1. Run Windows update on your computer and select the custom option. How
many updates are available? List them. Do the same for Offce update.
2. Review newspapers, magazines, and Web sites and fnd one case of an actual
security problem, such as theft of data. If possible, identify the value of any
losses and whether the criminal was caught. Summarize steps that could have
been taken to prevent the incident.
3. Find at least two biometric login devices and their prices.
4. Research the current access control methods provided by your favorite social
network site. Briefy describe how you would assign permissions to two sets
of data to two groups of people. Group A can see the frst set of items but not
the second, and group B can see the second set but not the frst.
5. Research the status of password-crackers. What length of password can be
broken in a reasonable time (say less than a month)?
6. Assume you are running a small business that needs to hire employees. Find
a site that you can use to do a simple background check on the applicants.
Explain why you chose this site.
7. The hacker group Anonymous uses software to stage denial-of-service
attacks against various targetsoften political. Unlike most attackers, the
group sometimes publicly invites people to sign up and add your computer
to its group of attackers. Explain why it would be a bad idea to join this
organization.
8. Talk to a friend or relative who works for a large company and ask about the
types of monitoring that the company performs on its employees.
9. Use the Windows Task Manager to list all of the active processes on your
computer. Search the Internet to identify the purpose of at least ten of the
processes.
10. Assume you have a server with about 500 gigabytes of data. Identify the
hardware and software you could use to make backups. Be sure to specify the
price and estimate the amount of time it will take to back up the data.
Technology Toolbox
11. Write a short paragraph or spreadsheet. Encrypt and save it and e-mail it to a
teammate. Identify a relatively secure way to exchange the password.
12. Find out if your computer has a Trusted Platform Module (TPM) that could
be used with BitLocker drive encryption.
13. Most computer systems have administrator accounts that provide complete
access to the computer for at least one person. Briefy explain what security
issues this might cause.
14. Use the Web to fnd the best price on a security certifcate that you can install
on a business Web server. Assume that you will need the server for at least
fve years. To be safe, check your Web browser to ensure that the certifcate
authority is listed in the Trusted Root certifcates.
15. If you have the appropriate network permissions, or using your own
computer, create a Marketing group and three users on your computer. Create
a folder and set the permissions so that the Marketing group can access the
fles in the folder. Add your instructor to the group and include a test fle that
he or she can read.
16. Check the security permissions on your computerparticularly the My
Documents folder (or wherever you store most of your fles). Is the folder
secure or should you set different permissions?
Teamwork
17. Create a subdirectory on a computer that enables you to set access rights.
Select a user or group and set permissions so members of that group can read
the data but cannot change it. All other users (except yourself) cannot read
the data.
18. Conduct a small survey of students (not in your MIS class). Find out how
often they back up their data, the last time they updated their operating
systems, and how many of them have been infected by a virus in the last 6,
12, and 24 months.
19. Create a simple chart in a spreadsheet and encrypt it. Send a version to each
person on your team using a different password/key so that they can add a
couple rows of data and return it. Combine the results into one spreadsheet.
Comment on any problems you encountered.
20. As a group, create a list of questions that you would ask a potential employee
who is interviewing for a job as a computer security expert at your company.
21. Each person should describe a movie scene or TV show episode that involve
some type of computer security or attack. Explain whether the event is
realistic. Swap the descriptions with team members and have each person add
a list of actions that could be taken to prevent the attack.
22. What privacy problems might exist at Rolling Thunder? What rules or
procedures should you enact to avoid problems? Write a privacy statement
for the company.
23. If Rolling Thunder Bicycles adds an Internet site to order bicycles and deal
with customers, what security procedures should be implemented to protect
the data?
24. Research the costs and steps involved in setting up a secure Web server for
Rolling Thunder that can be used to sell bicycles over the Internet.
25. Write a disaster plan for Rolling Thunder. Identify how the backup tapes will
be handled and the type of system you will need if a natural disaster hits.
26. Identify and briefy describe the top security threats that would be faced by
Rolling Thunder. Outline the primary steps you would take to reduce the
risks.
27. Particularly in terms of security, would Rolling Thunder be better off just
hiring a company to build and run an e-commerce site?
Additional Reading
Bequai, August. Technocrimes, Lexington, MA: Lexington Books, 1989.
[Security Pacifc and other cases.]
Harriss, H. 1999. Computer Virus Attacks Have Cost Businesses $7.6 Billion
in 1999. Computer Economics Report dated June 18, 1999. (http://www.
info- ec.com/viruses/99/viruses_062299a_j.shtml) [Increased costs of virus
attacks.]
Faltermayer, Charlotte. Cyberveillance, Time, August 14, 2000, p. B22.
[Worker monitoring statistics.]
Forno, Richard and William Feinbloom. PKI: A question of trust and value,
Communications of the ACM, June 2001, vol. 44, no. 6. [A good summary of
the diffculties of trusting public key certifcate authorities.]
Feig, Christy. Medical privacy rules to take effect, CNN, April 12, 2001.
http://www.cnn.com/2001/HEALTH/04/12/medical.privacy/index.
html?s=2 [Notice of new federal medical privacy rules.]
Ashley Fantz and Atika Shubert, WikiLeaks Anonymous Hackers: We Will
Fight, CNN Online, December 9, 2010. http://articles.cnn.com/2010-12-09/
us/hackers.wikileaks_1_julian-assange-arbor-networks-websites?_s=PM:US
[Anyone can join a botnetbut that is not a good idea.]
Government Technology, Ex-San Francisco Network Engineer Convicted of
Network Tampering, April 28, 2010. http://www.govtech.com/security/Ex-
San-Francisco-Network-Engineer-Convicted-of.html. [Conviction of Terry
Childs for refusing to give up the administrative username/password to the
citys network.]
Gregg Keizer, New Firefox Add-on Hijacks Facebook, Twitter Sessions,
Computerworld, October 25, 2010. [Announcement of Eric Butlers
Firesheep add-on that captures all unencrypted wireless traffc.]
Oakes, Chris. Privacy Laws Aim to Protect the Hunters as Well as the Hunted,
International Herald Tribune, March 23, 2001. [Good analysis of European
Union privacy controls.]
Post, Gerald, Improving Operating System Security, Computers & Security,
6(5), 1987, 417-425.
Thurman, Mathias (pseudonym). What to do when the Feds come knocking,
Computerworld, June 4, 2001, 62. [Situation where hacker used a stolen
laptop to attack other systems, and computer logs helped show the employee
was innocent.]
Whiteside, Thomas. Computer Capers: Tales of Electronic Thievery,
Embezzlement and Fraud, New York: Crowell, 1978. [Early cases of
computer fraud and abuse.]
http://csrc.nist.gov/encryption/aes/ [Reference source for AES algorithm.]
http://www.visabrc.com/doc.phtml?2,64,932,932a_cisp.html [Reference
to Visa CISP security.]
Cases: Professional Sports
Te Industry
Professional sports raise billions of dollars a year in the United States and around
the world. Actually, European professional soccer teams usually top the lists in
terms of revenue. In the United States, baseball teams brought in $3.6 billion in
revenue in 2001 (MLB 2002), basketball teams $2.7 billion in 2003 (NBA
2004), and football teams almost $5 billion in 2002 (Ozanian 2003). In 2010,
baseball brought in $6.1 billion, basketball $3.8 billion, football $8 billion, hock-
ey $3 billion, NASCAR $0.9 billion, soccer (top 20 teams worldwide) $5.2 billion
(Forbes 2011).
These numbers include gate receipts as well as other revenue sources such as
television payments. Of course, the teams also have enormous costs in the form
of player (and coach) salaries. In total, sports franchises are the most popular seg-
ment of the entertainment industry. If you add in the amount spent on gambling
(where it is legal, of course), sports are incredibly popular.
Information systems play several roles in sports management. Coaches, play-
ers, and scouts use information systems to track performance and opponents, store
game fles, diagram plays, and communicate information. IT is also used in the
front offce to sell tickets and merchandise. Like any other business, administra-
tors have to derive fnancial information and evaluate customers and suppliers.
Because of the popularity of the Internet and the role of television, sports teams
have also begun to implement sophisticated networks within the stadiums for use
by high-end customers.
Since the teams generally use networks to share information during games, se-
curity is critical. Any public networks have to be built separately from the team
networks. Fans might be frustrated if a public network crashes or is hacked, but
a team could be severely crippled if its main coaching system goes down or is
compromised.
Additional Reading
Badenhausen, Kurt, Yankees Soar, Mets Plunge on List of Baseballs Most
Valuable Teams, Forbes, March 23, 2011.
Baseball: http://www.forbes.com/lists/2011/33/baseball-valuations-11_land.html
Basketball: http://www.forbes.com/lists/2011/32/basketball-valuations-11_land.
html
Football: http://www.forbes.com/2010/08/25/most-valuable-nf-teams-business-
sports-football-valuations-10_land.html
Hockey: http://www.forbes.com/lists/2010/31/hockey-valuations-10_land.html
Forbes, MLB, April 15, 2002.
Forbes, NBA Valuations, February 9, 2004.
Ozanian, Michael K., Showing You the Money, Forbes, September 15, 2003.
Case: Professional Football
Football requires a large number of players, and that means a team needs a large
number of coaches. The information technology system becomes relatively more
complex to handle the increased number of users and machines. Brian Wright, IT
director for the Chicago Bears, notes that the NFL in general is looking closely at
security and how best to protect the information in our business. One of the things
we identifed was the need for better user authentication (Vijayan 2004). The
team installed a USB authentication key that staff members use to log onto the
network. They must also enter a PIN. The strength of the dual-factor authentica-
tion is that the PIN is relatively easy to remember, but outsiders cannot hack into
the network because they would need the physical USB key.
The Denver Broncos focused on providing more information for fansparticu-
larly the fans who watch the game from the luxury box seats. The IT department
installed fat panel touch screens and a high-speed network into every box. Using
the GamePlus system, fans can touch the screen and bring up the view from any
camera in the stadium. Rick Schoenhals, IT director for the Broncos, notes that
I just know technology is gaining a bigger place in sports venues and sporting
events. Were opening a new venue and we dont want to fnd out someone who
opened next week is doing these great things with technology and were not doing
it. We cant be complacent in any area, including technology (George 2002). The
GamePlus system consists of 135 screens. The Broncos are trying to offset some
of the costs by selling corporate sponsorships that display logos throughout the
game. Because Denver was the frst team to implement this type of system, the
costs were relatively high.
Local area and wide area networks are important to improve communication
within football teams, and many teams were early adopters. The Carolina Pan-
thers installed a 100-mbps network in 1997. Roger Goss, MIS manager for the
Panthers, comments on the speed by noting: We needed that because the coaches
create complex graphics, like playbook diagrams and game plans, and download
them from servers to workstations (Wallace 1997). The team also uses frame-
relay links to other teams and to the NFL offce. The connection is used to share
statistics and to notify the league about trades and waivers. With 83 users, the sys-
tem is heavily used on Monday to Wednesday when the coaches are creating and
distributing game plans, and scouting reports arrive for the next opponent.
Sports teams are increasingly aware that fans want up-to-the-minute informa-
tion on many aspects of the game. In 2000, the Washington Redskins created a
Mobile Flash application to offer team news via Web-enabled cell phones and
PDAs. Fans can sign up to receive daily e-mail messages about trades, statistics,
and player injuries. The Redskins were the frst NFL team to implement the wire-
less systemalthough others provide e-mail newsletters.
When the New England Patriots built a new 68,000-seat stadium in 2002, the
network was a critical element. The voice and data networks alone cost about $1
million. The network uses Nortel Passport switches with a gigabit Ethernet back-
bone. It links 80 luxury suites and provides more than 2,000 ports at 100 mbps.
The network is important to the teams and the fans. It is also important for renting
the stadium to businesses during the week. Pat Curley, the IT director, notes that
no other stadium has this setup. It makes it very exciting, and very challenging
(Cummings 2002). The network supports the various coaching and scouting sys-
tems used by teams. The Patriots scouts use notebook computers with comments
and data uploaded to the central servers for the coaching staff. The coaching sys-
tem also stores digitized video so coaches can watch specifc plays. Pat notes that
the coaching system runs on a separate LAN. She says weve designed every-
thing to be separate so that people who need Internet bandwidth, such as the suite
users and press, will have access but it wont confict with or steal our bandwidth.
Plus, its more secure (Cummings 2002).
After observing some of the problems that arose in New Orleans with Hurricane
Katrina, Bill Jankowski, senior director of IT for the Baltimore Ravens decided it
was time to establish a formal backup system for his teams data. Signing a con-
tract with AmeriValut Corp., he transferred 200 GB of data to their servers, and
every night ships update changes to their secure systems. For $3,000 per month,
AmeriVault stores all of the teams video clips, injury data, and purchase data
from the teams SQL-Server based transaction processing system (Fisher 2006).
Data security often extends beyond the digital world. Minnesota Vikings player
Michael Bennet learned that the hard way in 2003. He received credit card state-
ments with bills from various convenience stores. But the card did not really be-
long to him. Instead, an off-duty police offcer working as a security guard for the
team had stolen various documents from players and applied for credit cards in
their names. The identity thief was caught in part from video surveillance tapes
taken at one of the stores where he used the card. Several other well-known ath-
letes have had problems with identity theft as well (ESPN 2003).
The high visibility of American football and television seems to attract atten-
tion. Just prior to the 2007 Super Bowl, the Web sites of the Miami Dolphin Sta-
dium and the Miami Dolphins team were hacked and loaded with malicious code
that attempted to infect any computer visiting those sites. The infection was frst
identifed by customers who had the Websense security software installed on their
PCsit would not allow them to visit those sites because it detected the mali-
cious code. Eventually, the owners of the sites were notifed and they cleaned up
the attack. Dolphins spokesman George Torres said that we are working on the
technology side to review all the code and do whatever we need to, on a security
basis, to prevent this from happening again (McMillan 2007). The code had links
to servers in China that attempted to download Trojan programs on the visiting
PCs. The code took advantage of a known exploit in the browser with patches is-
sued by Microsoft.
The Dallas Cowboys were one of the earliest adopters of information technol-
ogy and analytical tools. Much of the technology and data is now available to all
teams, but the Cowboys are still leading the way at storing and handling data. In
2006, the team installed Isilon Systems servers to handle 19 terabytes of video
dataenough to hold two entire seasons worth of game video online. Video tapes
of the games arrive from the NFL on Monday, Tuesday, and Wednesday. Robert
Blackwell, director of coaching video for the Cowboys noted that by the end of
the day on Thursday, every game is converted to digital format and stored in the
system. By the end of the year, the system holds video for 263 games. Mr. Black-
well plans to continue adding storage so all video can be kept online and instantly
available to coaches. The team also has a backup Isilon system that it carries to
training camp (Fisher 2006). In the meantime, the NFL might want to talk with the
Big-12 college football teams, including the Texas Longhorns. Instead of fying
video tapes around the country, in 2006, the organization selected XOS Technolo-
gies to transfer digital copies of the videos to each team via the Internet. Each uni-
versity in the Big 12 Conference can download entire games in about two minutes
per gigabyte of data, or approximately two hours for one game flm. Raymond
Thompson, VP of product marketing at XOX Technologies said that you wont
fnd anyone in the Big 12 who doesnt love Internet Exchange because of the time
saved and the ease of getting it done (Rosencrance 2006).
The Dallas Cowboys built one of the newest football stadiums in 2010-2011.
The stadium hosted the 2011 Super Bowl and the giant scoreboard /replay screen
above the feld is only one of the most noticeable technology features in the sta-
dium. In addition to the scoreboards, clocks, retractable roof, and lighting, the IT
system runs point-of-sale terminals, ticket sales and so on. The stadium has 3,100
fat-screen televisions to show game scenes throughout the stadium, plus Wi-Fi,
IP phones, and 300 IP security cameras. Housed in the stadium is a HP-based data
center with 127 HP blade servers, 100 terabyte SAN, and high-speed network.
OK, the data center runs team and stadium operations plus operations for team
owner Jerry Jones 35 other companies. But, the system has a dedicated staff of
13 IT workers (Taft 2011). Other stadiums are adding information features. The
Meadowlands installed Wi-Fi capacity to support fans uploading photos and vid-
eo. For Jets games, fans can access a FanVision device that feeds game stats and
customized replays (Patterson 2011).
Questions
1. How can well-known stars protect themselves from identity theft?
2. What threats exist for portable computers used for scouting and how can
those risks be minimized?
3. Would fans pay for mobile access to games? How much? What type of
network would be needed to handle the data?
Additional Reading
Copeland, Lee, Redskins Tackle Wireless Access, Computerworld, October 31,
2000.
Cummings, Joanne, Network of Champions, Network World, July 22, 2002.
ESPN, Vikings QB Culpepper, four Others Were Victims, August 20, 2003.
Fisher, Sharon, Football Tackles Disaster Recovery, Computerworld,
November 10, 2006.
George, Tischelle, Football Fans See Games From A New Angle, Information
Week, January 30, 2002.
McMillan, Robert, Sites Scoured of Malware after Offsides Hit,
Rosencrance, Linda, Texas Longhorns, Other Big 12 Teams, Now Swapping
Game Films Digitally, Computerworld, September 18, 2006.
Taft, Darryl K., IT & Network Infrastructure: HP Technologies Drive the Dallas
Cowboys Stadium, eWeek, January 14, 2011.
Patterson, Thom, Super Stadiums of the NFL, CNN Online, February 7,
2011.
Vijayan, Jaikumar, Chicago Bears Boost Network Defenses, Computerworld,
May 28, 2004.
Wallace, Bob, LAN Blitz Sharpens Panthers Claws, Computerworld,
September 29, 1997.
Case: Basketball
The summer of 2003 was not a good year for the electricity grid in the northeast-
ern United States. A combination of errors caused the international loop around
Lake Erie to overload and shut down the power grid for the entire Northeast. Jay
Wessel, the IT director for the Boston Celtics, was attending a picnic at 4:30 P.M.,
and he notes that when the grid crashed, every cell phone and pager in the place
went off (Kontzer 2003). Because most of the blackout was focused in New
York, he lost only some e-mail servers. Other teams had entire systems shut down.
But, it convinced Wessel to look into ways to provide backup in case his facility
lost power. The problem he encountered is that most backup facilities were also
located in the same areas. He notes that it doesnt really do me any good to back
up my data 10 miles away (Kontzer 2003). Like many other businesses, Wes-
sel is also looking for solutions to the problems with virus and worm attacks. By
keeping systems patched, his losses have been minimal, but he is frustrated with
the frequency of patches required.
The Womens NBA teams have added a new twist to marketing basketball.
When the Portland Trail Blazers acquired a new team in 2000, management
turned to a CRM system to target sales to season ticket holders. They also merged
demographic data from external marketing lists and Ticketmaster. Tony Cesarano,
database marketing manager, notes that the system signifcantly improved the ef-
fciency of the sales team and was able to sell 6,400 season tickets in four months.
He notes that in the past, we would have manually keyed the WNBA sales infor-
mation into a Microsoft Excel spreadsheet, which would have been time-consum-
ing and could have introduced inaccuracies into the database (Baron 2000). The
Trail Blazers are using customer relationship management (CRM) tools to boost
sales of season tickets for the mens team. The system has consolidated the data
that used to be scattered on spreadsheets of the 50 sales and marketing employ-
ees. It tracks individual ticket holders as well as corporate customers. One of the
big gains was to minimize overlap, where multiple salespeople often called on
the same prospects. The system has proved useful in managing requests by ticket
holders to change their seat locations. It can quickly bring up unsold or released
seats as well as track priority values of important customers. The organization
ultimately expanded the system to their Web site. Cesarano notes that the process
is complicated: Well have to manage a much larger volume of data coming off
the Web. Well also have to fgure out how to clean up the data, to make sure that
were using only useful, legitimate, and accurate data (Baron 2000).
The NBA in total boasts a fan base of over 50 million people for the 29 teams.
In an attempt to boost the level higher, the NBA installed a customer relation-
ship management system from E.piphany. Bernie Mullin, senor vice president of
marketing, notes that were going to have a 360-degree view of the fans and
customers of the NBA. Weve never had that before at the league or team level
(Songini 2001). The system pulls data from all of the customer-interaction points:
ticket sales, All-Star nomination ballots, the NBA store Web site, individual team
databases, and the NBA store in New York. The main focus is to place the data
into a data warehouse and provide analytical tools to sell more tickets. The plan is
to track sales by various events to answer questions such as whether certain teams
provide bigger draws, or if some months or days draw bigger crowds. The main
NBA offce in Secaucus, New Jersey, has about 1,000 employees. The NBA has
several data collection and online systems to maintain. Two staffers attend each
game with touch screen laptops to collect data. The data is transferred back to the
main system at headquarters in real time. It can be used by fans to create custom
highlight reels.
In 2007, the NBA created a site for fans in the Second Life Web site with games,
interactivity, and community features. Visitors can gather in an arena to watch a
3D diagram of a game as it is being played in real life. The site also has a copy of
the NBA Manhattan store to sell virtual merchandise for players avatars. Through
their avatars, fans can also play games, such as HORSE and a slam-dunk contest.
The Orlando Magic also implemented CRM software to help boost ticket sales.
A primary focus of their system is to identify and track customer complaints. If
customers have problems, the system can direct them to the appropriate vendor
and monitor vendor compliance with contracts. Julie Gory, fan relations and retail
manager, notes that we are a watchdog department that looks at things from a
fans perspective. Whatever happens from when the fan leaves their driveway
everything from parking, the cleanliness of the restroomswhen we hear of is-
sues, we make notes and input them to GoldMine and can review them on what-
ever basis we want (Songini 2002). Gory decided not to use the NBA CRM soft-
ware because it was too expensivethe NBA was going to charge them about
$100,000. The system was also more complex and harder to use.
Video is critical to coaches and players. Digital video provides enormous ben-
efts over tape because it can be edited and indexed quickly. The New Jersey
Nets were the frst team to implement a comprehensive system from Ark Digital
Technologies. The system lets coaches quickly grab the clips they want based on
various statistics. Ark CEO Alan Kidd notes that during halftime, coaches could
show clips of the frst half by type, for example, jump shots or post positioning
(Kreiser 2002). Ark has also created a digital video coaching system. The system
can show basic drills. Ultimately, it can be coupled with the game-day videos so
players can go to a Web site and compare their performance with the drills (www.
arkdigitalsystems.com).
In 2006, the Celtics began actively using analytical software from SratBridge
to analyze ticket sales and prices. Daryl Morey, senior VP of operations for the
Celtics uses the tool to fll the 18,600 seats by changing prices and offering pack-
age deals. He said that until we had this tool, it was very diffcult to create dy-
namic packages, because our ticke providers didnt have a rapid way to see which
seats were open. Now we can actually see in real time every single seat and how
much it is sold for. The analytical tool has also revealed that in certain sections
fans prefer aisle seating, so the organization has learned to focus on marketing the
interior seats for sales promotions (Havenstein 2006).
In 2007, the Atlanta Hawks and Thrashers implemented a new payment sys-
tem for season ticket holders. Those who signed up for a Chase credit card can
use their cell phones to pay for food or merchandise at 200 POS terminals in the
arena. Additionally, their phones Web browsers can read RFID tags embedded in
posters in the concourses, directing them to a site where fans can view and down-
load game information, videos, and promotions (Mitchell 2007).
By 2010, the NBA had its own 85-person IT department with a CIO. The team
is responsible for game-time issues handling networks and computers, as well as
maintaining the NBA Web site and New Jersey tech center on a daily basis. For
the fnals games, video and stat loggers tag video clips with metadata so any-
one can quickly fnd video clips based on various criteria. CIO Michael Gliedman
noted that We come up with a whole set of metadata that can be used in lots of
ways, including by video game developers (Brown 2010).
The NBA produces and distributes an enormous amount of video, with 1.9 bil-
lion videos viewed from the NBA.com Web site in the 2010/2011 season. (Los
Angeles Times 2011). NBA Digital supports season passes to watch games online
and also provides the NBA Game Time app for Android and Apple phones which
displays game statistics and scores.
Questions
1. What privacy issues arise from the NBA using a CRM system to track
customer purchases?
2. How can the CIO of a basketball team provide backup facilities for a small-
to-medium-sized network?
3. How long will it take for digital video coaching technology to be
implemented at the college level?
Additional Reading
Baron, Talila, Team Scores With Apps That Net Ticket-Buyers, Information
Week, February 21, 2000.
Brown, Bob, NBA Finals Showcasing Technology as Well as Basketball Skills,
Network World, March 11, 2010.
Goff, Leslie, NBAs IT Team Makes Play With Web, CRM Initiatives,
Havenstein, Heather, Celtics Turn to Data Analytics Tool for Help Pricing
Tickets, Computerworld, January 9, 2006.
Kontzer, Tony, Data Backup: Rethinking The Unthinkable(s), Information
Week, August 18, 2003.
Kreiser, John, Virtual Coaching, Information Week, July 22, 2002.
Los Angeles Times, NBA, a Hit in Online Video, is Looking to Grow in Social
Media, April 15, 2011.
Songini, Marc L., NBA Shoots For Data Analysis, Computerworld, May 28,
2001.
Songini, Marc L., Orlando Magic Shoot for Customer Satisfaction,
Wagner, Mitch, The NBA Builds A Second Life Virtual Playground with
Potential For Real Fun, Information Week, May 5, 2007.
Case: Baseball
Sports fans, particularly in baseball, are often crazy about souvenirs. They will
collect almost anything. So, it is not surprising that outdated copies of baseball
contracts found their way to an e-Bay auction set up by Scott Gaynor, a sports
memorabilia dealer. Bob Tufts, who played for the San Francisco Giants and the
Kansas City Royals in the early 1980s, found that his contracts were among those
offered for sale. While there is nothing illegal per se about the sale, there was
an important twist: As employee contracts, most of the documents contained the
social security numbers of the players. With some big-name players in the group
(read money), the risks were quite high. Bob Tufts observed Im shocked to
fnd out how easy it is for people to get their hands on fles like these (Rovell
2003). With a home address, the social security number, and a guess at a mothers
maiden name, criminals could have created fake bank accounts. The commission-
ers offce requested that the auction be stopped and the contracts returned. Some
of the bids had reached $200 before the items were pulled.
The Internet is increasingly important for attracting fans to all sports. Since
baseball has been number-intensive for years, Web sites are an important source
of data on games and players. In 2001, all major league baseball teams agreed to
consolidate and standardize their Web sites. All of the sites are now run from the
MLB servers. All sites were given a common look. An interesting aspect to the
change is that revenue from sales on the site is split across all of the teams. Bud
Selig, the commissioner, observed that the most signifcant part of our whole
Internet activity was the unanimous vote to share the revenue. That was a very
dramatic change in thinking, because disparity is the biggest problem we have
[Wilder 2001]. Bob DuPuy, chief legal offcer in the commissioners offce adds
that Of course, we agreed to share revenues that dont currently exist. I dont
think we could do it at a later date when there was revenue disparity (Wilder
2001). The site is operated by a separate company created by the team owners
and run in New York City. The new organization was initially funded with $1
million a year for three years from each of the 30 teams ($30 million a year). In a
somewhat risky move, the MLB site sells video access to games. The risk is that it
could alienate traditional television broadcasterswho provide a substantial part
of baseball revenues. But the site is still subject to local blackout rules, and the
picture quality is not even close to television standards. So, its market would most
likely consist of out-of-area fans.
Season tickets present interesting problems for clubs and fans. About 80 percent
of them are owned by corporationsparticularly the luxury suites. Robert McAu-
liff, CEO of Season Ticket Solutions, notes that season ticket holders miss 25
to 50 percent of games every year (Rosencrance 2001). The challenge for teams
is that flling seats increases revenue through sales of food and souvenirs. The
challenge for fans is that they want to effciently use their investments. The Phoe-
nix Diamondbacks, as well as other teams, have purchased software that enables
season ticket holders to manage their seats. The Web site allows ticket holders to
check on who is using the tickets, and which dates still have seats available.
Web casting ball games is an interesting legal and marketing area. Are people
willing to pay for Web broadcasts? Do they want to select specifc games? (Cur-
rently, MLB offers a subscription but only to specifc games.) Would such a sys-
tem encroach on television broadcasts? Is there a proft? Many of these questions
remain unanswered, and it is not clear that anyone is seriously addressing them
yet. In 2000, William Craig and George Simons created a company and the iCra-
veTV Web site in Canada. The company picked up signals from 17 broadcast sta-
tions in the United States and Canada, digitized the signals and offered them for
free on its server. Rebroadcasting signals is legal in Canada. But it is a violation of
U.S. copyright laws. The company was quickly sued by several U.S. sports agen-
cies as well as the Motion Picture Association of America, because the company
did not block the signals to U.S. customers. The company quickly made an out-of-
court settlement and shut down the site (McGeever 2000). So, many questions re-
main about whether fans want to see Internet, or even cell phone video of games.
And, if so, how much they might be willing to pay and whether those fees would
offset any lost television advertising revenue.
With its experience and in-house staff, MLB has become the leader in Web
casting large events. In 2006, more than 5 million people watched March Mad-
ness college basketball games on the Webthanks to MLB technology. MLB
sells the services to at least 25 clients, providing a proftable stream of revenue for
MLB. Bob Bowman, CEO of MLB.com, observed that content publishers were
being underserved. A lot of companies are having a hard time picking out a digital
strategyMusic has iTunes, but that is not a total solution. So it was obvious that
we had something if we used it right. (White 2006). Mr. Bowman and others are
concerned that the ISPs might make life more diffcult for themthe phone com-
panies have stated that they want to charge fees for carrying bandwidth-intensive
programs like streaming video. Mr. Bowman notes that Just receiving a live feed
and sending a compressed version over the Web is diffcult. We didnt anticipate
how in-depth it all was. The 30 MLB teams fnanced MLB.com with an infusion
of $60 million and jointly own the organization. The site relies heavily on Akamai
Technologies to distributed the content to its 18,000 servers worldwidereducing
the bandwidth demands on any one server.
MLB.com charges $80 a season for fans to watch almost any live game they
choose. The site also uses software from Open Text to store and tag all of the con-
tentproviding customers with access to video clips, statistics, and other data in
a non-linear format. That is, fans can search for and select events to see, including
highlight clips and summaries of games. Justin Shaffer, VP and chief architect
for MLB.com noted that we wanted to take advantage of new media in order to
provide a better experience for baseball fans worldwide via MLB.comAs we
continue to catalog and distribute exciting game footage for our fans, our partner-
ship with Open Text ensures that we can easily provide those fans with the types
of products that match their needs, whether its game summaries, highlights or
other specialized cuts of video (Rosencrance 2005). In 2009, MLB launched its
own network on cable TV. Video from every major league game is routed to the
Secaucus NJ studios, compressed into HD and standard defnition video and then
routed to satellites for transmission by cable companies (Mark 2009).
Once fans are at the stadium, they need to pay for things like souvenirs, food,
and beer. And no one wants to stand around fshing for change and signing credit-
card slips. The San Francisco Giants were an early adopter of contactless payment
readers. Newer credit cards have RFID chips embedded in them and customers
simply hold the card a couple of inches from the reader to activate the payment.
Not many customers have the RFID cards yet, but it simplifes payments for those
who do have them. Ken Logan, IT director, also plans to issue season ticket hold-
ers with a special card that gets them into the stadium and handles payments at the
concessions by linking through a database back to the holders credit card. So far,
Logan said the only drawback is just keeping people from spilling a beer on it.
They get pretty dirty and sticky (Mitchell 2007).
Any business wants to track its customers, and baseball is no exception. Teams
routinely send out e-mails and promotional information to ticket holders. In 2011,
a staffer for the New York Yankees sent a standard newsletter to several hundred
season ticket holders. The only problem is that a spreadsheet was attached to the
message that contained personal data for 18,000 season ticket holders. The spread-
sheet did not contain credit card data but it did have phone numbers, addresses,
names, seat numbers, and Yankee account numbers (McMillan 2011).
One of the more amazing technologies to be introduced to baseball was tested
in 2009 by Sportvision, a Bay Area company that developed the yellow frst-down
marker for football. The new camera and software system is designed to automati-
cally track the location of every ball and every player on the feld. It will have the
ability to generate data on how fast felders respond to hits, how hard they throw
the ball, and other statistics that have never existed before. The data could intro-
duce big changes in how defense is evaluated in baseball (Schwarz 2009). The
system was frst tested in San Francisco and uses four high-resolution cameras to
capture three-dimensional data. Sportvision tested the FIELDf/x system in the San
Francisco stadium in 2010 and was preparing to deploy it to other stadiums.
Umpires play a key role in any baseball game. As much as fans like to com-
plain about some of the calls, the talent and knowledge of the umpires is amazing.
It takes training and practice to stay up-to-date as a major league umpire. So MLB
and IBM teamed up to install special Web-based software to help umpires. Um-
pires use WebSphere technology to build mashups of the game, calls, and players.
They can leave notes for umpires at the next game. The portal lets them view
videos of certain plays, help analyze controversial calls, and provide detailed ex-
planations of rulings (Boulton 2008).
Questions
1. How are player privacy rights different from average citizens?
2. What procedures can be implemented to help protect players from identity
theft?
3. What are the benefts and drawbacks to centralizing the team Web sites for
MLB?
4. How could a team provide individual customized access to baseball games
via the Internet?
Additional Reading
Boulton, Clint, IBM, MLB Connect on WebSphere Web 2.0 Deal, eWeek, July
18, 2008.
Mark, Roy, MLB Network Launch Powered by Motorola, eWeek, January 9,
2009.
McGeever, Christine, Webcaster Under Fire, Computerworld, January 21,
2000.
McMillan, Robert, NY Yankees Staffer Accidentally e-Mails Customer List,
Mitchell, Robert, No Contact: Could Smart Phones Spur Contactless Payment
Card Adoption? Computerworld, June 11, 2007.
Rosencrance, Linda, Watching the Bottom Line From Box Seats,
Computerworld, August 6, 2001.
Rosencrance, Linda, MLB.com Scores with Open Text, Computerworld, April
26, 2005.
Rovell, Darren, Confdential Information Pulled From Online Auction, ESPN,
June 17, 2003.
Schwarz, Alan, Digital Eyes Will Chart Baseballs Unseen Skills, The New
York Times, July 10, 2009
White, Bobby, Major League Baseball Steps Out As Coach in the Game of Web
Video, The Wall Street Journal, March 27, 2006.
Wilder, Clinton, Redefning Business: A New Game Plan, Information Week,
April 9, 2001.
adopted by rivals?
sectors?
350
Chapter Outline
Transactions and Enterprise
Resource Planning
6
Chapter
Introduction, 352
Data Capture, 353
Point of Sale, 354
POS Advantages, 358
Process Control, 359
Electronic Data Interchange (EDI), 360
Extensible Business Reporting Language
(XBRL), 364
Elements of a Transaction, 366
Vendor Perspective, 366
Customer Perspective, 366
Transaction Fees, 367
Government Perspective, 368
International Issues, 369
The Role of Accounting, 370
Input and Output: Financial Data and
Reports, 371
Purchases, Sales, Loans, and
Investments, 372
Inventory, 372
The Accounting Cycle, 373
Process: Controls, Checks, and
Balances, 375
Human Resources and Transaction
Processing, 375
Input: Data Collection, 376
Output: Reports, 377
Process: Automation, 377
Production Management, 378
Production Issues, 379
Distribution and Inventory Control, 380
Integration in Business, 382
Enterprise Resource Planning, 385
International Environment, 387
Financial Accounting, 387
Logistics, 388
Human Resource Management, 388
Integration, 389
Supply Chain Management, 390
SCM Changes the Focus, 391
SCM Challenges, 392
Integration across Systems, 394
Customer Relationship Management, 395
Multiple Contact Points, 395
Feedback, Individual Needs, and Cross
Selling, 396
CRM Packages, 397
Summarizing ERP Data, 398
Digital Dashboard and EIS, 399
How Does an EIS Work?, 400
Advantages of an EIS, 401
Transaction Accuracy: Sarbanes-Oxley, 402
Summary, 404
Key Words, 406
Exercises, 407
Cases: Automobile Industry, 411
351 Chapter 6: Transactions and ERP
Ford Motor Company
How do you manage a huge organization? How can you integrate data from hun-
dreds of divisions scattered around the world? Bill Ford, great-grandson of Henry,
turned around the struggling manufacturer in the early 2000s. Along the way, he
made an amazing discovery: it is more important to sell vehicles at a proft, instead
of trying to just sell more at any cost. With hundreds of facilities, Ford, like the other
automobile manufacturers, has an immense job of ordering parts, keeping factories
scheduled, identifying cars and features that customers actually want, tracking pro-
duction, monitoring sales, and designing for the future. Having pioneered the mass-
production systems of the 1900s, Ford is only now learning to build a more fexible
manufacturing systemwhere one plant can produce different car makes. Informa-
tion technology plays a key role in making the factories and the entire operation
more effcient.
Although the company is publicly held, the family exerts signifcant infuence on
it. After former chairman and CEO Donald Petersen rubbed the family the wrong
way, he retired prematurely. Alex Trotman took the position in 1995 and hoped to
retain the Ford family support. On January 1, 1999, the disappointed Ford family
replaced Trotman as CEO with Jacques Nasser. Nasser came from the international
side of Ford, which was experiencing strong sales. In 2001, a series of crises, capped
by the Firestone tire recall on the Explorer led to high tensions in Ford and in the
industry. On October 29, 2001, Bill Ford, Jr., fred Jacques Nasser and installed him-
self as CEO. On September 5, 2006, the company replaced Bill Ford with Alan Mu-
lally from Boeing.
How do you process the data from transactions and integrate the operations of the
organization?
How do you effciently collect transaction data?
What are the major elements and risks of a transaction?
Why are transactions more diffcult in an international environment?
How do you track and compare the fnancial information of a frm?
What are the transaction elements in the human resources management system?
Can a company become more effcient and productive?
How do businesses combine data from operations?
How do you combine data across functional areas, including production, purchasing,
marketing, and accounting?
How do you make production more effcient?
How do you keep track of all customer interactions? Who are your best customers?
How can a manager handle all of the data in an ERP system?
How does the CEO know that fnancial records are correct?
Why do most companies avoid cloud computing for ERP systems?
Introduction
How do you process the data from transactions and integrate
the operations of the organization? This question has been asked re-
peatedly for centuries. In fact, it has driven the organizational structure of com-
panies. What were frms like in the 1800s or earlier? Most were small propri-
etorships. As frms became larger, how could they handle the thousands of daily
transactions? Hierarchical structures evolved to collect data at the point of sale,
summarize it, and move it up the chain of command to the CEO. Figure 6.1 shows
how transactional data is exchanged with customers, suppliers, manufacturing,
and banks, and summarized for higher-level decisions. Think about the problems
that arise if groups within the company (accounting, production, marketing, and
so on) all have different databases. When you have computers today that can cap-
ture all of the operations data automatically, do you still need a hierarchical struc-
ture? Instead, an integrated system running on a database can be used to share data
with everyone.
Whenever two people make an exchange, it is called a transaction. Transac-
tions are important events for a company, and collecting data about them is called
transaction processing. Examples of transactions include making a purchase at a
store, withdrawing money from a checking account, making a payment to a credi-
tor, or paying an employee.
Because of the importance of transactions, companies built systems to track the
various aspects, including accounting, human resources, and production. As com-
puter hardware, software, and database technology improved, it because clear that
companies needed integrated systems that held all of the data or enterprise-level
systems. These tools are complex, but can collect data on all aspects of the organi-
zation, including access for employees, managers, customers, and suppliers. The
Figure 6.1
Transactions and Enterprise Resource Planning Integration. The goal is to collect all
operations data and share it throughout the organization. Links to suppliers and banks
improve integration and provide more options for the frm.
Central Computer
Sales Reports
POS
Cash
Registers
Customers
EDI
Supplier
Bank
Process Control
Additional
Stores
Strategy
Tactics
Operations
Warehouse
Inventory
Management
CEO Information
integration of data from all areas of the organization from marketing, production,
human resources, and accounting leads to new methods of doing business.
Data Capture
How do you effciently collect transaction data? The basic compo-
nents of a transaction-processing system are illustrated in Figure 6.2. The focus is
twofold: accomplishing the transaction and capturing data. Data capture consists
of gathering or acquiring data from the frms operations and storing data in the
computer system. Entering data into the computer can be time consuming and dif-
fcult. For instance, banks have invested heavily in automating the collection and
recording of transaction data. Yet because many transactions are based on paper,
clerks still spend considerable time entering data. First, tellers enter the data into
their terminals. Then a bank staff reads the dollar value written on checks and
deposit slips. The bank staff works through the night, typing the amount into a
machine that codes the number on the bottom of the check so it can be read by
other computers.
As the volume of transactions increased, businesses looked for faster and more
accurate ways to get data into the computer. Four basic methods are used to col-
lect data, depending on its source. The data-collection method consumers are most
Trends
Because of legal ramifcations, businesses have always collected data about
transactions. Accounting systems play an important role in collecting and
analyzing transaction data. Through the 1960s, most business computers
were primarily producing basic accounting reports. Raw data was punched
into the computer by hand, and the computer produced totals and updated
the general ledger. In effect, the computer was used as a giant calculator to
automate the production of printed reports similar to those used before the
advent of computers. The primary reason for using the computer was speed
and accuracy. It was justifed because it was cheaper and less error-prone
than hiring thousands of people to produce the reports.
As computer capabilities increased in the 1970s, the most important
change was to use the computer to collect the raw data. In retail sales, the
cash register was replaced with a computer terminal and a bar code scanner.
Whenever a customer purchased an item, the transaction data was immedi-
ately sent to the main computer. This automation eliminated the need to hire
a person to enter the data at the end of the day.
The 1980s and 1990s resulted in more integration. Transaction data was
made available to managers throughout the company. One goal was to com-
bine the systems across the company into an enterprise system that enabled
managers to examine all aspects of the business.
By the late 1990s, integrated systems could support data for the entire
organization. Accounting packages handle all accounting information tasks,
HRM systems handle employee-related chores, and manufacturing systems
deal with production. These enterprise-level systems were designed from the
ground up to handle all of the major information tasks of a company. Be-
yond simple access to data, integrated systems make it possible to change
the way the company operate.
familiar with is point of sale (POS), where the sales register is actually a com-
puter terminal that sends all data to a central computer. On assembly lines, robots
and manufacturing equipment can collect data, such as quality control measures,
and return it to a computer. Typically the computer also can send control instruc-
tions to these machines. This exchange of data between manufacturing machines
and computers is known as process control. The third way to collect data au-
tomatically involves the exchange of information with organizations outside the
frm, especially suppliers and customers. Instead of dealing with paper records
such as purchase orders, it is possible to send orders electronically through a pro-
cess called electronic data interchange (EDI). The fourth method is to have the
customers select products and enter data directly into the electronic commerce
Web site. This fourth method offers the potential of reducing errors and costsif
the Web sites are built carefully and if the customers are motivated and capable of
selecting the products themselves.
Point of Sale
Several devices have been created to capture data at the point of the sale. Some
companies rely on keyboards to enter data, but high-volume areas have switched
to bar code scanners. All consumers are familiar with bar code scanners that read
the universal product codes (UPCs). The scanner reads the code and sends it to the
computer, which looks up the corresponding name and price. The computer prints
the receipt, stores the sale information, and automatically decreases the inventory
count.
Figure 6.2
Data should be captured as soon as it is generated. The operations level generates
data from many sources. Modern systems collect the data in a central database and
make it immediately accessible to managers.
Sales
Process Control
Banking and Finance
Workers
Another type of scanner is used by the U.S. Postal Systemoptical character
recognition (OCR)to read handwritten zip codes, allowing mail to be processed
and sorted faster. Even so, the post offce hires thousands of workers to type in data
that the scanners cannot read. Banks use a process called magnetic ink character
recognition (MICR) to process checks and deposit slips. MICR readers are more
accurate than straight OCR because they pick up a stronger signal from magnetic
particles in the ink. A few companies are using speech recognition technology
to enable workers to enter data by speaking to the computer. Speech recognition
enables the users to enter data while leaving their hands free to do something else.
The newest data-acquisition technology is radio frequency identifcation
(RFID). Bar codes have two major limitations: (1) they hold a limited
amount of data, and (2) even the best scanners have trouble reading
them. Basic bar codes can hold 10 digits, but 5 of those are used to
identify the manufacturing company, which leaves only 5 digits for
the actual product. As a consumer, you have certainly encountered bar
codes that are diffcult to read. RFIDs are small chips about the size of
a grain of rice that consist of an antenna and some transistors. When
the scanner sends a radio signal to the chip, the antenna receives
it and uses the radio waves to power the transistors and return the
stored data by modifying the scanner wave. For several years, people
have implanted the chips in their dogs in case the pet becomes lost and
loses his or her collar. The capacity and range of RFID chips depend on
the frequency (low, high, or ultra-high). The ultra-high frequency chips
can be read from the greatest distance (12 feet with a maximum of 40)
Reality Bytes: Pay By Phone
In 2011, several companies began a push to fnd a mechanism to use cell phones to
make everyday payments. The catch is that several technologies are available, but
no one can agree on which method to use. As Omar Green from Intuit points out,
Theres a lot of money at stake if its done right. Some examples include Star-
bucks, which has an application to display a pre-loaded Starbucks card on a smart
phone that the cashier can scan. The Bump application can transfer PayPal cash from
one phone to another. Laura Champers of PayPal noted that the company processed
$700 million in mobile payments in 2010a tiny fraction of the $70 billion it han-
dles in total. FourSquare also offers methods to transfer money to other phones. But,
none of them are likely to be adopted by typical retail vendors. Instead, the credit
card companies are trying to extend their reach through Near Field Communication
(NFC) which uses a chip similar to RFID with a shorter range. Put the credit card
number on a chip and put the NFC chip into a cell phone and customers can wave
the phone in front of a scanner. It is really just a credit card in a different format and
could be handled by taping a card to the back of a phone. The main wireless carriers
(Verizon Wireless, AT&T, and T-Mobile) also are trying to get into the game by cre-
ating the Isis network to work with Discover Financial. Some of the largest banking
companies have also been working on alternatives. But, without the support of the
credit card companies, it is likely to be diffcult to gain any traction with retail stores.
Adapted from Laurie Segall, Your Mobile Phone is Becoming Your Wallet, CNN
Online, January 19, 2011.
but hold only 12 bytes of data. The high-frequency chips are typically
scanned at less than a foot (with a maximum of 10 feet), but can hold
from 8 to 10,000 bytes of data. The ability to read the tags from a dis-
tance provides a substantial advantage in supply chain or warehouse
environments. Figure 6.3 shows that you no longer have to scan each
item from a few inches. You can walk down the aisles with a reader
and quickly pick up and distinguish the RFID tags from hundreds of
products at a time. The main drawback to RFID at the moment is that
the tags are relatively expensivearound $0.20 each. But if you have
to count inventory in a store or warehouse several times a year, the
RFID tags could save you substantial time and money, plus provide
improved accuracy. The scanners also have limitations on the number
of items they can recognize at one time. For example, it is doubtful that
a single scanner could read the RFID tags from an entire truckload of
products at one time.
Wal-Mart and the U.S. Department of Defense are pressing suppli-
ers to put the tags on all items shipped to Wal-Mart. Using their domi-
nant market position, Wal-Mart is asking suppliers to bear the cost of
the tags. However, at this point in time, it is probably not economically
feasible to include the tags on small, low-price items. Instead, the tags
are typically placed on pallets or boxed shipments. RFID tags at this
level will probably not be useful in the store itself, but it will help to
identify items received and track them through the supply chain deliv-
ery system.
C
B
E
antenna
Capacitor:
Collected energy
Transistors:
Data
RFID tag
RFID reader
Radio/microwaves
Data:
Alter the waves
Figure 6.3
RFID capture. The antenna on the tag captures the power from the radio signal to
release the data in the transistors and modify the wave signal so that the stored ID
can be retrieved.
Ultimately, if tags become cheap enough to place on all items, re-
tail stores might be able to radically alter their format. For instance,
scanners could be embedded in shopping carts or doorways. Customers
would simply select the items, and get an immediate price and running
total. Payment could be virtually automatic as well, so customers could
skip the checkout lanes. Cart-based systems could also suggest related
products or direct customers to aisles with similar products. Sound far-
fetched? Albertsons (a grocery chain) began experimenting with these
tools in some of its stores in 2003. Most customers were reluctant to
use hand scanners and scan everything themselves, but if RFID tags
could be used to simplify the process, it might be more useful.
Some privacy advocates have argued against RFID tags on the
grounds that if all of your clothes contained these tags (as originally
proposed by Benetton), it would be possible to track individual people.
However, the limited range of the scanners, coupled with the ability to
deactivate or remove the tags, can reduce or eliminate this problem.
Two-dimensional codes (squares) are used in some operations. UPS
was one of the frst to use them to route and track packages. Today, the
QR code standard developed by Denso Wave Incorporated in Japan is
available for many purposes. Figure 6.4 shows a sample QR code that
was generated from an online Web site. Increasingly, cell phones with
cameras can be used to take a picture of a QR code and the encoded
data provides a link to a Web site with detailed information. Several
Web sites will generate QR codes for free based on the data or Web site
you enter. Although the extra data capacity can be useful for packages,
the power of smart phones makes the QR codes appealing to market-
ers. Products, displays, even billboards now carry QR codes to encour-
age people to obtain background information. Cities, such as New York,
are adding the codes to various places, such as bus stops, parks, and
public documents. Museums and public art displays are adding them
to provide background information about various pieces.
The QR code examples point out some important features of transac-
tion data. Primarily, the format has to be standardizedand it helps
RealityBytes:NewJob:ChiefDigitalOffcerofNYC
With the increased sales of smart phones, more people are carrying around devices
that be used to obtain and share information. The QR code is a useful method to tag
items so people with smart phones can quickly receive information about objects
carrying the code simply by taking a picture of the code. New York City hired Ra-
chel Sterne as a chief digital offcer to help fgure out where to put the codes. The
City uses the codes on building permits to enable people to link to the underlying
blueprints and construction records. Residents also suggested putting the odes on
mass-transit locations, such as bus stops so riders can use them to get up-to-date
schedules. Other suggestions included parks and art workto provide detailed data
to interested viewers.
Adapted from Zachary M. Seward, Where Should New York Place QR Codes?
The Wall Street Journal, February 24, 2010.
if the standard is free of royalty payments. Also, the codes need to con-
tain enough data to be useful. The old bar codes, and even RFID chips
do not contain much data. Larger capacity also improves the capability
of the code to handle error correctionwhere the data can be recovered
even if some of the code is destroyed or covered. The presence of low-
cost readerssuch as cameras on cell phonesensures the adoption
and spread of the code.
POS Advantages
Several advantages arise from using automated data entry. Directly capturing data
means fewer errors occur because the machines make fewer mistakes. However,
sometimes it is not easy to collect data at the source. POS systems also have built-
in error detection methods to make certain the numbers are read correctly. By
collecting the data immediately, it is easier to fnd and correct mistakes. If a clerk
using a POS system enters an incorrect product number (or the scanner reads it
incorrectly), the error can be caught immediately.
With POS data collection, the computer performs all necessary computations
immediately. Hence, the job is easier for clerks and fewer errors will occur. For
example, a retail store might give discounts to certain customers. With a POS
system, the employees do not have to keep track of the customers or discounts,
because the computer can look up the discounts and apply them automatically.
Similarly, prices are maintained and changed by the computer. To hold a sale, you
simply change the price in the computer (one place) and put up a new sign. Of
course, when there are thousands of items and prices, there are still plenty of op-
portunities for errors.
POS systems also can provide better service to customers. Because the sales
data is captured immediately, the managers and clerks always know the inventory
levels. If a customer calls to learn whether a product is in stock, the clerk can in-
stantly determine the answer. With most systems, it is possible to tell the computer
to hold that item until the customer picks it up. Some companies even connect
their store computers together. If you fnd that one store has run out of a particular
item, the clerk can quickly check the other stores in the area and tell one to hold
the item for you.
Figure 6.4
QR codes. A two-dimensional bar code that can contain thousands of characters, the
QR code was initially created to track auto parts by Denso Wave, Inc. in Japan. Most
cell-phone cameras can read the codes so it is often used to link viewers to Web sites
containing marketing or background information.
Process Control
Manufacturing frms often deal with a different type of data. Most factories use
machines that can be connected to each other and to computers. The computers
can exchange data with the production machines. If you want to alter your prod-
uct, you would need to change the manufacturing settings on several different
machines. If the production line has 10 machines, each with 5 control items that
need to be set, it could take several hours to reset the entire production line. And
then you have to test and verify every setting. Even a minor change in the product
means that someone has to set each of the machines correctly. By connecting the
machines to a computer, the computer can store the appropriate settings. When
you make a change in the product, the computer sends the correct settings to all
the machines. Computers are often used to monitor the progress of the produc-
tion line. The data is used to identify problem spots and to help the frm meet
production goals. Figure 6.5 illustrates the basic concept of individual machines
controlled from one location.
Technology also can be used to collect data from manufacturing machines.
With this communication, the computer can constantly monitor production levels.
Managers can keep track of hourly and daily production, and even track individ-
ual products. If a customer wants to check on the progress of a special order, the
manager can determine how much of the product has been produced and when it
is likely to be completed.
Process control computers can also be used to monitor quality in the manufac-
turing process. Sensors can automatically measure almost any characteristic. They
can check for items such as thickness, weight, strength, color, and size. These
Figure 6.5
Process control is the control of production machines from centralized computers.
The computers monitor data from the machines and make continual adjustments. The
central control enables designers to specify all production settings from one location.
Various Production machines: lathe, press, dryer, . ..
Control Terminal
Production data:
Quantity
Quality
Time
Machine states
Control settings
and commands
measurements can then be passed to a computer. If the computer notices a trend or
a major problem, it can notify the operators. In some operations, the computer can
send messages to the machine causing the problem and reset its controls to correct
the problem automatically.
Two basic diffculties exist with process control. First, the large number of ma-
chines makes it diffcult to establish standards, making it harder to connect the
various machines together. Second, production machines can produce an enor-
mous amount of data. Some machines can generate billions of bytes of data per
hour. This large amount of data requires effcient communication lines, high-speed
computers, and a large storage capacity. Despite these complications, process con-
trol can provide enormous advantages. It enables companies to change production
processes and alter products faster and more often. It provides better information
and control over quality. It enables manufacturers to create products that match
the needs of individual customers: mass customization.
Electronic Data Interchange (EDI)
EDI is a form of automated data entry that supports operations by transferring
documents between frms electronically. The essence of EDI is the ability to
transfer data among computers from different companies. The goal is to connect
to suppliers so that production orders can be sent automatically at substantially
lower cost than traditional paper-based systems. Two basic methods are used to
accomplish the transfer: (1) send the data directly from one computer to the other
or (2) send the data to a third party that consolidates the data and sends it to the
proper location. Early EDI implementations were based on direct connections as
individual frms experimented with the technology. In both methods, two impor-
tant considerations exist: establishing the physical links and transferring data in
a format compatible to all users. The frst issue has largely been solved with the
Internet. Today, EDI connections can be established through e-mail or Web site
links.
For EDI to work, each company must translate its data into a form that can be
used by the other companies. If one company like Wal-Mart or GM takes the lead
Reality Bytes: Mobile Payments a Tough Sell Without Credit Card Companies
Wireless carriers announced the intention to scale back plans for payments by phone
only a couple of months after the joint venture Isis was formed. The goal was to
create a new payment method involving cell phones, where carriers would take a
percentage of transaction revenue. The main problem was that the system would
compete directly against credit-card companies, notably Visa Inc. and MasterCard
Inc. Retail merchants and customers are unlikely to use a system that does not in-
clude the two biggest players in payments. Drew Sievers, CEO of mFoundry which
provides mobile banking technology, noted Not including the 800 pound gorillas of
the industry will make it very hard to succeed. But, the card companies do not need
the carriers, so they no longer have a way to take a cut of the revenue. A Nilson Re-
port notes that in 2010, 57.2 percent of debit or credit card purchases were handled
through Vias, and another 25 percent by MasterCard.
Adapted from Robin Sidel and Shayndi Raice, Pay-by-Phone Dialed Back, The
and requires suppliers to send data via EDI, then they are free to defne the base
transaction objects. Suppliers must translate their objects into the appropriate EDI
structure. Yet a supplier might need links to several customers. If each customer
used different EDI defnitions, the supplier must have a conversion system for
each link. Someday it might be possible to create standards for EDI connections,
forcing everyone to conform to one type of data defnition. Although there is some
progress in this area, frms with existing EDI systems will be reluctant to spend
the money to convert their data.
Data conversion might sound like an easy task, but it is complicated when the
transaction systems were created over long periods of time and were poorly docu-
mented. In many cases, the programmer might have to search major portions of
the corporate systems to fnd the appropriate data defnitions. Once the appropri-
ate data is found, it can be hard to modify. Existing programs might expect the
data to maintain specifc formats. Making changes to the data can require rewrit-
ing other programs.
The concept of EDI is closely tied to Supply chain management (SCM) which
revolves around purchasing, but also incorporates just-in-time delivery, searching
for competitive pricing, and controlling and monitoring quality.
Proprietary EDI
As displayed in Figure 6.6, most of the early EDI systems were created indepen-
dently: one large company required suppliers to provide data and accept orders
electronically. The goal was to cut the costs of transactions and speed up the or-
dering process. EDI arrangements also enabled manufacturers to improve quality
control and to implement just-in-time inventory systems. Suppliers were encour-
aged to adopt the EDI systems by threatening a loss of sales if the vendors did
not comply.
Reality Bytes: QR Codes from Ethical Bean Coffee Co. Connect to Customers
Ethical Bean Coffee Co. is a small three-store chain of coffee shops in Vancouver,
British Columbia. Similar in some ways to Seattle, Vancouver has a surplus of coffee
shops on every corner. Seeking a way to stand out, in 2010, Ethical Bean started put-
ting QR codes in its train ads. When customers use their smart phones to scan the QR
Code, a menu appears on their screens. Customers can order a cup of coffee while
riding the train and pick it up when they arrive at the Ethical Bean shop. CEO Lloyd
Bernhardt notes that business has doubled since adding the QR codes and that We
catch people who are on the go and dont have a lot of time. QR codes are easy to
generateseveral Web sites make it easy to enter a URL or e-mail address and will
generate the square bar code free. Scanbuy, Inc., a New York company that develops
QR code advertising estimates that 30 million U.S. people have a code reader on
their phone. Companies still have to print and distribute the codes in their advertis-
ing. Ryan Goff, director of social-media marketing at MGH, Inc. cautions that QR
codes are not the end-all, be-all. They may not exist in two years. But theyre a tem-
porary solution to the problem of How do you connect people to online things in the
real world?
Adapted from Emily Glazer, Target: Customers on the Go, The Wall Street Jour-
nal, May 16, 2011.
With proprietary systems, the lead frm establishes the standards in terms of the
hardware and the types and format of data to be exchanged. From the standpoint
of the lead frm, these controls ensure that they are able to connect to each supplier
with one standard technique.
To a supplier, proprietary systems created by one company can lead to prob-
lems. Imagine what happens when the supplier sells to several frms, and each frm
requires the use of a different EDI system. In addition to the hassles of providing
data in the proper format for each customer, the suppliers employees would have
to learn how to use several different systems. Purchasers face similar problems
unless all of their suppliers follow a standard.
EDI Standards and Te Internet
Multiple proprietary systems lead to confusion and higher costs. Consequently,
several companies have tried to push for standards that make it easier to share
data. Two primary standards exist for EDI messages. The UN sponsors the Edifact
standard; the United States defned the ANSI (American National Standards In-
stitute) X12 defnition. Figure 6.7 shows the overall structure of an EDI message.
A signifcant difference between the standards is in the numbering system used to
represent the types of messages, segments, and data elements. Figure 6.8 presents
a partial list of the segment types available in the X12 standard. The standards
also specify the exact format of the data required in each segment type.
Web sites are used to advertise and display information about products and
their availability. Search engines enable companies to fnd components and poten-
tial suppliers quickly. EDI transactions such as orders and request-for-prices can
Figure 6.6
EDI can be built from individual pair-wise links over proprietary connections. If the
majority of transactions are between two companies, this method will work fne. If
companies deal with many different suppliers or large customers, this method can
cause problems when each link requires conversion to a different format.
Supplier
Customer 1
Queries &
Orders
Invoices &
confirmation
Order
Database
& Accounts
Customer 2
Production
Database
& Accounts
Convert
Convert
Figure 6.7
EDI standards. UN Edifact and U.S. ANSI X12 standards are similar in format; each
message consists of segments and detailed data lists. Each message, segment and data
element are defned by numbers from a predefned list of possible transactions. There
are substantial differences in the numbering system used for the segments and data.
Message
Segment
Composite Data Element
Data Element
Code Lists
Figure 6.8
Sample segment codes for ANSI X12. A partial list of the codes used within X12 EDI
messages. Only the number is transmitted. Each segment specifes the format of the
additional data.
104 - Air Shipment Information
110 - Air Freight Details and Invoice
125 - Multilevel Railcar Load Details
126 - Vehicle Application Advice
127 - Vehicle Buying Order
128 - Dealer Information
129 - Vehicle Carrier Rate Update
130 - Student Educational Record (Transcript)
131 - Student Educational Record (Transcript) Acknowledgment
135 - Student Loan Application
139 - Student Loan Guarantee Result
140 - Product Registration
141 - Product Service Claim Response
142 - Product Service Claim
143 - Product Service Notifcation
144 - Student Loan Transfer and Status Verifcation
146-RequestforStudentEducationalRecord(Transcript)
147-ResponsetoRequestforStudentEd.Record(Transcript)
148 - Report of Injury or Illness
be handled over the Internet as e-mail messages. The Internet can also host secure
communication channels between two partners. These links can be used for high-
volume exchanges of data.
Extensible markup language (XML) was developed in the last couple of
years to provide better Internet support for EDI. At its foundations, XML is a tag-
based document that contains data. As shown in Figure 6.9, the tags indicate the
type of data contained within the document. The document can have a hierarchical
structure similar to the EDI standards, such as Order OrderItem Product. Vari-
ous industry groups have been establishing standard document formats (data type
defnition or DTD) for common documents within their industries. Many software
packages can read and write XML documents, so companies can use diverse hard-
ware and software and still communicate easily.
Figure 6.10 shows the sample XML document using a browser. XML is the
means to exchange data in a form that all computers can read. At least, that is the
promise and hope. XML is still evolving and is still relatively expensive to imple-
ment because it often requires customized programming on both ends to handle
the data. Software vendors are developing tools that incorporate XML automati-
cally, and suppliers will be able to purchase the software needed to handle the
multiple connections to customers.
Extensible Business Reporting Language (XBRL)
The Securities and Exchange Commission (SEC) is responsible for monitoring
fnancial aspects of businesses in the U.S. to protect investors and maintain fair,
orderly, and effcient markets. The SEC collects and publishes fnancial data by
<?xmlversion="1.0"?>
<!DOCTYPEOrderListSYSTEM"orderlist.dtd">
<OrderList>
<Order>
<OrderID>1</OrderID>
<OrderDate>3/6/2004</OrderDate>
<ShippingCost>$33.54</ShippingCost>
<Comment>Needimmediately.</Comment>
<Items>
<ItemID>30</ItemID>
<Description>Flea Collar-Dog-Medium</Description>
<Quantity>208</Quantity>
<Cost>$4.42</Cost>
<ItemID>27</ItemID>
<Description>AquariumFilter&Pump</Description>
<Quantity>8</Quantity>
<Cost>$24.65</Cost>
</Items>
</Order>
</OrderList>
Figure 6.9
XML for EDI. Data is sent in a standard format that is easy for computers to parse
and read. Industry groups have been establishing standard formats and tags for
exchanging EDI data within their industries.
frms, including annual reports, balance sheets, and income statements. The com-
mission has required frms to submit data electronically for several years. In 2011,
the commission moved to require most companies to submit data using a new
standard: extensible business reporting language (XBRL). XBRL is largely an
XML-based system for tagging specifc fnancial data. Background information is
available at www.xbrl.org, and www.sec.gov/xbrl.
The accounting discipline has been around much longer than computers have.
Although one goal of accounting is to create standardized views of company per-
formance, accountants and companies have evolved different terms over time. For
example, revenue, earnings, and income are all terms that have been used to rep-
resent the same concept, and any fnancial document might use different varia-
tions of these terms. These differences make it diffcult for computers to read and
extract data from fnancial statementseven if they are in HTML or other digital
format. XBRL is designed to encourage accountants to tag all of the data in a
statement with a standardized name so that it can be retrieved electronically.
Of course, fnancial reports are more complex than just lists of numbers. For
example, they include footnotes and descriptions. And even the data contains oth-
er attributes such as the time period and currency. XBRL has tags and options to
handle most of these issues. Figure 6.11 shows a simple example of a tag that is
used to display the Revenue for a specifed year. This tag can be embedded in a
typical XHTML document which can handle the overall layout and formatting.
The attributes on the <ix > tag provide the type of data (name=ifrs:Revenue),
the units or currency EUR, and the actual data to display (6,863,545).
Figure 6.10
XML document. XML is designed to transfer data between companies and
computers. You can defne any type of tag to describe the data.
Unlike typical XML, XBRL is not designed for transferring transaction or EDI
data. Instead, the tags are defned specifcally for reporting summary data. Yet,
this data is increasingly used by many people: investors, banks, fraud investiga-
tors, and so on.
Elements of a Transaction
What are the major elements and risks of a transaction? Transac-
tions are a critical foundation of modern economic societies. In many ways, trans-
actions defne and enable different types of societies and cultures. A transaction
consists of an exchange of a product or service for money. Consequently, there is
always a risk that something might go wrong in the exchange. Figure 6.12 sum-
marizes the transaction risks that are borne by the vendor, the customer, and the
government. The laws and culture determine how these risks are minimized.
Vendor Perspective
At heart, a vendor cares only about one fundamental aspect of transactions: re-
ceiving the money. This simple statement has several complicating factors: (1) the
payment might never arrive, (2) the payment might be fraudulent, (3) the custom-
er might repudiate the transaction and withhold the money, or (4) the government
might invalidate the transaction.
In older times, these risks were minimized through personal reputation of the
customer and money based on precious metals. More recently, in our mobile,
anonymous society, credit card companies have stepped in to assume much of this
vendor riskfor a price. The credit card companies and credit bureaus provide
identifcation and personal reference services. Under the proper conditions, the
card companies effectively guarantee payment to the merchant. The conditions
primarily consist of (1) keeping good transaction records and (2) identifying the
customer through either a magnetic swipe of the card or a signature.
Customer Perspective
In some ways, customers face a more complex set of risks. Their primary concern
lies in receiving the identifed value for their money. Specifcally, they want to
be sure that (1) they receive the product or service that they ordered, (2) they are
charged only the amount they agreed to pay, (3) the seller does not repudiate the
Figure 6.11
XBRL document element. Note that all of the data elements are tagged with
standardized identifers and with attributes that specify the time frame and currency.
Source: http://media.ifrs.org/ixbrl_example1_2011-03-25.xhtml
<ix:nonFraction
contextRef="y2012"
format="ixt:numcommadot"
name="ifrs:Revenue" unitRef="EUR"
decimals="0" scale="0"
id="XWAND00000326">
6,863,545
</ix:nonFraction>
transaction such as by claiming the product was stolen, and (4) the transaction is
legalfor example, not stolen goods that could be confscated.
In older times, these risks were small when customers shopped at local stores,
could physically examine the goods, paid in cash, and received a written receipt.
Identifcation and reputation of the merchant were critical. Even so, fraud was
a problem. Again, credit card companies stepped in to reduce much of this risk.
Today, credit card companies warranty products, ensure delivery, and validate the
merchants.
Transaction Fees
While often invisible to consumers, the fee for using a credit card consists of a
percentage of the transaction cost paid by the merchant. Of course, economics
shows that the price of the product refects a portion of this fee, so the consumer
and merchant both bear the cost. The cost depends on the size of the merchant, the
card company, and the size of the transaction. Typical fees are 1.5 to 5 percent,
with minimum costs of about ffty cents per transaction.
What about payments by check instead of credit card? In this case, the vendor
bears a relatively high risk that the check is invalid or will be refused for insuff-
cient funds. While there are laws against check fraud, the merchant would experi-
ence high costs to recover the money. Consequently, vendors generally contract
with specialist frms to validate and process checks. Again, the merchant pays a
fee to transfer this risk to another company.
Vendor
Customer
Government
1. Receive payment.
2. Legitimate payment.
3. Customer not repudiate sale.
4. Government not invalidate sale.
1. Receive product.
2. Charged only as agreed.
3. Seller not repudiate sale.
4. Legal transaction.
1. Transaction record.
2. Tax records.
3. Identify fraud.
4. Track money for other cases
(drugs and terrorism)
Credit card company
accepts risks for a fee.
Figure 6.12
Transaction risk factors. Risk elements and causes vary depending on the transaction
method. Laws and business procedures are created to minimize these risks, so new
laws are needed when transaction methods change.
Government Perspective
When discussing transactions, many people forget about the perspective of the
government or society. Yet several government organizations have a strong inter-
est in transactions to protect various members of society. The primary interests
include (1) an auditable record of transactions and fnancial statements for the
protection of investors, (2) a record of taxable transactions for the collection of
sales tax, (3) identifcation and tracking of fraudulent transactions, and (4) general
tracking of money used in transactions to monitor other types of crimes, such as
drugs and terrorism.
Most government interests are established by laws and administrative rules that
are enforced through manual audits of fnancial documents. Some data is collected
from transaction partners, such as the $10,000 cash reporting rule in the United
States. On the other hand, governments are rapidly becoming aware of the juris-
dictional problems inherent in electronic commerce. The most prominent issue is
the inability of states to force out-of-state frms to collect sales taxes for them.
After a few embarrassingly huge fraud cases in the upper echelons of compa-
nies in the early 2000s, Congress got tired of hearing CEOs trying to avoid crimi-
nal charges by saying they did not know anything about the problems created by
their underlings. So, Congress passed the Sarbanes-Oxley Act that requires exec-
Reality Bytes: You Cannot Buy That
Most people are comfortable buying just about anything using credit and debit cards.
But, many people are not aware that the card companies prohibit them from buying
certain items. Currently, the list is relatively short, but the point is that the card com-
panies do have this power and no oversight or appeal process exists. What items are
banned? The basic list is gambling chips (in casinos), online gambling, online por-
nography, donations to at least one non-proft organization (WikiLeaks), and medi-
cal marijuana. The online gambling case is understandable, because online gambling
has been declared illegal by the U.S. government. Medical marijuana falls into the
same category. Although some states have made it legal to buy small amounts of
marijuana, the federal government still outlaws the purchase of marijuana, so card
companies are concerned about their liability. Still, some customers want the option
to use their cards for all purchases. The issue of gambling chips is much harder to de-
fendand the card companies have not even tried to explain. Note that they do allow
gamblers to withdraw cash using their cards at casino ATMswith exorbitant fees
and interest charges. The online pornography issue is defended by noting that the
number of chargeback claims on those transactions is huge, so the card companies
are trying to reduce fraud costs. Still, it should be possible to institute more secure
methods for people who want to use their cards. Blocking donations to WikiLeaks is
far more disconcerting. Yes, the U.S. government leaned on banks and card compa-
nies to block transfers to the site. But, it was not a U.S. organization, and WikiLeaks
has never been tried in a court for any legal violations. As John M. Simpson from
the non-proft Consumer Watchdog put it You ought to be able to use a credit card
for any legal purchase. It seems to me that credit card companies are imposing their
moral values on the world.
Adapted from Quentin Fottrell, What Your Credit Card Won't Let You Buy, The
utives to take responsibility for all fnancial reports. The top executives have to
certify that the accounting reports accurately represent the status of the frm. The
act has scared many executivesnot enough to give up their huge salaries and
pensionsbut they are beginning to search for ways to validate the transaction
and fnancial systems. It is no longer suffcient to just record transaction data
now frms have to institute processes and ensure that people follow them to guar-
antee that the correct data is being recorded.
International Issues
Why are transactions more diffcult in an international envi-
ronment? In the early stages of e-commerce, several people suggested that the
Internets global reach would make it easy even for small businesses to sell prod-
ucts internationally. However, several factors interfere with international sales.
Figure 6.13 summarizes some of the major points involving sales into other na-
tions. Jurisdiction for dispute resolution is a major issue. Nations have many dif-
ferent laws and cultures. Tactics (and content) that might be commonplace in one
country could be illegal in another nation. Gambling is a classic example. The
United States might ban gambling on the Internet, but how can it enforce that law?
On the other hand, if a U.S. citizen has a dispute with an overseas casino, there
would be few options to protect the consumer. Similarly, fraud and Web attacks
can easily originate from several different nations. Varying laws and lack of coor-
dination make it diffcult to identify and prosecute the perpetrators.
Privacy and payment methods present additional challenges. The European
Union has privacy rules that are substantially stricter than those of other nations,
and it does not want international companies transferring personal data to comput-
Figure 6.13
Some international issues with transactions. The basic issues of language, exchange
rates, and tariffs are straightforward. Processing shipments through customs can be
challenging. Bigger problems arise if something goes wrong because of the different
jurisdictions.
Shipping
Currencies
Languages
Customs and taris
Jurisdiction for disputes
Dierent laws and systems
Verify seller and purchaser
Payment methods
ers outside of the Union. Identifying vendors and customers is also problematic
when nations utilize different payment mechanisms. Although the major credit
cards are accepted in many nations, many people do not use them. For years,
transactions between international businesses were largely handled with special-
ized documents issued among large international banks. With the increasing im-
portance of the Internet, leading to international sales to individuals, these issues
become more diffcult to solve. Credit card companies and banks largely protect
consumers, but international crime makes the task more complex. More impor-
tantly, vendors are largely unprotected so many companies are reluctant to sell
directly to international customers.
Te Role of Accounting
How do you track and compare the fnancial information of a
frm? Accounting systems are important because they extend throughout the
company and because they focus on money. They are used to collect data and
evaluate performance. Accounting systems also enable managers to combine the
many divisions into an integrated picture of the entire company. Accounting sys-
tems also provide controls over the data to ensure accuracy and to prevent fraud.
The primary purpose of accounting is to collect the fnancial data of the frm, en-
sure that it is accurate, and create standard reports. It is hard to capture all of the
elements of an accounting system in one illustration, but Figure 6.14 summarizes
the essential components of an accounting system. The accounting transaction
system can be examined in terms of inputs, outputs, and processes.
If you are not going to be an accountant, why do you need to understand ac-
counting, and why review it in an information systems book? The short answer is
that the accounting information system provides the foundation data for the frm.
Reality Bytes: More Ways to Transfer Money Online
American consumers rarely use checks or cash. Credit and debit cards are still im-
portant in transactions, but online payment methodsparticularly PayPalcontinue
to grow in popularity. In May 2011, three of the largest banks in America announced
a new service to make electronic payments. The banks involved (Bank of America
Corp., Wells Fargo & Co., and J.P. Morgan Chase & Co.) are in a joint ventured
named clearXchange. Mike Kennedy from Wells Faro and chairman of the new ven-
ture, noted that Customers want to move payments from paper to electronic meth-
ods, so if we can meet our customers fnancial needs, they will be better customers
with us. The service will rely on the cell phone and is aimed at person-to-person
transfers. Money will be transferred directly from checking accounts to another ac-
count based on a linked e-mail address or phone number. Initially, the service was
released in Arizona and carried no transaction fees; but banks might eventually add
fees for the service. Small businesses could also use the technology to avoid paying
processing and merchant fees for credit and debit cards. A survey by consulting frm
Aite Group LLC noted that in 2010 U.S. households made $865 billion in payments
between individuals in 11 billion transactions. Dividing by the number of households
yields an average of $7,500 and 96 transactions per household.
Adapted from Robin Sidel and Amir Efrati, Big Banks Join Battle for Online Pay-
ments, The Wall Street Journal, May 25, 2011.
The longer answer revolves around the fact that the accounting process is increas-
ingly automated in companies. As a result, all managers have to understand and
analyze accounting data. You will have access to basic accounting data and reports
directly from the computer. It will be your responsibility to monitor the changes
and understand the effects.
Input and Output: Financial Data and Reports
Raw fnancial data is collected by the accounting department and stored in an ac-
counting journal. Modern accounting requires the use of a double-entry system
to ensure accurate data. In a double-entry system, at least two entries must occur
for every transaction. Generally, one entry records the effect of the money (e.g.,
cash, accounts payable, accounts receivable), and the other refers to a specifc
category (e.g., sales, offce expenses, commissions). Each entry includes the date,
amount of money, account number, the name of the person or frm involved, per-
haps a comment, and the name of the person making the entry. The journals pur-
pose is to record all the transactions.
Journal entries represent raw data. To be useful, this data must be transformed
into information. The frst step is to categorize the data by accounts or categories,
which is the purpose of the general ledger. The ledger is a collection of accounts
that break the data into specifc categories. Common categories include accounts
receivable, accounts payable, inventory, and cash. Although some standards
Suppliers
Customers
Management
Shareholders
Banks
and
Creditors
Departments
& Employees
Sales &
Accounts
Receivable
Produce
Management
Accounting
Reports
Produce
Shareholder
Reports
Sales & Receivables
Inventory Changes
Purchases &
Payables
Loans & Notes
Expenses
Equity
Shareholder
Reports
Management
Reports
Product
Inventory
Orders &
Accounts
Payable
Inventory
Management
& Fixed Asset
& Cost Acct.
Payroll &
Employee
Benets
Cash
Management,
Investments,
Foreign
Exchange
Supply &
in-process
inventory
Governments
Tax
Filing &
Planning
Strategic
& Tactical
Planning
Inventory Changes
Payables
Capital Acct
Sales Tax
Tax Filings
Tax data
Receivables
Inventory &
Assets
Payroll
Planning Reports
Planning
Data
Figure 6.14
Transaction processing is a major function of the accounting system. The accounting
system collects data throughout the company and produces consolidated (centralized)
reports that are used for planning and management.
exist, each company can defne its own chart of accounts, which allows owners
and managers to examine data on whatever categories are important to their frm.
Defning the chart of accounts is one of the important steps in managing a busi-
ness because it defnes the type of information you will examine on a daily basis.
For managers to make comparisons between divisions and other frms, ac-
counting systems produce standardized reports. Most companies produce balance
sheets, cash fow statements, and income statements every quarter. These reports
are produced following standard accounting rules to enable owners, managers,
and investors to compare the fnancial positions of various companies over time.
Purchases, Sales, Loans, and Investments
One of the primary purposes of accounting is to record the fnancial transactions
with external organizations. In addition to collecting the raw data, the accounting
system contains controls that minimize fraud by limiting access to the data. The
system also creates summary and detail reports to monitor key information.
Managers often build exception reports into the accounting system that are
triggered when some event occurs. If sales in some region suddenly drop, if there
is a major increase in the cash balance, or if inventories fall below a defned lev-
el, a message will be sent to the appropriate manager. The manager typically re-
sponds by searching the recent summary reports for a possible cause.
Inventory
Most organizations need to control inventory carefully. Retail stores fnd it hard to
sell items that are not in stock. Manufacturing frms need to receive and process
parts as cheaply as possible. Inventory control consists of knowing exactly what
Reality Bytes: Vending Machine Payments
Vending machines can be useful, but they typically have one big problem: They only
accept cash. Today, people buy most items with credit and debit cards. If people
could pay with cards or even their cell phones, it is possible that sales at vending
machines would increase. In 2009, U.S. sales fell 10 percent to about $20 billion
partly because of the recession. Operators would also gain data on specifc sales so
items would be less likely to run out and they could track the top- and weakest-
sellers. The technology exists, so the decision seems straightforward. The problem is
that the technology is relatively expensive. And most vending machines are run by
small frms who cannot afford the technology. John Mitchell Jr. runs one of larger
operations. Treat America Ltd. Of Merriam Kansas, runs about 12,000 vending ma-
chines. He notes that Youre catering to a population that might be as small as 30 or
40 people. The unique preferences of that population can drive sales signifcantly.
The company originally stocked machines based on average preferences of the re-
gion. In January 2010, Mr. Mitchell converted about 40 percent of his machines to
record real-time sales data; which showed that about 40 percent of the 45 slots were
dead spirals selling fewer than one item per week. Installing a wireless system
costs about $300 on a $3,000 machine, and vendor proft margins are typically only
one or two percent.
Adapted from Ilan Brat, Restocking the Snack Machine, The Wall Street Journal,
August 3, 2010.
items are available and where they are located. The system also needs to deter-
mine when to place new orders. It must then track the orders to make sure each
item is delivered to the appropriate location at the right time. With EDI, the inven-
tory control system can monitor current sales and automatically place orders with
the supplier.
Manufacturing frms use these systems to implement just-in-time inventory
control. The computer system monitors the current production requirements,
keeps track of deliveries, and electronically sends orders to the suppliers. The sup-
pliers then deliver the parts just as they are needed on the production line.
Automated inventory control systems also help identify and prevent theft. By
recording all movement of items from receipt to sales to shipping, management
knows exactly how many items exist. Consider a retail store like a bicycle shop.
The computerized inventory notes that there should be three Cateye computer
speedometers in stock. Yet when a customer asks to buy one, you notice there
are only two left. If there is no mistake in your inventory report, you conclude that
someone stole one of the items. Although the system did not prevent the speedom-
eter from disappearing, it does show which items are susceptible to theft. It also
helps control theft by employees, who will be less likely to steal if they know that
the items are carefully monitored.
Te Accounting Cycle
An important aspect of accounting systems is that they produce information in
specifc cycles. Firms are required to produce reports that refect the fnancial con-
dition of the frm at the end of every quarter. Accounting systems are based on
these requirements. For the most part, managers operate from quarterly reports,
with intermediate monthly reports for some items. Because of the volume of data
in the detail, most companies keep only current statistics and summary reports on
fle. Older data is shuffed off the system to make room for the current numbers.
As a result, managers may not have easy access to detailed data from prior years.
Reality Bytes: Chiquita ERP Problems
In 2006, Chiquita Brands, based in Cincinnati, began a multimillion-dollar upgrade
to consolidate its accounting systems and centralize budget planning and forecast-
ing. Operating worldwide with three semi-autonomous business units, the company
faced many challenges, including software bugs, staffng shortages, and resistance
from the business units. Even after four years, the company was struggling to get all
of the components to provide value. CIO Manjit Singh noted that The general per-
ception is, we got it to work. But it was painful, and the [managers] arent necessar-
ily looking at this as a successful implementation Once we start using it for fore-
casting, Im hoping the reaction to the project will turn more positive. The system
was based on Hyperion and is designed to consolidate data used for forecasting. The
goal is to remove e-mailed spreadsheets from the process and get fnancial planners
to use the centralized system so everyone sees the same data and same analyses at
the same time. Singh noted that Hyperion (now owned by Oracle) sold the project to
the fnance group and bypassed IT. IT was called in to support the project after the
contracts had been signed.
Adapted from Robert L. Mitchell, Inside Chiquita's 'Painful' Finance System Over-
haul, Computerworld, August 9, 2010.
Technology Toolbox: Selecting an ERP System
Problem: How do you select and evaluate ERP systems?
Tools: : Several ERP systems and even more ERP consulting frms exist to give you
choices and help you evaluate your options. Choosing a system is relatively diffcult
because of the costs. The cost of the hardware and software is high, but you face
larger costs in converting from your existing system and redesigning your business
processes. Once you have implemented a particular ERP system, it is diffcult to
switch to another vendor. So, you have to be careful to select a system that is going
to work for your particular organization. Today, most large companies already have
ERP systems, so the growth market lies with small and midsize businesses. Most
of the ERP vendors are actively targeting this SMB market. Businesses within this
category have to pay attention to the anticipated growth and scalability options of
the software systems. For example, even small businesses can gain the advantages
of ERP systems by leasing them from companies that host the software on Internet
serversyou pay only a monthly fee.
In terms of technical business features, most of the large ERP systems are somewhat
similaroffering the standard accounting, HRM, SCM, and CRM integration. How-
ever, you will fnd differences in ease of use, internationalization, and customization
options. For a smaller business, one of the more important areas to look at is the
quality of the predefned industry-specifc application. Most ERP vendors provide
shells that are customized for each industry, so it takes less time and money to set up
the system for your company.
Planning Stage Goals and Outputs
Initiation Estimate costs, establish objectives, select team
Initial planning Initial vendor list and basic features (hardware platform, fees,
internationalization,vendorsizeandstability,andsoon).
Requirementsgathering Identifybusinessrequirements.Detailedlistofspecifcations,unit
goals,andcriticalfeatures.Evaluationcriteria.
Demos and selection Productdemonstrations,ratings,andsitevisits.
Implementation Customize applications, convert data, restructure company
operations,defnenewprocesses,andtrainemployees.
Selecting an ERP system is a time-consuming task, and you have to carefully
evaluate the needs of the company. You will create lists of primary objectives and
generate detailed specifcations. You will endure hundreds of hours of vendor pre-
sentations and comparisons of databases, reports, and customization capabilities.
Then you have to implement the solution. Many companies work with a partner
consulting frm that specializes in one ERP system. This detailed knowledge simpli-
fes the conversion process because the consultant frm has experienced experts. But
even midsize frms take six months to a year; larger operations can require a year
or two. Most companies choose to implement ERP systems in stagesperhaps ac-
counting frst, followed by SCM and CRM. Even these packages can be split and
implemented in smaller pieces.
Quick Quiz:
1. Assume you work for a midsize construction frm that does about 30 percent of
its work internationally. Find at least three ERP vendors and outline the features
they provide.
2. Assume you work for a large retail clothing frm with stores in most U.S. states.
Identify the specifc accounting and fnancial features you would want in an ERP
system.
Process: Controls, Checks, and Balances
Double-Entry Systems
An important objective of accounting systems is to maintain the integrity of the
fnancial data. The goal is to prevent mistakes and discourage fraud. Double-entry
accounting provides a method to locate mistakes in data entry. If an amount is en-
tered incorrectly, the account totals will not balance.
Because many transactions involve outside organizations, mistakes can be
caught by sharing data. Every month frms receive a statement from the bank. The
totals can be compared to changes in the frms cash account. Similarly, compa-
nies typically send receipts when they receive payments from each other. Auditors
periodically send verifcation requests to suppliers and customers to make sure the
data was recorded correctly. EDI strengthens this approach, because transaction
data is transmitted in computer form among the companies.
Separation of Duties
Another type of control is the separation of duties. A manager in the purchas-
ing department might be responsible for choosing a supplier of parts. Only the
accounting department can authorize the transfer of money to the supplier. The
objective is to minimize fraud by requiring a potential thief to deal with multiple
employees.
Many banks take this concept a step further. They require employees (espe-
cially tellers) to take their vacations every year. Several instances of fraud have
been revealed when the employee was no longer at the job to keep the fraudulent
mechanism running.
Audit Trails
An audit trail is important to accounting systems. It enables investigators to track
backward through the data to the source. A cash fow statement might indicate that
the company has spent twice as much money this month as last. To fnd out why,
trace backward and fnd all of the raw entries that make up the number. Together
with dates and amounts, the raw journal entries can contain the identity of the per-
son responsible for the entry. By keeping this identifcation data, it is possible to
list every article that affects an item on a report.
Human Resources and Transaction Processing
What are the transaction elements in the human resources
management system? Every company has employees. Companies collect
hundreds of pieces of data for each employeesome for management purposes,
others because they are required by law. For years, the human resources (HR)
department focused on flling out and storing forms. The enormous amount of pa-
perwork alone begs for computerization just to cut down on storage space needed.
Computerized databases also enable managers to fnd specifc data on employees.
Early HR software emphasized these two benefts. Modern HR software is ex-
panding beyond simple forms to improving data collection and providing better
analyses. To illustrate the problems presented by large-scale transaction-process-
ing systems, consider the three areas of input, output, and processing.
Input: Data Collection
Figure 6.15 illustrates the basic components of a human resources management
(HRM) transaction-processing system. Note that the system is even more com-
plex because the data comes from all areas of the company. To understand how
the HRM systems became so complicated, begin with the obvious data that needs
to be collected: numbers related to the payroll. For hourly workers, the system
needs to collect and monitor hours worked. For many sales tasks, the system must
compute sales by employee to determine commissions. Professional service frms
often ask employees to track their time in terms of billable hours for work that
is charged back to clients. In all three situations, as the number of employees in-
creases, it becomes increasingly diffcult to collect all of these statistics and verify
their accuracy. It also becomes harder to fnd specifc pieces of data.
Think about paychecks you have received. In addition to the payment amount,
there could be 10 to 20 other numbers on the pay stub. Companies monitor and
report several types of payroll taxes, including federal, state, local, Social Secu-
rity, and health. Also, frms monitor employee benefts, such as health care and
retirement. Most frms also handle employee deductions for employee purchases,
savings plans, stock purchases, parking, meal plans, and other options. In some
situations, companies must garnishee wages and forward them to a third party.
Management
Managers
Employees
Customers
Government
Process
Payroll
Benets
Vacation
Compile
Merit
Evaluations
& Salary
Changes
Job
Applicants
Produce
Management
Employee
Reports
Screen
Jobs &
Applications
Produce
Government
Reports
Employee Data Files
Employee
Data
Evaluations
Salary
Merit &
Salary
Sales Data &
Commission
Employee
Data
Merit
Applicant
Data
Employee
Summaries
Job &
Applicant
Data
EEO Data
Government
Reports
Management
Reports
Figure 6.15
Most employees know that human resources management (HRM) deals with
payroll and benefts. But HRM also collects data and produces reports for a myriad
of government reports, oversees employee evaluations and job applications. The
department also handles training and education opportunities.
Human resource departments also track days taken for vacations, personal
time, and illness. In larger companies, HRM provides training courses and offers
testing of critical skills. Employee attendance and performance data is stored and
incorporated into evaluations.
With the increasing use of merit pay, the system must also track employee eval-
uations. Some performance measures are tied to productivity or output within the
employees department, so HR must relate employee work schedules to produc-
tion and quality measures. Most companies use a centralized HRM department to
advertise job openings and to screen the initial applicants, verify credentials, and
keep basic employment and hiring data.
Output: Reports
The human resources department also produces several reports related to payroll.
Along with printing checks, HRM must provide expense reports and forecasts to
the accounting system. Periodic reports are created for job vacancies and analyses
of employee performance and morale.
HRM departments also spend a great deal of time creating reports for vari-
ous government agencies. All companies must fle various economic reports deal-
ing with employment. Tax-withholding data must be fled regularly with federal,
state, and local agencies. HRM departments create equal employment opportunity
reports detailing characteristics of their workforce, job applicants, and hiring de-
cisions. Then there are various reports required by the Occupational Safety and
Health Administration (OSHA) regarding injuries and exposure to various haz-
ards. If employees need to be certifed, companies fle aggregate reports with the
various regulatory agencies. All of these reports have deadlines.
In addition to the standard reports, the human resources department is responsi-
ble for maintaining compliance with all relevant employment laws. Hence, HRM
staff must continually monitor the employment data and evaluate it for exceptions
and problems.
Process: Automation
The human resources department is a busy place. Keep in mind that the data and
reports apply to every branch of the company. Even standard items such as pay-
checks become complicated when the company is split into several divisions scat-
tered across the country. Also, remember that accuracy is crucial. Employees can
become upset if their paychecks are wrong. Errors with government reports can
lead to fnes and lawsuits. Equally important, companies with good HRM depart-
ments are able to offer additional benefts to employees. With a good information
system, they can offer cafeteria-style benefts where each employee selects a per-
sonal combination of benefts.
Small businesses have long complained about the burdens imposed by govern-
ment reports and data collection. To alleviate some of the hassles and expense,
several companies specialize in automating the data collection and report writing.
Consider payroll: Because of the constantly changing laws, many companies rely
on an outside agency to collect data and print the paychecks. One of the larg-
est providers is Automated Data Processing (ADP). Even if a company chooses
to maintain its own payroll records, it typically purchases the software from a
third party instead of trying to keep up with the annual changes using internal
programmers.
Several companies sell software that automates HRM data handling and pro-
duces government-required reports. From economics to equal employment to
OSHA reports, the basic HRM reports are being computerized. You still need to
collect the data in the proper format and convert it to the purchased software. In
addition to saving time in producing reports, the packages often contain the essen-
tial government rules and can analyze the data to spot potential problems.
Some newer technologies are being used to simplify data gathering. In particu-
lar, companies are searching for ways to make it easier for workers to deal with the
HRM department. A system created by PRC, Inc., uses touch-tone phones and a
voice-response system to enable workers to make changes directly to their base
information, like changing their address or tax withholding. Another approach is
to install PC-based kiosks and use the Internet, so that employees can look up
information, sign up for training classes, or modify their personal data whenever
they wish. Other companies are using similar software and the corporate network
to allow workers to perform basic HR tasks from their desks or from home using
the Internet.
Production Management
Can a company become more effcient and productive? Some-
times this question is easy to answerif you have been losing money. But per-
haps you are manager of an organization that has been successful and proftable.
Many companies have been successful for years and continue to operate the same
way they always have. This practice might not be bad in every case, but the con-
cept of continuous quality improvement dictates that you should always look
Figure 6.16
Production management. Production of goods and services generates considerable
information. Purchase orders are placed with suppliers, internal documents like
the bill of materials record production, shipping invoices record interactions with
customers. Quality control measures are recorded at several steps in the process.
Production
Information
System
Bill of Materials
Crank UL6500
Pedals LK3500
Stem UL6600
Saddle Selle
Purchase
Orders
Customer Order
Shipping
Receiving
Production
and Assembly
Quality
Quality
Quality
Suppliers
Customers
for improvementsno matter how successful you have been. New production and
information technologies are introduced continuously. Any one of these technolo-
gies might give you the ability to leapfrog your competition, or it might give your
competitors the ability to squeeze you out of the market.
As shown in Figure 6.16, information technology has several important uses in
production management. It is important to track orders, shipments, and all steps
in the production process. The bill of materials is an important concept in man-
ufacturing. It records the components that were used to manufacture a fnished
product. Data on it is used to trigger deductions in inventory of parts and signal
the addition of the fnal product to fnished inventory. These actions impact the ac-
counting ledgers, so recording the proper amounts and time is important. Similar-
ly, quality control measures are taken at several steps in the processparticularly
with the receipt of supplies and the shipment of fnished products. Detailed quality
measures during the production process are used to identify processes, machines,
and employees that need improvement.
Production Issues
Manufacturing in large scale presents several problems and issues to management.
As shown in Figure 6.17, consider the case of a clothing manufacturer with sever-
al factories. Each factory can produce multiple items and the company works with
many factories. The items are collected and shipped to hundreds or thousands of
customers and stores. How do you schedule effcient production? How do you
factories
Customers
or stores
Figure 6.17
Production challenges. Multiple factories produce many items that are shipped to
hundreds of customers. How do you schedule effcient production? How do you
ensure the right products go to the correct locations?
ensure that the right products are shipped to the correct stores? These questions
seem obvious, but they are critical. In the 1980s, the huge designer jean company
Gitano self-destructed because it was unable to answer these two questions.
Obviously, information systems are a key component to the answers. But you
need to make sure the processes are in place to collect the correct data in a timely
manner. This data has to be available throughout the company so managers can
monitor for problems and make the correct decisions. For instance, with only par-
tial data, it might appear that a factory is idle, while complete data might show
that it is scheduled to ramp up production as soon as supplies or new machines
arrive. Collecting data in a manufacturing environment can be a diffcult process.
Most workers are busy trying to build products and are not interested in taking
time to enter data into a computer. You need to defne the processes and jobs to
ensure that accurate data is also being recorded. You also need to automate as
much of the data collection as possible. Some machines can transfer data directly
into the information system, but often you need to assign the task to someone.
This step is particularly important when things go wrong. You also might have
to add bar-codes or RFID tags to products so they can be counted and recorded
automatically.
Once you have accurate data, you can usually purchase tools or hire experts to
analyze the production processes and fnd improvements. The feld of production
operations management has many tools and models to organize and optimize pro-
duction and scheduling.
Distribution and Inventory Control
A key mantra in todays business is that you do not want to hold inventory. Just-
in-time (JIT) production was designed to eliminate parts inventories held by
manufacturers. Instead, suppliers are asked to deliver parts exactly when they are
needed. In one sense, the process shifts the inventory onto the supplier, because
the supplier is penalized if it is late with shipments. But, ultimately, it reduces the
Reality Bytes: Controlling Enterprises with BPM
In the early 2000s, New York Attorney General Eliot Spitzer fled lawsuits against
several insurance companies for fraud and excessive fees. Politely stated, the entire
industry appeared to suffer from poor management with limited knowledge of how
fees were collected and disclosed. Integro Insurance Brokers was founded in New
York in 2005 with the idea of applying business process management principles and
software to the industry. Christophe Marcel, enterprise software architect stated that
there were a lot of questions about transparency and disclosure of where fees came
from. There has been an industrywide initiative for quality improvement processes
To do that effectively, you need BPM. Using tools form BEA Systems Inc, the com-
pany built client-services applications to provide up-to-date data on transactions, and
enable executives the ability to monitor all of the processes. Harcourt Assessment
Inc., a San Antonio-based developer of scholastic tests, is using BMP software from
Savvion Inc. to manage the process of developing test questions, using the system to
automate online processes for collecting and analyzing customer feedback.
Adapted from Heather Havenstein, BPM Is Helping Firms Control Critical Busi-
ness Processes, Computerworld, October 30, 2006.
Technology Toolbox: Designing an Executive Information System/Dashboard
Problem: How do you begin to understand the data collected by an ERP system?
Tools: Executive information systems or digital dashboards are designed to display
summary data in real time. The systems usually begin by presenting a graphical
overview using some primary measures of the organization. Executives and manag-
ers can quickly compare progress on key variables. They can then drill down and see
more detailed statistics and even original transaction data such as orders.
http://www.visualmining.com/ncs/projects/Examples/PerformanceDashboard/TransitMetrics.jsp
An EIS can be a powerful tool to follow the progress of the company. Most ERP
systems have some type of software that makes it relatively easy to extract data and
create an EIS. However, one of the most diffcult tasks is to identify the key items
that need to be displayed on the main form. Across industries, companies vary enor-
mously in the outcomes managers need to follow. Even within an industry, frms and
managers rarely agree on what items are the most important. Consequently, each
company needs to develop an EIS customized for its executives.
One approach to identifying the desired elements in an EIS is to follow the bal-
anced scorecard method proposed by Robert S. Kaplan and David P. Norton. They
recommend a complete analysis beginning from the top to identify the main strategic
goals and derive measures that ft into primary categories. For instance, you would
typically need measures on fnancial, customer, internal data, and learning and
change. The top managers cooperate to identify the main indicators of each variable.
Quick Quiz:
1. Assume you have been hired to help a regional law frm. What key elements
might go on the main EIS screen?
2. Assume you have been hired to help an airplane manufacturer; design the main
EIS screen.
3. To help design an EIS for a retail video store chain, briefy describe the screens
and steps needed to identify stores that are performing below par and fnd pos-
sible causes of the problems.
number of products held in inventory because everyone along the supply chain is
interested in minimizing the value of the stalled inventory. Instead, everyone fore-
casts production needs and delivers products by keeping them moving through the
supply chain.
As shown in Figure 6.18, the distribution center (DC) is an important part of
a modern supply system. Incoming bulk shipments are unloaded from the trucks
and placed on high-speed conveyor systems. Bar code readers in the center of the
DC read each box. Outgoing trucks are matched to a specifc customer and store.
The computer reads the customer order list and routes the specifed number of
boxes of each item to every truck. When the trucks are loaded, they deliver the
items to the designated customer and store. Additional savings are created by rout-
ing the same truck to the ship or rail yard freight docks to pick up an incoming
bulk shipment. These trucks then deliver the bulk shipments on the left side of the
DC in the diagram. Scheduling systems keep the trucks moving to maximize their
capacity and to reduce deadhead or empty loads.
Distribution centers might be run by manufacturers, customers, or third-party
transportation specialists. Wal-Mart is a leader in using the centers to reduce costs
and deliver exactly the needed products to every store. Ultimately, you also need
to match your shipment data against the customer receipts to ensure that your sys-
tem is accurate and to compensate for shipping losses.
Integration in Business
How do businesses combine data from operations? The easiest way
to understand the power of todays computer systems is to look at how integra-
tion has changed over time. Almost any organization has issues with integration.
Consider a basic manufacturing company that has several production plants and
sells into multiple markets. The managers need to track production data, costs,
and quality from each plant. Figure 6.19 shows that they also track sales by region
Factories
Customer
or stores
Figure 6.18
Distribution center. Factories deliver products in bulk to the DC. High-speed
conveyors read bar codes on the incoming boxes to split the products into smaller
shipments and route them to trucks headed for individual stores.
and use average costing to estimate net profts. ERP systems support considerably
more complex analysis and integration across suppliers and customers as well as
fnancial institutions. But the basic ideas are the same.
Assume that you are a manager responsible for producing the integrated report.
You need to collect the basic data, perform some simple computations, create the
report, and write a report for management discussing the results. The process of
creating this report has changed considerably over time. Figure 6.20 shows the
various methods by decade. It is important to understand all of the methods be-
cause it helps you appreciate the amount of work and steps processed by a modern
ERP system. Also, companies are at different levels of development, so you could
end up working for a company that uses any of these technologies.
In the 1960s and 1970s, computer systems were built for individual depart-
ments and areas within the company. In many companies, these systems became
islands. They were focused on one task and did not share data with each other.
For instance, the accounting department collected the basic transaction data and
produced the necessary accounting reports. Anyone in the company who wanted
to use this data relied on paper printouts of the standard reports. When spread-
sheets arrived in the 1980s, the basic accounting numbers were often rekeyed into
spreadsheets in other departments. Besides wasting employee time in retyping
numbers that were already stored on a computer, this practice caused more errors
from mistyping the data. Furthermore, consider that when the accounting depart-
ment changes the numbers, some users of the data might not get the updated ver-
sions, and people would attempt to make decisions on the basis of outdated data.
Producton
Plant Quantity Quality Costs Prior Month
Boise 260,721 9.6 $41,222,697 219,961
Decatur 171,371 8.75 $31,094,809 188,862
Yaounde 314,856 9.4 $29,119,967 406,859
746,948 9.25 $101,437,473 815,682
Wtd. Avg. 135.80
Sales
Region Quantity Avg.Price Sales Prior Sales Wtd. Cost Net
US East 225,091 $289.17 $65,965,373 $53,533,640 30,567,941 $35,397,432
US West 204,725 $273.20 $56,217,470 $42,018,759 27,802,185 $28,415,285
Canada 65,763 $192.00 $12,561,225 $12,327,456 8,930,786 $3,630,439
South America67,495 $133.50 $9,013,737 $11,281,515 9,165,996 ($152,259)
Other 47,313 $169.00 $7,991,533 $8,120,900 6,425,228 $1,566,305
610,387 $211.37 $151,749,338 $127,282,270 82,892,136 $68,857,202
$0
$5,000,000
$10,000,000
$15,000,000
$20,000,000
$25,000,000
$30,000,000
$35,000,000
$40,000,000
$0
$10,000,000
$20,000,000
$30,000,000
$40,000,000
$50,000,000
$60,000,000
$70,000,000
US East US
West
Canada South
America
Other
Sales
Sales
Prior Sales
Net
Figure 6.19
Integration with ERP. An ERP system can integrate almost all aspects of the company
into one huge database. Either directly or through an executive information system,
management can retrieve any report or piece of data by selecting options.
Notice that in the 1980s, the clerical support was replaced with PCs. As a man-
ager, you are expected to write your own documents and create your own charts.
In the 1990s, organizations began the steps toward integration. The basic trans-
action is stored in a DBMS, and networks provide access to the database. More
important, this integration makes it relatively easy to create automated tools that
extract the desired data, display results in charts, and enable managers to quickly
fnd the answers to basic questions. To some extent, this type of system means that
the company no longer needs you: the middle manager. You might be interested to
know that this trend was predicted in 1958 (Levitt and Whisler, Harvard Busi-
ness Review). It just took 20 years longer than they anticipated for technology to
be developed.
As shown in Figure 6.21, in the frst decade of 2000, even more powerful inte-
gration tools were adopted. The goal of these tools is to integrate all of the data in
the company, including ties to suppliers and customers. Again, the entire system
runs on a DBMS; the difference lies in its reach and its capabilities. Tracking cus-
tomers and working closely with suppliers are increasingly critical functions in
business. Just collecting the data and billing customers or paying suppliers is not
suffcient. You need to be able to analyze and compare the data. Who are your best
customers? Which ones cost you the most? Which suppliers provide the best qual-
Figure 6.20
Middle management over time. The methods used to create integrated reports have
changed over the past three decades. With simple transaction systems, managers
computed the totals, drew graphs, and had secretaries type the report. With the
adoption of personal computers, middle managers (or their secretaries) reentered the
data into spreadsheets and used a word processor to print the fnal report. With an
integrated system, top managers can use a personal computer to query the database,
draw the graphs, and produce the fnal report.
3 reports
(printed)
weekly reports
1970s
Sales
terminals
Sales
terminals
Sales
terminals
Sales
terminals
Sales
terminals
Sales
terminals
Central
computer:
create reports
Central
computer:
create reports
Manager:
Integrate, graph
analyze
Manager:
Integrate, graph
analyze
Secretary:
type & revise
Management
3 reports
(printed)
weekly reports
1980s
Management
personal computer
sales
data
sales
data
sales
data
1990s
DBMS:
On-line data
Personal Computer
Weekly reports &
ad hoc queries
(applications)
Management
query
data
ity and are the most reliable? Answers to these questions can mean the difference
between success and failure.
Enterprise Resource Planning
How do you combine data across functional areas, including
production, purchasing, marketing, and accounting? Enter-
prise resource planning (ERP) is the current state of the art in integrated
information in business systems. The systems incorporate data from fnancial ac-
counting, logistics, and human resource management. The feld is dominated by
large, expensive software packages from companies such as SAP, Oracle, Law-
son, and Microsoft Dynamics. The systems use databases, processes, and rules to
provide up-to-the-minute data on the major fnancial issues in a frm. One of the
key points of ERP systems is that they run on top of a DBMS; hence, all of the
data is centralized and accessible via DBMS queries and reports.
As shown in Figure 6.22, ERP systems handle all of the fnancial accounting
systems. They also emphasize purchasing, human resource management, and in-
vestment management. The systems are tailored for specifc businesses and can
focus on areas such as manufacturing, research and development, and retail sales.
One of the primary strengths of the ERP systems is that they were designed to
handle data for large companies operating in an international environment. In the
late 1990s, many companies chose to install commercial ERP systems, instead of
trying to modify their existing systems to handle the year 2000 problem.
Computer use in most companies began with transaction-processing systems.
Because transaction systems are structured and there is considerable experience at
Figure 6.21
Integration with ERP. An ERP system can integrate almost all aspects of the company
into one huge database. Either directly or through an executive information system,
management can retrieve any report or piece of data by selecting options.
sales
data
Sales
terminals
Sales
terminals
DBMS
Executive Information
System
Personal Computer
or PDA
or Browser
Management
query
data
Suppliers
Banks
Production/
Service
Enterprise Resource
Planning
this level, it is a logical place to begin. However, it is also tempting to treat each
transaction separately: (1) payroll services can be purchased from a specialized
data processing company, so the data will be handled separately from the other
corporate data; (2) a sales order-processing system might be constructed indepen-
dently of the inventory control system; (3) process control systems to handle man-
ufacturing tend to be isolated because the data (e.g., robotic control signals) are
different from the data used in the rest of the company; (4) similarly, the corporate
accounting system is often developed as a stand-alone product. Journal entries are
created by copying data in reports produced by other systems. Although each of
these transaction systems offers management advantages to their respective de-
partments, it is diffcult for managers to use data from other departments. Also,
independent systems make it diffcult for executives to share data and evaluate
interrelationships between the departments.
The amount of data integration needed in a company often depends on the man-
agement structure of the frm. Some frms are highly decentralized, so that each
business unit makes its own decisions and functions independently of the others.
Typically in these situations, only accounting data (proft/loss) are integrated and
reported to upper management.
On the other hand, some organizations are much more integrated. In your eco-
nomics courses you were shown the difference between vertically and horizon-
tally integrated frms. Consider a vertically integrated frm such as an oil company
that functions at different levels of production (including oil exploration, drilling,
transportation, storage, and retail sales). Although an oil exploration team may not
need access to daily fuel sales in New York State, they do need to forecast future
demand for oil. Likewise, the retail sales division does not need to know the daily
costs associated with drilling for oil, yet they might need to track deliveries and
communicate with the corporate offce.
Consider a horizontally integrated frm such as Wal-Mart with retail stores
in many different cities. It achieves lower costs by combining the buying power
of all its stores. By coordinating sales, warehouses, and distribution, Wal-Mart
can negotiate better prices with manufacturers. Moreover, Wal-Mart reduces op-
erating costs by standardizing management practices (and information systems)
Figure 6.22
Primary ERP functions. Each area has dozens of additional detailed functions. The
real key is that all of the components are integrated. For example, any data entered
into the HRM system is immediately refected in the fnancial reports.
Accounting
All transaction data and all fnancial reports in any currency
Finance
Portfolio management and fnancial projections
Human Resources Management
Employee tracking from application to release
Product design and manufacturing life cycle
Supply Chain Management
Purchasing,qualitycontrol,andtracking
Customer Relationship Management
Contacts, orders, and shipments
across all the stores. By integrating information from all stores, it is easier for
Wal-Mart to forecast customer demands. Also, by networking the store infor-
mation systems, managers who experience higher sales of certain products can
request shipments from stores that are not selling the item as rapidly.
Manufacturing frms can gain additional benefts from integrating data. Benefts
like just-in-time inventory, total quality management, and mass customization can
exist only with the tight integration of data. The National Bicycle Industrial
Company of Japan illustrates how integrated data is used to provide customized
products to mass markets.
International Environment
Several ERP features are important to frms operating in an international environ-
ment. First, all menus and reports should be available in several languages, so
clerks and managers can use the language they prefer. Second, the system should
handle currency conversion automatically, so managers can view reports in any
currency. Similarly, conversions should be capable of being fxed at a point in
time, so that when items are transferred, they can be valued at the exchange rate in
effect at that time, even if the rate changes later.
A more complex feature for the international environment is the ability to pro-
duce reports following the rules of individual nations. For example, a company
with subsidiaries in many nations would need to produce reports that follow the
rules (e.g., depreciation) for each specifc nation, and then produce consolidated
reports following the rules of the home nation.
A third complicating factor arises from taxes. In addition to the rates, the rules
and procedures vary by nation. The rules are particularly important for payroll and
beneft applications. A good enterprise application automatically incorporates the
rules for each nation and state.
Financial Accounting
The accounting system is a core feature of an ERP. Eventually, all transactions
must be recorded in the general ledger accounts. The accounts fulfll the standards
required by each nation. They are used to create the standard accounting reports.
The systems provide fexibility by enabling managers to create their own subac-
counts and subledgers, which are used to create reports on additional topics. An
important feature of the accounting system is that standard accounting reports can
Reality Bytes: Paying for the Train
Long Island Rail Road trains do not accept credit cards. And the tiny train station
in Amagansett in the Hamptons does not have ticket vending machines. Conduc-
tors will accept cash, but most young riders do not carry much cash. So they get a
bill, which they might or might not pay. One rider, Renee Osgood, noted that You
dont always plan ahead to make sure that you have cash on you. I never really carry
around cash, anyway. In response, LIRR president Helena Williams noted that the
railroad is increasing the number of signs, asking riders to carry cash.
Adapted from Andrew Grossman, Cashless Hop Rides on the LIRR, The Wall
Street Journal, July 27, 2010.
be generated at any time for any section of the company. The ERP system auto-
matically uses the most up-to-date data.
In addition to standard fnancial accounting, the systems manage assets and
provide common treasury functions such as cash management. The systems also
provide basic audit trails and other accounting controls. To make them easier to
use, most ERP systems provide enterprise (or executive) information system (EIS)
capabilities. Managers can examine data at virtually any level of detail. From
summary values, they can drill down to more detail.
Logistics
Logistics consists of the operations required to purchase materials, deliver them to
the warehouses and factories, and sell and distribute products. It incorporates tra-
ditional MRP analysis, quality control, accounts payable, and accounts receivable.
In todays manufacturing companies, logistics is an important component of
just-in-time inventory and demand-driven production. Using an integrated sys-
tem, the marketing department gets up-to-the-minute data on customer demands.
Marketers can cooperate with designers and engineers to develop new products.
The specifcations can be transferred to the production machines and raw material
orders can be generated for vendors. Purchasing and payments can be tracked and
generated over EDI networksincluding the Internet. As orders are generated and
inventory levels change, the accounting data is automatically updatedproviding
instant analysis of proftability.
For service-oriented companies, logistics involves service management tasks.
The ERP systems can track customers, identify repeat customers, monitor service
contracts, help salespeople with call management, and handle automatic billing
and accounts receivable issues.
Human Resource Management
Payroll is a complicated function, particularly in a multinational environment in-
volving different rules and currencies. Even in a single state, the issues of benefts,
state and federal rules, and legal issues arising from child support make handling
payroll a complex task.
Reality Bytes: Marin County ERP
County governments perform many tasks that are similar to common businesses.
ERP software exists for government agencies, but it typically must be heavily cus-
tomized. In 2006, Marin County in California signed contracts with SAP and De-
loitte Consulting to build a new ERP system to replace the old legacy system. By
mid-2010, Marin executives were unhappy with the system and fled lawsuit against
Deloitte Consulting, claiming that only 50 percent of the functionality is available
and working. David Hill, director of the countys IT services department argued that
In order to make it happen with SAP, we would need to re-architect it, redesign it,
and that would be quite expensive. The budget analysis estimated fxing and sup-
porting the SAP system would cost at least $34.7 million over the next ten years;
which starting over with a new tool would cost $26.2 million.
Adapted from Chris Kanaracus, Marin County to Rip and Replace Ailing SAP Sys-
tem, Computerworld, August 24, 2010.
Todays HRM departments handle such additional tasks as recruitment, train-
ing, travel, and organizational planning. Each step must be documented and re-
quires a variety of federal and state reports. In addition to these basic tasks, most
of the major ERP systems enable HRM departments to offer Web access to basic
data. For example, employees can use the Web to check on their taxes, change
their withholding status, and sign up for beneft plans and training sessions.
Integration
Integration is probably the most important feature of the ERP systems. All the
data is stored in a central database; hence, data is entered only one time (but into
a double-entry accounting system). All reports are generated from the base data.
Custom queries and reports can be generated through the DBMS.
Consider a simple example. A manufacturing plant takes an item from inven-
tory. The system instantly adjusts the inventory quantity on hand. It also updates
the fnancial value of the inventory holdings on the general ledger and any sub-
ledgers that utilize that fgure. New orders can be triggered automatically with the
orders and payments sent through common EDI mechanisms. All of the changes
are made automatically. When managers request reports, the new data is automati-
cally incorporated and displayed using current currency conversions.
The key point to remember is that all of the transactions and accounts are in-
tegrated. Managers can request reports by using any combination of data at any
timeand each report will use the most up-to-date information.
Database
Headquarters
(England)
Database
Subsidiary
(Spain)
Financial data
General ledger
Payroll
(others)
Manufacturing
Product details
Inventory
(others)
Use inventory item.
Update inventory value.
Check for reorder point.
Order new item through EDI.
Update Accounts payable.
Project cost report.
Daily production reoprt.
Weekly nancial status.
Cash ow.
Budget versus actual cost.
Figure 6.23
ERP integration. Although data can be distributed, it is still integrated across the
organization. Changes in one item (inventory) cause changes in all related databases.
Reports are generated from current data.
Most of the major ERP systems also utilize distributed hardware and software.
Hence, the database can be split into many pieces stored in different locations. As
changes occur in one location, they are automatically distributed across the net-
work to the other locations. The company can add a subsidiary with its own pro-
cessing support. Yet, all of the new data is readily accessible to managers through-
out the company.
Figure 6.23 provides a simple example of data integration. When a factory uses
an inventory item, the system reduces the current inventory count. It also changes
the inventory valuation in the general ledger. The item usage might trigger a pur-
chase through the EDI system, which must also be recordedalong with the ac-
counts payable change. Since the databases are shared across the organization, all
changes are automatically included when new reports are generated.
Remember that all of the modules are integrated. So manufacturing schedules
developed in the production module automatically provide data to the payroll sys-
tem and personnel systems. Then the fnancial data (e.g., wages) is linked back to
the general ledger, which provides updated data for all fnancial reports.
One important catch with an ERP system is that it requires changes to the way
the company operates. In many cases, these changes can be goodfor example, it
forces everyone to follow the standard accounting procedures. In other cases, the
ERP is too infexible and interferes with the way the company operates. Managers
have to carefully evaluate the trad-eoffs of integration and fexibility.
How do you make production more effcient? Supply chain manage-
ment (SCM) concentrates on the production side of ERP. It begins with logistics
(purchasing and receiving components), through manufacturing confguration,
and into distribution of the products. As shown in Figure 6.24, these factors in-
volve a signifcant portion of the company, requiring integration of a huge amount
of diverse data.
The key to understanding the value of SCM is to go back in time again to see
how manufacturing evolved. From the 1920s through the 1970s, companies in
many industries recognized the importance of economies of scale or mass produc-
tion. The automobile industry presents the classic example. Producing thousands
Reality Bytes: Applicant Tracking
About 61 percent of North American companies use applicant tracking software
to scan and sort job applications. According to a survey by HR research frm Bersin
and Associates, even small and mid-sized businesses have begun adopting them. The
frm notes about 55 different vendors sell applicant tracking system (ATS) software.
The largest vendors include Taleo, silkroad, Oracle, iCIMS, Lumesse, and Kenexa.
Job applicants use a Web site to upload a resume which is then scanned for key
words and routed to the appropriate location if it fnds enough matches. An interest-
ing aspect to the software is that many of them do not fully support PDF fles. So job
applicants should typically submit a resume in Word format.
Adapted from Kelly Eggers, Resume Debate: Word v. PDF, The Wall Street Jour-
nal, May 25, 2011.
of identical cars enables the company to spread the huge fxed costs across a large
baseleading to lower average costs. The huge scale enabled the car companies
to negotiate better prices with suppliers and dealers, reducing costs even further.
So, in the name of lower costs, the companies produced thousands to millions of
identical items. They relied on the marketing departments for two critical pur-
poses: (1) forecast consumer preferences in advance, and (2) convince consumers
they need the products that were built. So, the car salesperson says, Sure, we
could order a car for you, but it will cost more and take months. You would really
be happier with this car and you can drive it home today.
Of course, mass production has the potential for mass disaster. If you predict
incorrectly, or cannot convince customers to buy the existing product, you end up
dumping the products at sale prices. Remember that you have to clear the way for
next years models. More critically, mass production means that it is impossible to
please all of the consumersleaving a niche open for your competitors. A niche
in the small-car market enabled Toyota to become one of the largest producers in
the world.
SCM Changes the Focus
So how does SCM help? It can change the entire system. Mass production begins
at the supply side and builds products as cheaply as possible to eventually sell to
consumers. With a truly integrated supply chain, it is possible to start with the cus-
tomers. The marketing department identifes exactly what each customer wants.
The customized orders are entered into the system, and the engineering depart-
ment evaluates the order, makes design changes as needed, and schedules produc-
tion. Manufacturing knows each desired production date and organizes products
to minimize production costs. Component orders are placed with suppliers elec-
Vendors Customers
Design &
Engineering
Marketing
Distribution
Manufacturing
Purchasing
Administration & Management
HRM Accounting Finance MIS
Just-in-Time
Quality control
& Custom orders
Quick
response
Designs and Quality
Mass Customization Demand-pull
Design feasibility
& production costs
product planning
partnerships
& joint
development
partnerships
& joint
development
planning
& monitoring
planning
& monitoring
Figure 6.24
Total data integration begins with the vendors, tracks data through all operations of
the frm, and incorporates information from customers. Each area in the frm has
easy access to data from any other location. This integrated data is used to make
better decisions by enabling managers to focus on the big picture instead of on local
solutions.
tronically. On the day of production, as components arrive, they are scanned into
the system and routed to the appropriate location. At the same time, payments are
scheduled with the banks. As the parts arrive at each machine, the central process
controller confgures each machine correctly as the product moves through the as-
sembly line. At the end of the production line, the product is labeled and shipped
to the appropriate customer. The customer can be notifed electronically and billed
automatically.
Imagine how much easier it is to be in marketing now. Instead of convincing
customers to buy what you have produced, you frst fnd out exactly what they
want, and that is what you sell them. It is not quite that simple, but at heart, that is
the principle. Mass customization.is the process of manufacturing products de-
signed for specifc consumers but using mass-production techniques to keep costs
low. It can be done only if you have an integrated system. In many cases, it is not
possible to produce exactly what the customer wants (after all, most customers
want the world for free). And sometimes you have to produce for groups of simi-
lar customers instead of a single consumer, but the principle is the sameuse the
integrated technology to provide as much customization as possible. For example,
it might never be possible to produce a truly customized automobile economi-
cally, but customers might be satisfed with more choice of options.
JIT production is an example of how integrating information can change a pro-
duction process. Japanese manufacturers created this method with low-tech kan-
ban signals that conveyed information along the production line to the suppliers.
Today, ERP systems can provide immediate information from all stages of pro-
duction. This information can be made available to suppliers so they can schedule
deliveries. The overall goal is to reduce production costs and eliminate bottle-
necks by reducing the need for huge inventories.
SCM Challenges
SCM and integration are powerful tools. You might expect all frms to be using
them heavily. But SCM has two serious obstacles: (1) cost and (2) establishing
connections and agreements with outside frms. The issue of cost can be over-
comealthough the tools are expensive, if they can save enough money, the cost
is acceptable. But that explains why smaller frms with fewer purchases might not
be able to afford the technology. The second issue of building connections across
multiple frms is far more diffcult.
A key issue in SCM is connecting to suppliers electronically. It can also require
connections to distributors and transporters. Each of these hundreds of companies
could have different hardware and software. Plus, all companies involved have to
worry about security issues. Sharing data is a more diffcult problem than simply
blocking all access. B2B e-commerce is an interesting solution proposed to solve
some of these problems. In the late 1990s and early 2000s, several B2B auction
sites were set up to coordinate purchasing and tracking of products within an in-
dustry. For example, Covisint was established by several automobile manufac-
turers to handle transactions for automobile components. Few of these electronic
auction sites survived the dot-com crash. However, some are still supported and
used in major industries, such as automobiles and steel.
To a typical consumer, purchasing seems like an easy process: you fnd a prod-
uct and buy it. As shown in Figure 6.25, the reality is considerably more complex.
Consider the purchase of a new item. First the buyer has to identify the purpose
and individual features desired. Then the buyer identifes potential suppliers and
sends a request for proposal (RFP). The suppliers select products that match the
desired features and make proposals and bids for the sale. The buyer evaluates
each request, selects a supplier, and negotiates a contract. From this point, the
process is similar, even for repeat orders. The buyer generates a list of specifc
products, prices, and quantities, along with a desired delivery date. This purchase
order is sent to the supplier, who verifes the orders and checks to make sure prod-
ucts are in stock and available for delivery. Any problems with prices or delivery
times require notifying the buyer and negotiating changes. Eventually, the prod-
ucts are shipped to the buyer along with an invoice detailing the items delivered.
Of course, the buyer has to verify the quantity and quality of the items received.
Any discrepancies are recorded and the seller is notifed. Some items are returned.
Ultimately, the seller sends a bill to the buyer detailing the amounts owed. On
the accounting side, the appropriate accounts payable entries have to be made.
Finally, the supplier has to be paid, and needs to monitor receipt of payments and
match them against the bills.
One goal of an ERP system is to automate as many of these tasks as possible.
Figure 6.26 shows a version of an ERP screen for a buyer or product planner. Oth-
er screens enable suppliers to connect and electronically pick up purchase orders
or check on the status of deliveries and payments. All data is entered into a shared
database. From the buyers perspective, notice the scorecard that quickly indicates
the status of several key variables. Additional tools on the page enable the buyer
to check on the current inventory status of individual products. For example, he
Figure 6.25
SCM Purchasing. Purchasing is considerably more complicated than it appears.
It often requires several interaction steps with supplierseven for existing
relationships. Monitoring prices, quality, and timeliness is critical to both sides.
Identify product features
Request for Proposal
Evaluate suppliers
Negotiate contract
Identify exact items and costs
Correct errors
Receive products
Match receipts against order
Return items
Pay supplier
Provide product details
Negotiate contracts
Verify orders and times
Correct errors
Schedule delivery
Monitor shipments
Handle returns
Monitor payments
Buyer
Suppliers
can quickly see if the company has too many items in stock for a particular prod-
uct. Additional tools provide the ability to examine payments, shipments, prob-
lems or exceptions, and perform analyses on vendors and products.
Integration across Systems
The buyer-seller relationship points out a major challenge that remains to be
solved: how do you exchange data electronically between buyers and sellers? To-
day, both the buyer and seller probably have ERP systems, but they might be from
different vendors. Even if they are from the same vendor, they might not format
data the same way. So, the buyer will ask the seller for a person to log in and
check on the supplier data. Similarly, the seller will ask the buyer to log in and use
its system.
The major work in the 2000s by the ERP vendors has related to connecting sys-
tems and sharing data electronically. XML is one tool that was established to help
companies exchange data. As shown in Figure 6.27, ideally, each company would
implement an interface that would accept and read XML documents sent from its
partners. The machines could quickly parse the documents and transfer them into
the required internal format. Internal functions could then schedule deliveries,
check on shipments, or transfer payments. All of the data within an XML docu-
ment is in a specifed format and tagged so that a computer program can quickly
read the fle and identify the data and its purpose. Several companies, including
IBM and Microsoft, are developing technologies to use XML to make it easier to
share data across companies.
Figure 6.26
Purchasing manager perspective. Notice the scorecard to indicate important
performance variables. This portal page includes links to tools that help manage day-
to-day tasks.
How do you keep track of all customer interactions? Who are
your best customers? Although customers are important to all businesses,
the Internet and wireless applications add new dimensions to managing customer
relationships. One problem is the expanding number of customer contact points,
from sales representatives, to call centers, to Web sites, and wireless connections.
Customers expect merchants and suppliers to remember actions and decisions that
were made earlierregardless of the method of contact. Consequently, companies
need integrated systems that instantly provide all details of customer contacts. The
new technologies also provide innovative methods to keep in touch with custom-
ers and identify their specifc needs to sell additional products and services. Sev-
eral software tools have been developed to improve customer relationship man-
agement (CRM). Most ERP systems have added (or purchased) CRM systems to
provide an integrated contact and sales management system.
Multiple Contact Points
One of the greatest challenges facing a company today is the multiple sources of
contact points with customers. Most of the original systems designed to handle
these interactions are separate. Salespeople often keep their own records, Inter-
net support systems may not be totally connected to the sales fulfllment centers,
and faxes are rarely integrated into the online customer fles. But customers as-
sume that when they talk to one person, that person has records of all the prior
interactions.
Figure 6.27
The role of XML. XML is used the transfer messages to diverse systems. Each
system contains tools to read and write XML documents and convert the data into the
appropriate internal database tables.
Buyer
Supplier
ERP: Oracle ERP: SAP
Messages
and data
<xml>
<order>
</order>
</xml>
At frst glance, it appears that it would be straightforward to build an integrated
application to hold all customer interaction data. Of course, it would be a lot of
data and would take time to build the application. But the real challenge lies in
getting everyone to enter all of the data. Consider the situation of a salesperson
who has invested time and collected substantial data on product preferences and
customer work environments. That information gives an advantage to the sales-
person. Why would the salesperson be willing to share it?
Customers with multiple divisions and many different product tracks also add
complications to CRM. The system has to be able to track transactions, questions,
and issues by a variety of factors (date, product, company, person, and so on).
The system also needs a sophisticated search routine so users can fnd exactly the
pieces of data required.
Feedback, Individual Needs, and Cross Selling
The main purpose of CRM systems is to provide individual attention to each cus-
tomer to improve sales. By tracking prior purchases, you understand the status of
your customers. By providing new channels of communication, you improve the
ability of customers to provide feedback to comment on products and services and
to make suggestions for improvements. By identifying patterns in purchases, you
can develop new ideas for cross selling. If a group of customers tends to purchase
several products, you can search the CRM database to fnd customers with only
part of the solution and have your salespeople demonstrate the advantages of the
entire suiteusing the other customers as examples and references.
The fip side to CRM is that collecting and coordinating substantial data about
the customer can lead to privacy problems. As long as the data is secured and used
internally, few problems arise. But frms still need to be sensitive to customer
wishes about unsolicited contacts. In fact, customer privacy requests need to be
part of the CRM system. The issues are more complex when the selling frm has
multiple divisions, and each one wants to push new products to existing custom-
Reality Bytes: Customer Relationship Management at Dow
Most people do not think about the mundane tasks that computers handlesuch as
sorting data. Computers have been doing it for so long, that it is taken for grant-
ed. And sorting is relatively standardfor small amounts of data. But sorting large
amounts of data is defnitely not trivial. In 2010, computer scientists at UC San Di-
ego set a world record for sorting when they were able to sort one terabyte of data in
60 seconds. A full minute seems like a long time to wait for a sort, but a terabyte is
a ton of data to move around. The task required 52 computers with dual quad-core
processors and 24 gigabytes of RAM, plus 16 500 GB disk drives. The team also set
a record for the fastest data sorting rate. The system sorted one trillion data records
in 172 minutes (almost three hours). Alex Rasmussen, the lead graduate student on
the team noted that Sorting puts a lot of stress on the entire input/output subsystem,
from the hard drives and the networking hardware to the operating system and appli-
cation software. Many business transactions and reports require sorted data, includ-
ing search results and product recommendations.
Adapted from Daniel Kane, Data World Record Falls as Computer Scientists Break
Terabyte Sort Barrier, UC San Diego News Center, July 27, 2010.
ers. The marketing staff needs to use the CRM system to coordinate and monitor
all contacts.
Wireless applications provide even more options for CRM. Your salespeople
can stay in constant contact with the corporate database. They can retrieve current
shipping status or detailed customer information during a sales call. They can for-
ward questions or comments, which can be analyzed and answered immediately.
CRM Packages
ERP systems often have a CRM component that provides a unifed view of cus-
tomer data. For example, as Figure 6.28 shows, sales managers can quickly check
on the status of customer purchases, returns, or any other transaction involving
customers. When salespeople call on customers, they need to be able to quickly
retrieve all of the recent transactions with that customer. Perhaps the customer
had problems with a recent shipment, or perhaps they just placed a new order. The
salesperson can review those activities before calling on each customer. Then the
salesperson can investigate any problems or check on current status and provide
immediate answers to customer questions. One of the biggest challenges with a
CRM system is collecting the necessary data. An ERP system will have all of the
basic transaction datafrom the perspective of the main company. But will sales-
people enter all of their personal data into the system?
Figure 6.28
Sales and CRM from a sales managers perspective. Keeping all the customer
interaction data in one location makes it straightforward for the salesperson to
understand all of the companys associations with each customer.
Another important aspect of CRM is the ability to provide direct customer
service. As shown in Figure 6.29, ERP systems have self-service portals that en-
able customers to check on order status, pay bills, order new products, contact
salespeople and account managers, check on company news and events, or ask
for additional help. Although the page does not directly support XML, it does
make it easy for customers to check on various transactions 24 hours a day. By
encouraging customers to connect directly to your system, you obtain more ac-
curate data and reduce costs. Of course, if your customer is the 800-pound gorilla
in the industry, it is more likely that your salespeople will be spending time on the
customers ERP site.
Summarizing ERP Data
How can a manager handle all of the data in an ERP system?
ERP systems result in huge databases containing an enormous amount of data.
Since the data represents all aspects of business transactions, it contains incredibly
valuable information needed to make key business decisions. But, as a manager
or executive, how do you fnd this information? Since it is stored in a relational
DBMS, you could write queries to retrieve data. But, after a little time, you will
see that you often need the same data every dayso you can compare the prog-
Figure 6.29
PeopleSoft CRM from a customers perspective. The portal into the ERP system
enables customers to check the status of various items, order new products, and
contact the company.
ress of common items. Consequently, many ERP systems have special tools to
retrieve this standard data and display it in a graphical format.
Digital Dashboard and EIS
One of the major challenges to any information system is to make it easy to use.
This process is complicated because the decisions need to be provided to upper-
level managers who have little time to learn complex applications, yet deal with
huge amounts of diverse data. One approach is to build a portal that displays key
data and graphs on one page. The page retrieves data from a data warehouse, the
Internet, or even machines within a factory and displays graphs and warnings.
Toolkits exist to help build this digital dashboard. The older term for this ap-
proach is an executive information system (EIS). Figure 6.30 shows a sample
digital dashboard that with various charts designed to provide a quick overview of
several key functions.
Figure 6.30
Digital dashboard example. Like the gauges on an automobile dashboard, the digital
dashboard presents a top-level graphical representation of the status of various
elements. Managers identify variables and choose from charts, gauges, icons, and
tickers to display the data. Each item should be selectable to drill-down and provide
more detail.
Much like the dashboard on a car, the purpose of the main screen is to pro-
vide an overall picture of the status of the frm or a division or production plant.
Managers can select the specifc division and make comparisons to yesterday, last
week, last month, or other locations. If there is a problem or a decision to be
made, the executive can drill down to get more detailed data by pointing to an-
other object. For example, if the main screen shows that current sales in the west
region are low, the executive can focus on the west region and bring up the last
few quarters of sales data. The EIS will graph the data to highlight trends. The
manager can then dig deeper and bring up sales by departments or check the sales
performance of each employee, highlighting unusually high or low sales fgures.
By pointing to customers, executives can get current profles on the main custom-
ers and examine their recent purchases.
How Does an EIS Work?
For an EIS to work, it must be connected to the transaction-processing system or
data warehouse, since it is the source of the data. Many of these systems are cre-
ated with special software that simply grabs data from the corporate databases.
In one sense, the EIS is a complex model of the frm. Figure 6.31 illustrates how
executives can visit different divisions on the computer and retrieve the data
immediately. For the EIS to be useful, the computer model must be a faithful rep-
resentation of the actual company.
As a model, the EIS display has inputs of raw materials and people. Outputs
are typically measured by traditional accounting standards of profts, costs, and
growth rates. The EIS maintains budgets and forecasts and can compare them to
actual values. The functions and processes are determined from the individual
departments. For instance, there could be a production model that describes the
manufacturing output. An EIS at McDonnell Douglas has a graphics screen that
displays portions of airplanes as they are being built. As a wing is completed, it is
drawn onto the computer model.
Reality Bytes: Cash is No Longer King
In the U.S. the Federal Reserve System is in charge of issuing and controlling cash.
A study on the stock of small-denomination bills in use indicates that the use of cash
reached a peak in the mid-1990s. Between 1995 and 2006, noncash transactions had
increased from 250 a person to more than 300. Much of the shift has been to debit
cards. But, it has been diffcult for small vendors (or friends) to accept payments
with cardsparticularly without a storefront. A few companies, including PayPal,
Intuit, and Square are trying to make it easier for people to make small payments
without cash. For example, Square has a small credit card reader that plugs into the
earphone jack of an iPhone or iPad. Processing costs for credit cards are relatively
high and sometimes require monthly contracts. Square relies on per-transaction costs
of 15 cents plus 2.75-3.5 percent of the transaction value. Apple and several other
companies are trying to fnd new ways to transfer money. But any mechanism must
be supported by both customers and vendors.
Adapted from Claire Cain Miller and Nick Bilton, Cellphone Payments Offer Alter-
native to Cash, The New York Times, April 28, 2010.
Advantages of an EIS
The primary goal of an EIS is to provide easy access to corporate data for the ex-
ecutives. Instead of waiting for a report, the top executives can retrieve the data
as soon as it is available. Also, because all the data is accessible from the same
system, it is easier to examine data from different departments to produce a better
view of the big picture. Another useful feature is that the executives use of the
data is nonintrusive.
Imagine that you are CEO of a company, and you do not have an EIS. The
monthly reports have just indicated that one of the warehouses is not running
smoothly. You want to fnd out what the problems are. You suspect the ware-
house manager is part of the problem, but you need to be sure. What do you do?
The most direct approach is to go visit the warehouse. But what happens when
you show up at the warehouse? It is likely that the manager and the workers will
change the way they work. Your attempts to collect data have altered the way the
system runs, so you will not get the information you wanted.
Other options include sending other people or asking for additional information
via the chain of command. Although useful, these methods will be slower and the
information you receive will be colored by the perceptions of the people collect-
ing the data. For example, many people will try to focus on what they think you
want to hear.
Production
Distribution
Sales
Central Management
Executives
Data
Data
Sales
Production costs
Distribution costs
Fixed costs
Production Costs
South
North
Overseas
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
1993 1994 1995 1996
South
North
Overseas
Production: North
Item# 1995 1994
1234 542.1 442.3
2938 631.3 153.5
7319 753.1 623.8
Data
Data
Data for EIS
Figure 6.31
Executive information system. As industries become more competitive, managers
search for ways to evaluate and improve the overall operations. Enterprise systems
collect data from across the frm and make it available to managers and top
executives. Executives can start with an overview of the frm and drill down to
various levels and departments to get more detailed data.
The EIS minimizes these problems by providing instant access to the corporate
data. The executives can produce reports and examine departments without inter-
fering with the operations of the company. Graphs can be created automatically.
The executives can set up different scenarios or simulations. Most of these activi-
ties are accomplished simply by pointing to objects on the screen.
Transaction Accuracy: Sarbanes-Oxley
How does the CEO know that fnancial records are correct? Af-
ter several huge corporate scandals in 2001 and 2002 (Enron, Worldcom/MCI,
Tyco, and so on), Congress decided to try and stop some of the problems. In par-
ticular, high-level executives were pleading innocent claiming that they did not
know about the problems. So, they passed the Sarbanes-Oxley Act, which requires
top executives of public companies to sign a statement attesting to the validity of
the annual fnancial statements. So, how does the CEO know if the transaction
data and reports are correct?
The solution proposed by the accounting standards board (AICPA) is to require
more independent audits and reviews of the audits. Accounting audits examine all
aspects of the transaction-processing system: from data collection, to data storage
security, to accuracy of calculations and reports. Auditors also check physical as-
sets (particularly bank accounts) to ensure that the numbers listed on the fnancial
statements tie to real-world data. As shown in Figure 6.32, bank accounts are a
critical element of auditing because they are maintained by external companies.
Similarly, inventory and customer sales are at least sampled to ensure that they
match against the internal fnancial records. These external numbers help anchor
the internal statements. For example, it is hard to falsify the books if you have to
match the cash balance at the bank. But not impossible. Somehow the Italian dairy
company Parmalat managed to list $5 billion in cash in a Cayman Island account
from 2002 to 2004 that did not exist.
Reality Bytes: Debit-Card Fee Cut
Overall, banks are in charge of the debit card and credit card systems. Techni-
cally, MasterCard and Visa are now independent companies, but they were created as
agents of the largest banks to handle processing for credit and debit cards. Retailers
who accept the cards pay fees to use the cards. In the U.S., fees for debit cards aver-
aged from 1-2 percent of the transaction cost. These costs added up to an estimated
$20 billion a year for U.S. merchants. And they were about twice as high as in Eu-
rope and other nations that regulate the fees. In 2010, Congress passed a law to limit
the bank interchange fees charged for debit card purchases. The law took effect in
July 2011 and essentially cut the fees in half. Banks expressed concern about losing
a revenue stream and searched for ways to cut customer reward programs and in-
crease other fees. Merchants would gain through lower costs and could either reduce
prices or increase profts. Most economists were betting merchants would keep the
extra proft.
Adapted from Miguel Bustillo, Merchants Win Debit-Card Fee Battle, The Wall
Auditors also focus on the transaction and reporting processes. Do standard
procedures encourage the timely and accurate recording of all data? Do people fol-
low the standard procedure? The generally accepted accounting practices (GAAP)
were created to establish standards and guidelines for accounting procedures. To-
day, one of the purposes of accounting software is to encode these practices into
the system. In one sense, corporate accountants and managers are given less fex-
ibility. Yet from managements perspective, the software encourages everyone in
the frm to follow the proper procedures. Of course, some companies and man-
agers will fnd these procedures too restrictive. Particularly if the company uses
highly individualized practices and nonstandard production techniques.
One of the strengths of an ERP is the ability to defne and enforce rules that
apply to transactions. Accountants and auditors build the rules to maintain data
integrityoften through separation of jobs. For instance, you can require a sales
manager to verify all sales data entered by salespeople. Expenses can be automati-
cally submitted to review by managers, and rules can be added to identify and fag
unusual activity or high values. Receipt of supplies can be automatically checked
against purchases Vendor identifcation and payments can be handled relatively
automatically. Todays auditors spend considerable time evaluating ERP rules and
procedures to ensure standard practices are followed to protect the integrity of the
fnancial data.
Cloud Computing
Why do most companies avoid cloud computing for ERP sys-
tems? It is tempting to think that cloud computing would be a natural ft for ERP
systems. The software is complex, and cloud systems could automatically main-
Financial
Statements
Cash
Inventory
Sales
Bank
Customer
Inventory
Some nancial numbers
are tied to the outside
world.
Validate these to anchor
the statements.
Figure 6.32
Financial audits. Verifying the quality of the data requires tying elements of the
fnancial statements to externally validated numbers. Inventory can be physically
counted. Customer sales can be sampled, and bank data must match the internal
records.
tain critical backup facilities for data and servers. Companies could access their
data using simple computers with Web browsers from anywhere in the world. Be-
cause all companies need basic accounting and HRM software, it should be pos-
sible for vendors to provide online systems. And, several companies do provide
these cloud-based services. NetSuite is one of the leading independent providers.
The main vendors, including Oracle, SAP, and Lawson also provide options for
hosting their software online. The primary beneft to online hosting is that compa-
nies need fewer initial resources and staff to maintain the hardware and software.
Most of the costs become variable costslargely based on the number of users.
However, many companies have avoided using cloud computing for their pri-
mary accounting ERP systems. Some of the reasons include concerns about secu-
rity, privacy, and reliability. Cost, and the perception of cost, is also an important
issue. Most ERP systems require customizationregardless of whether the sys-
tem is online or run internally. These costs can be a signifcant portion of ERP
design and implementation. Managers are also leery of variable costs based on the
number of users. Most managers expect (wish) their company to growso they
anticipate increases in the number of future users. With in-house, fxed systems,
this growth generally does not cost extra money for the ERP (until hardware has to
be scaled up with large increases). With cloud-based systems, managers anticipate
increased costs every year, so they perceive costs as lower with their own system.
Large companies generally built their own systems and hired personnel to
maintain them before the cloud-based systems became available. It would be dif-
fcult for them to justify switchingparticularly if switching requires recreating
the customized components. Consequently, most cloud systems target small and
mid-size businesses. Yet, small businesses are likely to have minimal account-
ing systems and rely on simple systems and probably a single accountant. They
are unlikely to see much beneft from an online system that requires continual
monthly payments.
Another factor in holding back online systems is that browser-based ERPs are
relatively new. Plus, mobile Internet connectionsparticularly in the U.S.have
been relatively slow and expensive. As ERP software, networks, and mobile tech-
nologies improve, more businesses might see a value in using online systems.
One area that has attracted companies is CRM. Salesforce.com has been rel-
atively popular and experienced high growth rates. However, many companies
treat their Salesforce data independently from their ERP data. It is largely used as
a contact-management system. Salespeople use the system to share contacts and
notes with others in their company. But, this data is probably not connected to in-
house data including production and shippingunless people manually transfer
the information. Still, these cloud-based CRM systems appear to provide benefts
to companies and salespeople so they continue to use them until more integrated
tools become available.
Summary
Every organization must perform certain basic operations: pay employees, manu-
facture products or services, pay bills, monitor revenue, and fle government re-
ports. Operations are relatively structured, short term, and easy to computerize.
They form the foundation of the company. MIS supports operations by collecting
data and helping to control the underlying processes. Transaction-processing sys-
tems are responsible for capturing, storing, and providing access to the basic data
of the organization. The goal is to capture the transaction data as soon as possible.
Common collection methods include point-of-sale devices, process control, elec-
tronic data interchange, and electronic commerce Web sites. Because data is the
foundation for all other decisions, transaction-processing systems must maintain
data integrity and minimize the threats to the data.
Even when the frm has production and sales in multiple nations, ERP soft-
ware makes it possible for managers and executives to get a complete up-to-date
picture of the entire company. Equally important, this level of integration makes
it possible to change the way the company operates. Instead of relying on mass
production to lower costs, companies can focus on the customer and the market.
Working from the customer perspective, the frm can produce the items and op-
tions desired by the market. This type of production is possible because the in-
formation is instantly available to production and to suppliers. Integrated data on
customers improves relationships with customers by enabling companies to iden-
tify client needs and respond to problems faster.
Integrating accounting, production, sales, and employee data makes it easier to
identify bottlenecks, identify and reward the best employees, and gain a competi-
tive advantage. Ultimately, the key to ERP systems is not the softwareit is the
integration and reengineering of the business operations that make your company
better. The tools provide the ability to change the way the frm operates, and to use
these capabilities to fnd a better overall solution. Whether the frm manufactures
products or is a service frm or governmental agency, you can fnd new ways to
operate when everything is integrated.
A Managers View
How do you keep up with the thousands or millions of transactions in a
modern organization? Information systems have been developed to handle
internal data from accounting, human resource management, and produc-
tion. Enterprise systems integrate the many pieces into data that is consistent
and accessible across the organization. Increasingly, it is possible to interact
electronically with other companies. These systems make it easier to iden-
tify and interact with your customers and suppliers. The integration makes it
possible to alter the way the company is structured and managed. Creating
agile frms that use mass customization to deliver exactly what the customer
wants at costs comparable to those experienced with mass production.
Key Words
Web Site References
General Travel Reservations
Expedia www.expedia.com
Kayak www.kayak.com
Orbitz www.orbitz.com
Sabre Travelocity www.travelocity.com
Discounts
Bestfares www.bestfares.com
Cheap Tickets www.cheaptickets.com
Priceline www.priceline.com
TravelHUB www.travelhub.com
IndividualAirlines(U.S.)
Alaska/Horizon www.alaskaair.com
American www.aa.com
Delta www.delta.com
Frontier www.frontierairlines.com
Jet Blue www.jetblue.com
Southwest www.southwest.com
United www.ual.com
US Airways www.usairways.com
accounting journal
audit trail
bill of materials
chart of accounts
continuous quality improvement
customer relationship man-
agement (CRM)
digital dashboard
distribution center (DC)
drill down
electronic data interchange (EDI)
enterprise resource planning (ERP)
exception reports
executive information system (EIS)
extensible markup language (XML)
extensible business report-
ing language (XBRL)
general ledger
just-in-time (JIT)
magnetic ink character rec-
ognition (MICR)
mass customization
optical character recognition (OCR)
point of sale (POS)
process control
radio frequency identifcation (RFID)
request for proposal (RFP)
supply chain management (SCM)
transaction processing
Review Questions
1. Describe four methods of data capture.
2. What is the role of XML in EDI transactions?
3. What is the purpose of XBRL?
4. Explain how to a vendor all payment methods have costs.
5. Why are customers unlikely to adopt a new payment mechanism?
6. How do accounting and fnancial systems form the backbone of ERP
systems?
7. Why are human resource management systems such an important component
of ERP systems?
8. How are distribution centers used to reduce inventory, and what information
is needed for them to operate effciently?
9. What is the role of a DBMS in an enterprise resource planning system?
10. What are the primary features and capabilities of an enterprise resource
planning system?
11. How does supply chain management enable a frm to change its structure and
focus?
12. Why is customer relationship management so important? Why does it require
an integrated system?
13. Why is an executive information system or digital dashboard so important to
an ERP system?
14. If you start a new company, would you use cloud computing to handle your
accounting system? Why or why not?
Exercises
1. Choose 10 Web sites that sell items or content. Create a table and identify the
payment methods that they support.
2. In 2010 and 2011, a few cell phone companies tried to create a new standard
to enable people to pay for items with their cell phones. Why did that project
fall apart? What alternative, but similar method is being pushed?
3. Identify the primary government laws and rules that are designed to protect
the governments interest in transactions.
4. Assume that you are running a small manufacturing business. To reduce
costs, you want to contract with an overseas manufacturer to build a key
component. How would you locate a company do to the work? What
transaction details and data sharing will you have to set up to make the
process work?
5. Find a balance sheet and income statement for a company (Try the EDGAR
system at www.sec.gov). Rebuild the two statements as separate pages in a
spreadsheet.
C06E06Base.mdb
6. Create the basic forms needed to use the database in a transaction system.
7. Create the spreadsheet in Figure 6.19 that shows the initial integration. The
database fle contains the basic data. You should build queries and import the
data into a spreadsheet to create the analysis and charts. It is best if you use
dynamic links.
8. If you have access to an ERP system, take a specifc role (salesperson,
production manager, customer, supplier, etc.) and use the system to enter data
and check on the current status variables for your role.
9. Choose an industry (for example, automobiles, bicycles, or spare parts).
Research the industry and identify what aspects follow mass production
versus what operations follow mass customization strategies. Identify
whether the industry sells predominantly to consumers or other businesses.
10. Find two cloud-based accounting systems that could be used by a business
with revenues of about $1 million a year and 50 employees. If possible,
estimate the costs of the systems. Pick one and explain why it is better than
the other.
Technology Toolbox
11. Find information on at least two accounting or ERP packages that could be
used for a business with 100 to 150 employees. Identify the strengths and
weaknesses of each package. Are the packages tailored to specifc industries?
12. You need to select an accounting or ERP package for a small service
business, such as a law frm or accounting frm. Identify the major features
that you want to see in the accounting system.
13. You have been hired as a consultant to a small graphics design frm with 20
employees. Design a digital dashboard for the owner-manager to monitor the
important variables of the company.
14. Compare the purchasing capabilities offered by two ERP vendors. Visit
their Web sites and view the demos. If possible, include copies of the
main purchasing forms from each vendor. Summarize the similarities and
differences.
15. Find an organization that uses a digital dashboard-executive information
system. Identify the major components on the main screen, and briefy
explain how the system is used to drill down to additional information.
Teamwork
16. As a team, fnd at least 10 Web sites that deal with international customers
and point out the features that they provide that might not exist or are
different on typical American Web sites.
17. Assume you want to create a customer relationship management system for
a company that sells supplies and products to hair salons. Each team member
should fnd one CRM product then identify the strengths and weaknesses
for this task. Combine the results and select one tool to recommend to
management.
18. Each team member should research one ERP system. Identify the
internationalization features available. Assume you are running a midsize
company with offces and operations in several European nations. Describe
the benefts provided by these international features. Make a recommendation
on which product to buy.
19. As a team, research the status of XBRL in fnancial reporting. Identify which
organizations require its use, which tools can generate XBRL reports, which
tools exist to retrieve and analyze XBRL data, and so on.
20. You have been hired as a consultant to a midsize agricultural company that
packs fruit and ships it to retail outlets. The fruit is grown under contract at
regional farms. About 65 percent is sold to grocery and chains and fruit stores
in the United States, while the rest is shipped overseas to high-end stores and
restaurants. Design an ERP system with a digital dashboard for this company.
Assign individual components to each team member. Be sure to identify
potential problems and risks that might arise.
21. Identify the major transaction-processing components in the system.
22. What features are used in the database and the forms to ensure quality of
transaction data?
23. For each major transaction type, identify the sequence of steps that are
performed and determine which ones are time critical.
24. For each major transaction type, estimate the frequency of the transaction and
the volume of data involved.
25. Using the existing forms, perform the following tasks for Rolling Thunder:
a. Take a bicycle order.
b. Assemble a bicycle.
c. Create a new purchase order.
d. Record the receipt of purchases.
26. Extract sales and cost data by model type and create a spreadsheet to
analyze it. (Hint: Use the Extract Data form.) Write a short report discussing
proftability and any trends. Include graphs to illustrate your comments. Your
spreadsheet should look at monthly sales by model and monthly material
costs by model. Be sure to compute proft margins and examine percentages.
27. Assume that Rolling Thunder is experiencing problems with quality control.
Suddenly there are several complaints about the components. Write a report
describing all of the data and reports we would need to help us resolve these
problems.
28. Top management needs an analysis of purchases and deliveries from vendors.
Begin by using queries to extract the appropriate data to create a basic
spreadsheet. Write a report analyzing the data; include graphs to illustrate
your points.
Vendor
Purchases
Order total $
Percent of
vendor total Received $
Receipts %
of purchase
Avg. # days
to deliver

29. Describe how an ERP solution could improve operations at Rolling Thunder.
What operations would you implement frst?
30. Describe how Rolling Thunder Bicycles might use a supply chain
management system. In particular, how would it integrate with its suppliers?
31. Describe or create a subsystem that could be used by the head of sales
management to monitor and improve customer relationships and sales.
Additional Reading
Anthes, Gary H. When fve 9s arent enough, Computerworld, October 8,
2001. [Challenges of transaction processing at Visa.]
Bleakley, Fred. Electronic payments now supplant checks at more large frms,
The Wall Street Journal, April 13, 1994, p. A1, A9. [Costs of handling
checks.]
Disabatino, Jennifer. The technology behind the problem, Computerworld,
October 22, 2001. [Transaction processing diffculties in providing airline
passenger data to the FBI.]
Harreld, Heather, Extended ERP reborn in b-to-b, Infoworld, August 27, 2001.
[Updated perspective on ERP and new applications.]
Krill, Paul, Data analysis redraws CRM landscape, Infoworld, November 30,
2001. [CRM vendors and capabilities in sales analysis.]
Loshin, Pete. Transaction processing, Computerworld, October 1, 2001. [Basic
concepts of transactions.]
Songini, Marc L., PeopleSoft project ends in court, Computerworld, September
10, 2001. [Serious problems with an ERP installation at Connecticut
General.]
Trombley, M. Visa issues 10 commandments for online merchants,
Computerworld, August 11, 2000. [Visas attempt to get vendors to protect
credit-card databases, and Gartner Groups estimate of online fraud.]
Want, Roy, RFID: A Key To Automating Everything, Scientifc American,
January 2004. [The strengths and limitations of RFID with a good description
of how it works.]
Weston, Randy, ERP users fnd competitive advantages, Computerworld,
January 19, 1998, p. 9. [Summary of ERP benefts.]
Winslow, Ron. Four Hospital Suppliers Will Launch Common Electronic
Ordering System, The Wall Street Journal, April 12, 1994, p. B8. [EDI for
hospitals, including costs.]
Cases: Automobile Industry
Te Industry
Automobiles remain one of the most expensive, complex products sold to con-
sumers. Manufacturing requires thousands of steps with parts acquired from hun-
dreds of suppliers. And people buy cars for more than transportationthey are
purchased based on image and style. As Ralph Szygenda, legendary CIO of GM
puts it: GM is ultimately in the fashion business, and to win, the company needs
the best cars and trucks, which we have, with more to come (Szygenda 2003).
Perhaps he was a bit optimistic. North American sales for 2003 dropped to 16.6
million vehicles. GM had the largest market share at 28.3 percent, followed by
0
50
100
150
200
250
300
350
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
B
i
l
l
i
o
n

$
Annual Revenue
General Motors
Ford
DaimlerChrysler
Toyota
-0.30
-0.25
-0.20
-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
R
a
t
i
o
General Motors
Ford
DaimlerChrysler
Toyota
Ford (20.7), DaimlerChrysler (12.8), Toyota (11.1), and Honda (8.1) (Warner
2004). The astonishing number is found in the 1960s when GMs market share
was over 50 percent. However, the declining share is not all that it seems. Check
the Web sites of the big three companies and you will quickly see that they have
bought up several other brands over the past years. Dig a little deeper into GMs
annual report and you will fnd that the company at one time owned shares in
Subaru, Suzuki, Fiat, and Daewoo, among others. Ford has similar arrangements
with other companies. By 2010, the total U.S. market share for the big three man-
ufacturers (GM, Ford, Chrysler) was only 45 percent.
Company 2010 YTD/May Company 2010 YTD/May
GM 19.1 Hyundai 4.4
Ford 16.9 Mazda 2.1
Chrysler 9.4 Subaru 2.3
Toyota 15.2 Mercedes-Benz 1.9
Honda 10.5 BMW 1.8
Nissan 8.1
Market Share: http://online.wsj.com/mdc/public/page/2_3022-autosales.
html#autosalesE
Most of the frms in the industry also faced a huge problem they referred to as
legacy costs. Over the years, the heavily unionized American frms promised huge
pensions and health care benefts to their employees. Now, they face billions of
dollars in fxed cost payments to retireescosts that were never fully funded from
prior operations. So, a substantial percentage of current sales revenue went to pay-
ing these costs.
Competition
In general, the industry is highly competitive. It has been estimated that the De-
troit big three have an excess production capacity of 2.5 million vehicles (Plun-
kett Research). Total U.S. sales fell slightly in 2003 to 16.6 million new vehicles.
Compounding the problem is the fact that many of the sales have been driven by
deep discountseither in price or in fnancing deals. At one point, in 1993, Ford
pushed so hard to make the Taurus the number one selling car, that it was giving
discounts of $2,600a considerable sum at the time. Most observers believe Ford
was losing money on every car (White 2004).
At the same time that the American car companies appear to be suffering, it
should be noted that only 40 percent of Americans can actually afford to buy new
cars (National Independent Dealers Association). In 2001, the Bureau of Trans-
portation Statistics reported sales of 8.6 million new passenger cars (not trucks)
and 40.7 million used passenger cars (and 11 million new bicycles) (www.bts.
gov). Used truck numbers are not reported, but in terms of new vehicles, cars
make up about 50 percent of the total number of vehicles. In any case, sales of
new vehicles account for about 17 percent of automobile sales. Globally, about 60
million vehicles were sold in 2003 (GM annual report).
Sales of American vehicles have continued their decline. In 2007 Ford expe-
rienced an 8.1 percent drop in sales, Chrysler Group sales fell 1.4 percent, and
GM sales dropped by 21 percent. In contrast, most import manufacturers reported
double-digit increases in sales (Lam 2007). Ford and GM both reported that part
of the decline was due to a reduction in sales to car rental feets. The manufactur-
ers are trying to move away from these low-margin sales. Sales were also down
for most other months in early 2007. The domestic manufacturers announced
sales promotions for July. Also in 2007, Daimler, the German parent company an-
nounced the sale of Chrysler to a private equity frm. The U.S. auto industry also
faced huge pension-fund and health care costs that were promised to workers and
retirees decades ago.
Technology and Management
The ultimate question the automobile industry faces is: How do you manage such
a large, complex company? In 2003, GM had 326,000 employees worldwide, and
revenues of $185 billion (annual report). The company ran 32 plants in North
America alone (Garsten 2003). By 2010, the number of employees was down to
202,000 (2010 annual report). With multiple divisions, thousands of designers and
engineers, and a recalcitrant union, how can anyone possibly determine what is
going on, much less control it? As one of the largest companies in America for
over 60 years, GM is an interesting case in management and mismanagement. The
company was initially created by combining several independent brands. Alfred P.
Sloan, who organized and ran the company in the 1930s, knew that with the infor-
mation system available at the time, he could never control an organization that
big. His decision was to let the divisions run somewhat independently, but report
to a set of policy committees and central staffs in a system he called decentral-
ized operations and responsibilities with coordinated control (Taylor 2004). All
auto manufacturers have struggled in the last couple of decades with the question
of how to control the huge organizations. As technology has improved, new op-
tions have been created that provide for increased central control across the entire
organization. Tools such as global networks, CAD-CAM systems for engineers,
workgroup and collaboration systems for managers, and enterprise resource plan-
ning systems for operations and fnance make it possible to share data. Yet install-
ing these systems is only the frst step. It is also necessary to change the way the
company runs. And that can be diffcult in an organization used to fexibility and
defned ways of functioning that evolved over decades.
The automobile manufacturers have tested many variations of information
technology over the past decade. The industry spent billions of dollars on telemat-
ics (integrating information systems into the vehicle). It has spent billions more on
supply chain management and enterprise resource planning systems. Today, most
of the $3 billion a year that GM and Ford spend on IT goes to operationskeep-
ing the existing systems running (Piszczalski 2002). The companies have still not
achieved total integration of their systems, and might never reach that state. But
they are still interested in tools that can solve specifc problems and provide gains
in quality, cost reduction, and bringing new models to market faster.
In 2007, Ford announced a partnership with Microsoft to install an integrated
in-car communication and entertainment system. The system has USB and Blue-
tooth ports for establishing connections to most consumer electronic devices. It
also supports voice-activated callingincreasingly important as more states ban
hands-on cell phone use in cars (Rosencrance 2007).
Te Crash of 2008
The recession of 2008 was the last straw for GM, and the company fled for bank-
ruptcy on June 1, 2009. In the process, the U.S. and Canadian governments bailed
out GM (and Chrysler) with billion-dollar infusions of cash. In exchange, the two
governments gained partial ownership rights to the two companies. Ford had for-
tuitously refnanced several loans right before the crash and was able to avoid
the fnancial crunch and did not fle bankruptcy or require government subsidies.
In the cleanup process, GM shed most of its legacy costs, by creating a new
company that holds mostly newer assets and dropped the pension fund obliga-
tions. In exchange, the unions gained control over most of the new GM shares. In
November and December 2010, the new GM issued an IPO and became the new
GM (2010 Annual Report). Chrysler was sold to Fiat, the Italian car company.
Ford remains as an independent company. Both GM and Ford sold off or closed
several car divisions. For example, GM closed Hummer, Pontiac, and Saturn, and
discarded Saab; leaving the company with Chevrolet, Buick, Cadillac, and GMC
in the United States. Ford sold Jaguar and Land Rover to Tata Motors and closed
the Mercury division.
The other big changes in the industry stem from the rising price of gasoline
in 2010hitting over $4 per gallon in most states in 2010 and 2011. This price
level encouraged people to search for cars with better gas mileageparticularly
hybrid gas/electric cars. Although the American companies had some hybrid cars
available, the Toyota Prius was the industry leader by a long shot. In 2011, GM
began selling a hybrid car that could run 40 miles on electric power, and Nissan
introduced the Leafan all-electric car. Huge uncertainty remained over whether
customers would buy all-electric cars and whether it would be possible for the
cars to have a long enough range to be useful as a primary vehicle.
The third overall trend is that total sales outside the U.S. exceed the value of
sales within the U.S. By 2011, China had become the largest automobile market in
terms of the number of vehicles sold. For example, GM sold 8.4 million vehicles
worldwide in 2010 and 2.2 million in the U.S. The total industry sales in the U.S.
were 11.8 million, and total worldwide sales were 73.6 million (GM 2010 Annual
Report). Worldwide, Toyota was the largest automobile manufacturer in 2010.
Additional Reading
Boudette, Neal E. GM Revs Up Fuel-Cell Development, The Wall Street
Garsten, Ed, UAW May Sacrifce Jobs For Benefts, The Detroit News,
September 8, 2003.
http://www.niada.com/Industry_Information/ind_10key.htm
Lam, Judy, Ford Sales Decline 8.1%; Chrysler Posts Slip in Sales, The Wall
Piszczalski, Martin, Auto IT Moves Back to Basics, Automotive Design &
Production, October 2002.
Plunkett Research, U.S. Automakers Suffer Declining Market Share/Their
Response is to Restructure and Seek Greater Effciency including Cutting
Costs, http://www.plunkettresearch.com/automobile/automobile_trends.
htm#1, 2004.
Rosencrance, Linda, Ford, Microsoft in Sync with In-car Digital System,
Szygenda, Ralph, Its a Great Time to Be In I.T., Information Week, January 27,
2003.
Taylor III, Alex, GM Gets Its Act Together. Finally, Fortune, March 21, 2004.
Warner, Fara, Show Floor News: Detroit to Focus on Tastes as Its Market Share
Decreases, The New York Times, January 5, 2004.
White, Joseph B and Norihiko Shirouzo, At Ford Motor, High Volume Takes
Backseat to Profts, The Wall Street Journal, May 7, 2004
Case: Ford
Ford (ticker: F) has presented several interesting cases in management over the
years. Driven initially by Henry Fords effcient adoption of mass production, the
company dominated the automobile industry. Through all of the changes in the in-
dustry, technology, and fnance, the company has largely remained under the con-
trol of the Ford family (40 percent of the voting stock). In October 2001, the fam-
ily ousted Jacques Nasser and Henrys great-grandson, Bill Ford became CEO.
The frm struggled for a year, trying to recover from its earlier problems. One of
the major changes implemented by Mr. Ford was an emphasis on proftsinstead
of raw sales numbers. Toyota is close to overtaking Ford and becoming the num-
ber two car company in the United States. Mr. Fords comment: I dont want to
be the biggest. I want to be the best. With lower sales, Ford surpassed GM in
profts for 2003 (White and Shiorouzo 2004).
Revenue Decisions
The automobile industry is a classic case of mass production. Building thousands
of identical cars and trucks reduces costs. But then you have to sell them. Does
everyone want the exact same car or truck? In 2001 and 2002, Ford began to re-
evaluate its strategy, with a focus on identifying the source of profts. Why build
300,000 Tauruses if the company loses money on each car? In the process, the
company found that it needed to track exactly which vehicles are selling, what
features are in demand, and the details of regional preferences. Bill Ford promot-
ed Lloyd Hansen to a corporate vice presidency in charge of overseeing revenue
managementidentifying exactly which products should be built. The company
began analyzing data from existing sales channels. More important, it combined
consumer Internet research data to identify regional choices. Mr. Hansen says
that the company was surprised to fnd consumers in sunny states were interest-
ed in four-wheel drive pickups. There was always this gut-feel paradigm that
people who live in Texas, Florida, and California dont need a 4-by-4 (White
and Shiorouzo 2004). By sharing the data with dealers, Ford increased shipments
of the more expensive trucks. Ford uses the same information analysis to target
discounts more precisely. With truck sales, Ford was able to increase revenue by
$934 per vehicle from 2003 to 2004 (Taylor June 2004).
Think about the decisions required for revenue management. The key is that
you have to decide exactly which products to produce tomorrow. You have dozens
of cars and trucks, each with several trim variations. You have dozens of plants.
You have some data on sales and profts. Should you produce 1,000 Ford Explor-
ers that might generate $4,000 per vehicle in profts, or should you run 1,000 Fo-
cuses with lower proft margins. If you do run the Explorers, how much discounts
should you give to get them sold? Keep in mind that this is a $25 million decision
that you have to make every day (White 2002).
Now, make the process even more complicated: what are your rivals going to
do? If Cadillac offers a rebate on the Escalade, how will that affect sales of the
Navigator? Should you match the discount, or accept the possibility of lower sales
at higher proft margins? What is the probability that Cadillac will offer that re-
bate? More importantly, if you respond, will they retaliate and offer even larger
discounts? Few models are actually this sophisticated today, but in reality, you
need to forecast a huge number of variables to make these decisions.
The system of mass production makes it diffcult to respond to changes in de-
mand. Forecasts are helpful, but Ford has worked hard to add fexibility into its
production system. In particular, it is working to build different versions of cars
from the same plant. At one time, each plant was designed to produce only one
type of car. Today, plants need to be able to switch production so that when de-
mand for crossover wagons increases, the company can stop producing sedans and
build the high-demand cars instead. Toyota and Honda pioneered these practices,
and Ford is still trying to catch up, but the company is making progress. Many of
these changes are being implemented in the new Chinese plants. The goal is to put
in only enough capacity to meet current demand, but enable the plants to expand
quickly if demand takes off (White and Shirouzo 2004).
Of course, it always pays to improve effciency in production. Fords plant in
Cuautitlan, Mexico, produces 300,000 to 400,000 cars and trucks a year. It uses
a just-in-time production system, where suppliers deliver parts as they are need-
ed on the lineinstead of holding them in inventory. Until 2002, the partially
completed vehicles were tracked through manufacturing and painting using paper
identifcation sheets. But this manual system generated huge numbers of errors
from lost, damaged, or switched sheets. Ford called on Escort Memory Systems
from California and CAPTA from Mexico to convert the system using radio fre-
quency identifcation (RFID) tags. Dealing with the electronically dirty manufac-
turing foor and the 220 C paint oven temperatures was a problem. The com-
panies used high-temperature RFID tags with 48-byte memory and 1,200 bytes/
second transfer rate. The tags were attached to the pallets that carry the vehicle
bodies through the plant. Antennae were embedded in explosion-proof contain-
ers in the foor under the conveyor belts. As each skid passes over an antenna, the
system reads it and feeds the information to the central control system (Johnson
2002). In 2004, ford installed WhereNet Corp.s real-time parts-locating system in
its F-150 truck plant. The system changes production from a push-based inventory
supply to one that pulls products onto the line when an operator pushes a button.
The system knows which parts are needed and delivers them to the job site just as
they are needed (Rosencrance 2004).
In a fashion market, producing new models quickly is critical. Ford has reduced
its development time by 25 percent from the late 1990s and is getting better by 10
percent a year. Still slower than Toyota, but Ford was able to produce a third of its
new models in 30 months or less (Taylor February 2004).
Turnaround
Ford weathered the recession and liquidity crisis that took down GM and Chrys-
ler, but it faced many of the same pressures in terms of lost sales and restructuring.
CIO Nick Smither arrived at Ford before the 2008 crisis, but Ford had already
slashed costs and dropped models. Smither noted that I have always said a career
in IT is a career in change management. But no one could have anticipated this
level of change (Johnson 2010).
One of the changes at Ford was to simplify production and orderingby re-
ducing the number of confgurations. For example, up to 2008, a customer could
order a full-size SUV with one of 76,000 confgurations. By 2010, the list was cut
to 1,500. Not exactly simple, but easier to produce. Ford also reduced the number
of vehicle platforms, from 27 down to 15; and it wants to use those for all global
production. IT costs have also been slashed, and Smither focused on removing
excess capacity and duplication. It dropped from six data centers down to four and
cut the application portfolio by 40 percent. The company added new collaboration
tools and built standardized systems to support integrated design, procurement,
sales, marketing, and manufacturing processes. It also developed new in-vehicle
communication and entertainment systems. All while cutting IT staffng by 20
percent (Overby 2010). The CIO now reports to the CEO and meets weekly with
corporate executives to restructure the entire company into a single system. For
example, a new WebQuote system is used by all engineers in Ford to defne prod-
ucts and solicit quotesfrom any division around the world. Everyone sees the
same data and same suppliers.
In cooperation with Microsoft, IT also developed the in-vehicle Sync system
that integrates and controls everything from the radio, to seats, to the navigation
system. It is largely voice-controlled and is highly praised by most customers.
Questions
1. How does Fords information system help them reduce costs and add fexibility to
the production line?
2. What data is provided by the ERP system that can be used to make tactical and
strategic decisions?
3. What type of data does Ford need to know to build its forecasts and plan produc-
tion levels? How does the ERP system provide this data?
Additional Reading
Johnson, Dick, RFID Tags Improve Tracking, Quality on Ford Line in Mexico,
Control Engineering, November 2002.
Johnson, Maryfran, IT Fuels Fords Dramatic Turnaround, Computerworld,
August 31, 2010.
Overby, Stephanie, How a Global IT Revamp is Fueling Fords Turnaround,
CIO, August 30, 2010.
Rosencrance, Linda, Ford Deploys WhereNet Wireless System in New Truck
Plant, Computerworld, June 30, 2004.
Taylor III, Alex, Detroit Buffs Up, Fortune, February 9, 2004.
Taylor III, Alex, Bills Brand-New Ford, Fortune, June 28, 2004.
White, Joseph B. Automobile Industry Is Using Science Of Revenue
Management to Sell Cars, The Wall Street Journal, May 5, 2002.
White, Joseph B. and Norihiko Shirouzo, At Ford Motor, High Volume Takes
Backseat to Profts, The Wall Street Journal, May 7, 2004
Case: General Motors
General Motors (ticker: GM) began life as a diverse companya collection of
manufacturers loosely controlled through a central organization. It really has not
changed much over the years, but it keeps trying. Even the bankruptcy fling and
new GM did not actually change the management structurealthough it did
bring in a couple of outsiders as CEOs.
Product Management and Marketing
The old GM listed nine major brands on its home page (without Buick). And
that list does not include several European and Asian brands that are fully or
partly owned by GM, such as Opel, Vauxhall, Daewoo, Fiat, Isuzu, and Suzuki
(Plunkett Research 2004). Imagine trying to monitor the dozens of companies,
product lines, and markets. One step GM is taking is an attempt to build a central
European database encompassing clean data from Spain, the Netherlands, and the
UK. Most of the work is being done for its Daewoo division. The goal is to create
a consistent list of its 700,000 customers and prospects and make it available from
its Amsterdam headquarters. Daewoos UK direct marketing and e-commerce
manager, Matthew Young, notes: In Europe, it was apparent that we had no con-
tinuity or integration of style. This will also plug the hole in our cost effciency
(GM Daewoo 2004). The new GM was down to four domestic brands, plus
ownership in several international companies (2010 Annual Report).
Production and Collaboration
Production of automobiles is changingparticularly as the automotive technol-
ogy changes. For example, the gas pedal in the 2004 Corvette is not mechanically
connected to the engine. Instead, it is an electronic connectionmuch like the
fy-by-wire Boeing 777 jet. Many cars have 30 wireless sensors in them, with
plans for new models with over 50 devices in them. Tony Scott, chief technology
offcer at GM, notes that we spend more per vehicle on software than we do on
steel, and that software has to survive the 10-or-more-year life of the car (Sul-
livan 2003).
Implementing new technologies into automobiles is a complex task. Advances
have to be evaluated, tested, and priced to match the level of the automobile. To
handle the changing technologies and track which ones are being used in each ve-
hicle, GM adapted a technology road mapping process developed at Saab (at one
time a subsidiary of GM). The basic process places the model year on the X-axis
of a chart, and the Y-axis represents increasing performance. Individual projects
are drawn on the chart to show if they are funded or unfunded, and when they
started. Collecting the data for charts across the organization was not easy. Even-
tually, a database-driven Web site was created to hold the information about exist-
ing and planned projects, containing the description, budget, development stage,
key personnel, cost, and planned applications. Even then, most divisions were re-
luctant to enter the appropriate data. So, GM forced divisions to enter data by
explaining that if there was no data in the database, then the project must not be
important, and it would not be funded. The process was useful for some groups,
but avoided by others. The good news is that everyone could see how the various
projects scattered across the company were related. For example, electronic steer-
ing was going to rely on the 42-volt electrical system (older cars use a 12-volt
system), so the timing of the two projects had to match. The system also showed
several redundant projects, which caused conficts because everyone wanted the
other project to be canceled (Grossman 2004). A powerful extension of the road
mapping process was to extend it to suppliers. For example, by overlaying Mo-
torolas road map, engineers could plan for future capabilities.
Automobile manufacturers are really assemblers and integrators. Each plant
has thousands of suppliers who provide everything from individual bolts and glue
to fully assembled cockpits and seats. Each of these companies has additional sup-
pliers. How can anyone track the orders and solve problems if something goes
wrong several layers deep into the supply chain? Product life cycle management
(PLM) software presents one answer that is being used in the automotive world.
The system enables GM to track the source of all components used in a product
during its lifetime. It also provides information to the supply chain participants.
Terry S. Kline, GMs global product development process information offcer,
notes that in product development, IT projects like PLM have helped pare a bil-
lion dollars of cost over the past three years. But Kline also quickly notes that the
system has also helped reduce the time-to-market from idea to product from 48
months to 18 months (Teresko 2004). About 80 percent of a car and as much as 70
percent of the design content comes from outside the automobile company. PLM
integrates the business information with the engineering (computer integrated
manufacturing) data and shares it with other members of the supply chain.
Making the thousands of decisions and setting strategy is incredibly diffcult.
Rick Wagoner, the CEO, noted in 2004 that we didnt do everything right over
the last 12 years. About three times during that period you think youve got it,
and then something else comes up. We made huge progress in 92, 93, 94, but
then we started to pay the price because we had, out of necessity, underspent (on
new products) during that tight period and didnt do that well on the revenue side.
Then, in the last three or four years, weve run the business very well, but the pric-
ing has been tough. So my lessonthese arent complaints but an observation
is, Dont ever think youve got it licked, because you probably dont. This is not a
one-step game. This is a multiple[-year] thing, and its hard, and you learn as you
go along. The key thing is to remember what youve learned so you dont have to
relearn it (Taylor 2004).
In early 2007, GM signed a $1 billion fve-year contract with AT&T to provide
global IP services supporting voice, data, and video communications around the
world. The system provides a global virtual private network for GM based on
Multiprotocol Label Switching (MPLS) that supports routing packets by priority
values. Ralph Syzgenda, CIO at GM, stated that The real end goal is that every
employee has the same type of capabilities no matter where they are in the world,
and in fact, when they wake up, they dont need to know where they are in the
world. It just works. (Thibodeau 2007).
Technology in cars is also increasing. In 1990, a typical car ran software con-
sisting of one million lines of code. By 2010, the number is expected to increase to
100 million lines. Software and electronics in 2004 constituted one-third the cost
of the car. GM is concerned about managing the software development process
and the liability. Anthony Scott, chief technology offcer for GMs Information
Systems and Services organization said that People love to sue GM when they
get into an accident or have some sort of a problem. As more and more software
comes into the car GM is going to expect [software developers] to participate at
some level in that liability. (McMillan 2004).
Ralph Szygenda is already a legend at GM. As its frst CIO in 1996, he inher-
ited several problems. GM had tried to simplify its IT functions by outsourcing
computer management to EDSof course, the company wanted to keep the prof-
its, so it purchased EDS. But the IT departments were still not running smoothly.
Among other problems, it took the company 48 months to develop a new car. Szy-
genda went several steps further into outsourcingby moving to a multivendor
approach, forcing them to compete against each other, yet still cooperate to pro-
vide products that worked together. In the process, he slashed annual IT spending
from $4 billion to $3 billion, and installed 3-D virtual reality modeling software
along with a collaborative infrastructure. Model development time dropped to 24
months. The time it takes to deliver a special-ordered vehicle dropped from 70 to
30 days. To control the outsourced components, Szygenda kept 2,000 IT manag-
ers in house. He observes that when I frst started doing this, people said you
couldnt outsource critical aspects of the company, and I agree with that. Those
2,000 [managers] are the critical part, but the hands, arms, and legs of building IT
can be outsourced (Zarley 2003). Szygenda uses the 800-pound gorilla clout of
GM to keep suppliers in line and encourage them to adopt standard procedures.
A frustrating problem for customers arises when they take a vehicle to the deal-
er for a repair, but the dealer does not have the repair parts in stock. Dealers do
not like to carry large inventories because the carrying costs are too high. GM
maintains an inventory of 1.3 million parts and dealers typically stock from 5,000-
12,000 parts. In the past, GM relied on the judgment of individual dealer parts
managers, who waited until the end of the week to submit a batch parts order.
The practice also caused problems for GM distribution centers as workers rushed
to fll orders from hundreds of dealers arriving at the same time. To reduce the
problems, GM installed a Retail Inventory Management (RIM) system that tracks
daily parts usage at each dealer. The data is sent to the regional distribution center
and replacement parts are shipped to the dealer. A pilot test involving 350 dealers
in Florida resulted in a smoother work load with an order variance of 2 percent
compared to 21 percent before installing the system. The system also uses analyt-
ics to recommend which parts to stock at each dealership. GMs Saturn division
had used a similar system for years, but it was a challenge to build the new system
to integrate thousands of dealers using diverse systems (Thibodeau 2006)
Technology has not always been successful at GM. In the 1990s, it installed a
computerized system for dealers that required them to order cars 90 days in ad-
vance, and it made it diffcult to change orders. In the meantime, GM announced
that it wanted to take over 700 dealerships and run them centrally. Then, the order-
ing system crashed and sales plummeted for a month. The system was eventually
fxed, but the two decisions alienated dealers. Jack Smith, ex-CEO of GM, notes
that these and other problems were exacerbated by market manager Ron Zarrella,
who was brought in as an outsider from Bausch & Lomb. Smith notes that the car
business is unique, and an outsider might say, Thats really a dumb way to run.
But the unfortunate thing is, that is the way it runs. There are some things you just
have to live with (Taylor 2004).
GM, and other companies have begun experimenting with RFID tags to track
parts through the supply chain and factory. For example, GM uses active RFID
tags to store job information to indicate which parts need to be installed in each
vehicle. But the company is still focused on using bar codes to track parts. Bar
codes are substantially cheaper and are not subject to electrical interference from
metals and chemicals on the production foor. Engineers do believe passive tags
could be useful within the car, making it easy for repair people to quickly scan for
the exact part installed deep in a vehicle (Weier 2006).
As one of the largest organizations in the world with plants and offces in 200
countries, 280,000 employees, thousands of dealers, and millions of customers,
GM has an identity problem. GM has tried federating identity management sys-
tems from several companies, including tools from Novell, Siemens, Oracle, and
Sun Microsystems. But James Heaton, global director of identity management
said that a lot of vendor solutions halfway work for GM, but when we try to scale
them, they break. We break almost everything we touch the frst time we try to use
it. Even creating a system that handles password resets is challengingparticu-
larly with multiple languages. If a user forgets a password, the system asks per-
sonal questions to identify the user and reset the password. When the system was
installed, GM saw problems spike in Portugal and Brazil. One caller complained
I did everything the system told me to until I came to the validation questions
about my goat. I dont have a goat. Obviously there was a translation error in the
system, but it takes time and resources to spot the problems and fx them. The pro-
cess is complicated because GM has 2,481 applications that require authentication
(Babcock 2007).
In 2009, GM tried an experiment by auctioning a new car on eBay. The site
allowed California buyers to place online bids for cars from 225 of the 250 deal-
ers in California. A J.D. Power & Associates study had noted that 70 percent of
buyers use the Internet for research before buying a car. The initial auction was
successful, but most dealers already use the Internet and even eBay for sales
particularly for used cars (Barris 2009).
Near the end of 2009, Terry Kline became the CIO of General Motors when
Ralph Szygenda retired. One of Klines frst goals was to upgrade desktops to
Windows 7 (from XP), to make the systems easier and cheaper to manage. He
was also exploring options to use cloud computing to reduce management costs
of running new servers. But, he expressed concerns about the pricing models
(Thibodeau 2009).
IT played some interesting roles in the development of electric vehiclessuch
as the Chevy Volt for GM. Because EVs are such a big shift for consumers and
manufacturers, new design issues arise that do not have years of data or prior
solutions. GMs approach was to build multiple sensors into all of the early test
vehicles. As employees drove the vehicles, the data was collected and reported to
the designers. IT built new tools to retrieve and analyze the data. The designers
used the data to evaluate tradeoffs and make decisions on the batteries, motors,
and user interfaces (Brandon 2011). IT was also heavily involved in designing
tools for monitoring battery and charging levels. Most EVs have the ability to re-
port various status items to users through Web sites and smartphone applications.
Questions
1. How does GMs decentralized structure affect the implementation and use of
ERP systems?
2. How does an integrated production and design system help GM design new
cars?
3. Why is an integrated production system so important for building custom
vehicles?
4. Will GM eventually be able to build a higher percentage of custom-ordered
cars?
Additional Reading
Babcock, Charles GM, Boeing Faced Uphill Battle To Reach Global Identity
Management, Information Week, June 28, 2007.
Barris, Mike, GM, eBay to Test Online Car Sales, The Wall Street Journal,
August 11, 2009.
Brandon, John, Car Tech: Electric Vehicles Get an IT Assist, Computerworld,
April 12, 2011.
Grossman, David S., Putting Technology on the Road, Research Technology
Management, March/April 2004.
McMillan, Robert, GM CTO Sees More Code on Future Cars, Computerworld,
October 20, 2004.
Plunkett Research, U.S. Automakers Suffer Declining Market Share/Their
Response is to Restructure and Seek Greater Effciency including Cutting
Costs, http://www.plunkettresearch.com/automobile/automobile_trends.
htm#1, 2004.
Precision Marketing, GM Daewoo UK Leads Drive to First pan-Euro
Database, March 26, 2004.
Sullivan, Brian, GMs Tony Scott On High-Tech Driving, Wireless Cars And
Software Crashes, Computerworld, May 8, 2003.
Taylor III, Alex, GM Gets Its Act Together. Finally, Fortune, March 21, 2004.
Teresko, John, The PLM Revolution, Industry Week, February 2004.
Thibodeau, Patrick, GM Turns to IT to Fix Parts Supply Chain,
Thibodeau, Patrick, GM Aims for Global IP System with New AT&T Deal,
Thibodeau, Patrick, Hard Times Arent Slowing GMs New CIO,
Computerworld, November 16, 2009.
Weier, Mary Hayes, Nissan Plant Embraces RFID, Information Week,
November 27, 2006.
Zarley, Craig, Ralph Szygenda, CRN, December 15, 2003.
Case: Chrysler-Fiat
In 1999, when former Chrysler CEO Bob Eaton introduced the redesigned Jeep
Grand Cherokee, he held up a bag that he claimed contained all of the parts that
were carried over from the earlier design. He seemed proud of this huge design
change. But the costs had to be enormous. Toyota saves billions of dollars by reus-
ing as many components as possible. Reuse also leads to more fexible production
and design, enabling Toyota to produce new vehicles in a few months instead of
years (Taylor 2004). In 2007, Daimler spun-off the Chrysler Group by selling it
to a private equity frm Cerberus. Daimler owned Chrysler for slightly under 10
years, and fnally had to abandon it to restore proftability. In March, 1998 Daim-
ler paid $36 billion for Chrysler and technically sold it for $7.4 billion, but the
Wall Street Journal calculated that it would actually cost Daimler $650 million
in cash to dump Chrysler. In the 2008 recession and liquidity crisis, Chrysler sur-
vived only through funding by the federal government. Eventually, the bulk of the
company was sold to Fiat. The goal was to use Chrysler automotive technology
(largely engines) to build larger cars in Europe and use Fiats knowledge and pro-
duction to build smaller cars for sale in the United States.
Technology in Manufacturing
In 2004, DaimlerChrysler had 104 manufacturing plants in 37 countries. Commu-
nication among the 362,000 employees was challenging across those distances.
The company used videoconferencing and held more than 100 meetings a day
with the technology, saving it as much as $14 million a year in travel expenses.
The engineering teams used Web portals to share work and check on business in-
formation (Niccolai 2004).
In 1998, after the merger of Daimler and Chrylser, Susan Unger was named
CIO and head of global IT operations of the new company. In the middle of the
dot-com hype, she noted that she had a different message, which was that in the
e-world, you have to have business value. It was kind of funny at the time be-
cause I was a voice in the wilderness talking that way. Because of this focus, the
company worked on B2B initiatives. Unger notes that many other companies have
had to write off their Internet projects, but we have had none because we were
kind of religious about ROI (return on investment (Gibson 2004).
One of the big tools that DaimlerChrysler implemented was DELMIA from
Dassault Systemes AG. Unger points out that as the engineer designs a vehicle,
they can use the tool to create a virtual manufacturing environmentincluding
your work cells, your line, your equipmentand simulate your workers installing
a wiring harness or whatever else. The multiple divisions (Daimler, Chrysler, and
Mitsubishi) all use an e-engineering portal to share CAD information.
DaimlerChrysler was also a big user of the Covisint automotive auction sys-
tem. Sue Unger noted that in 2003, the company used the system for almost 200
separate bids, saving it between 15 and 20 percent compared to traditional prices.
More important, the system simplifed the purchase process, reducing the time
from months to days (Gibson 2004).
Auto Sales and the Internet
In the middle of the dot-com boom, some people were forecasting that all car
sales would be made online within 6 to 10 years. Today, most sales are still made
in personpartly because people want to see, touch, and drive the cars. Yet Sue
Unger reports that 90 percent of buyers use the Internet for research before shop-
ping (Gibson 2004).
Although DaimlerChrysler used incentives to try and gain market share for its
primary models, Mitsubishi Motors was being hammered in the market. Their
sales approach in 2002 did not helpwhen the company targeted frst-time young
buyers with a zero-zero-zero campaign. Essentially, even poorly qualifed buy-
ers could get new cars with nothing down and no payments for one year. Unfor-
tunately, most of the cars had to be repossessed at the end of the year. In 2004,
Mitsubishi tried to recapture market share through marketing. This time with a
new twista return to Internet advertising. In 2001, the company abandoned Web
ads as nonproductive. In 2004, the company is spending $6 million on the Web
alone. DaimlerChrysler increased its Web spending by 30 percent in 2003 (Man-
galindan 2004). Although the Mitsubishi campaign represents only 5 percent of its
total advertising, the company believes the Web ads are more effective than other
approaches. The key: targeted ads and quick ties to dealers. Tom Buczkowski of
Silver Creek, New York, is one who responded to the new approach. Research-
ing SUVs at the Autobytel site, he was presented with an ad from Mitsubishi. By
signing up for a test drive, he won a $15 gas card. At the Mitsubishi site, he built
a customized Outlander. The details and the personal data he entered were sent to
the nearest dealer in minutes. The next day, the dealer called him and suggested
a test drive. Mr. Buczkowski was impressed with the response If you get a re-
sponse from somebody a day after the request, thats something you dont fnd too
often anywhere else. He eventually bought a Mitsubishi.
Given the size of auto manufacturers and the number of dealers, providing com-
munications with the dealers was a critical aspect to selling cars. DaimlerChrysler
included 5,300 U.S. dealerships and another 1,500 worldwide. The manufactur-
er provided 240 applications to the dealers ranging from support and training to
sales. Prior to 2003, most applications were standalone programs running with
little or no control on dealer servers. Richard E. Earle, CIO at the Chrysler Group,
led a group to replace all of the applications with Web applications running on a
central server and accessible through Web browsers. With 300 to 500 internal de-
velopers working on the project, the DealerConnect system was built in a couple
of years. Serving 300,000 users in 10 languages, it servers 8 million pages a day
(Ulfelder 2005).
After the buyout by Fiat, Chrysler provided little fnancial and production in-
formation. The company focused on selling existing models until new, smaller
cars could be designed by the Fiat team. In 2008, Chrysler IT did move more of its
IT operations offshore to be handled by the Tata Consultancy Services (Thibodeau
2008).
In 2008, Chrysler tested an offer of a Wi-Fi hot spot as an option in its ve-
hicles. The system would provide Wi-Fi connections within the car and use cel-
lular phone service to connect to the Internet. At a base price of almost $500 plus
$29 a month for service, it is not clear how many customers would purchase the
option. Within a year, many smartphones began providing similar servicesand
they would not be tied to the car.
Questions
1. How do integrated design systems solve typical problems that arise?
2. Check out the Web sites by the automobile manufacturers. Why are they so
bad? As a customer, what additional features would you want to see?
Additional Reading
Gibson, Stan, DaimlerChrysler CIO on the Road to IT Success, eWeek, May 3,
2004.
Mangalindan, Mylene, After Wave of Disappointments, The Web Lures Back
Advertisers, The Wall Street Journal, February 25, 2004
Niccolai, James, DaimlerChrysler Sees Savings in RFID Tags, Computerworld,
March 17, 2004.
Taylor III, Alex, Detroit Buffs Up, Fortune, February 9, 2004.
Thibodeau, Patrick, Chrysler Moves More IT Work to Offshore Giant Tata,
Ulfelder, Steve, IT Playbook: Great Works, Computerworld, January 3, 2005.
Case: Toyota
Since the gas crises of the 1970s, Toyota has been gaining U.S. and world market
share on American automobile manufacturers. Toyota had the lead in small-car
design. Over time, the company leveraged this knowledge of consumer preferenc-
es to improve quality. The company was also the frst to adopt lean manufacturing
techniques to reduce costs and formal integrated quality control to take the lead in
quality. Eventually, the company introduced the Lexus division to capture a large
share of the high-end market. By 2007, the company was approaching the num-
ber-two position in sales. With common Japanese kerietsu ties to suppliers, Toyota
pioneered just-in-time inventory deliveries. These inventory- and quality-control
measures require considerable communication with suppliers. Toyota has adopted
computer technology to facilitate communications. Toyota is trying to grow pro-
duction capacity but with an emphasis on fexibility. Using its ties to suppliers and
dealers, the company can increase or decrease production at any plant on short no-
tice. It also tries to standardize vehicle platforms and keep the line fexible enough
so that it can switch each line to a different product depending on the demand
(2006 Annual Report and Duvall 2006). As with other manufacturers, Toyota uses
advanced CAD systems to design new vehicles and parts. With this technology,
Toyota has been able to reduce design time for a new vehicle from four years
down to between one and two years. Toyota also centralizes all of its data. The
CIO and executives have access to all global data when making decisions (Miller
2006). Toyota briefy took over the number two spot in U.S. sales, but its momen-
tum was derailed when the tsunami hit Japan in 2011, causing disruptions in the
supply chains for many components.
Technology in Sales
From 2000 on, the world auto industry had an excess production capacity, leading
to intense competition; particularly in the United States. Online Web sites such as
Edmunds and Vehix also made it easy for customers to search for cars and pric-
es from multiple dealers. Proft margins at dealers fell for several years. To help
maintain overall profts, dealers had to get bigger to increase volume and spread
the overhead costs. Many also carry multiple brands of vehicles. Gulf States Toy-
ota is a distributor for 145 Toyota dealers in fve states. The company processes
more than 190,000 vehicles a year at an 84-acre facility in Houston, TX. The com-
pany faced a problem in 2004 where they either needed to add more buildings or
fnd a way to get vehicles through the facility more effciently. WhereNet Corp
provided an RFID-based vehicle tracking system. New vehicles delivered by rail
car are tagged with an active RFID sensor and the VIN is entered into a database.
The facility has 40 wireless LAN access points and 74 RFID readers. As vehicles
enter and leave the facility, the data is picked up and entered into the database
(Rosencrance 2005).
Toyota USA also needs to communicate with dealers and distributors. An early
communication system developed in the early 1980s was based on IBM AS/400
technology. The system handled parts ordering and warranty claims, but all other
communications needed to be handled by phone or fax. With 1,100 Toyota and
Lexus dealers, Toyota needed a better way to communicate and improve dealer
productivity. The Dealer Daily system was created to handle all common commu-
nications. It is a Web-based system running over a virtual private network (VPN)
that links the dealers to Toyota headquarters. Dealers can order parts, fle warranty
claims, and manage fnancing and insurance for customers. Dealers pay for access
to the system with a monthly fee based on the bandwidth they choose. Dealer Dai-
ly can also deliver critical news and information directly to any user at any dealer-
ship. Toyota is also working on a build-to-order system that will be integrated into
Dealer Daily (Rosencrance 2002).
Led by the Prius, Toyota is one of the leaders in hybrid vehicles. In 2011, the
company began expanding the role and variety of electric and hybrid vehicles
making Prius a separate division. Toyota also struck a deal with Salesforce.com
to create a private social network for owners of its hybrid and electric vehicles.
The Toyota Friend network will enable customers to share data with Toyota and
other owners. It will provide notices for maintenance schedules and other issues.
It will also allow customers to share driving and maintenance tips with each oth-
ersomething that is commonly done now through a variety of public Web sites
(Jennings 2011).
Toyota faced a couple of public image problems when it had to recall millions
of vehicles in 2010. For many years, Toyota was perceived as a quality leader,
and recalls and negative press affect that image. To take the next step at maintain-
ing high quality, Toyota developed a business intelligence (BI) tool to monitor
six years worth of product quality data pulled from internal systems. The sys-
tem consolidates and standardizes the data, enabling engineers to compare and
flter data on confgurations, customer call centers, warranty claims, and service
centers. Tom Trisdale, manager of Toyotas Technical Information & Diagnostics
Group noted that The great thing is, everything is available from one source. We
can have results very quickly (Vijayan 2011). By using off-the-shelf software
from Endeca, the IT department built the system within a few months. Similar to
other BI projects, 45 percent of the time was spent on cleaning and standardizing
the data to ensure consistency.
Questions
1. Why is information technology so important for lean just-in-time production?
2. Why does Toyota use online searches for vehicles in stock instead of building
cars to order?
3. Given the sophistication of Toyotas systems, why have defect rates and
design problems increased?
Additional Reading
Duvall, Mel, Whats Driving Toyota? Baseline, September 5, 2006.
Jennings, Ralph, Toyota Plans Social Network for Electric-Vehicle Buyers,
Miller, Jon, How Toyota Uses Information Technology (IT) for Kaizen, Panta
Rei, June 21, 2006. (http://www.gembapantarei.com/2006/06/how_toyota_
uses_information_technology_it_for_kaizen.html)
Rosencrance, Linda, Dealer Daily: Toyotas Communication Pipeline,
Rosencrance, Linda, Toyota Distributor Turns to WhereNet for Vehicle
Tracking, Computerworld, August 12, 2005.
Vijayan, Jaikumar, Business, IT Collaboration Drives New BI System at
Toyota, Computerworld, May 4, 2011.
adopted by rivals?
sectors?
428
How do you make money on the Internet?
What types of products are sold online?
How do Web-based services work and why do they change the world?
How can customers pay for products and why do you need new payment mechanisms?
How do frms get revenue from Web ads and how do customers fnd a site?
How do you create an EC Web site?
How do portable Internet connections (mobile phones) provide new ways to sell things?
When do consumers and businesses pay sales taxes on the Internet?
Does the Internet create a global marketplace?
What are the costs for cloud computing?
Chapter Outline
Electronic Business
7
Chapter
Introduction, 429
Selling Products Online, 432
Business to Consumer, 433
Business to Business, 438
The Production Chain and
Disintermediation, 440
Three Stages of a Purchase, 443
Price Competition, 444
Services and Web 2.0, 446
Social Networking and Consumer
Services, 446
Digital Products, 448
Business Services, 449
Network Services, 451
Search Engines, 454
Payment Mechanisms and Risk Mitigation
in e-Commerce, 454
Payment Risks, 455
Payment Mechanisms, 456
Advertising, 460
Traditional Media and Name
Recognition, 461
Web Advertisements, 461
Privacy, 466
Developing and Hosting Web Sites, 467
Simple Static HTML, 468
Web Auctions, 470
Amazon WebStore, 471
Auction versus WebStore, 472
Web Commerce Servers, 472
Content Management Systems, 473
Application Service Providers, 474
Web Hosting Summary, 474
Mobile Commerce, 476
Taxes, 477
Global Economy, 478
Summary, 480
Key Words, 481
Exercises, 482
Cases: Retail Sales, 485
429 Chapter 7: Electronic Business
Wal-Mart
Introduction
How do you make money on the Internet? In the late 1990s, the early
days of the Web when millions of people frst went online, thousands of com-
panies were created to sell items on the Internet. Billions of investment dollars
were spent by these frms developing new ways to conduct business on the Inter-
net. Some managers loudly proclaimed the dawn of the new economy and the
death of bricks-and-mortar companies. Most of these early frms focused on
selling products to consumerssimilar to existing mail order companies. In 2000
and 2001, hundreds of these frms crashed. Some managers then proclaimed the
Internet dead. Yet, from the beginnings in the mid-1990s, sales using the Inter-
net have increasedeven faster than retail sales in general. The initial irrational
exuberance (then-chairman of the Federal Reserve Alan Greenspans term) and
consequent crash, highlight the importance of carefully addressing the question:
exactly how can you use electronic business to make money? Wrong answers can
be costly on both sides: wasting money on hopeless schemes or losing money by
not using the Internet.
In the mid-2000s, attention and investment dollars returned to the Internet. This
time focusing on frms specializing in new services, dubbed Web 2.0. Firms like
YouTube, Flickr, Shutterfy, and social networking sites FaceBook and Twitter led
the way by attracting millions of members. Following the frst-round Web sites,
many of these frms received millions of dollars from investors and many provide
free access to consumersrelying on advertising to provide a revenue stream.
There is no hard-and-fast defnition of Web 2.0 frms, but the general distinction
is that they provide specifc services to consumers. Instead of providing products
or information the sites generally provide a framework for interaction, and users
provide content.
How do retail stores handle transactions? Wal-Mart has grown to become not just the
largest retailer but the largest company in the world. Each store is huge and sells hun-
dreds of thousands of items a day. Each sale has to be recorded with absolutely no
errors. Moreover, the store has to monitor sales and choose which products to order.
Then it must monitor the progress of incoming items, compare the received items to
each order, and resolve problems and errors. Wal-Mart has succeeded because it uses
information technology to handle these operations at the lowest possible cost.
A retailer that focuses on low prices can never rest. It must continually search
for new ways to reduce costs, improve store management, and choose the correct
products to sell at the best price. Being the biggest provides several advantagesno-
tably in terms of negotiations with suppliers. For these reasons, Wal-Mart is one of
the companies leading the transition to radio frequency identifcation (RFID) chips.
Initially, Wal-Mart is asking major suppliers to place the passive tags on pallets and
major crates. The primary advantage is that the receiving docks can quickly scan in-
coming shipments and match the products to the individual orders. RFID tags carry
more information and are easier to read than traditional bar codes.
What is electronic business? E-commerce, or EC, can be hard to defne. On
the one hand, it could be defned as selling items on the Internet. But Figure 7.1
shows there are many aspects to business and many ways of using Internet tech-
nologies. Some writers refer to the broader concept as e-business. The main point
is that EC represents considerably more functions than just putting up a Web site
to describe or sell products. The fundamental goal of e-business is to increase
sales and reduce costs by improving relationships with existing customers, and by
reaching new customers and providing new services.
For at least the past 50 years, companies have followed and refned the modern
business practices. These business-to-consumer (B2C) foundations were created
to facilitate purchasing and distribution in a world with limited delivery systems
and high communication costs. For example, retail stores for decades were small
Trends
Largely because of transportation costs, consumers have been limited to
purchasing products through local retail stores. Even for products produced
elsewhere, it was generally cheaper to ship items in bulk to the retailer than
to ship to individual customers. Plus, manufacturers and wholesalers did not
want to deal with individual customers. They did not want to spend the mon-
ey to create customer-service departments to handle thousands or millions of
individual orders, returns, and complaints.
Eventually, shipping and transaction costs began to decline. Sears got its
start and made its original reputation as a nationwide mail-order frm. Cus-
tomers could now order thousands of products from a catalog and have them
delivered. Over time, Sears found it proftable to build stores in thousands
of cities and moved away from catalog sales. In the 1980s and 1990s, many
other mail-order frms expanded to provide thousands of products direct
from the manufacturer to the customer. While some people prefer shopping
this way, only a fraction of total sales are made through mail-order compa-
nies. In 1998, mail-order sales were $356 billion, or slightly over 3 percent
of GDP.
Around 1997, sales over the Internet started to become important, and by
1999, e-commerce was the hot topic in the nation. By 2001, over 50 percent
of U.S. households had access to the Internet and e-commerce, promising
access to millions of customers. Hundreds of dot-com frms were funded
with venture capital and early IPOs to defne new ways to interact with cus-
tomers and businesses over the Internet. Hundreds of paper billionaires were
created as stock prices were driven by expectations and hype. Some people
were saying that bricks-and-mortar traditional frms were dead. But in late
2000 and early 2001, hundreds of the dot-com start-ups failed, laying off
thousands of workers and crashing the technology sector of the stock mar-
ket. Investors are wary, but the Internet continues.
Cell phones began with slow, poor quality analog signals that were gradu-
ally replaced with the second generation digital phones by 2000. The third
generation wireless phone will handle data and Internet access as well as
voice. Portable data access can create new opportunities in mobile com-
merce to improve contact with customers. Entirely new applications can
evolve from virtually immediate contact.
local frms where owners recognized customer trends and ordered a specifc mix
of products. As manufacturing, transportation, and communication costs changed,
some frms were early adopters and began to change retailing. For instance, Wal-
Mart took advantage of size in purchasing and distribution to create giant retail
stores with a relatively standard mix of products. The Internet provides even more
possibilities for change. At a minimum, the Internet makes more information easi-
ly accessible to everyone. Now consumer can instantly search for products, obtain
reviews, compare prices, and even get feedback from friends. Mobile phones
particularly those with bar-code reader applicationsmake these features instant-
ly available, even while shopping in traditional stores. The accessibility of this in-
formation changes the relationships between consumers and sellers, which should
eventually create changes in the way businesses operate.
On the business side, just-in-time inventory and EDI were only the beginning
of a revolution in changing the ordering systems of manufacturers. Electronic
marketplaces and auctions enable businesses to fnd new suppliers and obtain sup-
plies on short notice. Sellers can fnd new customers and negotiate prices without
expensive sales visits.
You should remember one important point: EC is not just an issue with new
frms. In many ways, existing frms have the most to gain from EC, because they
can leverage their existing strengths. Business to business (B2B) is also critical
Figure 7.1
Electronic commerce. The Internet is commonly used for sales from business-
to-consumer and business-to-business. Salespeople can use it to collect data and
communicate with clients and the main offces. New Web-based businesses are being
created to provide data and services.
Large business
Small business/
supplier
Customer
Salesperson
The Internet
Orders, Auctions,
and EDI
Sales and
CRM
Service, orders,
and information
Web hosting and
Web-based services
Consumers
for existing frms because it can reduce costs and provide new options to man-
aging purchases. Finally, the interactive aspects of EC will become increasingly
useful for intranets to improve internal operations and facilitate human resources
management.
Figure 7.2 shows a broader classifcation of e-commerce that includes B2B and
B2C as well as consumer-led initiatives. The consumer issues C2B and C2C make
up a minor portion of e-commerce. It is not even clear that good examples of
these categories exist. Auctions sites, led by eBay, and social networking sites, are
probably the best examples of C2C sites. Yet, the bulk of sales on eBay actually
originate from small businesses, and consumers do not receive any of the rev-
enue on social networking sites. It is even more diffcult to fnd examples of C2B
sites. Probably the best examples are contract employee sites such as vworker.
com. Businesses (or sometimes individuals) post jobs online and potential work-
ers submit a price to perform the specifed task. Similarly, the job search sites
might loosely be considered as C2B.
Eventually, the entire world changes when commerce systems change. But will
the economy really change that far? Did the mass failure of the EC frms in 2000
and 2001 really mean that consumers prefer the traditional systems? These are the
questions that make life fun and add the risk to entrepreneurship. You must under-
stand the fundamentals before you can create your own answers to these puzzles
and help shape the future of business.
Selling Products Online
What types of products are sold online? Selling products online is a
natural extension of the catalog sales that existed years before the Internet was
created. However, Web-based businesses can provide more pictures and descrip-
tions. Customers can search for products and sizes. It is easier to update the data
to provide new products. It is also easy to change prices or to tailor prices for
different groups of customers. These factors alone would probably be enough to
encourage more people to purchase goods online, but would not likely result in
a huge gain in sales. Yet, as Figure 7.3 shows, U.S. Web sales increased by an
average growth rate of over 22 percent from 2000 through 2010, according to
Figure 7.2
E-commerce categories. E-commerce can be considered in the four categories shown.
However, B2B and B2C are far more prevalent than consumer-led initiatives.
Business Consumer
Business B2B
EDI
Commodity auctions
B2C
Consumer-oriented
Sales
Support
Consumer C2B
Minimal examples,
possible contract employee
sitessuchasrentacoder.
com
C2C
auction sites (eBay),
but many of these are
dominated by small
business sales
US Census Bureau data. In the fourth quarter of 2010, Internet sales comprised 5
percent of the total retail sales. Not enough to proclaim the demise of bricks and
mortar stores, but a higher growth rate than expected from simply replacing mail
order sales. Note that the Census data does not include travel, fnancial services,
or B2B sales. One possible reason for the high growth rate might be the ability of
the Web to help customers fnd products and compare prices from multiple ven-
dors. Or, perhaps people simply want to purchase items from the comfort of their
homes. However, even if the high growth rate continues, it would take at least 25
years for EC sales to catch up to retail sales. And, some products and services are
unlikely to ever be purchased online. On the other hand, the Internet is heavily
used for research even if the product is not purchased online (consider housing
and automobiles).
To put the EC sales numbers in perspective, it is useful to consider one other
piece of data. In the fourth quarter of 2010, Amazon reported sales of $12.95 bil-
lion in their annual report. Compared to the total EC sales for that quarter of $52.6
billion, Amazon alone had almost 25 percent of the online sales! Some of Ama-
zons sales revenue comes from services, but the percentage is relatively low. This
large share partially exists because EC sales are still relatively new, and partially
because Amazon handles retail sales for many other companies.
Business to Consumer
When asked about e-commerce, most people think of B2C. Everyone is a con-
sumer, and it is easy to think about purchasing products and services over the In-
ternet. But how is e-commerce different from traditional sales methods? And are
there some products that people will not buy online?
Figure 7.3
U.S. e-commerce retail sales. Census Bureau data show a 26 percent average annual
increase. Sales for 2006-Q4 were 3.4 percent of total retail sales. Fourth quarter sales
average about 31 percent of annual sales for e-commerce, compared to about 26
percent for total retail sales.
0
10
20
30
40
50
60
B
i
l
l
i
o
n

$
U.S. E-Commerce Sales
A couple of simple rules dominate marketing. First, consumers prefer instant
gratifcation. Given a choice, with everything else equal, consumers will choose
the product at hand over one that will be delivered later. Related to this rule, con-
sumers will often buy items on impulsesimply because they are available and
enticing. Second, consumers will prefer to pay a lower total price. While this rule
seems obvious, it can be diffcult in practice. Consumers need to know that the
products being compared are equal. They need to know what other products or
suppliers exist and their prices. And they need to be able to compare total prices
(including taxes and transportation). Third, consumers prefer to see and touch
many products before they buy them. The real challenge in marketing is analyzing
the trade-offs when all conditions are not equal.
Food, Clothing, Housing, and Transportation
In many ways, traditional products are the least likely to be successful in e-com-
merceat least initially. Consider the four basic items on which people spend
most of their income: food, clothing, shelter, and transportation. Distribution is
the most critical issue with food. In the 1960s, when people lived closer together
in central cities and small towns, neighborhood grocery stores provided basic food
items within a short distance of many people. Yet these stores were ineffcient be-
cause they could stock only a limited number of common items; and the delivery
and inventory costs were high. Ultimately, large grocery chains that used their
size to hold down costs, provide a large selection, and negotiate favorable terms
with suppliers replaced the local stores. Consumers had to travel farther to these
new stores, but they were willing to accept the distance to reduce costs. Small
local stores were driven out. The two initial leading frms to sell groceries over
the Internet were Peapod and Webvan. Customers placed orders on the Web site
and drivers dropped off the purchases at a scheduled time, usually the next day.
Customers paid a delivery charge for the service. Neither company was very suc-
cessful in terms of profts. The increased costs, delivery time, and lack of ability to
touch the items discouraged customers.
Reality Bytes: Amazon Expenses
In November-December 2009, U.S. e-commerce retail sales not including travel hit
a record $32.6 billion, or an increase of 12 percent from 2008. For the fourth quarter
of 2010 (October-December) Amazons revenue was $12.95 billion, for an increase
of 36 percent. Think about those numbers for a second. Even assuming the U.S. sales
for October were similar to the November and December values, Amazon captured
over 25 percent of the U.S. online sales (not counting travel). To support its continu-
ing expansion, Amazon added 13 distribution centers in 2009, giving it a total of
52 sites. These costs have impacted Amazons proft, and the companys profts in-
creased only 8 percent compared to the 36 percent increase in revenue. Some of the
growth is fueled by digital products such as e-books, which now account for more
sales than print books at Amazon. In other cases, simple convenience and searches
encourage customers to shop online. Meaghan Keane, a Fairfeld, CT nursing student
noted that she did all of her holiday shopping online because I was able to fnd all
these different items that I wouldnt have even known where to look for in stores.
Adapted from Stu Woo, Expenses Eat at Amazons Proft, The Wall Street Journal,
January 28, 2011.
A second way to look at EC and food is to realize that few families cook their
own meals. Takeout from a variety of restaurants and even grocery stores is a pop-
ular substitute. Many restaurantsparticularly pizzeriasdeliver food on short
notice with just a phone call. Would there be a reason to convert the phone sys-
tem to an Internet connection? A few places do this, but many people still prefer
phones. Theoretically, the Internet can provide menus and it is better at handling
multiple customers at the same time. Furthermore, it could be used to provide
feedback to the consumerin terms of status of the order and when it will be de-
livered. But most restaurants are small businesses, and they have resisted building
the infrastructure to provide these features.
Clothing offers more prospects for e-commerce. Selection is always an issue
with local stores. No matter how large the store, it can carry only a small, targeted
selection of styles and sizes. And larger selection means greater inventory costs.
So, there is room for an EC frm to sell a wide selection of products across the na-
tion. In fact, several catalog mail-order frms concentrate on these markets. These
Reality Bytes: JCPenneys Gets Caught on the Dark Side
Search engine optimization (SEO) has become a popular topic among online ven-
dors. The goal is to obtain as high a ranking as possible on Google and other search
engines. If a site ranking closely matches a search term, it will be displayed at the
top of the page automatically; without paying any money to Google. Some people
spend a huge amount of time trying to determine how the search engines work. A
few people spend huge amounts of time trying to fgure out how to game the system
and obtain a high ranking for their pages. But, trying to deceive the search engines is
always a bad idea. Google succeeds only by returning accurate matches and decep-
tion can result in punishment. JCPenneys, an American department store, found out
the hard way in 2010/2011. For several months in 2010, JCPenneys was at the top
of the result lists for many products and even manufacturers. Apparently, the retailer
tried every trick in the book to get its site listed frst for hundreds of itemsusing
methods that crossed the line to the dark side of SEO. Doug Pierce, an expert in
searches, examined some of the companys practices and noted that Actually, its
the most ambitious attempt Ive ever heard of. This whole thing just blew me away.
Especially for such a major brand. Youd think they would have people around them
that would know better. Eventually, Google altered its algorithms and also punished
JCPenneys by forcing all of its entries to be buried several pages deep in the results.
JCPenneys fred its SEO consultant. Google uses on links from other Web sites to
help determine the content and importance of a site. Open Site Explorer helps iden-
tify when sites contain links to any given page. JCPenneys was paying people to
post links to its pages. Googles terms of service allow it to punish any company it
feels is not playing fairly. JCPenney also spent money on Google adsas much as
$2.4 million a month. In 2006, Google penalized BMW for using a black-hat SEO
strategy and basically removed the company from its search results. Overstock.com
was similarly punished with a two-month downgrade in 2011 when Google found
that Overstock was paying students to put links on their Web sites.
Adapted from David Segal, The Dirty Little Secrets of Search, The New York
Times, February 12, 2011; and Julianne Pepitone, Why Google Threw Overstock in
the Penalty Box, CNN Online, April 25, 2011.
mail-order companies have also been relatively successful at e-commerce. The
one problem is that clothing sizes are not quite standardized. Hence, many shop-
persparticularly womenprefer to try on clothing before purchasing it. Some
leading sites like Lands End have implemented electronic virtual models that en-
able customers to select items and see how they might look. But many people
prefer touching the individual garments frst. Still, the ability to search for styles
and sizes makes the Web a useful tool. And many consumers learn to trust certain
brand names, so they search the Web for those specifc brands.
Finding a home is always a diffcult task. A medium-size city might have thou-
sands of homes for sale at one point in time. It is hard to fnd and compare all of
the details. The Internet has helped in some respects. Searching is an important
strength of the Internet. Several real estate databases exist online to retrieve house
listings based on a variety of items. However, almost no one would buy a house
without seeing it in person; so the role of the Internet is limited. At frst glance, it
would seem that real estate might be a prime opportunity for e-commerce. Real
estate commissions are often priced at 6 percent of the sale price, which can be a
high value for expensive houses. So there should be strong incentives for remov-
ing the commissioned agent from the middle of the transaction. But buyers often
prefer to use agents. Real estate agencies control most of the housing market data
and they have resisted moving to an online worldparticularly one that might
result in a reduction in the commission rates. However, the Zillow.com site offers
several interesting features that might enable it to encroach on the traditional retail
listings.
Reality Bytes: Nestle v. Red Cross Tweets
Corporate marketing started to fall onto the social networking bandwagon by 2010.
The adoption of Twitter made it temptingly easy to create a social image. Companies
hired individuals and marketing frms to spread the news about the companies online.
But, image is something that can be feeting or at least easily affected by even simple
mistakes. In 2010, Greenpeace convinced people to food Nestles Facebook page.
Greenpeace objected to Nestles use of palm oil which allegedly impacts habitat for
orangutans. The organization encouraged supporters to food Nestles Facebook page
and use profle pictures of altered Nestle logos. The person running the Nestle page
responded by posting statements about taking down the negative comments. The re-
sponse angered thousands of more people and the foods escalated. Eventually, the
Nestle representative offered an apology, but the incident did not help the companys
image. In contrast, someone @RedCross posted an accidental tweet Ryan found
two more 4 bottle packs of Dogfsh Heads Midas Touch beerwhen we drink we
do it right #gettngslizzerd. An offcial response came quickly from @RedCross
with the tweet Weve deleted the rogue tweet but rest assured the Red Cross is sober
and weve confscated the keys. Dog Fish beer fans latched onto the #gettngslizzerd
hashtag, and breweries and pubs in many states launched beer-for-blood offers, in-
creasing awareness and donations to the Red Cross. Other examples of good and bad
responses exist, and new ones are likely to arise. Few rules exist in social networks,
and tastes change. But, building a loyal following and using light humor are gener-
ally good ideasparticularly when something goes wrong.
Adapted from Julianna Pepitone, 6 Painful Social Media Screwups, CNN Online,
April 7, 2011.
Transportation is more interesting because it can be a product (automobile) or
a service (airplane or subway ticket). Airlines have done well with direct e-com-
merce sales. Yes, they have offended the traditional distribution channel (travel
agents), but the people who travel the most have been willing to purchase tickets
directly to save money and gain control over their choices. Today, almost all air-
line tickets are sold online.
Automobiles are more interesting. Few people are willing to buy new automo-
biles over the Internet, but Web sites offering searches for specifc used cars are
popular. Part of the difference is that consumers have greater bargaining power
with used automobiles than with new ones, so the increased information on avail-
able cars gives them more leverage. If one owner or dealer offers a high price for
a specifc car, you can quickly fnd another one. New cars are more challenging
because of the strong relationship between the few manufacturers and the dealers.
Most people want to test-drive a car before they buy it, so the manufacturers
have a strong incentive to keep the dealer network. Many insiders also believe that
the salesperson is critical to selling cars, by overcoming objections and talking
people into buying cars when they hesitate. So, if dealers and showrooms are nec-
essary, what is the value of e-commerce in selling a car? Both General Motors and
Ford have experimented with online sales of new cars, with minimal success. Cur-
rently, the auto manufacturers prefer to sell through the dealers. A few states actu-
ally have laws that prohibit auto manufacturers from selling directly to the public.
While the manufacturers want to keep the advantages of the dealers, they also
want to fnd a more effcient method to distribute cars. Manufacturers have talked
about an Internet-based build-to-order system, where customers could select op-
tions and cars would be made to order. A few larger regional dealers would main-
tain basic inventory to support test-drives. The advantage of this system would be
to reduce inventories and enable manufacturers to build only the number of cars
that are needed. The main drawback is the diffculty in confguring assembly lines
quickly enough to hold costs down.
Reality Bytes: Even the Best Fall Down Sometimes
As an early adopter and leading e-commerce company, Amazon runs one of the most
sophisticated cloud-hosting services on the Internet. Several big-name Web-based
companies rely on Amazons servers to run their operations. On April 21, 2011, Am-
azon suffered a collapse of its internal network in its East Coast facilities. The dis-
ruption knocked out or slowed service to Internet service frms such as Foursquare,
Reddit, and Quora. Despite early optimism, it took Amazon engineers several days to
fully restore service. Amazon posted a detailed explanation of the problem after the
site was restored. Interestingly, the problem arose partly because of the redundancy
built into the EC2 facilities. A misconfgured router shifted traffc to a slow internal
network and datasets were unable to connect to their replica backups. Even when the
router confguration was fxed, essentially every dataset tried to re-establish a new
backup copyfooding the data controllers. With this knowledge, Amazon began
rebuilding its systems to prevent similar problems.
Adapted from Shara Tibken, Amazon Cloud Snafu Disrupts Websites, The Wall
Street Journal, April 21, 2011 and http://aws.amazon.com/message/65648/.
On the other hand, used cars present a different situation. Buyers have em-
braced the Internet as a method to locate used cars and compare prices. Sellers
fnd it a useful tool to avoid the high prices of newspaper advertisingparticu-
larly since it is inexpensive to place photos on Web sites.
The point of these examples is that without substantial changes in behavior,
people will continue to prefer the traditional sales mechanisms. Particularly since
these mechanisms have proven to be relatively effcientor closely controlled by
a group with a strong interest in maintaining the current system.
Product Characteristics for Online Sales
So, is the world really so bleak that B2C e-commerce is doomed to fail? What
about Amazon.com, one of the larger retailers on the Internet? By selling com-
modities, they avoid the problem of choosing products. The latest best-selling
novel is the same regardless of where it is purchased, so consumers do not need
to see and touch it before purchasing it. Besides, Amazon.com can offer a wide
selection of titles. It can also make it easier to fnd books and does a good job of-
fering related products that might interest the consumer. But the concept is still
hobbled by the distribution problemit takes time to deliver a book or CD to
the customer. So, online vendors like Amazon.com compensate by offering larg-
er selections than those found at a traditional local store. Amazon also discounts
some items to encourage people to accept the delivery delay. But, to achieve prof-
its, around 2003, Amazon began charging list price for mass-market paperback
bookswhich represent a substantial percentage of most book sales. Amazon has
also been one of the leaders in the switch to digital bookslargely by offering
their own e-book reader (Kindle) at declining prices.
To fnd successful B2C strategies, you need to look at the features that the In-
ternet provides that are superior to traditional stores. The most important ones are
the: (1) search technologies, (2) ability to quickly compare multiple products and
vendors, (3) low costs for large amounts of information, (4) ability to reach a wide
audience, and (5) ability to tailor responses to individual customers.
The essence of a proftable Web site for products is to identify items that can
beneft most from these features. For example, specialty products can be hard to
fnd. A few frms could reach a national audience using EC and capture most of the
market by being easy to fnd and offering competitive prices. Another approach is
to offer products with many options that require customization. Dell has been suc-
cessful at selling computers over the Internet by making it easy to confgure and
compare prices and by using modern production technologies to hold costs down
so that the EC solution highlights their prices.
Look at the EC sales data one more time and realize that almost one-third of the
sales take place in the fourth-quarter, year-end, holiday shopping season. People
are searching for specifc items. They might want unique items, or they might
search based on availability or the lowest price. Of course, the types of items that
ft these categories are also items that are likely to be purchased online during the
rest of the year.
Business to Business
Business-to-business e-commerce has the potential to be more important than
B2C e-commerce. First, businesses tend to buy repeat items in bulk, so they do
not need to test-drive or touch every product. Second, medium and large busi-
nesses already have high-speed connections to the Internet and rely on computer
systems in their daily operations. Third, costs are becoming a driving factor, and
technology can reduce the transaction costs and the number of errors.
Extended EDI
By simply offering the ability to sell products to other businesses, the Internet can
be used for EDI. For materials and components that are purchased on a regular ba-
sis, EDI software can connect across the Internet to automatically monitor inven-
tory and send orders to the appropriate company. For less frequent purchases, a
buying frm could set up software to scan servers for prices, select the appropriate
items, and place orders automatically.
Currently, few systems work this effciently. Most require a human to collect
data and place orders. Web sites make it easier to collect data on prices and avail-
ability, but every site has different search methods and different purchase screens.
EDI software helps by following standards, but the companies involved must in-
stall and confgure it. XML is a more fexible method of sharing data with sup-
pliers and customers. Eventually, the technology can be married with expert sys-
tems to provide more automated intelligence to handle ordering basic items and to
monitor the progress of standard orders.
Some companies, such as Offce Depot, have leveraged the Web to lock-in
business customers. Companies sign up with Offce Depot and employees place
orders for offce supplies directly on the Offce Depot Web site. Offce Depot de-
livers the products on a daily basis and bills the company for all purchases. It is
not exactly EDI, but a simpler method to connect directly to customers. Most ERP
systems provide these capabilities, so almost any business can create Web sites
that enable business customers to order directly.
Auctions
From an economic perspective, B2B auctions are one of the most exciting tools
created through e-commerce. In the past, companies purchased materials and sup-
plies from a complex set of distributors and wholesalers, driven by in-person sales
calls. Within this context, most manufacturers dealt with only one or two suppliers
for each part. To hold down transaction costs, it was simpler to establish long-term
relationships with a limited number of companies. Of course, it made it harder to
ensure that the buyers were getting the best price. Competition helps hold down
prices during the initial contract negotiations, but if anything changes in the en-
suing year or two, it is diffcult to renegotiate the contract. But in exchange, the
buyer gains a more stable environment.
Economic theory shows that well-run auctions are the most effcient way to
establish an effcient market price. To be well run, the auctions have to be open to
the widest range of participants, and everyone must have complete information on
the items and prices. Several industry-specifc auction sites have been established.
One of the more successful sites involves the steel industry. Signifcant amounts
of steel are still sold directly to manufacturers on long-term contracts. However,
the auction sites make it easier for suppliers to unload specialty and overstocked
inventory in a spot market. The auctions also give manufacturers the ability to
monitor the spot market and availability of steel products, so that they can quickly
pick up additional quantity.
For commodity items, auctions can hold down prices by improving competition
and making price and quantity data available to all participants. Auctions are also
useful for specialized products, when it is diffcult to determine an asking price.
Some companies have found that they can obtain higher prices for their products
when they sell them at auction. Also, auction prices can change easily, so if there
is a short-term jump in demand for your product, you will be able to capture the
additional proft. Of course, you might have to accept lower prices the next day.
Te Production Chain and Disintermediation
What problems might arise from shifting sales to the Internet? To understand the
issues in EC, you must frst understand the production chain. Shown in Figure
7.4, the strategic effects are important to EC. One of the key aspects in B2C e-
commerce is the ability to bypass entire sections of the production chain. Consider
the situation of airlines. In the 1960s and 1970s, airlines created giant reservation
systems to handle fight bookings. The system consisted of the airlines massive
central computers and databases and travel agent terminals connected by a custom
network. It was too expensive for customers to connect directly. Also, the systems
were hard to use and travel agents needed special training. Agents were paid a
commission based on the value of the fights booked through the reservation sys-
tem paid by the airlines. With the advent of frequent-fyer miles, airlines encour-
aged consumers to book fights with the airline itself, bypassing the travel agent
and saving the cost of the commission. But it is diffcult to search for fights using
parts
supplier
parts
supplier
parts
supplier
warehouse warehouse
supplier supplier supplier
tool
manufacturer
Manufacturer
workers
wholesaler
wholesaler
distributor distributor
distributor
retail store
retail store retail store retail store
Consumers
Figure 7.4
Production chain. The production chain is important to EC because EC offers the
ability to skip elements of the chain. It also makes it easier to reach new customers at
any level in the chain.
the telephone. The Internet changed everything. Several travel sites and the airline
sites themselves make it easy to fnd fights, compare prices, and purchase a ticket
any time of day without the assistance of an overworked travel agent or salesper-
son. Tickets are merely electronic reservation numbers, where you simply show
appropriate identifcation at the airport. On the production chain, the airlines (as
service providers) bypassed the intermediaries to sell directly to the consumera
process known as disintermediation.
As shown in Figure 7.5, for a few years, most of the airline industry
relied on a few companies to provide data and reservation services to
various Web sites. These companies were the same reservation systems
originally created by the airlines, including Sabre created by American
Airlines, and spun off. In 2010, the disintermediation extended further
when American Airlines, in a dispute with the reservation companies
(including Sabre), pulled all of their data from the online systems and
required customers to book online directly with the airline. Southwest
Airline had long followed a similar practiceto avoid the fees paid to
the reservation companies. One effect of this change is that it is more
diffcult for consumers to fnd and compare fight costs. If all airlines
require individuals to connect directly, a search would require check-
ing every airline separately. Eventually, a consolidator might be able
to search and display data from multiple sites, but the airlines will be
reluctant to pay the fees, so the consolidator would have to fnd an-
other way to make money.
Figure 7.5
Airline industry disintermediation. Partly because airlines sell a perishable service,
and partly because of the role of information (reservation), they have led the way in
fnding ways to connect to customers. Web sites initially supplanted travel agents,
but the Web sites were controlled by the original reservation companies. To reduce
fees, airlines (led by Southwest and American) bypassed the reservation systems and
require customers to work directly with the airline.
1960s-1990s
Airline (American)
Reservation
system
(Sabre)
Travel agent
Customer
2000-2010
Web Sites
(Expedia, Orbitz,
Travelocity)
2010-
A similar process can occur in manufacturing. Instead of wholesalers, purchas-
ing agents and suppliers are forming B2B auction sites. Companies can sell or buy
products and materials on a variety of Web sites. Instead of searching out buyers
or sellers, businesses simply submit a bid on a Web site that covers the desired
product.
Today, it is rare for a company to be vertically integrated across the production
chain. For example, most manufacturers rely on other frms to handle distribution
and retail sales. In a sense, they choose to outsource these functions because of
the costs. Over the last 50 years or so, frms have worked to become more eff-
cient within their niche. As shown in Figure 7.6, e-commerce has the potential to
change these relationships. By reducing the costs of dealing with individual cus-
tomers, it becomes possible for frms to circumvent the retailers and sell directly
to the end consumer. Since retail price markups can be in the range of 100 percent,
the manufacturer has a strong incentive to sell directly to the public to capture
some of this additional proft.
However, particularly in these early days of e-commerce, many manufactur-
ers are reluctant to remove the retail role. Taking sales away from the established
retail channel could alienate the retailers, who are currently responsible for almost
all of the sales. If retailers decide to drop your product, and consumers are slow to
switch to direct purchases, you may lose the market.
Interrelationship with the existing retail distribution channel is a critical factor
in any e-commerce strategy. Removing an intermediary can increase your profts
Manufacturer
Retailer
Consumer
Production Chain
E-commerce
Web site
Figure 7.6
Disintermediation. The traditional production chain evolved because separate frms
provided specialized functions more effciently. E-commerce reduces costs and has
the potential to circumvent various intermediaries. But disintermediation is a risky
strategy.
and can be used to reduce prices to capture more market share. But is the interme-
diary a critical component? Will your sales remain if you remove it? Are custom-
ers ready to switch to direct purchases? Some frms attempt to do both: keep sales
through traditional channels and also provide direct e-commerce sales through a
Web site. But they minimize competition and appease the retailers by charging the
suggested retail price on the Web site. So, consumers can often fnd the product
cheaper at a local storewhich is willing to offer a discount on the list price.
In some situations, the reliance on the retail channel leads to strange conclu-
sions. In early 2001, Compaq was producing a popular PDA with a color screen.
But Compaq had trouble with production and demand was substantially higher
than supply. Consequently, few retailers carried the product. Some of the busi-
nesses that were able to obtain the product did not bother to sell it at their local
stores. Instead, they auctioned the units individually on eBay for several hundred
dollars over the list price. Would Compaq have been better off auctioning the
PDAs directly to consumers?
Tree Stages of a Purchase
How do you use the Internet to market goods and services? From a marketing
perspective, the buying process has been defned in three basic steps: (1) prepur-
chase information gathering, (2) the purchase itself, and (3) postpurchase support.
Figure 7.7 lists these steps with examples. The Internet can support each of these
areas. The level of support in each area depends on the industry, type of prod-
uct, and capabilities of the customers. Note that the process differs depending on
whether the customer is a retail consumer or another company. For example, sales
to other companies are generally repetitive and typically require additional docu-
mentation and billing procedures.
In many ways, the purchase issue is a minor component. EC purchases cur-
rently offer only minor benefts compared to traditional phone orders. The main
beneft to EC lies in providing additional support to customers before and after
the sale. In particular, intelligent Web sites supported by expert systems can help
customers select options and products or solve problems. For fxed development
costs and relatively low monthly fees, the online systems can provide 24-hour
Prepurchase Purchase Postpurchase
Static data sites
Promotion
Product specifcations
Pictures
Schematics
Pricing
FAQs
Interactive sites
Confguration
Compatibility
Complex pricing
Transmission security
User identifcation
Product selection
Payment validation
Order confrmation
Service
Problem tracking
Sales leads
Resolve problems
Answerquestions
Product evaluation
Modifcations
Tracking customers
Figure 7.7
Electronic commerce. Websites are commonly used to support the three main phases
of marketing.
support. Sales can be increased by providing more detailed information, helping
customers customize their selections, and using an expert system to build cross
sales. Costs are reduced because the system is automated. Sales and costs can
be further improved by providing after-sale support. Expert-system guided sup-
port can help customers solve problems faster. Any product design or production
problems can be reported directly, giving you the chance to fx the product before
it ships to more people.
Price Competition
A primary concern expressed by many frms investigating e-commerce is the is-
sue of price. The Internet makes it easy for people to search vendors and com-
pare prices. This process is particularly easy for products that are the same (such
as books, videos, and electronic equipment). For example, Google (www.google.
com/products), Bing (click the Shopping tab), and several other sites offer search-
es for products that display prices and availability for a variety of products. As
shown in Figure 7.8, several cell phone apps even provide the ability to snap a
picture of a bar code and look up online prices. Consumers now have easy meth-
ods to obtain more information.
However, merchants are concerned that people will compare sites based only
on price. Availability of the item will also make a difference. Why are merchants
concerned about price competition? First, the existing retail product chain was
originally created so that retail outlets could provide personalized service and
product information to customers. Vendors survived and grew based on their
ability to provide customized information and support to local areas. Competing
purely on price and availability changes the rules and requires a different type of
merchant system. Second, if customers look only at price and availability, it is
easy for a new frm to enter the market. The new frm simply slashes prices and
sells products at a loss to attract customers. Of course, in the long run, the frm
will failbut so will the other frms. Economists can assume that managers and
owners are intelligent and will eventually learn to charge a price that does provide
a proft. But it might take time for this new economy to evolve. This fear of ir-
rational frms does have an element of truth. Amazon gained its market position
primarily by offering substantial discounts on books. In fact, Amazons fnance
Figure 7.8
Barcode search. Most smart phones have barcode scanning apps that use the camera
to capture the barcode, send the number to a Web server, search for the item and
return sites and stores with prices for that item.
http://scan.jsharkey.org/
Web search
Prices and more
offcer made the remark in 2000 that he was surprised people were criticizing
Amazon for not making a proft. He said that Amazon never intended to make a
proft on sales. However, after the crash in e-commerce stock prices, Amazon has
worked harder to cut costs, and has increased prices on many items to achieve a
proft.
Initially, the largest impact of Internet price competition will be on the retail
frms. By minimizing the aspect of location, the primary strength of local frms is
eroded. If customers are willing to wait for products to be delivered, then there is
no longer a point in having thousands of small local stores. But that if is huge.
The ultimate economy will depend on consumer preferences between price and
the ability to receive a product immediately.
A ruling by the U.S. Supreme Court in 2007 could ultimately affect the degree
of price competition on the Internet. In Leegin Creative Leather Products v.
PSKS Inc., the Court ruled 5-4 that in many cases manufacturers could enforce
minimum retail prices for their products. Although the case did not directly in-
volve the Internet, one of the issues manufacturers face is that customers might go
to local stores to check out products and get advice and service. But the customer
might then turn to the Internet to purchase the product at the lowest possible price.
Effectively, the local store does the work but loses the sale to a store that provides
minimal service. The effect of the controversial ruling remains to be seen, but the
most likely impact is fewer discounts available for customers. Or at least it will
be more diffcult to fnd the discounts.. to obtain more information.ture the
barcode, send teh icesut the airlines will be reluctant to
A few e-commerce frms have attempted to use the interactive features of the
Internet to set prices dynamically. In an experiment, Amazon.com charged dif-
ferent prices to different customers. It appears to have been a relatively standard
attempt to statistically evaluate price sensitivity to various products. However,
when customers learned that others had obtained the same product for a lower
price, they complained. Yet, in traditional stores, customers routinely are charged
different prices for the same itemsfor example, through coupons or negotiation.
As shown in Figure 7.9, part of the fear is that the Internet might someday be used
to force people to pay the highest amount they are willing to pay, as opposed to
Reality Bytes: Dr. Peppers Social Networking
It took Dr. Pepper several years to build a base of 8.5 million fans on Facebook. But,
the investment pays off. The company uses the group for test marketing. Each day,
the company puts out two messages on its Facebook fan page and then it tracks and
evaluates the responses. New York based Code & Theory created custom code to
measure how many times a message is viewed and how often it is shared with other
users. Robert Stone, director of interactive media services for Dr. Pepper Snapple
Group Inc. notes that We mine the data to understand what is appreciated, and what
is not. One thing the company learned is that diehard Dr. Pepper fans like edge
one-liners. For instance, the company heavily promoted the phrase If liking you is
wrong, we dont want to be right. Through Facebook, Dr. Pepper gains fast feed-
back to concepts, along with free marketing when fans spread the ads to their friends.
Adapted from Geoffrey A. Fowler, Are You Talking to Me? The Wall Street Jour-
nal, April 25, 2011.
perfect competition price. At a perfectly competitive price, many customers pay
less than the maximum amount they are willing to pay, providing them with a
consumer surplus. If vendors can charge the maximum price to each person, the
company can capture some, or all, of the surplus resulting in higher prices for
consumers.
Services and Web 2.0
How do Web-based services work and why do they change the
world? As households and businesses obtain higher-speed connections, new
Web-based services become possible. You have probably heard about some of
the consumer-oriented sites such as YouTube, Facebook, and Flickr. Other sites
offer more sophisticated services to businesses. For example, several companies
provide online accounting or even ERP services such as customer relationship
management. The purchasing companies pay a monthly or annual fee to use the
applications. All of the processing and data storage is handled by the providing
company, leaving the customer frm free to ignore the technical issues and focus
on running its own business. Some companies, such as Google, provide applica-
tions such as a word processor, spreadsheet, and communication systems that can
be used by both individuals and entire organizations. A third important category
of services relates to the Internet itself: providing network access, developing new
applications, and hosting applications on servers.
Social Networking and Consumer Services
YouTube, Flickr, FaceBook, Twitter, and similar social networking Web sites led
the service revolution for consumers. These applications provide new tools for
people to interact with each other, offering more reasons to spend time online.
Most of the consumer-oriented sites do not charge customers to use the applica-
Figure 7.9
Dynamic pricing. The ultimate goal is to set individual prices for each consumer
to capture the maximum price each is willing to pay, as opposed to the perfect
competition price, where everyone pays the same price, and some customers gain
because they were willing to pay more.
P
Q
D
S
Perfect competition
price
Price consumer is
willing to pay
consumer
surplus
tions. Instead, they make money by selling ads. Consequently, the sites that grow
the fastest and attract the largest number of users stand to make the most money.
Giving free access helps attract users, but the sites still need to provide a service
that users actually want. There are still plenty of opportunities to invent new ser-
vices that will attract users, but as always, the challenge lies in fnding an appli-
cation that is useful to millions of people. On the other hand, as users gain even
higher-speed connections, particularly mobile ones, new opportunities become
feasible.
Some Web sites are actually services in disguise. For example, 1800fowers
sells fowers, but the company does not grow or ship the fowers. It contracts all
of the details to other frms. What the main site really does is keep track of special
event days, such as birthdays and anniversaries. It basically provides a reminder
service and makes it easy to order products corresponding to those dates.
The challenge with consumer-oriented service sites is to make money. The
choices are (1) charge for the service, (2) sell related products, (3) sell advertising
space, and (4) sell the service to another company. Many of the failed dot-coms
chose option three. When the advertising market crashed, they could not cover
their costs and went out of business. Option four is discussed in the business ser-
vices section. Selling related products is probably the easiest solution today. In
this case, the service is simply another feature that will attract customers to your
site. But the additional costs can make it harder to compete on the basis of price,
so you have to be certain that customers really do value the service. Figure 7.10
shows the basic process of generating money through ads. The Web site (such as
Facebook) contracts with Google to display ads on pages. No money is received
at this point. Ads are displayed along with the content. If a user clicks on an ad,
MySpace
Google Ads
Advertiser
Content
Ad
$
$
Figure 7.10
Free content supported by ads. Sites can offer free content by displaying ads
delivered through Google or Yahoo. When users click on an ad, the advertiser pays
Google, which pays a portion to the original site.
Google tracks the click and directs the user to the advertisers page. At the end of
the month, the advertiser pays Google, and Google pays a small percentage to the
original site owner. The site usually receives only a few cents per click, so it takes
thousands or millions of clicks to make money. But the site has almost no costs for
carrying the ads, so it is easy to experiment with sites to see what brings in money.
Financing and building an ad-based application is a challenge. Revenue does
not arrive until the site succeeds in attracting thousands or millions of users. In the
meantime, you need millions of dollars to build, host, and market the new service.
Digital Products
This category blurs the lines between products and services. You can buy music
as CDs, movies on DVDs, paper books and newspapers. If you buy these items as
traditional products, they need to be shipped to you. But all of these products can
be sold and delivered as digital products or services that you download immedi-
ately. For example, some music sites, such as Rhapsody, provide the music as a
service because you only rent the music. If you drop your subscription, the music
will no longer play.
Digital products are a feld where e-commerce will eventually dominate. Al-
ready, many products are stored in digital format: music, news, books, movies,
software, and games. In late 2000, many consumers found how easy it was to dis-
tribute digital music over the Internet using MP3 fles and Napstereven though
it was illegal. Digital content over EC meets two of the main consumer criteria:
it is instantly available at any time, and costs should be lower since distribution
costs are small. Furthermore, digital content is more portable than traditional CDs,
DVDs, and books.
The main challenge to digital content revolves around intellectual property
rights and laws. Digital content can be easy to copy and redistributedepriving
the owners (artists) of any reward. The risk is that free distribution of digital con-
tent would remove all incentive for artists and authors to invest time for which
they receive nothing.
Reality Bytes: Fox Broadcasting Sells More Online Ads
Fox Broadcasting, a division of New Corp., announced plans to increase the number
of online video ads. The goal is to match the number of commercials in traditional
TV broadcasts. Pricing of commercials would also change as the broadcaster strives
to combine TV and online audiences into a single group. Toby Byrne, the president
of Foxs ad sales noted that This is what our business is becoming. There shouldnt
be a different commercial experience by virtue of which way you choose to watch
our content. For early 2011, Foxs traditional prime-time audience share for 18-49
year old viewers had dropped 5.4 percent to 4.6 million. One goal of the new pric-
ing strategy is to capture the number of viewers who watch the show online instead
of through broadcasts. Fox is following in the footsteps of the CW joint venture
between CBS Corp and Time Warner Inc. who adopted the same approach a year
earlier.
Adapted from Sam Schechner, Fox Shifts How it Sells Online Video Ads, The
Several companies (particularly Apple, Microsoft and RealNetworks) have
developed digital rights management (DRM) systems to prevent unauthorized
copying. Microsoft and a couple other companies have systems in place for books.
Most systems take advantage of the Internet. When a consumer purchases a digital
product online, the purchase is recorded in a digital rights management server and
issued a unique ID. From this point, the systems vary. Some work by periodically
checking the Internet server as the product is used to verify that it is an autho-
rized copy. In some systems, the generated ID can work only on the computer for
which it was frst created, so giving the fle to someone else does not allow it to be
played or viewed. Some systems enable users to transfer rights to another person;
others do not.
Another challenge with digital products is the payment mechanism. The trans-
action costs on credit cards and checks are too high to enable low-price purchases,
such as buying one song for a few cents. Until micro-payment systems become
accepted, it is diffcult for sites to charge for content. Currently, subscriptions are
the most common solution. One of the more successful sites is The Wall Street
Journal, which charges an annual subscription fee to several hundred thousand
subscribers. And even the Journal admits that it has diffculty preventing people
from paying for one subscription and sharing it (although that would be in viola-
tion of the subscriber agreement).
One diffculty with any protection system like the digital rights management
schemes is that it is diffcult to stop someone from breaking the system. Early
software vendors in the 1980s learned the lesson that copy protection schemes
were routinely defeated and removed. However, the digital millennium copy-
right act (DMCA) in the United States makes it illegal to break copy protection
schemes. As a relatively new law, it remains to be seen whether this condition
can be enforced. Also, the law does not apply to people outside the United States.
DVD vendors pursued this issue in 2000 and 2001 when a group broke the en-
cryption scheme used to slightly protect DVD movies. The Motion Picture As-
sociation of America (MPAA) has sued several companies in the United States
for even linking to sites that list the decode algorithm, but the code remains on
thousands of sites around the world. A major question exists in terms of whether
copy protection schemes can survive. In 2007, Steven Jobs at Apple began push-
ing the recording publishers to enable Apple iTunes to sell music without DRM.
In general, customers do not like DRM because it limits how the content can be
used. In Apples case, songs purchased on iTunes can be played only on the Ap-
ple hardware; and DRM-protected songs purchased from competing sites cannot
be played on that equipment. Protecting content has been one of the biggest hin-
drances that has delayed the introduction of many services, including download-
able videos and books.
Business Services
Providing online services to businesses is similar to providing services to consum-
ers, except that the services differ and it is easier to charge other companies for
using the application. Services can be as simple as providing direct applications
such as ERP or e-mail. However, the Web infrastructure has the ability to support
distributed services, where companies provide various pieces of a Web page
which appears to the consumer as a single integrated page.
Applications
Many of the applications are aimed at small to medium businesses that lack the
resources to pay huge upfront costs to purchase expensive applications. For in-
stance, several companies provide online ERP services. Smaller companies con-
fgure their accounting systems and employees access the system through Web
browsers. Since everything is Web based, it does not matter where the applica-
tion is hosted. And, the smaller company can rely on the service provider to keep
the servers running, provide backups, and monitor security threats. Salesforce, the
leading customer relationship management system runs purely as an online ser-
vice. Companies pay by the month or the year based on the number of employees
using the system. Companies can expand or contract their usage by purchasing
access for additional employees.
Google has purchased several companies to provide even more online service
applications to businesses (and consumers). An e-mail and calendar communica-
tion system was the frst main application. Companies simply pay an annual fee to
maintain all of their e-mail on Google servers. They get to use their own domain
name (as opposed to gmail.com), and Google handles all of the servers, backup,
and security issues. Google (and other frms) also offer online service applications
to handle word processing, spreadsheets, and slide show applications. Although
they lack some of the features of standalone packages, basing them on a central
server makes it easy to share the data created with other team members. For in-
Figure 7.11
Distributed services. Firms can offer digital services over the Internet to other
companies, such as document translation or reservation handling. With XML, the
process is relatively automatic and can be billed for one-time and limited uses.
Company 1
Company 2
The Internet
Original
document
Translated
document
Internet Service
e.g., automated
document
translation
stance, someone in marketing can write a marketing plan. By simply assigning
permissions, other employees in the company can work on the same flefrom
anywhere in the world.
Distributed Services
XML and other standards create opportunities for a new type of B2B e-commerce.
Web sites can provide specifc automated services that can be sold to other busi-
nesses. One example would be a Web site like Altavista (babel.altavista.com) that
has an automated document translator. Figure 7.11 shows the basic concepts. The
key is that the services are automated and simply called from your Web site. This
arrangement is actually more of a peer-to-peer system than a client-server technol-
ogy. You can create a Web site that uses services from many different companies.
For example, you might create a Web site that pulls current stock price data from
one site, performs some complex fnancial calculations on a second site, and con-
verts currencies using a third sites exchange rates. All of these activities happen
behind the page, so your users see only the fnal application.
The main advantage to this type of system is that experts can build objects and
maintain and run the services on a Web site available to anyone willing to pay the
service fee. In many cases, the service fee could be a small per-usage value. So
you could build a composite application that has state-of-the-art features and pay
only for the actual usage of these features. The other alternative would be to li-
cense these technologies on an annual basis for a higher fee and run them on your
own server, where you continually need to maintain and upgrade the services.
The technologies to support these services are still being developed. XML and
the simple object access protocol (SOAP) are two important technologies. SOAP
is a method of describing and activating services across the Internet. Ultimately,
to make it easier to fnd services on the Internet, companies will want to register
with a directory. In many ways, Google is a primary Web service because it pro-
vides searches, and more importantly, delivers ads that generate money.
Network Services
A third category of Internet services consists of the network connections. As a
consumer, you see a portion of this industry because you have to pay an Internet
service provider (ISP) to connect to the Internet. Businesses also have to pay an
ISP, but they have additional expenses that are provided by other companies. To
create a Web site, you need to register the domain with a registrar paying an an-
nual fee, purchase a security certifcate for an annual fee, pay a hosting company
to run your Web servers for a monthly fee, and pay someone to design, develop,
and maintain your Web softwareusually for a fxed fee plus maintenance costs.
Several companies compete in these various businesses, and the industry brings in
billions of dollars a year.
Internet Service Providers and Hosting
Providing Internet access to households and businesses is a big business. Think
about how much you pay for monthly Internet service and multiply that num-
ber by several million. The business also has large economies of scale, both in
providing services to users and providing facilities for businesses to host Web
sites. Consequently, the industry has consolidated over the past decade. However,
there is still a relatively intense battle for providing service for the fnal mile to
individual households. Chapter 3 describes the various options and costs, but the
Technology Toolbox: Paying for Transactions
Problem: As a business, how do you get paid?
Tools: Many mechanisms have been proposed to handle payments, from cash to
credit cards to cell phones.
Payment Method Fixed Cost Fixed Fee Discount Fee Fraud/Insure
Cash Low except for
security
$0.00 $0.00 Physical security
Check-physical $20/month $0.25 1.7% Included
Check-electronic $20/month $0.25 2.5% Included
Credit card-
physical
$10/month
Minimum $25
$0.25-$0.50 1.6% Covered:0.08%
fraud average
Credit card-
electronic
$30 - $50/month
Minimum $25
$0.25-$0.50 2.6%-4% Not covered:
0.25%fraud
average
Debit card Setup/key pads $0.35-$.055 0% - 2% None
PayPal None $0.30 2.2%-2.9% Covered
for physical
shipments
For decades, checks were the dominant payment mechanism in the United States.
Only recently has it been possible to use electronic checksessentially a direct debit
to your checking account based on the routing numbers. Each method has potential
advantages and drawbacks. As a businessperson, you might be tempted to accept as
many methods of payment as possible, so you do not shut out potential customers.
However, many of the methods have setup and fxed costs, so it can be expensive to
accept everything.
Any payment method essentially has three main costs to the business: (1) fxed
setup costs, (2) transaction costs as a fxed number and a percentage of the price, and
(3) the expected loss from fraud or the cost of an insurance system. Often, there is an
implicit cost to train employees in the proper procedures as well, which increases as
you try to accept more variations of payments.
The transaction costs and risk are all paid by the seller. Merchant banks usually
charge a monthly document or connection fee, and sometimes a one-time setup fee.
A fxed cost per transaction is common, as well as a percentage of the sale. Custom-
ers often prefer credit cards, although in 2002, 15 percent of U.S. purchases were
made with debit cards, and that number is increasing rapidly. Debit cards are useful
to brick-and-mortar merchants because of the lower transaction costs and minimal
risk. However, they are rarely accepted for online commerce because of the potential
risk to consumer accounts. Checks and credit cards carry similar transaction costs
todaywhen you factor in the insurance coverage for insuffcient fund checks. One
challenge with most of the payment mechanisms is that merchant banks usually re-
quire a monthly minimum fee charge. If your frm has few sales through the system,
you will still have to pay $25 per month. For small companies just starting out in
business, this fee can be expensive.
Quick Quiz:
1. Why have consumers rejected most electronic payment mechanisms?
2. What additional fees are charged for international transactions?
3. What happens if a customer refutes a charge?
primary choices are telephone (dial-up, DSL, or FiOS) and cable modem. Cell
phone or mobile access is gaining in popularity but cannot deliver the same high-
speed service of cable modems. Providing business connections generates even
more revenue, with prices for a T1 line running about $400 a month. Most of the
providers in this market are large and diffcult to compete against. However, many
small wireless frms have arisen to provide niche services in small markets. For
instance, you could pay $400 a month for a T1 line and install a wireless modem
and sell access to the Internet to your neighbors for a monthly fee. You can break
even (not counting initial hardware costs), if you fnd 10 neighbors willing to pay
$40 a month. Of course, you have to handle billing, marketing, and some legal
and technical issues. So, you probably will not make much money, but a T1 line
provides bi-directional speeds of 1.544 mbps, which can be useful if your neigh-
bors want reliable speeds for video conferencing or other activities.
Web hosting is also a large industry. It has several major providers, led by those
who have built huge data centers around the nation and world. The large ISPs,
telephone companies, also provide hosting services because they have direct ac-
cess to their networks. However, many of the large data centers cater to larger
companies. So intermediaries have evolved who purchase space in the data cen-
ters and then resell services to smaller companies.
Design and Development
Compared to the other areas of Web infrastructure, design and development is
highly fragmented with thousands of frms and millions of individuals working
on applications and Web sites. The existing and potential demand is huge, both
for individuals and for frms specializing in developing Web sites and applica-
tions. Chapters 12 and 13 discuss the various information technology careers and
methods of organizing and managing the IT processes. For now, it is important to
recognize that many Web-based careers exist, with a variety of skills ranging from
artistic to programming, security, and management. It is relatively easy to estab-
lish a Web design frm and substantial demand exists for these services. Startup
frms might need to focus on low-price jobs to gain experience and respect. Busi-
nesses can fnd several frms that will help with everything from art design, to
programming, and custom marketing strategies.
Facilitators
Creating a commercial Web site requires a fair amount of effort and knowledge.
Several companies advertise plans to put you on the Web for a low cost. These
systems might work well for small businesses with a limited number of custom-
ers. However, few of them will scale up easily if your business suddenly grows.
Creating and managing a Web site can be a complex processmade increasingly
more diffcult due to security issues. Consulting and specialist frms have arisen
to help you understand these issues and design large, secure, complex Web sites.
Other facilitators have arisen to help with more mundane tasks. For example,
most cities have small businesses that will help people sell items on eBay. In gen-
eral, it is relatively easy to set up an account on eBay and sell various items. How-
ever, many people have only a few items to sell, do not want to take the time to
learn the details of how eBay works, and are uncertain about their technical skills
to handle the process. For a percentage of the revenue, facilitator frms will handle
the complete process, including taking photographs, pricing the item, posting the
information, validating the purchaser, and shipping the item.
Search Engines
Most people have used search engines to fnd information on the Internet. The
searches are not always successful and tend to return a large number of sites un-
less the key words are specifc and relatively unique. Nonetheless, search engines
are an important method for potential customers to fnd your site.
One of your frst objectives is to get your site listed and indexed on the major
search engines. Each system uses a different process to search and categorize your
site. All of the search engines have a Web page where you can register your Web
site by entering the Internet address and a description of the purpose of the site.
Eventually, your site will be added to the search engine list. Once your site is list-
ed by one of the major services, the other search engines will also fnd it earlier.
Some companies offer to register your Web site with the search engines for
a fee. While this process might be convenient, it is rarely necessary. The good
search engines eventually fnd your site even if you do not register it at all. Some
people claim to know tricks to make your page appear at the top of the search en-
gine listings, known as search engine optimization (SEO). Do not believe them.
In some cases, the advice is harmfulsearch engine developers know most of
the same tricks and will automatically punish Web sites that use the blatantly bad
ideas. Read the Web site descriptions at the search sites for more useful advice.
Basically, make sure your main page contains a precise description of the sites
purpose. Include key words that consumers are likely to search for. Be as accurate
as possible. Think like a customer and try searching for other sites. Look at the
key words you used. Include them on your Web page. The problem with SEO
frms is that the search engines continue to change their algorithms, and trying
to optimize your site for one version can result in a huge drop when the search
method changes. The ultimate goal of a search engine is to identify exactly what
the user wants to see and provide just those pages. So your best objective should
be to carefully defne exactly what your site and pages contain so the search en-
gine can match those pages to specifc searches. The main focus should be on the
page content and descriptions, meta tag keywords, and links from legitimate sites.
In some cases, there is a way to get your site listed higher up in the search
engine results. Many of the sites accept advertising payments to give higher prior-
ity to your site. You will have to carefully evaluate the costs and benefts of this
approach compared to other advertising strategies. Most search engines also rank
sites based on their linkages from other Web sites. So, if you can get your site
listed on established pages, with an accurate description, it can be found faster by
the search engines and will often be listed higher on the results page. Remember
that most people will not scroll through more than a couple of pages of search
results to fnd your page.
Payment Mechanisms and Risk Mitigation in e-Commerce
How can customers pay for products and why do you need new
payment mechanisms? Credit cards remain the most popular method of
paying for items and services purchased on the Internet. In other transactions,
debit cards have become more popular, but they are basically processed as credit
cards for Internet transactions. Consumers like credit cards because they are com-
fortable with them and because banks have largely removed the risks to consum-
ers. Chapter 6 describes the various transaction risks to the consumer, business,
and government.
Payment Risks
The transaction risks in e-commerce are similar to those of traditional commerce,
but with a couple of twists because of the network connection. The Internet is an
open network where messages can be intercepted and transferred at will. Con-
sequently, it is challenging for the merchant and customer to verify the others
identity. Similarly, both merchant and vendor need to be concerned about the
transfer of money and digital products. Because these two issues stem from the
same cause (the insecure network), they have both been solved through the use of
strong encryption methods. These techniques are described in Chapter 5. They are
commonplace on the Internet today, and consumers and vendors face minimal risk
from the interception of data. Figure 7.12 shows that encryption can protect the
transmission and storage of credit card data, but vendors still assume several risks
because the only method they have to identify customers is by the credit card data.
Some vendors attempt to reduce this risk by shipping products only to the home
address corresponding to the customers credit fle.
The risks of nonpayment or nondelivery are more diffcult to solve. They have
been particularly challenging in an international environment where governmental
jurisdiction and enforcement are not effectively defned. For the most part, con-
sumers are still protected if they use a credit card to pay for a product. However,
the consumer may still fnd it hard to prove that a product was not received. On
the other hand, e-commerce businesses have virtually no protection from fraud.
The credit card rules specifcally exclude mail orders, telephone orders (MOTO),
and Internet orders. The card companies will assist merchants in identifying inval-
id cards, but will not guarantee the transaction. This issue is important because the
Gartner Group (Computerworld 8/11/2000) estimated that credit card fraud is
Vendor
Customer
Encryption protects
transmission of data and
veries identity of vendor.
Consumer is protectged
by credit card company.
It is critical that vendors
protect their databases.
Vendor is not protected by
credit card and has only
weak methods to verify
customer identity.
p
ro
d
u
cts o
r services
Encrypt (Database)
En
cryp
t (cred
it card
d
ata)
V
erify ven
d
o
r id
en
tity.
Figure 7.12
Risk mitigation in e-commerce. Encryption verifes the identity of the vendor
and protects the data in transmission. Credit card companies accept risks for the
consumers but not for the vendors. Huge vendor databases of credit card data are
tempting targets and need to be protected by the vendor.
12 times higher for online merchantswith about 1.1 percent of all online trans-
actions fraudulent.
Technically, it is relatively easy to use encryption to verify the identity of the
merchant and the customer in any transaction. In fact, the common Web encryp-
tion system works because the merchant buys a digital certifcate from an encryp-
tion company. Customers could obtain similar certifcates or digital signatures,
but they have little incentive to do so, since the credit card companies protect
them. With todays encryption systems, transmission risks are relatively minor.
Two far more serious risks are (1) theft of consumer data from the vendors com-
puter and (2) alteration of the purchase documents by either the merchant or the
customer (repudiation).
The risk of theft is real and has happened to several vendors. The potential tar-
get is huge: thousands or millions of validated credit card numbersall accessible
via the Internet. The only effective solutions are for vendors either to keep the
card numbers off-line or to encrypt them and bury the encryption key.
The second risk arises because digital orders are easy to alter. The solution is
to create electronic orders that cannot be altered. Again, this solution requires en-
cryption. In this case, the customer and the vendor both need a digital certifcate.
When both parties encrypt the order, it cannot be altered later.
Payment Mechanisms
Payment mechanisms must change along with the changes in transactions. Years
ago, when purchases were made locally, currency was the primary method of set-
tling transactions. Eventually, as banks stabilized and gained respect (and gov-
ernment guarantees), transactions were settled by checks. Business transactions
(particularly internationally) are often settled with letters of credit from banks. In
the United States, many payments have migrated to credit cards and debit cards.
Figure 7.13
Payment mechanisms. Credit cards do not protect merchants and have high
transaction costs, so they cannot be used for low-price items. Several systems have
been created to provide the desired characteristics, but customers have not yet been
willing to adopt them.
Smart Card
5400-1111-0000-
Name
Credit card drawbacks
High transaction costs.
Not feasible for small payments.
Do not protect the merchant.
Characteristics needed
Low enough costs to support payments less than $1.
Secure transmission.
Authentication mechanism.
Easy translation to traditional money.
Alternatives
Mobile phone bill.
Smart cards.
Digital cash.
From a consumer standpoint, credit cards are easy, available, provide short-
term loans, and offer protection from fraud and errors. Figure 7.13 shows the
drawbacks to credit cards and lists the characteristics desired from a new payment
mechanism. From an e-commerce merchant perspective, credit cards offer only
minimal support and are expensive. From the perspective of a mobile-commerce
merchant, credit cards will be unacceptable because the transaction costs are too
high to support small payments.
From a theoretical perspective, e-commerce payment mechanisms should be
easy to create. In fact, dozens have been proposed or started in the past few years.
None of them have garnered enough support to be successful. Some, like PayPal
process payments, but most of those payments today are based on credit cards.
Figure 7.14 shows the main participants in a credit card transaction. The frst
important concept is that all of them charge for their servicesand those fees are
billed to the merchant. The second issue is to examine what happens if something
goes wrong with a transaction. For instance, if a customer fails to pay bills, the
issuing bank is stuck with the charges. Similarly, if a merchant skips town with-
out shipping products, the merchant bank is on the hook. Both banks cover these
risks by charging higher fees and limiting who they provide services to. So, small
startup merchants might have problems fnding a merchant bank and will prob-
ably have to pay higher fees.
Several companies have proposed alternative payment mechanisms, but most
of them failed. PayPal (owned by eBay) has survived, but it is largely a credit card
processing system now. Google and a few other Web sites also offer credit card
Figure 7.14
Credit card industry. Merchants can accept cards only if they are associated with
a merchant bank and a card processor. Both of them charge fees. New, untrusted
merchants, usually pay higher fees because the bank takes some of the risk.
Issuing Bank Merchant Bank
Customer Merchant
VISA,
MasterCard,
AmEx, Discover,
JCB,
Security
Database
Payment data
Product/service
Card Processor
Authorization data
Payment data
processing. The main difference between these companies and traditional card
processing is that the merchant never sees the credit card data. Customers enter
the data through the PayPal or Google site and merchants receive an e-mail noti-
fying them of the payment. By not handling the actual credit cards, the merchant
faces fewer security issues, but pays higher transaction costs.
The main challenge with credit cards is the relatively high fees. The per-trans-
action costs make it diffcult to accept credit cards for an amount less than $10.
Mobile commerce offers one possibility for handling small transactions: put the
cost on the customers cell phone bill. Most phone bills already contain lists of
small transactions, and the total monthly fee is high enough to pay using tradi-
tional mechanisms. As long as the vendor builds up enough credits with the phone
company, the transaction costs on that side will be reasonable. Currently, the big-
gest potential drawback is the limited security. Phone numbers are easy to fnd,
although a 4-digit PIN would improve security a little. Requiring physical access
to the phone will provide more security, until the phone is lost or stolen.
For on-site transactions, mobile phone companies are beginning to offer near
feld communication (NFC) payment methods. NFC is similar to RFID in that
it uses tags and readers that pick up data at very short range (a couple of inches).
Essentially, cell phones have a chip (or perhaps a chip on a sticker) that contains
encrypted identifers. As shown in Figure 7.15, waving the phone next to a ter-
minal transmits the ID and the terminal subtracts the payment amount from the
consumers account. This account could be a prepaid account or perhaps tied di-
rectly to a bank account similar to a debit card. In advanced systems, the terminal
can send a message to the phone with the receipt and details of the transaction.
Some systems require a PIN as well as the phonewhich is useful if the phone is
Figure 7.15
Near feld communication payment. The cell phone contains a tag with an identifer
that is read when placing it close to a terminal. Sometimes a PIN is added for
security. The terminal subtracts the money from the consumer account which can be
a prepaid account or a debit account.
Bank
Customer
Terminal
Identifier + PIN
inches
price
Message receipt
Prepaid account
Debit account
Technology Toolbox: Choosing Web Server Technologies
Problem: How do your Web servers handle interaction and database tasks?
Tools: Web servers require special code engines to handle the interaction with users
and the database. Three basic systems have evolved over time: Java (J2EE), PHP/
PERL/PYTHON, and Microsoft .NET. Websites today are built using one of these
three technologies. There are
major differences in the under-
lying technologies and philoso-
phies of the three approaches,
and some developers treat them
as religious issueswhich is
a way of saying that proponents
of each often argue over the
relative merits of their choice
of technology. As a manager,
why do you care about the three
technologies? The main answer
is that you will probably have
to make a decision between
them if you develop a website. All three have evolved considerably over the past few
years and will continue to improve.
Of the three, Microsofts .NET is the newest and has some features that do not yet
exist in the other methods. It is also considerably faster and more effcient at pro-
cessing code and connecting to databases. It is also the most expensive. Java is prob-
ably the second most powerful technology. It has the advantage of being a standard
that is supported by many vendors. Systems developed using standard Java code can
be transferred to run on Web servers built using a variety of hardware and operating
systems. Not only is the server code inexpensive, but the hardware and Web server
software are inexpensive. The other approaches use scripting languages that have
diverse features. Websites built on these systems can be run on a variety of hardware
and software platforms. In all three cases, you need to work closely with the hosting
company to ensure that the proper hardware and software packages are installed.
Building interactive websites generally requires custom programming to tell the
server how to handle the customer responses. Development environments exist for
Java and .NET that help programmers write the code. Most applications are built as
Web forms where the user enters the data. This data is then transferred to the server,
and the code behind the form examines the values. Based on the responses, the data
can be stored in a database, e-mailed to someone, or additional pages can be returned
to the user.
The differences between the systems lie in the capabilities of the underlying code
and the amount of effort it takes to write the code and connect to database systems.
Increasingly, vendors are developing libraries of code for the three systems that pre-
package common procedures. These packages enable developers to build new appli-
cations faster and with fewer errors
Quick Quiz:
1. Why would programmers become so attached to one system?
2. What are the advantages of choosing the most popular server technology?
3. What are the dominant costs of creating a Web site?
lost. The method is popular in Japanparticularly to pay for subways and trains.
Apple and Google have both begun offering the chips in some phones. However,
vendors need to upgrade their terminal systems and it is not clear yet whether the
transaction fees will be low enough to justify adopting this payment method.
In summary, electronic payment mechanisms are still in their infancy. It could
take years for a standard to evolve and be accepted by enough merchants and cus-
tomers to be important. In the past, governments have borne the costs of creating
and printing money. Today, most seem unwilling to become involved, and they
have left the mechanisms to the private sectorwhich is more focused on devel-
oping a system that provides profts to the issuing authority, instead of developing
a system that would be widely accepted.
Even for traditional businesses, bill presentation and payment mechanisms
can make life easier and save money. Service businesses in particular can beneft
because they tend to bill clients on a regular basis. The goal is to send all bills
electronically and provide a simple method for clients to transfer funds to your ac-
count. If clients are large businesses, they may wish to use existing EDI or bank-
ing systems (wire transfer) to transfer money. Smaller companies with smaller
payments will generally prefer the newer online payment systems because they
have lower costs to the customers.
The electronic systems save time because the bills can be created automati-
cally from the in-house billing system. They do not have to be printed or mailed.
Customers can feed the data into their own accounts payable systems and create
electronic payments with only a few review steps. The payment data, and any
disputes, can be processed electronically and automated. Many systems work di-
rectly from checking accounts, effectively transferring an electronic check. As of
2007, the U.S. checking system supports completely digital copies of checking
data, processing transactions within the existing system at relatively low transac-
tion costs.
Advertising
How do frms get revenue from Web ads and how do custom-
ers fnd a site? Advertising has become an important source of revenue for
Web sites. To understand and take advantage of advertising, you need to look at
the process from three perspectives: (1) the advertiser, (2) the publisher Web site,
and (3) the consumer or ad viewer. You probably have the most experience as the
user, but do you recognize the three main types of ads in use today? As with other
components of the Web, consolidation among companies has reduced the com-
plexity of advertising. Today, you can deal with two or three companies to handle
most of your advertising needsboth as a publisher and an advertiser. In fact, for
2010, the Interactive Advertising Bureau (IAB) statistics showed that the top ten
Web advertising frms were responsible for over 70 percent of the total advertising
revenue.
The three main types of Web ads are: (1) Small text links provided by Google
or Microsoft that provide publishers with a small fee when users click on them;
(2) Banner ads that contain images and fash video, where publishers are usually
paid simply for displaying the ad to visitors; and (3) Independent links to third-
party Web sites where the site pays the publisher based on the number of visitors
generated.
Traditional Media and Name Recognition
To a consumer, name recognition of a company is an important element of buy-
ing products over the Internet. Trust is particularly critical when the consumer is
not dealing face-to-face with the merchant. Depending on your target customers,
it might be necessary to build this name recognition through advertising in tra-
ditional media (television, radio, or newspapers). Some early start-ups chose the
splashy, but expensive, method of buying television spots during the Super Bowl
to reach a large audience. If you do use traditional advertising, make sure that
your Web site name is easy to remember and easy to type. Avoid words that are
commonly misspelled.
Web Advertisements
Web advertising offers some potential advantages over traditional advertising.
Ads can be delivered to specifc audiences and to some extent controlled so that
people continually see new ads. Increasingly, Web ads deliver on their original
promise of tracking responses by measuring the effectiveness based on the click-
through rate, or the number of people who actually click on an ad and go to a site
to get additional information. The original ad model matched the offine model by
simply displaying banner ads on a page and waiting for people to click an ad.
Figure 7.16 shows the estimated total advertising revenue by year as reported
by the IAB. Advertising fell from 200-2002 because of the dot-com crash and then
accelerated when Google introduced keyword advertising. Google has since pur-
chased several other advertising companies, including DoubleClickone of the
leading portals for banner advertising. Along the way, advertisers developed more
Figure 7.16
Web advertiser quarterly revenue. Revenue fell in 2001-2002 with the dot-com bust,
but began to recover in 2003 as advertisers found new ways to attract customers.
Sources: http://www.iab.net/resources/ad_revenue.asp and Google quarterly reports.
0
1
2
3
4
5
6
7
8
9
B
i
l
l
i
o
n

$
Web Ad Revenue
Revenue
Google
sophisticated ads using Flash and video. All of the parties have also increased the
ability to track customers and monitor the performance of all online advertising.
Still, the comparison to revenue earned by Google shows a strong correlation with
the growth of online advertising. By 2008, Google was also generating some rev-
enue from other sources, including hosting e-mail and Google docs, so Googles
total revenue now exceeds the total online advertising revenue. Estimating total
advertising spending is a challenge, but some reports suggest that online spending
represents about 9-10 percent of the overall U.S. total. Based on changes in the
industry, much of the online spending has come at the expense of advertising in
newspapers and magazines.
Banner Advertising
If you want to advertise on a site, how do you get started? If you have a popular
Web site, where do you fnd advertisers? Like the traditional world of print, radio,
and television advertising, the daily issues of handling the ads, monitoring place-
ments, fnding clients, and billing can be time consuming and expensive. Most
sites choose a third party to perform these tasks, and DoubleClick is by far the
largest such company. Of course, DoubleClick takes a portion of the ad revenue
for its services. Figure 7.17 shows the intermediary role played by DoubleClick.
The third party also simplifes the process for advertisers, since it would be diff-
cult for a company to contract with hundreds of sites to place ads. In 2007, Google
purchased DoubleClick from a private equity frm, to consolidate its role in the
advertising market, but it still operates under the original name and Web site. In
Figure 7.17
Advertising consolidator. DoubleClick provides advertising services by functioning
as a central point for both advertisers and publishers. DoubleClick handles the details
of delivering the correct ads, tracking impressions and collecting and distributing the
money.
User Web browser
Publisher Website
Advertisers
DoubleClick/Google
content
Ad
request
page
page +
ad link
Browse
info
ad
negotiate sites
negotiate ads
Rotate ads
Track hits
Collect money
Distribute payments
Track customers
2004, DoubleClick earned about $300 million in revenue, so despite its position,
it receives only a small percentage of the total advertising spending. Although it is
convenient for advertisers and publishers to deal with a single central intermedi-
ary, it is relatively easy for users to block ads from the central server. Check the
Internet for information on how to add the DoubleClick sites to your hosts fle and
your computer will stop receiving any information from them.
An interesting twist on banner ads is that video games are now beginning to
place ads. Games that connect online can download new ads. Generally, the ads
are embedded within the game, such as billboards on car-driving games. But at
least one pizza company added a link to a game so that players can click an icon
within the game and order a pizza for delivery.
Advertiser and Publisher Perspectives
In the world of traditional banner ads, similar to the offine world, advertisers will
place ads based on the demographic characteristics of the potential viewers. Fig-
ure 7.18 summarizes the perspective of advertisersthe ones who pay for the ads.
They generally want the ads to be seen and to generate click-through responses.
There is some argument that click-through rates are not an effective measure of
advertising. Possibly the effect of an ad is to build an image or a brand name.
Consumers might not need to purchase something immediately, but they might
remember the ads later and use them to accept the validity of the company.
Increasingly, advertisers want to establish a relationship with the potential cus-
tomers. Consequently, audience size and demographics are important to adver-
tisers. In particular, advertisers are fnding the most success in targeting specifc
sites. For instance, automobile manufacturers fnd it worthwhile to advertise on
sites dedicated to automotive topics (such as vehix.com, Edmunds, and autobytel.
com).
Many Web site publishers would like to get a share of the advertising revenue,
but keep in mind that according to the private Internet advertising bureau (IAB),
about 75 percent of online ad revenue goes to the top 10 Web site publishers. So,
there is not much left for your share. The frst catch is that you need a substan-
tial volume of visitors to get anyone to consider your site. Probably at least 25,000
unique visitors a month, and 1 million would be a more likely minimumsince
advertisers prefer larger audiences. Figure 7.19 shows some of the key points
from a publishers perspective.
One of the most diffcult issues is obtaining the demographic data. You need
some mechanism to identify your Internet users and to obtain some personal data
Figure 7.18
Web advertiser perspective. Advertisers want the biggest target audience possible.
They need demographics about the Web site visitors, and they monitor response
rates.
Wantviewerstoseethead.
Wantviewerstoclickthroughtothemainsite.
Wanttocollectcontactinformationfromviewers.
Needtomatchsitedemographicstotargetaudience.
Monitorresponserates.
Cost.
from them. Of course, this data raises many privacy questions. Most sites fnd that
they have to reward customers in some fashion to get them to provide personal
demographic data, but it is often amazing how little is required to get customers
to respond. Common tactics include random drawings for prizes or free trinkets.
The other approach today is to have focused Web sites and then match the demo-
graphics against the existing public data. For example, antique car sites attract a
particular demographic, while baby sites attract another group.
Google Keyword Ads
In 2002, Google changed the online advertising world by introducing click-
through ads tied to searches. Look again at Figure 7.16 and it is clear that the
entire increase in online advertising since 2002 has been due to Google ads. The
basic concept is straightforward, but to use the system effectively, advertisers
must make some complex decisions. Advertisers choose keywords. When a user
enters that keyword in a Web search, Google displays an ad from a company that
purchased that keyword. Users win because the ads are likely to match what they
are searching for. Advertisers win because users are more likely to respond to the
adand advertisers pay only when a user clicks on the ad.
One twist is that prices for keywords are not fxed. Instead, advertisers pur-
chase keywords based on an auction process. The company that bids the highest
price for a specifc keyword gets listed higher in search results. Since thousands or
millions of people might search for a keyword, advertisers need to control expens-
es by specifying a maximum budget, such as a maximum daily amount. When the
budget is spent, your ads will no longer appear. So the highest-price bid can vary
throughout the day, week, or month depending on the bids and budgets of com-
petitors. Advertisers must choose keywords, the bid price, and the budget amount.
As shown in Figure 7.20, Google provides support for selecting keywords and
estimating bid prices and daily budgets. In particular, Google shows the average
number of daily searches for the keywords you select so you get some idea of the
popularity of each word. Google also shows the current average prices being paid
by your competitors for each keyword and the approximate position your ad will
get. You can use this information to set a daily budget. Be carefulsome words
Figure 7.19
Web publisher perspective. There is money being spent on advertising, but your
rates depend on volume and the ability to provide detailed demographic data. The
daily tasks of sales and providing the ads are often handled by a third party like
DoubleClick.
Income
Cost per thousand viewings ($1 - $50)
Need volume (25,000 or 1,000,000 per month)
Need demographics
Tasks
Ad rotation software
Tracking and monitoring
Ad sales staf
Billing
Tird Party: DoubleClick
can be expensive. Paying a higher price will put your ad near the top of the list,
but it will quickly blow through your budget. Pricing your ad too low means that
it will not make the frst results page. Increasingly, people will never look beyond
the frst page of results or ads. On the other hand, if customers never see your ad,
they will not click on it and you will not have to pay anything to Google. So, per-
haps there is a strategy of bidding low and waiting for all of the other budgets to
run out. Ultimately, results also depend on the number of competitors. In the end,
you have to experiment with different words at different times to see which cus-
tomers actually click your ad and how many then complete a purchase. Oh, you
also have to write the adbut it is easy to look at the format of other ads to see
what features are attractive and which ones are not.
Google AdSense
Another Google twist was the introduction of AdSensewhich is designed for
publishers of other Web sites. Basically, anyone can place Google ads on their
Web pages. The process works the same way as AdWords, but Google matches
the keywords to specifc sites (instead of waiting for a search). Visitors to your
site see ads that are placed by Google. When visitors click an ad, the advertiser
is charged through Google and Google shares a portion of the revenue with your
publishing Web site.
One of the strengths of the Google approach is that it is relatively easy for both
advertisers and publishers to join the program. And both can experiment without
Figure 7.20
Estimating budgets for Google AdWords. Google provides estimates on the number
of daily searches for each keyword you select. It also looks at the average bid price
for those keyword ads and estimates a daily cost and ad position.
long term contracts and can begin with low numbers and small budgets. In other
words, it is easy to experiment and tweak your involvement until you fnd a com-
bination that works. Even small advertisers and Web sites with a small number of
viewers are welcome. In many ways, Google democratized the online advertising
world. Almost anyone can now advertise or receive money from ads even with
low traffc volumes.
Click Fraud
One potential problem with Googles approach is the opportunity for click fraud.
In theory, you could enter a search at Google for your chosen keyword; then re-
peatedly click on your competitors addriving up its costs and forcing it off the
list when the budget is reached. Google claims to have software in place to rec-
ognize these attempts, and claims that fraud rates are relatively low. For example,
ad clicks from the same IP address in the same day are supposed to be charged
only one time. Fraudsters have to be a little more sophisticated than sitting at their
desk clicking away. And most people are probably relatively honest and would
avoid such behavior. Nonetheless, you need to monitor your usage data and click
through rates to watch for potential problems.
Privacy
Privacy is the important fip side of advertising. The more serious privacy prob-
lems that have arisen were due to issues with advertising. The problem is the
trade-off faced by advertisers. Companies want to target ads as closely as possible
to people who are likely to care about and purchase the product. Hence, advertis-
ers want to know a considerable amount of information about current customers
and viewers of various Web sites. Yet collecting this data creates a loss of privacy
for the customers.
DoubleClick instigated one of the broader privacy problems. By routing so
many ads through its servers, DoubleClick is able to track the Web pages vis-
ited by each of millions of Web browsers. At one point, DoubleClick wanted to
sell this information along with demographic data on the individual consumers.
Reality Bytes: Sales Tax Revenue Declines
Citizens often want governments to provide services such as parks, education, roads,
police, and so on. Many of them do not realize how cities and counties fund those
operations. In most states, property taxes are the primary source of revenuefol-
lowed closely by sales taxes. The Census Bureau reports in 2008 that on average 23
percent of U.S. state and local taxes come from sales taxes. The recession of 2008
resulted in declines in purchaseshence a reduction in sales tax revenue. But, the
National League of Cities reported that municipal sales tax revenue declined in six
of the ten years from 2005-2010. Part of the problem is that cities and companies
overbuilt the number of malls. The amount of retail store space increased by 126
percent from 1970-2010 while population increased only 52 percent. But, city lead-
ers are concerned that online sales ultimately are a problem because they will reduce
sales tax revenue collected through traditional stores.
Adapted from Miguel Bustillo and Kris Hudson, Faded Malls Leave Cities in the
Lurch, The Wall Street Journal, June 8, 2011.
Most consumers are not happy when a company tracks the sites they visit without
informing them of the process. The basic premise of tracking demographic and
customer data is that, by knowing more about the customer, it is easier to provide
specifc ads and data that might appeal to the customer and, in essence, fewer
junk ads the viewer does not want.
Most advertising sites, including DoubleClick, use third-party cookies to track
sites visited by as many people as possible. When you visit a site that contains a
DoubleClick ad, DoubleClick inserts a cookie on your computer that records the
time, Web site, and ad displayed. Any other site that uses DoubleClick returns that
tracking cookie. Because thousands of sites use DoubleClick, the company can
generate an extensive summary of your online activity. In 2010, the Wall Street
Journal ran an extensive series of articles on the area of Web advertising and pri-
vacy. Consumers can take some actions to protect their privacy. For example,
most browsers have privacy settings to block third-party cookies. These settings
have to be made manually. Browsers also have more advanced private modes
which block almost any data and cookies, but allegedly in response to advertis-
er complaints, the developers (particularly Microsoft and Google) have made it
somewhat diffcult to enter that special mode. In 2011, Microsoft announced that
it would support an opt-out privacy flter that would give people some ability to
prevent Web sites from tracking any information about them. It remains to be seen
if these tools will be implemented or if they will be useful.
Developing and Hosting Web Sites
How do you create an EC Web site? Once you have decided that you
want to participate in EC, you need to fgure out how to create the site and get it
hosted. In part because of the expense of maintaining a high-speed Internet con-
nection, several companies have been created to provide alternative Web-hosting
options. These hosting companies already have high-speed Internet connections,
Web servers, databases, and management staff. They provide a variety of leasing
options to host your site. One of the most important decisions to make regarding
the Internet is where to locate the Web fles. A variety of choices exists, and each
method has different advantages, costs, and drawbacks. The choice of Web-host-
ing method depends on several characteristics of your business. Companies will
often start with one option and move to other selections as they expandparticu-
larly small frms or start-ups.
Today, companies rarely host their own Web sites. Several huge companies
provide comprehensive capabilities with servers located around the world. Even
if you want total control over your own servers, several companies provide co-
location services, and with high-speed Internet connections, it no longer matters
where servers are physically located. Even large companies that have the exper-
tise to run servers typically make use of these large server farms. Virtual machine
technology has made it even easier to use hardware maintained by other compa-
nies. The virtual servers can be confgured and moved to any physical hardware.
Even if you want a retail Web page to display the current inventory level (to indi-
cate if an item is in stock), the Web server can connect to your internal databases
across the Internet. Still, there are times when companies will want to run their
own servers. The bigger question is when you can use existing software and when
it is necessary to create custom software to run on the servers.
Simple Static HTML
The most basic Web site consists of simple HTML pagestext and some images.
These pages are generally fast to load, require minimal support from the server,
and are relatively easy and inexpensive to host. For example, many Web provid-
ers offer free Web space. Most developers try to hold Web pages down to about
50KB per page, including graphics. The goal is to keep download times to an ac-
ceptable level even for slow connections. These basic Web sites cannot interact
with the viewer. For example, you cannot accept form data or process credit cards.
Similarly, the Web pages cannot interact with your internal databases. Generally,
a single set of pages is made available to everyone, with little or no customization
for individual users.
Although these relatively simple pages are easy to create, it is diffcult to
change them and keep them up to date. All changes must be made individually,
and the developer must keep track of the details.
Consider the steps involved in creating a retail Web site with static HTML
pages. You write a page that describes each product, including price and photo.
You create a style for the site, adding fonts and colors. You can link the pages,
probably using some type of index or a start page that lists major categories. With
free hosting, it is rare to have a personalized search engine, so customers will
need an easy method to fnd the products you are selling. Figure 7.21 shows the
resulting hierarchical structure that you will have to follow. Each product will
ultimately be displayed on a separate page. Each of these pages must be created
individually, and links created by hand. To change any content, such as prices or
Main Web Page
Categories
...
Category 1
Product photo
... ...
Product 1
Description
Price
Photo
Product 2
Description
Price
Photo
Product 3
Description
Price
Photo
Product n
Description
Price
Photo
Category 2
Product photo
... ...
Category 3
Product photo
... ...
Figure 7.21
Retail website with simple pages. Each product will generally be on a separate page,
but each page must be created individually and linked by hand.
descriptions, you must go to the desired page and make the change. This method
may work for a small number of products, but as the number of products increas-
es, it will become diffcult for the developer to keep content and links up to date,
and virtually impossible for users to fnd anything.
The other challenge with this Web structure is that the static Web server will not
collect customer data, and because it cannot process form data, it is challenging
to process a sale. One method is to use e-mail, where the customer enters product
choices into a message that is e-mailed to your order-processing department.
The Buy Now or Buy Me button is one of the easiest methods to accept on-
line payments. Several companies, such as goemerchant and Google, offer these
checkout services. You set up a merchant account with one of these services and
place a couple of lines of HTML text on your page. When users click the dis-
played Buy Now button, the item is added to a shopping cart that is maintained
by the service provider. As shown in Figure 7.22, to purchase the items, the user
is taken to the checkout form run by the service provider. All credit card and pay-
ment data is handled by the providerthe merchant never sees the customers
fnancial information. This process makes it easier for the customer to trust the
merchant, and the merchant does not have to handle the security problems and
costs associated with collecting and storing credit card data. Because all of the
interaction and processing is handled by the service site, the merchant site can run
as simple HTML pages. Of course, this convenience carries a pricethe service
provider collects a fxed fee per transaction and a percentage of the sales revenue.
However, if you decided to handle your credit card processing by yourself, you
would still have to pay the credit card company and a card processor frm. So the
net cost to using Google is relatively low. Plus, you can get a service that does
Merchant Web site
Product
Description
Price
Buy Now
Shopping Cart
Item Price

Total
Check Out
Credit Card Data
Name
Address
Phone
Card Number
Submit
Card Processor Site
Customer
Notication
(Accept/Reject)
Notify merchant
Figure 7.22
Order processing with static pages using a third-party processor. Clicking a link
brings the customer to the card processor to enter card data and make the purchase.
not charge fxed monthly fees, enabling you to start small and pay only when you
actually have sales. However, these services tend to provide less vendor protec-
tion than you get by using a credit-card processor. Many seller complaints against
PayPal are listed on various Web sites with claims that PayPal does not support
vendors. It seems to be a risky PayPal strategy, because vendors are the ones who
pay the fees and choose the provider.
The main drawback to this method is the diffculty in updating the data, and the
related challenge for the customer in fnding specifc products. Because the pages
are static, you have to edit each page separately if you want to change data. And it
is diffcult to add search and flter conditions to help customers fnd specifc prod-
ucts. Static pages work when you have a limited number of products that tend to
remain the same and do not require customization.
Web Auctions
Some companies are interested in selling only a few individual items. If you want
to experiment with the Internet, have a few unique items, or need to clear out a
couple of items, it usually does not pay to set up a separate Web siteprimarily
because it would be hard to attract customers on a part-time basis. EBay.com is
the best-known general auction site. The system operates similarly to newspaper
classifed advertising, but interactively enables potential buyers to bid on items.
As highlighted in Figure 7.23, anyone can buy or sell products. The Web site lists
the products and tracks the bids. At the end of the bidding period, the seller con-
tacts the high bid and arranges shipment and payment.
One of the major diffculties with individual sales is authentication and ensur-
ing that the transaction is handled properly. The buyer runs the risk that the seller
is dishonest. If the seller ships the item before receiving payment, the buyer may
never pay up. Most sellers are unwilling to accept this risk, so they generally re-
quire the buyer to send payment before the item is shipped. However, the buyer
runs the risk of fraud. In 2000, eBay recognized the importance of this problem
and added the ability for sellers to accept credit cards from any buyer (called
Billpoint).
Uncertain prices
Can set reserve price
Good for unique items
Eciency depends on
Full information
Adequate number of
participants
Figure 7.23
Auctions. Good for unique items where you do not know what price to set. Economic
effciency depends on the number of participants and full information.
Several layers of pricing exist at auction sites. Generally, the seller must pay a
fee to list the item and a second fee once it sells. If the item does not sell, the seller
pays the frst fee. The fee amounts depend on the value of the item being sold and
on the options you choose. Read the fee schedules closely. For an average, fgure
$1 to list an item and about 3.5 percent of the value when it is sold. If you choose
the credit-card billing option, the seller will generally pay another 3.5 percent of
the selling price. As eBay has grown in size, the other auction sites (such as Ya-
hoo) have left the industry. Economic theory states that relying on a single provid-
er of auctions might affect the charges because of a lack of competition. However,
a single site makes it easier for buyers and sellers to fnd each other.
Amazon WebStore
Amazon.com was one of the early e-commerce sites to enable small merchants
to start Internet sales. As shown in Figure 7.24, the system (Amazon WebStore
which used to be called MarketPlace) enables merchants to sell items by setting a
fxed price. Listing items for sale is similar to the auction process. The seller pays
a fee for listing the item and a fee based on the selling price when the item is sold.
The fees change over time, but are loosely $1 per item plus 6 to 15 percent of the
items value. Amazon also imposes some limits on the prices you can charge for
various items. The company provides a bulk loading program to transfer all of the
product descriptions and prices.
Amazon is essentially a Web mall that supports a number of sellers. Since the
products show up in a regular search at Amazon, many customers do not know
they are dealing with a third-party supplier. Consequently, although the costs are
relatively, high, the sellers gain instant credibility from the Amazon name. Other
companies offer merchant mall services, but few have the name recognition and
customer base provided by Amazon. Amazon also provides a system for vendors
to create their own store sites.
Cameras
Description
Price
Checkout
Catalog
database
Search
Transaction Processing
Amazon.com handles credit.
Sends order info to merchant.
Merchant ships item to consumer.
Consumer
Product search.
Choose vendor.
Pay for item.
Vendor Transfer
Description
Price
Scanned image
Contact info
Figure 7.24
Amazon.com WebStore. Vendors list individual items. Consumers see WebStore as
the store and search for a product. Amazon.com can process the credit card-based
purchase, and the vendor ships the product to the consumer. Vendors pay fees for
listing items, selling them, and using the credit card processing.
Auction versus WebStore
If you wish to sell a small to medium number of items, then auctions and Amazon
are a good option. The transaction and shipping costs tend to rule out low-priced
items. On the other hand, buyers are somewhat leery of high-priced items from
unknown sellers. For example, most people might consider buying a rare coin
from a dealer but would be more reluctant to deal with an unknown independent
seller.
If you have a small number of intermediate-priced items to sell, should you
choose an auction site or Amazon? Auctions are particularly useful for unique
items or items where you are uncertain of the value. The WebStore system has
a good search engine and works well for small retailers who wish to maintain a
continued presence on the Web. Auctions present a slight uncertainty in the fnal
price. However, economic theory observes that an auction that is based on free in-
formation, and attracts all the relevant participants, will result in the highest price.
One computer manufacturer that traditionally sold only through distributor chan-
nels tested this theory in 1999 by offering a limited number of machines at an auc-
tion site. All of them sold for higher prices than could have been obtained through
traditional outlets. Keep in mind that if you have several products, you can always
try multiple outlets and test the response.
Web Commerce Servers
Transactions on Web sites are often handled by Web commerce servers. These
software programs provide all of the features needed to run a commercial Web
site. As shown in Figure 7.25, the site can be run by a specialty hosting company,
Web servers
Database
Commerce Server Shell
Web/Commerce Hosting Company
Customers
Merchants
Your Web site
Products
Shopping cart
Sales
Load database
Images
Descriptions
Prices
Customize site
Figure 7.25
Web commerce servers. The Web hosting company provides the servers, Internet
connection, and commerce software that processes transactions. The merchant loads
the database with product descriptions, images, and prices. The commerce software
provides a typical retail experience.
or you can run it on your own servers. The software is used by companies like
Amazon to host multiple sites. You can also lease it from specialty hosting frms
just to provide services for your company. The capabilities of the site depend on
the features available in the software. As a vendor, you should examine the fea-
tures of different site products to see which ones you need. Software systems are
available from several companies to run on your own servers or as a hosted ser-
vice. Dot Net Nuke is a popular open-source system for Windows-based servers
and Joomla is sometimes used on Linux-based servers.
The Web hosting company provides a server, a connection to the Internet, an
account for the hosting software, and credit-card processing. The Web commerce
server hosting software provides displays of your products, the ability to cus-
tomize your stores displays, a search engine for your products, and the ability to
process transactions securely. Some of the hosting software packages offer de-
tailed customization options, while others are more restrictive; but all of them
provide the basic elements you need to create and run a retail Web site. The Web
hosting company generally charges the retailer a setup fee, a fxed monthly fee,
and a transaction feeparticularly for processing credit cards. The fees are highly
variable and each company bases them on different parameters (such as number
of products versus number of transactions). The main strength of the Web com-
merce software is that you load product information into a database so it is easy to
update and easy for customers to search. The software contains all of the code to
build a relatively sophisticated Web siteyou simply customize it with your own
designs. Today, you can build many elements of a commerce server using compo-
nents produced by the open source community.
Content Management Systems
If you have a small business with a Web site that sells only a few items, it is
relatively easy to create and manage the Web site. On the other hand, you might
have hundreds of products and need to change descriptions, photos, or prices on a
regular basis. Or you might have thousands of pages of content that need to be up-
dated. For example, a service organization might include descriptions of services
or analysis of recent events. Keeping pages up to date can require the participation
of dozens or hundreds of workers to create and edit the content.
A catalog management system is designed to help you maintain a database
of products for sale on Web sites. You enter product descriptions and prices on
a local system and it updates the Web site catalog. The system has to recognize
and know how to communicate with your commerce server software. The main
strength of the catalog management system is that it focuses on the products and
provides tools to group them into categories and change prices. It can then auto-
mate the Web server updates.
Web sites that focus on text pages instead of products are often more compli-
cated. The pages can be written by hundreds of different employees. You want
all of the pages to follow a standard style and to ft into your Web sites structure
and search system. You cannot let hundreds of employees edit the Web site direct-
lyit will be next to impossible to maintain consistency and security. A content
management system is designed to make it easy for nonprogrammers to update
content, maintain a consistent style, and keep track of changes to a Web site. The
systems generally allow workers to create content using traditional editors such as
Microsoft Word, and then convert the text into HTML and store it in the appropri-
ate pages. They also track the revisions, so you know who changed each page and
when it was modifed. Most of the systems also maintain version control so you
can switch back to an earlier version of a page if something goes wrong.
Application Service Providers
An application service provider (ASP) is a Web-based business that provides
a specifc service to other businesses. The service is very specifc, and might or
might not involve interactions with customers. For example, as shown in Figure
7.26, one company provides online accounting services for small businesses. For
a monthly fee, you can enter all of your transaction data and generate standard re-
ports. Other ASPs act as intermediaries in providing services. For instance, a few
major companies provide Web-based real estate listing services. Other companies
provide online reservations services for service businesses.
If you can fnd an ASP that provides the services you need, it will generally be
easier and cheaper to use the services of the ASP than to create your own Web site.
Competition should eventually give you greater choices in price and quality. Even
if an ASP does not exist for the service you wish to provide, it might be possible to
convince a frm to adapt their products or create a new service.
The other way to look at ASPs is to fnd your own application service that you
can provide to other organizations. It takes time to create the application because
it is almost always custom programming. But, once the application is running, it
generates money on a regular basis. And you might be able to collect advertising
revenue from the sitewhich also rolls in every day.
Web Hosting Summary
As shown in Figure 7.27, once you know the type of e-commerce site you wish to
run, you need to select a hosting option. Today, most companies outsource their
Web sites to specialty data-center companies that have direct high-speed connec-
Businesses that lease the use of the application.
Store data.
Analyze data.
Facilitate company interaction.
Business Application
e.g., Accounting
Figure 7.26
Application service provider. The Web server runs an application that holds data for
other businesses. The data may be exclusive to one business (e.g., accounting); it may
be used to interact with other businesses (e.g., supplier auction site); or businesses
may interact with consumers (e.g., real estate).
tions to the Internet. But, you can choose from thousands of companies with many
different options. You need to begin the process by identifying the primary fea-
tures that you need. For example, if you need your own brand identity, you will
want your own URL, so Amazon WebStore might not be the best solution. If you
need custom programming, you will have to fnd a hosting company that provides
support for the languages and databases you need. If you need Web commerce
software, you will want to test the catalog management system to ensure it is easy
to use and compatible with your existing systems.
Web hosting companies typically charge fees based on the disk space used by
your site and by the monthly data transfers. The monthly data transfers become
one of the critical issues and tend to affect the price. If your site contains lots of
video and large photos, or if you expect millions of visitors, you will need to look
at hosting packages that provide the ability to transfer huge amounts of data each
month. The monthly data transfer number represents the cost of the bandwidth
that you are using. It is a diffcult number to estimate for startup frms. You can
try to estimate it based on the expected number of visitors, the average number of
pages viewed, and the size of the page in bytes. But the expected number of visi-
tors can be hard to estimate. Once the site has been operating for a few months,
you can get a better idea of the data rates. Until your site gets mentioned on a na-
tional radio or TV show and the number of visitors skyrockets.
Web hosting companies provide several levels of servicewith increasing
prices for each type of service. The simplest hosting consists of basic HTML pag-
es, with no database connectivity and limited programming. You might or might
not get your own name as a Web address; but you can probably fnd this level of
service at almost no cost. You can add more features to basic sites to add more
e-commerce options. For example, database connectivity makes it easier to store
product details. Programming support is useful if you need to customize the Web
site. Preprogrammed shopping carts and security certifcates are useful if you han-
dle payments yourself. Web commerce software simplifes the tasks of handling
product sales. Most companies provide these services on a shared serverwhere
Business Situation Hosting Options
Smallbusinesswithafewbasicitems. StaticHTMLwithaBuyNowbutton.
Uniqueitemsofuncertainvalue. eBayauction.
Many items but minimal confguration
issues.
Web commerce server hosted by third
party.
Manyuniqueitemsandmerchantidentity
isnotcritical.
Amazon WebStore
Uniqueservice. Custom programming, probably run on
ahostedserver.
Custom application with tight linkages to
in-houseapplicationsanddatabases.
Custom programming running on your
ownservers.
Figure 7.27
Web hosting options. Today, most companies host their Web applications on the
servers of specialty companies. Your job is to match the services you need for your
business.
hundreds or even thousands of other company sites are hosted on the same physi-
cal server that is owned and managed by the hosting company. Another option
is co-location, where you own the server hardware and install the software and
manage the system and security. Essentially, you just rent space and a high-speed
Internet connection from the hosting company. This option is useful if you are
concerned about security or interference from other Web sites, and have the per-
sonnel to manage the server and security issues.
Running your own Web server is a challenging task. The software and servers
are relatively easy to deal with. Security and monitoring are considerably more
complex. Plus, it is expensive to install high-speed Internet connections to your
premises. But, sometimes you have no choice and need the total control over your
applications. Before you take this step, be sure to hire experienced security ex-
perts. And keep in mind that even large frms like Yahoo host their servers at spe-
cialty sites. Google is an exceptionin addition to running its own Internet con-
nections, the company custom-builds its own servers.
Mobile Commerce
How do portable Internet connections (mobile phones) provide
new ways to sell things? In many ways, mobile commerce may not be
much different from the existing Internet. As shown in Figure 7.28, users have
cell phones and tablets with wireless connections to the Internet. However, for the
near future, these appliances will have small, even tiny, screens. With 3G wireless
networks, led by Verizon, the data transmission speeds have improved but are still
relatively slow. From a technical standpoint, companies will have to rebuild their
Web sites to support these devices. Each screen of data will have to be carefully
designed, most graphics will have to be downsized, and the page navigation will
need to be simplifed. Because the keyboards are hard to use, the Web sites will
have to reduce the amount of data to be entered. The most recent 4G services are
relatively fast (particularly LTE offered by Verizon) as long as congestion stays
low. Higher speeds lead to new possibilities, including video; which might open
more opportunities for sales and marketing.
The real beneft of wireless connections is that they create new opportunities. If
people are connected everywhere (or at least in the major cities), a vast amount of
Figure 7.28
M-commerce tools. As cell phones and tablet computers converge, people can
connect to any business every place they go.
HTC Evo Motorola Xoom Apple iPhone
information becomes available at any time. Stores could instantly provide detailed
product information to customer phones. For B2B sales, a salesperson could in-
stantly retrieve data on competitor products, build charts, and transfer the presen-
tation to the customers system. Payments could be handled with wireless trans-
fers of secure code, digital signatures, or even biometric data.
M-commerce will have an impact even on businesses that choose to stay with
traditional sales methods. Consider a customer shopping for a product at a re-
tail store. The customer can instantly get comparative price quotes from a dozen
other sources and make an informed decision about whether to purchase the item
electronically and wait for it to arrive, or to pay a premium to take the product
home immediately. Several companies offer comparative shopping sites that make
it easy for customers to check prices on their cell phones. With GPS chips, cell
phones also provide the ability to expand location-based applications, including
advertising. For example, Google maps enables companies to advertise business-
es, such as restaurants, that are displayed on a map when the customer searches
for an address. With a little additional effort, the restaurant could offer coupons or
make it easy for customers to make a reservation or check waiting times.
A newer use of cell phones is to tie shopping into social networks. Several
companies encourage shoppersparticularly womento share potential purchas-
es with friends. For example, try on a piece of clothing virtually and let friends
vote. Even basic tasks such as reviewing products and recent purchases for your
friends to see might lead to increased sales.
Taxes
When do consumers and businesses pay sales taxes on the In-
ternet? Sales taxes present a problem in the United States. Most states tax sales
of products and services to obtain revenue to pay for public services. Many mu-
nicipal governments add their own sales taxes to cover costs of local services.
Consequently, over 7,800 separate tax districts exist, with the possibility of sev-
eral thousandbased on each city. The system works reasonably well for small
stores located in one district. But businesses that operate in more than one district
must register and fle tax forms within each district.
If the sales tax rate were the only issue, the situation might not be too bad. The
diffculties multiply because each district has different categories of products and
different taxable items. An item taxable within one district might not be taxed
within a second. A third district might place the item in a different category and
impose different taxes.
The interesting legal aspect of the taxes is that they are defned as use taxes
on the citizensand are supposed to be paid by consumers regardless of where
the product was purchased. Since consumers are generally slow to volunteer pay-
ments, the states require businesses to collect the taxes and forward them to the
appropriate agencies. This situation causes problems when the business is locat-
ed in a different state. At various times, states have attempted to require out-of-
state frms to collect the taxes, but the U.S. Supreme Court has always ruled that
the U.S. Constitution clearly forbids the states from taxing interstate commerce.
The fact that it is a constitutional issue is important, because it would require a
constitutional amendment to change the situation. Congress has discussed creat-
ing a simpler tax system, but it is unlikely that it will pass as a constitutional
amendment.
Local merchants often complain about the diffculty of competing with out-
of-state frms that do not have to collect taxes. However, these same frms could
sell into other states, so the issue could be neutral. Also, consumers who want to
touch the product and bring it home immediately are still going to buy from local
merchants. More important, state and local governments are concerned about los-
ing their tax base. If consumers shift more of their purchases to e-commerce, the
states will lose substantial revenue. For instance, consider the fact that Dell is one
of the leading retailers of personal computersrelatively expensive items. Yet
Dell has a physical presence in only a few states, so they do not collect taxes for
most of the sales. This multi-billion-dollar industry represents hundreds of mil-
lions of dollars of annual tax revenue to the states. On the other side, some large
companies, such as Cabelas (Idaho Statesman May 8, 2007), have argued that
their mail/Internet companies are distinct subsidiaries from their physical stores;
and sales from the Web site remain nontaxable, even when a physical store is
opened in a new state. Ultimately, states will have to increase other taxes to com-
pensate for this lost revenue. Since most economists consider an income tax to be
more progressive than sales taxes, the effect is not all bad.
Sales taxes on services are an even trickier issue. Since a large portion of the
GDP is based on services instead of product sales, many states have begun taxing
services. In 2003, Congress allowed the federal moratorium on Internet taxes to
expire. States are now free to charge taxes on Internet connections. Just as you
now pay several dollars a month in taxes and fees for telephone service, you might
be asked to pay more for your Internet connection.
Global Economy
Does the Internet create a global marketplace? E-commerce has the
potential to open up the global economy. Theoretically, anyone with access to the
Internet can purchase products directly from anywhere in the world. However,
actual practice cannot live up to the expectations of theory. Three major issues
still limit international trade: (1) transportation costs, (2) national policies, and (3)
payment and trust limitations.
Transportation costs will always exist, but they can be relatively high for in-
dividual orders. International bulk shipments are considerably more economical,
Reality Bytes: How Do People Communicate?
Texting, or short-message service (SMS), grew to huge volumes in the 2000s.
In the second half of 2010 alone, U.S. cellphone users sent more than 1 trillion text
messages. Although the number seems huge, it actually represents a slowing of the
growth of text messages. Eelco Blok, CEO of Dutch provider Royal KPN NV, noted
that the companys youth-oriented brand saw an 8 percent decline in SMS messages
in the frst quarter of 2011. As the number of smart phones increases, vendors are
adding apps to send text messages over Internet connections, bypassing the phone
company SMS system. This distinction is critical to the mobile phone carriers.
AT&T and Verizon, the biggest carriers, charge $20 a month for unlimited texting.
This huge price provides at least an 80 percent proft margin, compared to 35 percent
for data or voice. If customers switch from texting, carriers will need to make up the
lost revenue through data prices.
Adapted from Anton Troianovski, Carriers Sweat as Texting Cools Off, The Wall
so there will always be an incentive for retailers to purchase in bulk and resell
individual items. Transportation companies consolidate shipments to reduce costs,
but customers often want products relatively quickly, and shipping by air is more
expensive than shipping by sea.
Nations have many different policies and taxes regarding imports and exports.
Most shipments have to go through a customs agent. Even digital products carry
restrictions. A few nations attempt to monitor and control all Internet usageto
the point of insisting that all Internet traffc be channeled through government
computers.
Think about the consequences if every nation imposed its will on Internet sales.
While individual nations do have the right to control sales within their territory,
it would destroy e-commerce if every nation imposed its control on all Internet
sales. As e-commerce increases in importance, more of these issues are going to
arise, and frms will need to have technology ready to handle them.
Global e-commerce is evolvingfrom many nations, you can buy English rug-
by jerseys from British vendors. But, the payment mechanisms need considerable
work. Many U.S. Web sites will not accept overseas shipments because the risk
is too great. Many companies will not sell to nations in eastern Europe, southeast
Asia, and the Middle East because the risks are even higher.
Cloud Computing
What are the costs for cloud computing? Many of the e-commerce
hosting methods rely on cloud computing. Today, most of the major providers of-
fer virtual machineswhere the hosting company runs massive server hardware
and hosts independent virtual machines that you can lease. In true cloud comput-
ing systems, you pay only for the machines when they are used and you can usu-
ally expand the capacities by adding more servers on demand. As shown in Figure
7.29, the basic cost structure can include: a fxed monthly fee, charges based on
processing, monthly data storage costs, data transfer in and out of the server, and
fees for databases or specialized software.
Amazon is one of the most advanced providers of cloud services. The company
offers three major services: virtual machines in the Elastic Cloud (EC2), data stor-
age with Simple Storage Service (S3), and a couple versions of database access.
Figure 7.29
Cloud computing cost structure. Most costs are variable and it is usually easy to scale
up the capacity. Costs will increase but they can be matched to increases in usage.
Fixed monthly fee
Cost per processing
Data storage costs
Data transfer in and out
Database/software fees
Examples:
Amazon: Elastic Cloud (EC2), Simple storage servive (S3)
Microsoft: Azure and SQL Azure
Rackspace
Equinix
Microsoft offers similar capabilities with Azure and SQL Azure. Amazon includes
calculators on the Web site to estimate monthly costs based on expected usage
rates. Data transfer costs are a large component of the total bill, but they typi-
cally increase only as the site usage increases. Presumably, increased use and data
transfer leads to more sales revenue to cover the costs.
Summary
E-business is a complex topic. On one hand, the Internet and mobile commerce
simply represent new ways to interact with customers and handle transactions. On
the other hand, they have the potential to change the economy and society. The
Internet continues to expand into more areas business and consumers lives. Busi-
nesses at all levels need to fnd new ways to connect to customers, while staying
on the right side of the thin line of privacywhich keeps moving.
Businesses have many ways to make money on the Web and with
mobile commerce, but ultimately, revenue comes from sales of prod-
ucts, sales of services, or advertising. Although anyone can make mon-
ey with advertising thanks to Google AdSense, currently a handful of
frms receive the bulk of Web advertising dollars. So, you should prob-
ably think about fnding ways to sell products and services.
Advertising on the Web is straightforwardbut also controlled
through a few primary sites. Google is the main advertising point on
the Web and it controls most of the keyword search ads as well as ban-
ner ads through DoubleClick.com. You should be able to use Googles
keyword tools to fnd keywords, estimate the bid prices for them, and
establish a daily advertising budget.
E-commerce sales can be analyzed in three phases: prepurchase,
purchase, and postpurchase. The prepurchase phase consists primar-
ily of advertising, providing specifcations, and product confguration
or selection. Purchase largely consists of handling the transactions, in-
cluding verifying the customer, protecting the data transmissions, and
handling the money transfer. Postpurchase support includes service,
problem tracking, and cross selling.
Developing and hosting Web sites is a relatively complex business.
Most companies choose to outsource these activities to specialists. How-
ever, it is relatively easy for small businesses to create a Web site and
begin selling items online. The Buy Now button approach is relatively
inexpensive and easy to experiment with. Web auctions are useful for
unique items, particularly when the value is uncertain. Amazon Web-
Store is a good alternative for small businesses that want to sell prod-
ucts online to an established set of customers, without the costs and
hassle of building a separate brand name. More complex sites, such
as Web services are still built with custom programming, but several
frms can help develop and host these new services.
Mobile commerce is similar to e-commerce, but the wireless capabili-
ties can provide some interesting applicationsparticularly for B2B e-
commerce. Sales taxes and the Internet are a challenging problem for
states. Ultimately, states will have to alter their tax systems. Similar
problems arise on a global scale. Many issues involving customer au-
thentication, payment validation, and national control need to be re-
solved before global e-commerce can seriously expand.
Key Words
Web Site References
Cloud Hosting
Amazon (EC2) aws.amazon.com/ec2/
Amazon (S3) aws.amazon.com/s3/
Microsoft Windows Azure www.microsoft.com/windowsazure
Microsoft SQL Azure www.microsoft.com/windowsazure/
sqlazure
IBM www.ibm.com/cloud-computing
Savvis (CenturyLink) www.savvis.com
RackSpace www.rackspace.com/cloud
A Managers View
Making money on the Internet is not easy, but it is possible. Internet sales
are increasing faster than sales through traditional channels. The challenge is
to use the Internet to increase your sales while still maintaining existing re-
lationships. New services are being created and the sales of digital products
are expanding. M-commerce offers the potential to provide another shift in
business and societybut it will be a while before the hardware, networks,
and software become widespread enough to make a difference. You need
to understand how customers use Web sites to purchase products. Attract-
ing customers or making money through advertising requires a thorough un-
derstanding of Googles keyword sales method. Developing basic sites and
fnding hosting companies for them is relatively easy. Creating custom ser-
vice applications is still diffcult. Choosing the right hosting approach is a
key step in getting online with an affordable system.
application service provider (ASP)
auction
bill presentation and payment
business to business (B2B)
business-to-consumer (B2C)
catalog management system
click-through rate
co-location
commerce server
content management system
digital rights management (DRM)
disintermediation
e-business
intellectual property
near feld communication (NFC)
search engine optimization (SEO)
simple object access protocol (SOAP)
social networking
static HTML
Web 2.0
Review Questions
1. What types of products are easy to buy online and which ones are diffcult to
sell solely online?
2. What are the potential benefts and costs to disintermediation that can be
accomplished with e-business?
3. How does B2B differ from B2C e-commerce?
4. How does EC differ in the three areas of prepurchase, purchase, and
postpurchase?
5. How do Web 2.0 sites make money?
6. What are the primary payment risks in e-business and how do frms reduce
the risks?
7. What options are available for promoting a Web site?
8. What choices do advertisers need to make when dealing with keyword ads?
9. What options are available for building and hosting Web sites?
10. How do people use cell phones when making purchases?
11. Many people were concerned that by not requiring EC frms to collect sales
taxes, traditional frms would eventually lose business and the states would
suffer large declines in tax collections. Why did this scenario not happen?
Might it still happen in the future?
12. What problems make it diffcult for EC to be global?
Exercises
1. Choose a non-digital item that can be purchased online. Find at least three
sources for it and compare the prices. How can different prices exist for the
same item?
2. Using the Web or by talking to retailers, fnd a company that only sells
through retailers. Or, fnd a company that sells direct, but only at list price.
What characteristics of the product and industry support or encourage this
choice?
3. How can the Web site Twitter make a proft?
4. Choose an industry and assume you are running a Web site for a company
in that industry. Select keywords that you might use to advertise your Web
site. Estimate the amount you would be willing to bid for Google ads on the
keywords and specify a budget. Hint: How many people who click-through
an ad will become customers?
5. Select a Web site and identify the prepurchase, purchase, and postpurchase
elements.
6. Check with a Web advertising site and identify the cost to run a banner ad on
the site. Specify the type of ad and the number of times to run.
7. Obtain the fnancial statements from a publicly traded social network site and
identify the major sources of revenue and costs.
8. Research PayPal from the perspective of a vendor and briefy describe what
happens if a consumer claims that they never made a purchase that has
already been shipped.
9. What percentage of apps in the mobile market are for games, books, or
entertainment?
Technology Toolbox
10. Select one of the major Web server tools and identify how it handles multiple
languages.
11. Find a reference that identifes the percentage of sites based on Microsoft
Web servers.
12. Check your schools course catalog or talk with instructors in computer
science and MIS to see which Web technologies are taught at your school.
Also check a local community college catalog.
13. Use Dice.com to compare the number of jobs available for the three main
Web technologies.
14. Find current costs for processing credit cards through (a) PayPal, (b) Google,
(c) an independent merchant processor.
15. Find the current VISA merchant security requirements if your Web site
handles credit card data directly. Estimate the costs of complying with the
requirements if you are a small merchant.
Teamwork
16. Search for a common product such as a camera and choose three Web sites
with at least one of them from a small company. Compare the three sites in
terms of their support for purchase, pre-purchase, and post-purchase support;
and their use of social networks.
17. Have each person choose a different item for sale on eBay that has similar
items for sale at retail stores or other Web sites. Compare the fnal eBay price
to the other prices. Summarize the differences for the team.
18. Assume that a friend wants to start an online business selling jewelry. Have
each person fnd a hosting company that could handle the site. Identify the
costs and tools provided by each company. Share the results and choose one
to recommend to your friend. Briefy explain the basis of the decision.
19. Assume the team wants to create a new cell phone app. Find estimated data
on prices and sales for existing apps. Estimate the costs of developing and
marketing an app and use the data to estimate potential profts.
20. Choose a Web site. Evaluate all aspects of the site and write a plan to suggest
how it can be improved.
21. Identify at least three areas in which Rolling Thunder Bicycles could proft
from e-business. Be specifc, and explain what technologies would have to be
added (for instance, Web hosting).
22. Find at least fve sites on which it would make sense to advertise Rolling
Thunder Bicycles. As much as possible, identify the advertising costs and the
demographics of the site visitors.
23. Develop a plan for expanding Rolling Thunder Bicycles into international
sales. Be sure to identify any potential problems, and discuss how you will
deal with them.
24. Develop a plan for creating a Web-based system for connecting to suppliers.
What software would you need? How can you convince the suppliers to
cooperate?
25. The management of Rolling Thunder Bicycles cannot decide on a Web
strategy. For the three main approaches (simple HTML, auctions, and
commerce server), list the primary strengths, weaknesses, and costs as they
apply to this company. Make a recommendation and briefy explain your
choice.
Additional Reading
Borzo, Jeanette. Online Micropayment Systems See New Interest but Face Old
Hurdles, The Wall Street Journal, April 3, 2001. [Decline of e-commerce
advertising revenue and increase in selling of content.]
CNN, Study: Four sites account for half of Web surfng, June 5, 2001.
[Dominance of a few large frms in attracting users.]
Collett, Stacy. Amid turf battle, some middlemen call for truce with online
rivals, Computerworld, April 16, 2001. [Discussion of role of middlemen
and disintermediation in e-commerce.]
Fireswick, Kris. The e-fles, CFO, February 2001. [Comments and analysis of
some lead managers of failed dot-coms.]
Matthews, Robert Guy. Web Firm Lures Steel Giants, The Wall Street Journal,
September 2, 1999. [Early auction site for the steel industry.]
Moss Kanter, Rosabeth, The Ten Deadly Mistakes of Wanna-Dots, Harvard
Business Review, January 2001. [Discussion of the e-Commerce failures.]
National Mail Order Association, 1998 Mail Order Sales Results, 1999, http://
www.nmoa.com/Library/1998sale.htm. [Statistics on mail order sales.]
Regan, Keith. Study: Era of E-Commerce Profts Underway, EcommerceTimes,
June 20, 2001, http://www.newsfactor.com/perl/story/11381.html.
[Summary of McKinsey study on e-commerce profts.]
Shopping around the web, The Economist, February 26, 2000. [Special section
on e-commerce.]
Stone, Martha L., U.S. cell phone technology lags Japan,Europe, mediainfo, June
7, 2000,
http://www.mediainfo.com/ephome/news/newshtm/stories/060700n2.
htm. [Statistics on cell phone usage in several nations.]
Swisher, Kara. Web Retailers Faced Death; Now Can They Handle Taxes? The
Wall Street Journal, April 9, 2001. [Summary of issues on e-commerce sales
taxes.]
Cases: Retail Sales
Te Industry
The retail sales industry includes a broad variety of products, but regardless of
the product, several fundamental issues apply. You need to remember that retail-
ers play three critical roles in selling products. The frst is locationretail stores
succeed by displaying and delivering products where the customers are located.
The second is customer service. Many customers need help choosing products
and need reassuranceparticularly when buying large items. Retail stores also
0
50
100
150
200
250
300
350
400
450
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
B
i
l
l
i
o
n

$
Annual Revenue
Wal-Mart
Sears
SuperValu
Amazon
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
R
a
t
i
o
Wal-Mart
Sears
SuperValu
Amazon
handle returns and take care of any problems that might arise with a product. The
third major role played by retailers is to spread the sales risk. Generally, retailers
purchase items in bulk. If the items do not sell, the retailer marks down the prices
until they are sold. But, the retailer decides which items to carry and what fnal
price to charge, so the retailer takes the loss if products do not sell well. These
three services are important to manufacturers. As part of the economics of special-
ization, manufacturers can concentrate on building products, retailers can focus
on customer service, and both frms come out ahead.
Transactions
Tracking sales of individual items and recording transactions are critical factors
in retail sales. Originally, retail frms tracked fnal sales only for themselves. They
used the information to help decide when to reorder and when to cut prices. Since
the retailer sits in the middle between suppliers and customers, the retailer par-
ticipates in several types of transactions. Consequently, retailers built important
accounting and inventory systems. Recording transactions with customers is rela-
tively easy in theory. The challenge is to keep the costs as low as possible, which
is why retailers continually experiment with technology at the checkouts. The re-
tail sales transactions also have to be accurate and secure. Because retailers are
generally responsible for collecting sales taxes, the systems also have to maintain
adequate audit trails for the state examiners. Retailers also have to track purchases
from suppliers. The actual purchase is not that diffcult to track, but they also have
to monitor shipping and fnancing costs. Moreover, there are always special or-
ders, and someone has to match received items to orders, monitor quality, and
often track down missing or incorrect orders. Increasingly, retail stores also have
to build transaction systems to handle the fnancial aspects as well. As more cus-
tomers switch to digital payment systems (debit cards), retailers have to work with
banks to validate card numbers, verify account balances, and transfer the transac-
tion totals. The detailed fnancial system must also be able to go back and provide
documentation for possible fraudulent transactions.
Today, led by Wal-Mart, the emphasis is on the entire supply chain. A key aspect
to reducing costs and improving the proft margin is to minimize the items held in
inventory. Or, in retailer accounting language, maximizing inventory turn (over).
The ultimate goal might be a just-in-time replenishment system: just as a custom-
er buys an item from the shelf, a new one shows up at the loading dock. Of course,
this perfect system is almost impossible. But the proftable stores get closer every
year. One key is the ability to forecast exactly which products customers are go-
ing to buy, and then build a supply chain so that the right products are delivered
on time. Transaction information is a key element in supply chain management.
Retailers share their daily sales data with manufacturers, so manufacturers can
forecast demand and fne-tune their delivery times.
Supply chains often suffer from what is called the bullwhip effect. A small in-
crease in sales causes retailers to forecast an increase in demand and bulk order
more of the product from their distributors. If the increase arises from several
stores, the distributor also sees an increase in demand and orders more products.
In the meantime, the stores perceive a shortage that is not being flled by the
distributor, so they increase their orders even furtherto hoard the product and
ensure deliveries. The manufacturer sees a giant increase in demand and starts
cranking up capacity. The initial small increase in demand has been magnifed as
it moves up the supply chain. Ultimately, the manufacturer over-compensates and
the supply increases too far, raising costs for everyone and depressing prices as
the retailers mark down the item to get rid of it. Of course, the fuctuating prices
send the wrong signal to customers and can start the cycle all over again. One way
to minimize the bullwhip effect is to make sure that all participants have access to
the original data on customer demand and that everyone uses the same forecast.
Retailers generally support this approach and share their sales data with manufac-
turers (Hugos 2003).
Of course, a large retail store carries tens of thousands of products and has a
few thousand suppliers. So, the retail store needs to be able to integrate with thou-
sands of different systems, or at least provide data in some common format. The
massive volume of data also presents problems for storage, access, and transpor-
tation across networks. Ultimately, the stores that succeed in handling these data
volumes at the least cost are the ones that have prospered.
Additional Reading
Hugos, Michael, Essentials of Supply Chain Management [book excerpt],
Case: Wal-Mart
Wal-Mart (ticker: WMT) represents a classic business technology case. Empha-
sizing customer service and low prices, the company pushes its suppliers hard.
Founder Sam Walton realized early that computers would play an important role
in these goals. In the mid-1960s, computers were used to handle basic accounting
functions. Counting inventory was still painful because the tools were not avail-
able. It was not until 1983 that item-level bar codes were available on the majority
of products. Each product was assigned an identifer or stockkeeping unit (SKU)
and a standardized bar code. At that point, Wal-Mart installed point-of-sale ter-
minals and was able to track inventory based on the sales. In 1987, the company
used a satellite system to link all of the stores to company headquarters. The sys-
tem enabled the company to compile real-time inventory data (Dateline 2002).
To reduce the inventory levels, Wal-Mart knew it needed to enlist the cooperation
of its suppliers, so the company instituted the collaborative planning, forecasting,
and replenishment (CPFR) program to share data with its suppliers. Pete Abell, re-
tail research director at the AMR Research consulting frm, notes that Wal-Marts
margins can be far lower than other retailers because they have such an effcient
supply chain. He estimates that the company pays 5 to 10 percent less than its
competitors for products (Johnson 2002). The suppliers are willing to offer better
prices to Wal-Mart because sharing the supply chain management reduces their
costs as well. And it does not hurt that Wal-Mart is the biggest company in the
world with almost $250 billion in annual sales and more than 1.2 million employ-
ees (www.fortune.com global 500).
As Americas biggest employer and driven to be a price-leader, Wal-Mart seeks
new ways to reduce labor costs. In 2007, the company implemented a new com-
puter scheduling system that evaluates historical patterns and slots employees to
cover the busy times and then sends them home during periods with fewer cus-
tomers in the store. In the process, employees no longer have set schedule shifts.
Instead, they pick up a few hours each day. The system will also improve com-
pliance with federal wage and hour lawsperhaps reducing the lawsuits against
Wal-Mart. (Maher 2007).
Wal-Mart led the industry in improving its supply chain to reduce costs. Its in-
credibly effcient distribution centers are one part of the story. Bulk shipments
arrive from vendors at one side of the distribution center. For example, a large
truck might be full of identical boxes of laundry detergent. The contents of the
truck are unloaded and placed on a conveyor belt. The products are routed through
the distribution center to the other side where a mixed load of multiple items is
placed into another truck. This load might contain some laundry detergent, toys,
and clothes. The mixed load is then taken to a specifc store. The computer-con-
trolled distribution center knows exactly which items need to be shipped to each
store and loads the trucks accordingly. The system runs by reading the individual
bar codes and shifting products to different belts.
In 2003, Wal-Mart shocked many people when it announced that its top 100
suppliers would have to switch to radio frequency identifcation (RFID) chips in
2005. Initially, the RFID tags would apply only to pallet loads of products. The
tags provide some signifcant advantages over bar codes in the warehouse. They
can be read from farther away and do not require perfect line of sight. Multiple
tags can be read at the same time (if boxes are stacked together). They are less
susceptible to damage. An important issue is that they can contain more data. This
last factor is important because it would enable Wal-Mart to electronically record
and track the receipt of individual shipments. Bar codes allow it to track only the
specifc product. A person has to manually match the shipment with the specifc
order. The main drawback to the RFID approach is cost. Bar codes are essentially
free, since they can simply be printed on the box or a label. RFID tags in 2003
cost around 50 cents each. Wal-Mart is hoping that by forcing suppliers to adopt
the tags, the bulk demand will convince manufacturers to expand capacity and re-
duce prices through economies of scale. Wal-Marts target is 5 cents per tag, and it
anticipates that about 1 billion tags will be needed initially (Vijayan 2003).
It is risky to be on the leading edge of technology. Standards are not yet in
place and costs are relatively high. The lack of standards is a substantial problem
in the international marketplace. As of mid-2004, many Chinese manufacturers
were waiting for an RFID standard to be approved before agreeing to use RFID
tags. Consequently, some companies might end up having to add the RFID tags
after the product reaches the United States (Sullivan June 2004). Several groups
have also raised concerns about privacy issues if RFID tags are moved to indi-
vidual products. Although the tags can be read from only a limited range of a few
feet, some people expressed concerns about stores being able to track people by
tags placed in their clothes. And if not enough suppliers and distributors install the
hardware and software to take advantage of the tags, the entire experiment could
be an expensive trial. Some experts were suggesting that as many as 50 percent
of these trials would fail. Nonetheless, Linda Dillman, CIO of Wal-Mart, observed
that the retailer believed it needed to stay on the leading edge of technology. It
was important to try technologies that might help, even though the technology
eventually might not be useful (Sullivan July 2004). Initial trials with Kimberly-
Clark were successful. In April 2004, the company shipped pallets of its Scott pa-
per towels with RFID tags through Wal-Marts scanners. But Kimberly-Clark has
been working toward the RFID system since 2001 (Bacheldor 2004).
By early 2006, more than 300 suppliers were using RFID tags on products
shipped through fve distribution centers to 500 Wal-Mart stores. Carolyn Walton,
VP of information systems noted that out-of-stock items with RFID tags were
replenished three times faster than before and the amount of out-of-stock items
requiring manual intervention dropped by 10 percent. We see this as a continual
process in an effort to focus on bringing our customers what they want (Songini
2006). The company intends to push the use of RFID tags through all of its pro-
cesses. A typical truck delivery carries 7,000 boxes that have to be unpacked and
organized. Wal-Mart plans to provide employees with wearable devices that can
identify high-priority boxes that need to be unloaded immediately.
Collecting RFID data internally solves some problems, but Wal-Mart is push-
ing for tighter integration of the supply chain by asking suppliers to share data
on the Global Data Synchronization Network (GDSN). The GDSN system builds
pools of XML-based data for specifc industries and stores. Supplies and mer-
chants upload data and synchronize their own systems so everyone has the current
data (Songini 2007 and www.gs1.org).
In 2010, Wal-Mart rolled out RFID tags to track individual clothing items. Ini-
tially, the clothing items will contain removable smart tags. Raul Vazquez, the
Wal-Mart executive for the Western U.S. noted that This ability to wave the
wand and have a sense of all the products that are on the foor or in the back room
in seconds is something that we feel can really transform our business. Wal-Mart
also expects the technology to cut down on employee theft because it will be eas-
ier to see which items are missing. Wal-Mart subsidized some of the costs to the
suppliers to ensure adoption by its major suppliers (Bustillo 2010).
Tighter Integration with Manufacturers
Procter & Gamble sells $40 billion of household products every year. You know
most of their brand names. Despite the use of information technology and elec-
tronic ordering systems, Steve David, the CIO, pointed out that returns from retail
stores cost P&G $50 million a year. Problems arise from incorrect orders, dam-
aged products, and inaccurate shipments (Stahl 2003).
How can so many problems exist in an electronic system? One of the problems
is that every company has a different description and ID number for every product.
For example, P&G might have an internal ItemID for a tube of Crest toothpaste.
But Wal-Mart will assign a different SKU to each specifc size and favor. The
Wal-Mart database might even have a different description for the product than
that used by Crest. Both companies probably have even more numbers to describe
bulk orders or boxes of toothpaste. Relying on people to match the descriptions or
to fnd items in a catalog or Web site is time consuming and easily causes errors.
And manufacturers tend to continually introduce variations or new products and
change the numbers. The consulting frm A.T. Kearney estimates that 30 percent
of the information in catalogs is incorrect at any point in time (Konicki 2002).
P&G alone sells around 60,000 different items.
One answer is to ensure that everyone uses a single standard description and
ID of every product. But you have to get all of the manufacturers and retailers
to agree to use the system, and someone has to pay for its maintenance. The cur-
rent answer is the UCCnet registry. It is a not-for-proft subsidiary of the Uniform
Code Council (UCC). Ultimately, it makes money by charging manufacturers and
retailers an annual fee to use the system (www.uccnet.org). Each item listing con-
tains 62 pieces of product data. By placing it in a single location, all merchants
can synchronize their databases to the manufacturers standard. Errors are reduced
when everyone works from the same clean database (Meehan 2002).
With more accurate data, Wal-Mart can rely on EDI to place most of its or-
dersreducing the costs and time of having employees manually enter each or-
der. EDI has been around since the 1980s, but it carried a high cost because most
of the transactions were on private VAN networks that are expensive. In 2003,
Wal-Mart began encouraging suppliers to switch to an Internet-based EDI system.
To ensure security, most merchants are using the electronic data interchange In-
ternet integration applicability statement 2 (AS2) standard. AS2 provides security
through certifcate-based encryption. Merchants and producers can use the system
either through a Web browser or through low-cost client tools that will connect to
their ERP systems (Scheier 2003).
Competition
Competition in discount retail is stiff. Wal-Mart has literally grabbed most of the
market from other retailers. For comparison, look to Kmartwhich was domi-
nant through much of the nation before Wal-Mart. Because the company did not
have the supply chain management systems, and because it did not have the data
to forecast customer demand, its stores rarely had the products in stock that cus-
tomers wanted. Ultimately, it fled for bankruptcy protection because it could not
compete against Wal-Mart and Target. In 2000, the Kmart CEO announced that
he would spend $1.4 billion on technologymore than the company had spent
in an entire decade. The company burned through fve CIOs in seven years and
went without one for almost two years. The company also had to take a $130 mil-
lion write-off of supply chain software and hardware that did not work properly.
Although spending the money earlier on technology might not have saved Kmart,
it is clear that it has provided a competitive advantage to Wal-Mart (Sliwa 2002).
Smaller stores have found ways to use technology to compete with Wal-Mart.
Smart & Final, a California-based chain with 225 stores has installed in-store
technology to display prices and handle all data processing. The system analyzes
demand and can adjust product prices on-the-fy without the need to print new
labels or reprogram POS terminals. It also provides immediate sales data to sup-
pliers with real-time access to inventory. Bob Graham, vice president of stores
technology noted that using technology like this is the best way to go up against
a particular big company [Wal-Mart] that is taking over the world (Hall 2004).
In 2010, Wal-Mart began experimenting with a new technology within its
Sams Clubpersonalized discounts for individual customers. The eValues pro-
gram runs primarily from a kiosk inside the store. Customers log into the kiosk
and the system prints a set of coupons specifcally aimed at that customer with
discounts on individual products. The members can also get e-mails and view
coupons on the Web site. The system analyzes prior member purchases to predict
what products each customer might buy and how much discount should be offered
(Martin 2010).
Privacy
Wal-Mart collects a huge amount of transaction data, data that includes detailed
information on everything customers purchase. Unless payment is made in cash,
the Wal-Mart databases also identify the individual purchaser. This data can pro-
vide useful insights to manufacturers and marketing companies. In 2001, three
retailers (CVS, Kmart, and Wal-Mart) decided to pool their pharmacy sales data
and sell up-to-the-minute numbers to drug manufacturersin exchange for a few
million dollars in fees. However, Wal-Mart also made a decision to stop selling
general sales data to market research companies. A Wal-Mart spokesman notes
that our competitors were getting more out of the third-party aggregation than we
were, so it made more sense for us to stop (Rendleman 2001).
Financial Transactions
On March 31, 2004, Wal-Mart spotted a problem: 800,000 shoppers were over-
charged on their credit and debit cards. The transactions were double- or triple-
billed due to a hardware error. First Data Corporation, which clears fnancial
transactions for Wal-Mart, noticed the problem when reviewing one of its qual-
ity control logs. The company reversed the duplicate transactions, but Wal-Mart
put up notices to customers to double-check their statements (DAmbrosio 2004).
First Data blamed the error on a computer glitch, but did not provide details.
In 2010, Wal-Mart announced that it would replace all of its payment terminals
in the U.S. with a version that supports smartcard technology. Smartcards contain
a chip to provide additional information, including a PIN; and they are commonly
used in Europe. To date, U.S. banks and other vendors have resisted switching to
the smartcards; largely because of the costs (Vijayan 2010). It is not known if the
terminals will also support near-feld communications (NFC); which are likely to
be the primary method adopted for cell phone payments.
Online
Wal-Mart has several online systems in place, but they provide only a tiny fraction
of sales. Yet the site does get traffc. A test of the site for the 2003 holiday season
showed that it was being used somewhat over its capacitymaking it sluggish.
The site was ranked19th of 20 Web sites. Consumers had to wait an average of 54
seconds to complete an online transaction (Rosencrance 2003).
Somewhat surprisingly, in 2003, Wal-Mart introduced an online music ser-
vice that charges 88 cents per song11 cents less than the leading site by Apple.
Songs are being provided by Anderson Merchandisers, which acquired Liquid Au-
dio, one of the innovators in digital music. With a somewhat limited selection of
songs, it remains to be seen whether customers will use the service (Bartels 2003).
Wal-Mart also tried to compete with NetFlix and offer DVD rentals by mail for a
monthly fee. That operation was folded after a year of weak demand.
Over time, Wal-Mart recognized that its Web site was not an effective method
of selling products. Linda Dillman, executive vice president, stated at the end of
2005 that weve changed our dot-com presence. The company now views the
Web site as a marketing tool. Eventually, the retailer hopes to add online software
to help customers organize their data and prepare for trips to the store, particularly
in terms of the pharmacy and health care areas (King 2005).
Moving in a slightly different direction for Wal-Mart, its Sams Club division
started selling a digital health records system to physicians in 2010. The package
deal bundled Dell computers with software from eClinicalWorks. The package
includes software, maintenance and training and is priced based on the number of
doctors in a practice (Lohr 2009).
IT Development
Unlike many of todays large companies, Wal-Mart tends to develop most of its
own software. The company is leery of commercial packages because it does not
want to be held hostage by the software vendor. The CIO of Wal-Mart also em-
phasizes that developers have to understand the users tasks before creating soft-
ware. Just developing something does not mean it will actually be used. Conse-
quently, before the IT staff creates and deploys an application, developers have to
work in the real job. For example, if someone is going to rewrite a point-of-sale
application, he or she has to work for a couple of days as a cashier. That way,
developers learn what issues are important and what problems need to be solved
(Schrage 2003).
Questions
1. Why are some people upset about RFID tags? Are their fears justifed?
2. What does the UCCnet project do and why is it so important to retailers?
3. Why did Wal-Mart stop selling some transaction data and will it affect the
company in the future?
4. What is the role of Wal-Marts Web site? Is it a useful tool?
5. Should Wal-Mart continue to rely on developing its own software?
6. What are the benefts and costs to having developers work in a functional
area before creating new applications?
Additional Reading
Bacheldor, Beth, Kimberly-Clark On RFID Trial: So Far, So Good,
Information Week, May 17, 2004.
Bartels, Chuck, Wal-Mart Begins Testing Online Music Service, Information
Week, December 18, 2003.
Computerworld, Dateline: A Timeline Of Technology Advances At Wal-Mart,
September 30, 2002.
Bustillo, Miguel, Wal-Mart Radio Tags to Track Clothing, The Wall Street
DAmbrosio, Dan, 800,000 Cards Overcharged At Wal-Mart, Information
Week, April 5, 2004.
Hall, Mark, Retailer uses Technology to Defend Against Wal-Mart,
Johnson, Amy Helen, A New Supply Chain Forged, Computerworld,
September 30, 2002.
King, Julia, Wal-Mart to Add 250 IT Jobs, Expand Online Presence,
Computerworld, December 2005.
Konicki, Steve, Shopping For Savings, Information Week, July 1, 2002.
Lohr, Steve, Wal-Mart Plans to Market Digital Health Records System, The
New York Times, March 10, 2009.
Maher, Kris, Wal-Mart Seeks New Flexibility in Worker Shifts, The Wall Street
Martin, Andrew, Sams Club Personalizes Discounts for Buyers, The New
York Times, May 31, 2010.
Meehan, Michael, UCCnets Promise: Synchronized Product Data,
Rendleman, John, Customer Data Means Money, Information Week, August 20,
2001.
Rosencrance, Linda, Report: Some Retail Sites Strained By Online Shopping
Surge, Computerworld, December 3, 2003.
Scheier, Robert L., Internet EDI Grows Up, Computerworld, January 20, 2003.
Schrage, Michael, Dont Trust Your Code To Strangers, Computerworld,
September 18, 2003.
Sliwa, Carol, IT Diffculties Help Take Kmart Down, Computerworld, January
28, 2002.
Songini, Marc, Wal-Mart Details its RFID Journey, Computerworld, March 2,
2006.
Songini, Marc, Wal-Mart Pushing Data Synchronization Network,
Stahl, Stephanie, Editors Note: Effciency Leads To Lower Costs, Information
Sullivan, Laurie, China RFID Standards Are Eagerly Awaited By U.S.
Manufacturers, Information Week, June 21, 2004.
Sullivan, Laurie, RFID Will Go Through Growing Pains, Information Week,
July 2, 2004.
Vijayan, Jaikumar and Bob Brewin, Wal-Mart To Deploy Radio ID Tags For
Supply Tracking, Computerworld, June 12, 2003.
Vijayan, Jaikumar, Wal-Mart to Support Smartcard Payments, Computerworld,
May 21, 2010.
Case: Sears, Roebuck and Co. and Lands End
As one of the earliest mass merchandise stores, Sears (ticker: S) has a major place
in the history of business in the United States. Richard Sears opened his frst store
in Chicago in 1887 selling watches and jewelry. In 1896, he began shipping the
general catalog that made the company famous. In 1993, the company stopped
distributing the general catalog because of the production and mailing costs. Its
Craftsman tools and Kenmore kitchen appliances are household icons. In the
1980s and 1990s, Sears acquired several frms that were unrelated to its primary
mission. For example, in 1981 it acquired the brokerage frm Dean Witter as well
as the Coldwell Banker real estate company. By 2003, the company had divested
all of those frms to concentrate on sales (www.searsarchive.com).
Gary Comer founded Lands End in Chicago in 1963 with a couple of friends to
sell yachting hardware. Over time, the company expanded into luggage, moved to
rural Dodgeville, Wisconsin, and took off with direct sales of clothing through its
catalogs, telephone operators, and the best guarantee in the business (www.land-
send.com). In 2001, Lands End had revenues of about $1.6 billion compared with
$41 billion for Sears (Weiss 2002). In 2002, Sears purchased Lands End for $1.9
billion. However, as a wholly-owned subsidiary, the company remains relatively
independenteven down to its information technology systems. The merger has
had some interesting effectspartly of the success of Lands End and the number
of Lands End executives who gained power at Sears.
Lands End used technology to boost sales. In 2001, the company noticed that
it was losing sales because products were not in stock. So, in 2002, the company
built an inventory-management workbench. The data warehouse system uses an
analytic engine to monitor sales and create reports. It automatically alerts sales
managers when popular items need to be restocked. The system links to seven-
year sales histories in the Lands End IBM DB2 database. It even contains re-
gional weather data so it can determine why sales of raincoats or parkas increased
in an area. With its success, the company created another workbench to help with
business-to-business sales, and a third one to help schedule workers in the ware-
houses (Whiting 2003). In addition, the Lands End Web site introduced several
innovative features.
Online
As a direct merchant, Lands End presented several innovations. One of its more
powerful tools is its virtual model. Customers enter some basic measurement
data into the Web site and the system draws a 3-D model to match their body
type. From that point, customers can have the model display combinations of the
clothes to see how they might ft. Extending on its telephone-based personal ser-
vices, the Web site also has a personal shopper service. When new styles are re-
leased, the personal shoppers recommend combinations to shoppers based on their
preferences and prior purchases.
Lands End built its company by establishing strong relationships with suppli-
ers around the world. These ties enabled the company to offer custom clothing
through its Web site. In 2001, the company began offering custom Chino pants for
$54 online, compared with standard prices of $30-$40. By 2004, the company was
offering custom dress shirts or blouses, dress pants, and jeans as well. In addition
to basic measurements, the Web site asks customers to choose among basic body
types. The company then has the shirts or pants built to the specifcations in a few
days. The customization feature is useful because the top reason for returns is that
clothing does not ft properly (Swanson 2001).
Lands End quickly learned an important lesson about Internet sales: you still
have to provide good customer service, so the Web interface does not necessarily
save money. On the other hand, the level of service provided by Lands End is
amazing and generally exceeds that provided by most merchants. For example, a
customer called to buy clothes for his wife for Christmas. He had already ordered
a charcoal-gray blazer and slacks. He wanted to know if a skirt on a different page
would match. The customer service representative put him on hold and quickly
called a personal shopper, who pulled both garments from a rack and compared
them to see that they did not match. In 96 seconds, the service representative was
back on the phone with the caller with a suggestion for a different skirt that did
match and a decision to ship the skirt at no extra shipping costs, since the frst
items were ordered a scant 12 hours earlier. Although the order was placed over
the phone, the same level of service is available online. Bill Bass, vice president of
electronic commerce, notes that one of the great fallacies of the Internet is [that]
youll save on customer service costs because customers [will] serve themselves.
But reducing service costs is not really an important issue to Lands End. Instead,
the company would like to reduce the cost of printing and mailing its 250 million
catalogswhich accounts for 41 percent of its operating costs (King 1999).
The site also offers the ability to shop with a friend. A customer shares a Web
site with a friend in a different location. The two see the same products and can
communicate via a chat session. To Lands End, it is all part of customer service.
One of the tools that Lands End needed help with is the search engine. Ac-
tually, few Web sites have had good search engines. Although companies like
Google make their search engines available to commercial sites, they are designed
to search static Web pages and not product databases. Specialty search companies
such as EasyAsk and Endeca have created search engines that enable customers
to ask questions in natural languages (e.g., English). The system then searches the
database for the best matches (Sliwa 2002).
In 1999, Lands End launched a global Web site, building the site on IBMs
Websphere platform and partnering with Berlitz to handle the initial translations.
By 2001, 14 percent of the total sales were outside of the United States. The frst
sites they added were Japan, the UK, and Germany. A few months later, they add-
ed Ireland, France, and Italy by cloning the UK site. Sam Taylor, vice president of
international operations, notes that to launch the French site, it cost us 12 times
less than the UK site, and to launch Italy, it cost us 16 times less. Thats the beauty
of the Internet. Its so scalable (Sliwa 2001). He also observed that it is consider-
ably cheaper to reach customers via the Internet than through printing and mailing
catalogs. However, he also pointed out that the company should not have created
the Japanese site frst. Theres nothing worse than when your programmers are
making changes, and they look at the site to see if it works and they cant read it
[Sliwa 2001].
In 2009, watching sales shrink at traditional stores, Sears began pushing new
Web sites and new ways of purchasing products. For instance, the MyGofer.com
site lets customers order items and pick them up at a local store on the same day.
Common products include groceries, prescriptions, pet food, and some electronic
items (Bustillo and Fowler 2010). A similar process can be used to order items
and pick them up at a local Sears store. Executives say that online sales have in-
creased by double-digits, reaching $2.7 billion in 2008. Edward S. Lampert who
purchased and merged K-Mart and Sears is reportedly more willing to spend mon-
ey for online ventures than trying to save the local stores.
Te Merger
Sears knew that its clothing division was in trouble, even though they sold $4.7
billion of apparel a year. And with the divestiture of its credit card and other unre-
lated operations, the executives knew that clothing sales had to be turned around.
The merger with Lands End, one of the most successful direct clothing mer-
chants, gave them the opportunity to fx the problem. The Lands End executives
were given control over the Sears clothing department.
Although it was expected, trouble began brewing between the two cultures.
The Lands End group was used to making fast decisions informally. The Sears
bureaucracy had multiple levels of red tape. Sid Mashburn, Lands Ends vice
president of design and now design chief at Sears as well, refers to the main head-
quarters as the Battlestar Galactica and suggests giving out machetes at the
door to cut through the ingrained procedures. He might have been referring to the
29,000 pages of company guidelines that Sears once had (Merrick 2004).
When the Lands End merchandise was frst brought into the 870 stores, sales
were weak. Placing $139 cashmere sweaters next to $17 sweatshirts did not help
sales of either item. Customers were confused about where to fnd items. Also,
Lands End executives did not have data on what items would sell, or characteris-
tics of the shoppers. Several items had to be marked down at the end of the season
to clear out the inventory.
Jeff Jones, originally chief operating offcer at Lands End, asked if he could
purchase data mining software to analyze customer purchases. Executives said
that Sears did not use those tools. After some behind-the-scenes negotiations,
Jones was fnally able to get the tools. Using the software, he was able to learn
that the store needed to target more upscale customersthose with incomes of
$50,000 to $100,000 or more. Other merchandising experts from Lands End have
been brought in to retarget all of the clothing lines and identify the primary cus-
tomers. Mindy Meads led groups to picture an image of a representative customer
for each of fve brands of womens clothing sold at Sears. This routine exercise
and the data mining were things that Sears had neglected for years (Merrick 2004).
In March 2005, Kmart purchased Sears. The combined companies are run un-
der the Sears Holdings Corp. name. Karen Austin, the CIO from Kmart was made
CIO of the new company (Sliwa 2005).
Technology
In 2002, Sears purchased a huge storage area network from EM to handle a new
customer relationship management database. The system will include 95 terabytes
of new storage. It will be used to combine data from several inventory databas-
es and existing data warehouses. The system is designed to give the company
the ability to examine purchases by customers within the store and over time. As
Jonathan Rand, director of merchandise planning and reporting, commented, the
company needs to make sure customers fnd the merchandise and service they
want in our stores, while eliminating what they dont want faster than the compe-
tition (Mearian 2002).
Because the Lands End division continues to run as a separate subsidiary,
Sears did not attempt to merge their information systems. Their Web site and in-
ternal systems remain as separate units. This decision minimizes interruptions and
avoids causing problems with a system that has been successful.
At the same time, Sears realized that it needed to overhaul its own informa-
tion systems. In 2003, Alan Lacy, the CEO of Sears, observed that weve got too
many point-of-sale systems, too many inventory systems, too many this, that, and
the other thing, because we basically allowed for many, many years each busi-
ness to do its own thing, which were not going to do anymore (Sliwa 2003).
However, he also recognizes the importance of keeping the successful Lands End
systems separate. In early 2004, Sears announced that it was installing 37,000
Internet-enabled IBM cashier terminals. The connected systems will enable Sears
to provide additional types of customer service (McDougall 2004).
In 2004, Sears signed a $1.6 billion 10-year outsourcing contract with Comput-
er Sciences Corporation (CSC). The main elements of the contract cover desktop
services, servers, networks, and system management of the Sears.com Web sites.
Management of the servers running the fnancial reporting and sales systems re-
mains outsourced to IBM. The Lands End operations remain completely indepen-
dent (Weiss 2004). Less than a year after signing the contract, the company ter-
minated the contractalthough Sears and CSC fought over the cancellation fees.
Sales reports for 2007 indicated substantial declines for both Sears and Kmart
stores.
Sears also made a decision to recentralize its IT staff. Up until 2003, the staff
was split into the various business units such as human resources or credit. The
company also moved to standardize its choice of hardware and software vendors.
The overall goal was to reduce the number of disparate systems. Cheryl Murphy,
vice president of IT operations and engineering, observed that were pulling the
IT staffs together to get operational excellence and to drive that into the company.
We want to fnd one way of doing things (Rosencrance 2003).
Suppliers
Sears has been successful in its automotive division in building a relationship with
Michelin. In a pilot test with GlobalNetXchange, Sears provides detailed real-
time sales data to Michelin. Sears had planned a sale on Michelin tires in June
2002, but even before the sale began, Michelin spotted a problem. Sales earlier in
the year had been higher than expected, and the sale would push even more tires
out the door. Michelin was forecasting that it would be short 5,000 tires. Since
Michelin had direct access to the sales and inventory data, and knew which sizes
were selling, the company was able to increase production before the sale began.
Hank Steermann, senior manager of supply chain for Sears, commented that this
is a way to manage the supply chain that is good for Sears, because were fulfll-
ing our commitment to customers, and its good for the supplier, because theyre
selling more units than wed planned (Konicki 2002).
Questions
1. How can Sears use information technology to improve the sales of clothing
particularly the Lands End items?
2. What are the benefts to centralizing the information systems and the IT staff
at Sears?
3. Given the benefts of centralization, why does every Sears executive keep
saying that they will not integrate the Lands End systems with the Sears
systems?
4. If Sears wanted to combine data from the Lands End fnancial system with
the main systems at Sears, how could they do it without actually merging the
systems?
Additional Reading
Bustillo, Miguel and Geoffrey Fowler, Sears Scrambles Online for a Lifeline,
The Wall Street Journal, January 15, 2010.
King, Julia, Service Needs Drive Tech Decisions at Lands End,
Konicki, Steve, Sears, Michelin Test Supply Chain, Information Week, July 1,
2002.
McDougall, Paul, Ordering IT, Information Week, March 22, 2004.
Mearian, Lucas, Sears To Build Huge Storage Network For CRM,
Merrick, Amy, Sears Orders Fashion Makeover From the Lands End Catalog,
The Wall Street Journal, January 28, 2004.
Rosencrance, Linda, New Sears VP Of IT Operations Wants To Simplify IT
Environment, Computerworld, May 16, 2003.
Sliwa, Carol, Clothing Retailer Finds Worldwide Business on the Web,
Sliwa, Carol, E-Retailers Seek Improved Search Engines, Computerworld,
August 12, 2002.
Sliwa, Carol, Sears CEO Says Company Will Standardize Technology,
Computerworld, January 20, 2003
Sliwa, Carol, Kmart CIO gains top IT job at Sears Holdings, Computerworld,
April 4, 2005.
Swanson, Sandra, Lands End Brings Custom-Made Pants Online, Information
Week, November 6, 2001.
Weiss, Todd R., Sears Buying Lands End, Computerworld, May 13, 2002.
Weiss, Todd R., Sears Inks $1.6B IT Outsourcing Services Deal With CSC,
Whiting, Rick, The Data-Warehouse Advantage, Information Week, July 28,
2003.
Case: Super Valu and Albertsons
SuperValu (ticker: SVU) began as the Winston and Newell grocery warehouse
serving Minneapolis, MN in 1924. Over the years, the company developed a dis-
tribution chain, largely supplying independent and affliated grocers in the region.
Buying other regional distribution chains, the store hit $1 billion in sales in 1971
under the SuperValu name. In 1980, the company moved into the retail trade by
purchasing Minnesota-based Cub Foods, operating fve stores. From 1990 for-
ward, the company acquired increasingly larger retail-store chains; yet continued
to distribute food to unaffliated stores. By 1998, SuperValu handled 15 percent
of the nations food distribution. In 2006, the retail side expanded massively with
the purchase of approximately half of the retail stores (1,100) and distribution
network of Albertsons, making SuperValu the third-largest national grocery chain
(www.supervalu.com).
Albertsons was a regional grocery chain that grew into the third largest grocery
chain in America. Founded in 1939 by Joe Albertson, the company maintains its
headquarters in Boise, Idaho. Until its 1999 merger with American Stores, Albert-
sons expanded through internal growth and creating new stores. The merger with
American Stores gave it wider access in California and the Midwest. Some of the
acquired stores maintain the original name, such as the Jewel-Osco grocery/drug-
stores in the Chicago area. In 2003, the company had around 2,300 retail stores
and total revenue of $35 billion. The stores were supported by 17 distribution
centers [annual report]. Despite the mergers, Albertsons has seen total revenue
decline since the 1999 fscal year. In 2001, the company replaced the CEO by hir-
ing Larry Johnstonwho lost out in the GE competition to replace Jack Welch.
Johnston quickly replaced most of the executive team. His stated emphasis was
to reduce costs and streamline the company, but operating/administrative costs
in FY 2003 were higher than they were in FY 1999. Johnston also closed several
hundred storesdeciding to focus only on markets where the chain held the num-
ber one or two position. In 2006, Johnston sold half the company to SuperValu
and the other half to private equity frm Cerberus for eventual liquidation. John-
ston walked away with over $100 million in cash for driving a business into the
ground.
Competitive Treats
The biggest threat to SuperValu (and Kroger and Safeway and the other big gro-
cers) is Wal-Mart. Wal-Mart now sells more groceries than any other chain in
America. Its SuperCenter stores provide a one-stop market for customers, and the
company is rolling out more of them every day (Gose 2002). With Wal-Marts
supply chain effciencies and heavily advertised low-price strategy, customers
fock to the stores. Oh yes, and Wal-Mart employees are not unionized. The big
grocery chains in California have unionized workers who not only demand high
wages, but have gone on strike several times to try and force the wage and beneft
policies. With the collapse and sale of Albertsons, it will be interesting to see if the
unions can maintain their power to command higher wages (Peltz 2004). Grocery
chains gained some breathing room in 2007 when Wal-Mart announced that it
was slowing its expansion into the grocery feld. In 2006, Wal-Mart accounted for
21.1 percent of U.S. food sales, giving it more than twice the market of Kroger, its
nearest competitor. Wal-Mart decided to scale back its expansion plans to avoid
market saturation in rural areas and because of stiff competition in cities (Jargon
2007). On the other hand, British supermarket giant Tesco PLC announced plans
to place stores in Southern California and Las Vegaskey markets for the new
SuperValu.
By 2011, SuperValu operated 1,114 retail food stores and 381 discount food
stores, plus handled distribution to 899 discount stores and 1,900 independent
retail stores. The company runs 21 national distribution centers. The company
distributes and sells all standard grocery and drug-store items including nation-
al brands and private labels. The company had about 142,000 employees, with
88,000 of them members of unions. In 2009, SuperValu hired a new CEO; then
a new chief marketing offcer in 2010 and a new CIO in late 2010. The company
experienced a decline in sales for 2010-2011. Even same-store sales declined in
many marketswhich was primarily blamed on competitors with lower prices
(2011 Annual Report).
In February 2007, SuperValu almost got scammed out of $10 million. Two
e-mail messages purporting to be from American Greetings Corp and Frito-Lay
tried to get SuperValu to change the bank accounts for future payments to these
companies. In early March, SuperValu said it transferred $3.6 million to one of
the bank accounts, but then double-checked the changes and notifed federal law
enforcement on March 6. The FBI offce was able to shut down the fake accounts
and retrieve most of the money. It is not clear if the scammers were eventually
caught (Vijayan 2007).
Technology
Transactions processing is a key element to holding down costs in grocery stores
particularly when labor is expensive. Grocery stores were the frst to implement
bar codes and checkout scanners in the 1980s. Although bar codes provide better
data, they do not signifcantly increase the speed of the checkout process. Some-
one still has to unpack the cart, move each item by hand over the scanner (some-
times multiple times to get it to read), and then pack the items into bags. Albert-
sons, like other chains, has experimented with self-service checkouts. Custom-
ers with a small number of items run them over the scanner or weigh them and
place them into bags. Usually one clerk oversees four checkout stations. To reduce
theft, the station weighs the bags as items are added and compares the weight to a
known value. But the stations do not really change the checkout process, merely
shift some of the work to the customers.
Payment mechanisms are a challenge for grocers. Proft margins are extremely
tight and retailers make only a few cents on any item. Customers no longer like
to carry cash, but the fees to process credit and debit cards cuts into margins.
Check-processing costs at one time ran as high as 75 cents per check. Jacki Sny-
der, manager of electronic payments at SuperValu observed that in some cases,
the fees on one grocery order exceed $1, more than the supermarket proft on
the same order (Thibodeau 2000). Supermarkets including SuperValu are testing
new systems that use fngerprints to identify customers. The Pay By Touch system
links the identity to the customers credit or debit card and also verifes age for
purchases of alcohol or tobacco (Sullivan 2006).
Albertsons had just completed an installation of Oracles ERP system to run
operations before the sale to SuperValu. The company was also using PeopleSoft
to handle HRM tasks, so Oracles purchase of PeopleSoft will ultimately beneft
the company. Because of the number of stores, SuperValu at least initially is keep-
ing the chains on their existing IT systems. This approach minimizes the problems
of integrating data systems and keeps the stores running effciently, but it makes it
more diffcult to manage the divisions as an integrated organization.
By 2011, new CIO Wayne Shurts said the IT organization was changing focus
from cost-cutting to concentrating on strategic investments. His plans included
moving to Microsoft Offce 365, the cloud-based applications from Microsoft
priced at $10 per user per month. The bundle includes SharePoint and Web con-
ferencing in addition to standard offce tools (Nash 2011). The goal is to make it
easier for corporate directors to retrieve and share information and comments with
other directors and store managers.
Questions
1. Will enough customers use the handheld scanners to make them proftable?
2. If there are 150,000 different items in a store, and LCD tags cost $5 each,
how long will it take to recoup the cost of outftting an entire store with the
LCD tags? Remember that all employees are unionized, so assume average
wages are at least $10 per hour. Write down any other assumptions.
3. Will new checkout technologies be enough to help Albertsons stay
competitive against Wal-Mart? What can Albertsons do to stay alive?
Additional Reading
Gose, Joe, Supercenter Showdown, National Real Estate Investor, December
2002, Vol 44(12), p. 14.
Jargon, Julie, Wal-Mart Slowdown to Hit Food Sector, The Wall Street Journal,
June 5, 2007.
Nash, Kim S., 2011 State of the CIO: IT Departments Are Fueling Company
Growth Through Strategic Technology Investments, CIO, January 1, 2011.
Peltz, James F., How the Supermarket Strike Was Settled, The Los Angeles
Times, March 8, 2004.
Sullivan, Laurie, Touch Once For Groceries, Computerworld, February 13,
2006.
Thibodeau, Patrick, Older Systems Hinder E-payments, Computerworld,
September 25, 2000.
Vijayan, Jaikumar, Phishers Nearly Pull Off $10M Scam of Grocer,
Computerworld, October 29, 2007.
Case: Amazon
In 1994, with a handful of programmers and a few thousand dollars in worksta-
tions and servers, Jeff Bezos set out to change the retail world when he created
Amazon.com (ticker: AMZN). Shel Kaphan, Amazons frst programmer, assisted
by others, including Paul Barton-Davis, used a collection of tools to create Web
pages based on a database of 1 million book titles compiled from the Library of
Congress and Books in Print databases. Kaphan notes that Amazon was depen-
dent on commercial and free database systems, as well as HTTP server software
from commercial and free sources. Many of the programming tools were free soft-
ware (Collett 2002). In July 1995, Amazon opened its Web site for sales. Using
heavily discounted book prices (20 to 30 percent below common retail prices);
Amazon advertised heavily and became the leading celebrity of the Internet and
e-commerce.
Amazon was often held up as the darling of the dot-com era. Anyone could
sell books online, but Amazon showed that you had to be frst (or close to frst
with more money), to capture name recognition and market share. But, looking at
Amazons sales and losses, it is clear that Amazon never made any profts selling
books. By the mid-2000s, Amazon altered its strategies on book salesincreas-
ing the price of mass-market books and reducing its inventories to focus on the
best-sellers. With the release of the Kindle e-book reader, Amazon signifcantly
jump-started the sale of e-books. By 2010, Amazon was selling 15 percent more
e-books than even paperback books (Hamblen 2011).
Amazon sales continue to growaccounting for a signifcant portion of on-
line sales. Some of the sales come from partnerships with other sellersAmazon
simply handles the Web site and payment methods. Although sales were increas-
ing rapidly through at least 2011, costs were increasing faster. Most of the costs
were investments in new technologies and expanding data centers (Woo 2011).
Amazon also continues to sell digital music and in 2011, it expanded its offerings
of digital video. Customers who join the Prime club who pay an annual fee to get
reduced shipping costs will also get online access to thousands of videos. But, in
both music and movies, Amazon remains a small player compared to the Apple
and Netfix.
Selling Relationships
As a premier site on the Web, Amazon realized that it could leverage its huge
name recognition and trust. The company gradually signed deals with both large
companies and small retailers. Customers go to the Amazon site and search for a
product. The Amazon database retrieves the product descriptions and prices and
displays them in a consistent format. Amazon handles the payment details and
sends the order to the partners for fulfllment. Some customers never know they
are dealing with a third party. By 2004, 25 percent of Amazons sales were for its
partners. But some of the deals did not go as well as others. In 2004, Toys R Us
sued Amazon and Amazon countersued. The primary complaint was that Amazon
had promised Toys R Us an exclusive right to sell toys on the Amazon site. But,
Amazon signed on many small companies that also sold toys. David Schwartz,
senior VP and general counsel for Toys R Us stated that We dont intend to pay
for exclusivity were not getting. Amazons initial response was that We be-
lieve we can have multiple sellers in the toy category, increase selection, and offer
products that (Toys R Us) doesnt have (Claburn May 2004). By 2007, Amazon
was simply the largest marketplace on the Web.
Small merchants accelerated a shift to Amazons marketplace technology. For
example, John Wieber was selling $1 million a year in refurbished computers
through eBay. But increased competition and eBays rising prices convinced him
to switch to direct sales through Amazon. Similar small merchants noted that al-
though the fees on Amazon are hefty, they do not have to pay a listing fee. Plus,
eBay shoppers only want to buy things at bargain-basement prices (Mangalindan
2005).
Initially, Amazon built its Web site separately from its order-fulfllment system.
The separation existed partly for security reasons and partly because the technol-
ogy to connect them was in its infancy. One of the most important features offered
by Amazon was the affnity programwhere customers are shown lists of other
items they might want to purchasebased on an analysis of purchases by other
customers. It proved to be a useful technology for books, music, and movies
identifying related products that customers might otherwise miss.
In 2000, Amazon decided to overhaul its entire system. The company spent
$200 million on new applications, including analysis software from E.piphany,
logistics from Manugistics, and a new DBMS from Oracle. The company also
signed deals with SAS for data mining and analysis (Collett 2002). But, one of
its biggest deals was with Excelon for business-to-business integration systems.
The system enables suppliers to communicate in real time, even if they do not
have sophisticated IT departments. It provides a direct con-nection to Amazons
ERP system either through programming connections or through a Web browser
(Konicki 2000).
The technology and sales did not come cheaply. In 2003, Bezos noted that $900
million went to business technology; $300 million was spent on the fulfllment
centers; and $700 million on marketing and customer acquisition (Murphy 2003).
That last part largely represents selling books at a loss or offering free shipping
while trying to attract customers. Those numbers add up to the $1.9 billion debt,
but the company also lost another $1 billion.
New Services
Amazon requires huge data centers and high-speed Internet connections to run
its systems. Through vast economies of scale, Amazon is able to achieve incred-
ibly low prices for data storage and bandwidth. Around 2005, the company de-
cided that it could leverage those low costs into a new business selling Internet-
based services. The company offers an online data storage service called S3. For
a monthly fee of 15 cents per gigabyte stored plus 20 cents per gigabyte of data
transferred, any person or company can transfer and store data on Amazon servers
(Markoff 2006). Through a similar service (EC2), any company can use the com-
panys Web servers to deliver digital content to customers. Microsoft reportedly
uses the system to handle some large fles for downloads. The company essen-
tially serves as a Web host, but instead of paying fxed costs, you pay 10 cents per
virtual server per hour plus bandwidth costs. Amazons network can handle bursts
up to 1 gigabit per second. The system creates virtual servers, running the Linux
kernel, and you can run any software you want (Gralla 2006).
Amazon also runs the A9 search engine, so it began selling keyword advertis-
ing. Similar to Google, A9 provides advertising links to other Web sites and shares
the revenue. The company also provides mapping and other services that can be
integrated into your Web site.
Perhaps the most unusual service is Mturk. The name derives from an 18-cen-
tury joke where a mechanical chess-playing machine surprised European lead-
ers and royalty by beating many expert players. The trick was that a human was
hidden under the board and moved the pieces with magnets. Amazons trick is to
use human power to solve problems. Companies post projects on the Mturk site
and offer to pay a price for piecemeal work. Any individual can sign up and per-
form a task and get paid based on the amount of work completed. Amazon takes a
10 percent commission above the fee. For example, the company Casting Words
places audio fles on the site and pays people 42 cents to transcribe one minute of
audio fles into text (Markoff 2006).
Adam Selipsky, vice president of product management and developer relations
at Amazon Web Services observed that Amazon is fundamentally a technology
company; weve spent more than one and a half billion dollars investing in tech-
nology and content. We began by retailing books, but it was never in our business
plan to stay with that (Gralla 2006).
Amazon has been increasing its online cloud-based services. The main of-
ferings include the S3 service to store fles, but also the Elastic Compute Cloud
(EC2) and Amazon Web Services (AWS) which enable anyone to buy Web ser-
vices or any type of computer server and pay just by the amount of processing
time used and the data transferred. The company also offers the Relational Data-
base Service (RDS) which provides a cloud-based relational query system. In an
interesting twist, Netfix moved some of its main IT operations to Amazons EC2
and AWS services. Netfix noted that it reduced costs by moving off expensive
IBM hardware and they were growing so quickly that it was easier to use Amazon
to confgure and manage the servers and network bandwidth (Kanaracus 2010).
Questions
1. Who are Amazons competitors?
2. Why did Amazon create most of its own technology from scratch?
3. Will Amazon ever become proftable enough to recoup its initial
investments?
Additional Reading
Claburn, Thomas, Amazon.com And Toys R Us Are On The Outs,
Collett, Stacy, The Webs Best-seller, Computerworld, September 30, 2002.
Gralla, Preston, Computing in the Cloud, Computerworld, December 21, 2006.
Hamblen, Matt, Amazon.com Touts More e-Book Sales than Paperbacks,
Kanaracus, Chris, Netfix Turns From Oracle, IBM to Amazon to Save Cash,
Konicki, Steve, Amazon Taps Excelon To Redo Supply-Chain System,
Information Week, October 30, 2000.
Mangalindan, Mylene, Threatening eBays Dominance, More Online Sellers Go
It Alone, The Wall Street Journal, June 22, 2005.
Markoff, John, Software Out There, The New York Times, April 5, 2006.
Murphy, Chris, Amazon, The Services Firm, Information Week, June 10, 2003.
Woo, Stu, Expenses Eat at Amazons Proft, The Wall Street Journal, January
28, 2011.
adopted by rivals?
sectors?
505
How can information systems be used to support teams of workers?
How many different ways are there to communicate with team members?
Which method is best for each type of message?
How can several people work on the same documents?
How does an organization remember past events?
Where can you fnd a system that provides these groupware capabilities?
What if you do not have a server and want to collaborate?
What could you accomplish if fve people work on the same screen at the same time?
How do you collaborate and share data if everyone uses different software?
Chapter Outline
Teamwork
8
Chapter
Introduction, 506
Communication, 507
e-Mail, Instant Messaging, and Voice
Mail, 509
Web Pages and Blogs, 511
Scheduling and Project Management,
512
Conferences and Meetings, 513
Peer-to-Peer Network Complications,
515
Mobile Systems, 516
Collaboration, 516
Documents and Changes, 517
Version Control, 518
Wikis and Online Applications, 520
Information Rights Management, 521
Workfow, 522
Group Decision Support Systems, 523
Knowledge Management, 525
Organizational Memory, 526
Service Processes, 527
Communication and Scheduling, 528
Collaboration, 532
Workfow, 535
SharePoint Confguration, 537
Peer-to-Peer Collaboration, 540
Sharing Documents, 540
Immediate Interaction: Microsoft
SharedView, 542
Sharing with Multi-Touch Panels, 544
File Formats and Standards, 545
Cloud-Based Collaboration, 546
Summary, 549
Key Words, 550
Exercises, 551
Cases: Package Delivery, 555
506 Chapter 8: Teamwork
FedEx
Introduction
How can information systems be used to support teams of
workers? Companies are increasingly approaching problems and decisions
with teams, often built from members drawn from across the company. Workers
can be members of several teams at the same time, and teammates might work in
different parts of the country or around the world. Sharing information is a major
element of teamwork, and networks are needed to transfer data and share resourc-
es. But beyond simple sharing, Figure 8.1 indicates that workers need to work
collaboratively on a collection of project documents. Over the past few years, the
standard personal productivity tools have gained features to support collaboration
and sharing. Additional tools are available to stay in touch with teammates, dis-
cuss strategies, schedule meetings, and share research. As a manager, you need to
use these tools on a regular basis to become more effcient.
Teamwork is not an invention of the past couple of years. Most teams have
developed methods to handle the collaboration needs of the group. Many orga-
nizations still use paper-based systems. Paper documents are photocopied and
circulated for review. Individuals suggest changes and corrections, which are
forwarded to one person. Meetings are held in personpeople sometimes fy
thousands of miles for meetings, everyone takes notes, and some people dominate
the group discussions. Many people still rely on e-mail to send versions of docu-
ments. Someday, people will look back and consider these systems to be quaint
and ineffcient.
Teamwork revolves around two fundamental concepts: communication and
collaboration. Teammates need to communicate on many topics, and several com-
munication methods are available today. One challenge is to choose the right com-
munication method for each problem. Collaboration entails working together. To-
day, that means sharing electronic documents and tracking the revisions made to
each document. More than simply combining changes, it also involves discussing
why the changes were made. Collaboration also involves scheduling tasks and
maintaining project schedules.
Software tools have existed for several years to help with the basic commu-
nication and collaboration tasks. The strengths of todays tools are the integra-
How does technology help teams work together? Everyone has heard of FedExthe
company that invented overnight delivery of small packages. But FedEx is not alone;
it must compete with several other large companies to earn your business. Being on
time and at a reasonable price are important attributes to success. FedEx achieves its
success through communication and teamwork. Downstream workers have access to
information about incoming packages and can alter their schedules if necessary to
provide better service.
Since 2000, with several mergers, FedEx has grown beyond overnight deliveries.
The company now controls several package delivery and shipping carriers. It also
acquired Kinkos in 2004. Information systems are used to integrate these operations
and work closer with customers. Ultimately, to FedEx, teamwork entails identifying
customer needs and fnding new ways to solve problems.
tion and focus on teamwork. Today, you can automate the business processes by
automatically triggering messages and monitoring the schedules and the fow of
documents through the company. One goal of these systems is to track the deci-
sions, documentation, and discussions by everyone involved in a decision. If this
knowledge is collected and indexed carefully, it can become the foundation of the
organizations memory. If you need to solve a diffcult problem, you could turn to
this knowledge base to fnd similar problems, or suggestions from workers who
faced the same problems.
Communication
How many different ways are there to communicate with team
members? Which method is best for each type of message? Com-
munication is a critical element in teamwork. You need to schedule meetings, dis-
cuss options, share and revise documents, and identify opinions. Because com-
munication is so important to teams, several technologies are available. You can
choose between simple meetings in the hall, phone calls or voice mail, e-mail and
instant messaging, and online conferences, or full-blown team meetings. How do
you choose the proper communication method?
As shown in Figure 8.2, one of the issues in organizing your daily work life is
to manage the interruptions. Phone calls, e-mail messages, people walking into
your offce, and IM requests all break your concentration. One question you have
to answer is to identify which method works best for you (and your teammates).
Having multiple channels of communication gives you fexibility. Your goal is to
prioritize each technology and method so that you accomplish your goals, yet still
Figure 8.1
Teamwork. Businesses increasingly rely on teams of workers that require
coordination and shared documents. Several tools exist to help workers collaborate
on common documents, schedule meetings, and track changes.
1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31
Dept Costs Labor
Mrkt 445 87
Prod 227 45
HRM 554 32
Projected Sales
0
200
400
600
800
1000
West East North South
New
Maint.
Tools
Alpha Project
Goals:
Problems:
Departments:
Trends
Personal computers originally provided support for basic personal tasks
writing documents, computing totals, and drawing charts. In the 1990s,
companies realized the importance of networks for sharing data. In the pro-
cess, companies realized that networks can change the way managers work.
Instead of relying on rigid management hierarchies, small teams can be cre-
ated to solve problems as they arise. Communication across networks makes
it possible for teammates to be located anywhere within the company. Be-
yond the benefts of communication, can information systems do even more
to support teams?
In the 1990s, Lotus Notes revealed the demand for collaboration; but it
largely relied on e-mail to transfer copies of documents. In the 2000s, peo-
ple tried to use Web sites and blogs to communicate and work together, but
these tools are diffcult to use with multiple contributors and have problems
handling offce documents.
One of the biggest changes in Offce software in the past couple of years
has been the introduction of powerful tools to help groups of workers com-
municate and collaborate. These features were expanded and integrated into
Offce 2007 which provides SharePoint as a central communication point
and SharePoint Workspace/Groove for peer-to-peer sharing. A couple of
third-party companies have begun to offer collaboration tools, but they do
not yet achieve the integration Microsoft has provided in Offce.
Instant Message/Chat E-Mail
Phone Call
Physical Visit
Video conference
Figure 8.2
Communication interruptions. Which type of interruption is the least disruptive for
you? Need to fnish your work. Need to help the team.
have time to work with the group. If possible, you might try to get the group to
agree on the priorities of the main methods. Most organizations have a culture that
establishes these priorities, but sometimes you need to discuss them so that every-
one treats the channels with the same priority.
e-Mail, Instant Messaging, and Voice Mail
E-mail is a relatively common means of communicating. It has the fundamental
advantage of being asynchronous, which means that the sender and receiver do
not have to be connected at the same time. The recipient can pick up and reply to a
message at any time. In a group context, e-mail is even more important because it
is easy to send messages to the entire group. Moreover, the messages can be saved
and organized so that they create a thread or history of a discussion. E-mail is also
used to send documents as attachments.
As shown in Figure 8.3, e-mail systems such as Lotus Notes and Microsoft
Exchange were originally advertised as groupware tools because of the support
provided for attachments. The e-mail systems essentially become moving data-
bases of documents. One of the drawbacks to this approach is that multiple copies
of the documents are transferred continually with each message, which makes it
Figure 8.3
Problems with e-mail groupware. E-mail messages with attached documents are sent
to team members. But, people end up with multiple copies, they do not receive the
same messages, and it is hard to identify the correct version.
Design Team
Client HQ
New York
Finance & Management Team
Consulting HQ
California
Design Project
Singapore
Multiple copies
Different messages
Old versions
Which version is current?
Design
documents
changes
design
changes
contract
changes
modified
designs
diffcult to track versions and identify the current documents. Plus multiple people
might be altering the same documents at the same time, so someone has to merge
the fnal changes into a single document. E-mail is still an important tool for com-
munication, but its drawbacks make it diffcult to use as a standalone groupware
system.
Texting and other instant messaging tools have become popular in the past few
yearsparticularly with students. However, the technology can also be a time-
saver for groups of workers. Each team member can add the others onto the buddy
list. The ability to see if someone is online can save considerable time. If you
encounter a problem while working, you could immediately see who might be
available to answer questions. Yet if you are busy and receive an IM request, you
can still choose to ignore it until you have time for a break. But, it is easy for some
people to get carried away with text messages; which can make it diffcult for oth-
ers to concentrate on diffcult tasks.
Voice mail systems were created so that you can leave phone messages for peo-
ple who are out of the offce or talking to another person. Current voice mail sys-
tems have many advanced options, such as the ability to send the same message
to multiple people, receive notifcations when a message is retrieved, and archive
messages for later use. The biggest drawback to voice mail is that some people
leave long messages that you have to listen to before they get to the real point of
the call. Long e-mail messages are annoying, but at least you can quickly scan the
message to fnd the important elements. It is also diffcult to search stored voice
mail messages to fnd particular topics.
E-mail, voice mail, and IM are all push communication technologies. As a mes-
sage sender, you choose when to place the message and contact the receiver. The
other person did not necessarily ask to be contacted, so you are interrupting his
or her work. Of the three methods, e-mail is the least intrusive since the recipient
has to retrieve the messages and open each one. Because e-mail is less intrusive,
you have to remember that any particular message might never actually be read.
Reality Bytes: Google Apps
Applications to support teamwork are becoming more important every day. The
problem is that the applications might not have all of the features people want.
Google has been pushing to become a big player in the online market for collabora-
tion with its Web-based Google Apps. In 2010, Google had more than 2 million busi-
ness customers, including some large companies and several government agencies.
The price is attractive to many organizations at $50 per user per year. Google runs all
of the servers, provides 25 GB of e-mail storage per user and promises 99.9 percent
uptime with 24/7 customer support. But, Google Apps are not the same as traditional
desktop software. Aisle 7, a small health and wellness marketing company switched
to Google Apps in early 2009. Although the company has only 32 users, only 10 peo-
ple initially used Gmail. Within three months. Only two people were using Gmail.
Aisle 7 also said that over the course of using Google Apps for a year, the company
experienced four instances where the servers were down for several hours each. The
company eventually switched to Microsofts online service.
Adapted from Shane ONeill, The Trouble with Going Google: 4 Reasons Why I
Got Out, Computerworld, August 11, 2010.
It could be buried in a stack of junk messages, or even lost or destroyed. Most
systems enable you to ask for receipt confrmation when a message is opened. But
these systems are not always automatic. The user can often override this request
and tell the computer not to send the confrmation message. Voice mail messages
suffer from the same issues and have similar controls. The point is that as the
message sender, you need to understand the limitations of the tools. If you need
confrmation of a message, it is your responsibility to follow up and make sure the
message was received and understood.
Web Pages and Blogs
Web pages were initially created to share information among researchers. As a
common platform, you can make fles available to anyone on your teamregard-
less of where they are located or what equipment they are using. Most people are
familiar with Web browsers and pages. Several tools help you create Web sites
and update documents. Despite these tools, it is challenging to create Web sites
and keep updating the pages. As shown in Figure 8.4, Web pages are generally
seen as a client-pull technology. You create the Web pages and place them on
the site. Team members can view or retrieve the documents, but there is no good
mechanism for making changes or replacing the contents. In one sense, Web pages
are a broadcast technology, where you can provide the same data to many people.
Recognition of this feature makes it clearer when you should use Web sites for
communication.
Web logs or blogs were created primarily for individuals to use to create diaries
on the Internet. The technology was largely developed to overcome the diffcul-
ties in creating and updating Web sites. A blog consists of a site framework that
enables others to read your comments. A simple editing tool makes it easy to write
daily or hourly comments and immediately post them to the Web. The process is
simplifed so you can focus on the content. It is also designed to save all of your
Figure 8.4
Web pages and blogs. One-way communication that is easiest to create with a single
author. Comments can be saved and easily searched by coworkers.
Web server
Author
Coworkers
Web site
Blog
One-way communication.
Easiest with one author.
Users search for data.
earlier comments so people can trace the history. These tools could be used in a
business or teamwork setting. They would provide a record of the teams actions
and decisions. Each person could record current problems, solutions, and describe
issues for discussion. Later, if questions arise, you can search the logs to fnd pro-
posed solutions or learn exactly why a particular decision was made. Of course,
you would restrict access to the logs. The big question is whether the team mem-
bers will take the time to record their thoughts and relevant issues on a regular
basis. Blog software is easy to use and you can fnd shareware implementations
that can be installed for little or no cost; so it is relatively inexpensive to test the
technology.
Scheduling and Project Management
Scheduling is a challenge with any team project. It is even more complicated when
people work on several teams at the same time. Scheduling a meeting or confer-
ence is diffcult and often requires several messages back and forth just to fnd a
common time. Electronic calendar tools make this process a little easier. Figure
8.5 shows that teammates keep personal calendars that are shared with each other.
Each person can check to see if there is an open time slot or use advanced tools to
help fnd the best time for a meeting. Most e-mail systems support shared calen-
dars, but participants generally need to use a common e-mail client. For example,
Microsoft Exchange mail server provides tighter integration with the Outlook cli-
ent software. If you use one of these systems, you need to learn how to mark some
items as private, and you need to check the calendar throughout the day as the
schedule changes.
Figure 8.5
Scheduling. When calendars are shared, scheduling software can examine calendars
for each participant and look for an open meeting time. Schedules can change
throughout the day, so many people carry PDAs or cell phones to watch the schedule.
Calendar Server
(Microsoft Exchange)
08:00 Meeting
08:30
09:00 Project A
09:30
10:00 Meeting?
10:30
11:00 Dentist
08:00 Project B
08:30
09:00 Project A
09:30
10:00 Meeting?
10:30
11:00 Project C
08:00 Project B
08:30
09:00
09:30
10:00 Meeting?
10:30
11:00 Meeting
However, these tools can also make it harder to plan. If your schedule is con-
stantly changing, you can waste time jumping between projects. For example, you
might have a big project presentation scheduled for tomorrow, so you spend time
organizing your notes. Then the meeting gets shifted because a higher priority
event occurs. You have to rush around trying to collect and organize data for the
new meeting.
Project management tools were created to help manage and control groups of
workers on large projects. These tools are discussed in more detail in Chapter 13.
The basic concept is that you defne each task that needs to be accomplished and
how these tasks depend on each other. For example, in construction, you must frst
lay the foundation before you can frame a building. If an earlier task gets delayed,
it pushes back all of the dependent tasks. Project management software graphi-
cally shows the tasks and their dependencies. As tasks are completed or delayed,
the schedule is updated and everyone can see the changes.
Conferences and Meetings
Face-to-face meetings are expensive. Even if everyone on the team works in the
same building, a meeting requires everyone to stop their work, organize their pre-
sentations, and set aside most other communications during the meeting. If people
have to travel to a meeting, the costs are considerably higher. Of course, some-
Figure 8.6
Online Conferences. You pay a third-party server (Microsoft LiveMeeting or Cisco
WebEx) to broadcast the meeting to hundreds or thousands of participants. Some
interaction is supported, but large meetings work best with one-way communication.
Conference Server
Encoding Server
Internet
Microsoft Live Meeting, Cisco WebEx
Microsoft Communicator
times the major point of a meeting is to create a deadline that forces everyone to
organize their work and meet the project milestones.
As shown in Figure 8.6, several electronic systems have been developed to sup-
port online conferences and meetings. At the high end, videoconference systems
provide real-time voice and video connections so people in different locations can
hear and see each other. Videoconference systems require video cameras, micro-
phones, and television displays at each location. For more realistic conferences,
some systems use multiple cameras at each location or provide remote control
over the cameras. The primary drawback to the systems is the bandwidth required
to transmit realistic video data. Early systems used proprietary connections to en-
sure suffcient bandwidth. Newer systems can run across Internet connections, but
require special confgurations of routers to ensure the continuous availability of
the required bandwidth. The problems multiply if the video signals have to be
sent to multiple locations, so most systems operate point-to-point between two
locations.
As bandwidth connections improve, lower-cost systems are being developed
that enable you to set up videoconferences using personal computers and Web
cameras. Be sure to test the systems frst because low resolution cameras or slow-
speed links can cause more distractions than benefts. While these systems do
not give you the remote control capabilities of the big systems, they can support
meetings from your desk. They are particularly useful for meetings with two or
three participants. For example, systems such as Skype support HD Web cams to
provide more realistic images. Cisco and other vendors are using the term telep-
resence to denote high-speed, high-quality connections. But, each side needs a
connection with at least 1 mbps transfer rates. For additional fees, the systems
can support multiple video connections at the same time. Remember that although
broadband connections typically support the speed needed for incoming data, the
outgoing data speed can be much lower.
If you do not need video, it is relatively easy to set up a telephone conference.
Most phone systems have provisions to connect several people into one conversa-
tion. If your company system does not support these connections, you can pay a
fee to an external company to host your phone call. Each participant simply calls
a special number at the same time to join the conversation. Note that if you have
several people in the same room participating in a phone call, you will have to use
a special conference telephone. Polycom and Cisco are two of the leading ven-
dors. The phone has a speaker and robust microphone to pick up everyones con-
versations. More important, it contains circuitry to eliminate audio feedback. Ad-
vanced versions of the phones also support full duplex communications, which
means that you can interrupt someone speakingbecause the phone transmits and
receives at the same time.
In part because of the bandwidth problem, several companies have set up spe-
cial servers with high-speed network connections to enable you to run large con-
ferences online. Microsofts Live Meeting and Ciscos WebEx are two examples.
Essentially, Microsoft or Cisco acts as a hosting company. Your frm pays a fee to
use their serviceseither monthly or one-time, based on the number of partici-
pantswhich translates to bandwidth demand.
The systems are designed to handle as many as a few thousand participants.
The process is oriented to a broadcast approach. A video and audio feed can be
viewed by everyone. For reasonably sized groups, you can integrate telephone
conferencing to handle interactive questions. The system also includes a standard
chat tool if you do not want to deal with the telephone costs. Both systems sup-
port Web interaction. The presenter can share applications or the desktop with
other participants. Team members can be given control to run an application or
annotate a document. You can link in Web sites, where each participant can click
on links and fll out forms. A polling tool enables you to quickly create questions
and let the audience vote. The system also includes the popular electronic white-
board where each person can contribute and draw in a different color. Of course,
you will probably want to limit these interactive features if you are presenting to
hundreds of people.
Microsofts system also includes a feedback mechanism through a seating
chart. At any time, a participant can click the seating chart to request that the pre-
senter slow down or speed up, or to ask a question. The audience notes are dis-
played to the presenter on a color-coded status board. If you see many requests to
slow down (red squares), you know you are going too fast. A useful feature of the
system is that it records the presentation so you can save it on your Web site and
customers can replay it later.
Peer-to-Peer Network Complications
Chapter 3 explains that most companies use network address translation (NAT)
in their routers to isolate their internal networks from the Internetpartly to im-
prove security and partly to conserve the limited number of IPv4 addresses. In
summary, each computer within the companys network receives a unique IP ad-
dress that cannot be accessed by anyone outside the company. As shown in Figure
8.7 although security offcers like NAT, it makes it considerably more diffcult
to establish computer connections to individuals outside the companyinclud-
ing traveling employees, customers, and business partners. In most cases, the best
Company 1
Company 2
NAT
NAT
Communication
Server (xed IP)
Figure 8.7
NAT complications. Computers cannot connect directly to each other with network
address translation. Instead, a communications server with a fxed IP address
handles the connections. A company can run its own server (such as Microsoft
Communicator), or pay a third party such as Skype. The servers support voice, video,
and data sharing.
solution is to establish a server with a fxed IP address that centralizes the com-
munication links. The server has a fxed IP address and does not use NAT, so it is
accessible to people both inside and outside of the company. This approach pres-
ents problems for small businesses because it requires the purchase, installation,
and maintenance of more hardware. Small businesses will probably want to use
third-party companies to handle communication needs. NAT is the second reason
Microsoft Live Meeting and WebEx have grown in importance. Other companies,
such as Skype, offer online voice communications for the same reasons.
To improve security and reduce costs, larger organizations will probably want
to install servers to handle communications themselves. For example, Microsoft
provides the Communicator server and client software that integrates with Offce.
The server also centralizes the management and security issues, making it easier
for the corporate IT department to monitor and control the connections and use of
the technology. Users who have accounts on the Communicator server can make
voice calls using their computer or even a telephone. It also supports video con-
nections, instant messaging, and data transfers. Communicator is integrated with
Offce, so users can quickly determine which team members are online, send chat
messages, or make phone calls with the click of an icon.
Mobile Systems
Messaging and conferencing technologies are fairly well established for tradition-
al environments and meetings. Of course, new features are added every year, so
you need to keep up with the changes to see which options might be useful for
your particular situation. On the other hand, the telecommunications world has
changed markedly in the past decade. People now rely on cell phones, tablets,
and wireless Internet access, and want the ability to connect as they travel. As the
cellular networks improve, mobile devices are gaining the ability to handle voice,
data, and even video connections. For example, the iPhone offers a proprietary
video chat system, but it initially works only over WiFi connections. As availabil-
ity of 4G speeds increases, more people will be interested in using the technology
over cell phone data connectionsbut beware of the bandwidth caps and costs.
Collaboration
How can several people work on the same documents? This ques-
tion forms the heart of collaboration. One teamwork solution is to assign separate
Reality Bytes: Government Teamwork
In 2010, volcanic ash from Iceland drifted over Europe and shut down air traffc
for several days. (High ash concentrations can damage engines.) Jens Stoltenberg,
Norways prime minister was one of the millions of people strandedhe was stuck
in a New York airport. So he pulled out his Apple iPad and cell phone and worked re-
motely. He commented that There are good means of communication, I have close
contacts with my offce all the times, and there are a lot of activities in Norway
where we try to reduce the consequences of the volcano in Norway.
Adapted from CNN Online, Stranded Leader Runs Country by iPad, April 17,
2010.
issues to each person. Working independently, each person creates documents that
are then reviewed by the rest of the team. But some projects are not easily divided,
and sometimes groups of people have to work together simultaneously.
Networks and e-mail systems make it possible to send documents to each per-
son. But what if you need to identify the changes made by each person? Or what
if you need to change directions and revert to an earlier version of a spreadsheet?
And how do you handle issues in a hierarchical system where each person reviews
a document and approves it before it can be forwarded to the next authority? With
paper-based systems, these issues are generally handled by markups and hand-
written margin notes. Microsoft Offce contains several features that support elec-
tronic collaboration.
Documents and Changes
Sharing documents is relatively easy with LANs and the Internet. Keeping track
of when changes were made and who made them, or whether you have the current
version of a document, is more diffcult. One approach is to continually e-mail
new versions of documents to everyone in the group, but it is diffcult to ensure
everyone has the current versionor even knows where the current version is
located. Ultimately, the best approach is to store the documents on a central serv-
er. These tools are described in the following sections, but you might not have
them available; or some people might not be able to access them. Even if you do
have centralized tools to handle documents, you might want to track individual
changes.
Figure 8.8
Shared documents. By tracking changes, you can record the suggestions by each
team member. The toolbar makes it easy to accept or reject the changes as they are
merged into the fnal document.
Over the past few years, Microsoft has introduced several changes to the Offce
software suite. Some of the most important changes have been made to support
teamwork. One of the most useful features is the ability to track changes in Word
documents. Figure 8.8 shows an example. To see the power of this option, begin
by creating a new document and typing a couple of sentences. Then choose the
Review/Tracking/Track Changes option. Now add a new sentence and delete or
replace a few words from an existing sentence. As you make the changes, Word
adds markup notes. You can send the document to other people on the team and
all changes they make will be marked. When you get the document back, you can
quickly see the changes that were made. The Reviewing toolbar contains icons to
help you fnd the changes and accept or reject the changes with a single click. The
markup annotations are commonly used for Word documents. For spreadsheets,
you need a few more steps because the annotation is different.
The protect document options provide additional control over the document.
You have the ability to control what changes (if any) other people can make. For
instance, if you need to maintain consistency with other documents, you can pre-
vent people from adding styles. You can also limit which team members can edit
your document or allow them to edit certain paragraphs or cells in a spreadsheet.
Another option is to not allow people to change the actual document but give them
the ability to add comments. Adobe Acrobat has similar features that enable you
to specify the items that can be changed by other users. You can password-protect
the document and limit users to entering form data or allow them to modify the
document.
Excel uses the protect document option under the Review ribbon bar to restrict
editing to specifed people. The Sharing feature also tracks the changes that ev-
eryone makes to the spreadsheet. Once you have enabled sharing, you can use the
Track Changes option to highlight recent changes. The system also keeps a history
of all changes, so you can rollback changes to a prior versiona useful feature if
someone makes a major mistake or accidentally deletes important information.
The ability to track changes and restrict changes is a powerful tool for team-
work. At some point in creating work you need to lock down the changes. Even
if everyone initially contributes to a document, you need to freeze the changes so
everyone can work forward from the same point. If people keep going back and
changing costs or plans, it alters the rest of the project planning. The ability to
perform these tasks electronically makes it easier to share documents. You no lon-
ger have to worry about having multiple copies foating around and trying to fnd
them or trying to identify what changes were made.
Version Control
Version control is a powerful tool both for teams and for individual documents
because it automatically saves every revision. Most documents need multiple revi-
sionswhether they are text documents, spreadsheets, or artwork. Have you ever
created a spreadsheet, made dozens of changes, and then realized that you were
better off with the frst version? Multiply this situation by the number of team
members and throw in a changing environment. Your manager or client walks in
and says that last weeks version was much better. Version control systems save
multiple copies of every document. As shown in Figure 8.9, each revision is saved
independently along with comments and the name of the person who modifed the
document. At any time, team members can revert back to an earlier version. You
can also open an earlier version and copy just a few elements and paste them into
the current version.
Most version control systems work from a database. You check out a fle when
you want to change it. When you are fnished making changes, you check in the
fle and add comments about the changes. The comments can be almost anything,
but you want to add enough detail so others understand what was changed. The
members of the team should discuss the types of comments and level of detail
that you expect to see in the version control system. Ideally, the version control
system should be integrated into your fle server so that versions are automatically
maintained, or at least the team members are reminded to check out/check in the
documents.
Version control can be provided in multiple ways: (1) Embedding the versions
within the fle, (2) Windows keeps versions whenever a fle is saved, or (3) Using
a central database server to save versions. Microsoft Word used to support embed-
ded versions, but the option was dropped with Offce 2007. Excel still supports
a history of changes when you create a shared workbook. Embedded versions
are easy to create, but can be problematicthere are horror stories of companies
sending out Word fles to clients without removing the prior versions. Windows
supports versioning through its System Protection architecture. By default, it is
enabled for the main system drive, but you need to manually activate it for ad-
ditional drives or partitions. With System Protection enabled, Windows automati-
cally keeps a shadow copy of a fle whenever you save a new version. You can
browse to the fle, right-click it, and choose Restore previous version to pick the
version you need. SharePoint, described in another section, has a more powerful
system for version control, where all copies are archived in a central database and
can be retrieved and managed with version control tools.
Figure 8.9
Version control. Each revision of the document is saved individually along with
comments. At any time, team members can rollback the changes and revert to an
earlier version. Or, they might simply refer to an earlier version and copy sections to
be pasted into the current version.
North 235 312
South 334 416
East 116 268
West 645 879
North 235 312
South 334 416
East 116 315
West 645 879
North 235 294
South 334 418
East 116 145
West 645 806
Monday, initial version, Bob Jackson
Tuesday, new East forecast from marketing
Thursday, marketing implements a new
forecasting formula
North 235 312
South 334 416
East 116 315
West 615 798
Monday, CEO throws out marketing
formula, reverts to old and modifies West
Wikis and Online Applications
You have probably used the Wikipedia, the free online encyclopedia created by
Jimmy Wales and Larry Sanger. The key feature of the encyclopedia is that any-
one can create and edit content. What you might not realize is that the underlying
Wiki software is available for organizations to use for their own purposes from
MediaWiki. Once you install the software, your employees can use Web brows-
ers to create shared content. Drafts can be created and edited by multiple people.
Links can be built between projects. Employees can return later to search for in-
formation. Effectively, the system can be used much as a giant bulletin board or
Web site, but it is easier to create and edit text, images, and video. Depending on
the situation, you could even open up the Wiki to your customers, so they can
share their stories about how to use your products and services. As another exam-
ple, the HRM department might use a Wiki to create and manage job descriptions.
Managers can share the defnitions instead of creating new ones from scratch. Em-
ployees could add elements over time, making the descriptions more dynamic and
more accurate.
As shown in Figure 8.10, if you need to share more formal documents created
by word processors, spreadsheets, and slide shows, you could consider using the
online applications, such as those provided by Google. These Web sites enable
users to create and share documents on a Web server, where workers use brows-
ers. The Web site maintains the underlying software, so you simply purchase a
monthly subscription for each employee. More importantly, the fles are stored
in one location that can be shared with anyone else in the organization. Because
it uses basic Web technologies, authorized employees can read and edit the fles
from almost anywhere in the world. The application providers (e.g., Google or
Microsoft) provide automatic backup and recovery services. Their servers handle
In-house Wiki
(Job Descriptions)
Web-based applications
Documents and
Spreadsheets
Data and documents are
shared on a Web server.
Figure 8.10
Wikis and online applications. Wikis make it easy for people to create work on
information together. Web-based applications are designed to share complex word
processor and spreadsheet documents.
most of the traffc, so users can work with fairly simple client computers and mod-
erate-speed Web connections.
Information Rights Management
What would happen if your marketing forecast containing your top customer list
and expectations was delivered to your competitors? Or what if your employee
list with salaries, bonuses, and manager comments was distributed throughout the
company? Obviously, these are dangerous outcomes. How do you prevent them
from happening? As explained in Chapter 5, you can store the documents on a
server and assign access controls so only certain users can read or change them.
But how do you prevent one of the authorized employees from copying the fle to
his or her laptop or e-mailing it to someone else? The standard solution is to sim-
ply trust your employees. But what if you need a more secure solution?
Information rights management (IRM) is one answer to these questions.
The technologies are still evolving and few standards exist. The latest versions
of Microsoft Offce contain some tools to help you control access to documents.
As shown in Figure 8.11, essentially every chosen document is protected through
encryption, and the encryption keys are stored on a central server. The creator of
the document gets to specify the list of people who can open or edit a document.
These people are authenticated against the server list, and the server issues the
appropriate decryption key to open the document with the specifed permissions.
Regardless of how the document was obtained (from the server, by e-mail, or a
copy on a USB drive), the user must frst be authenticated against the informa-
tion rights server before the decryption key will be issued. Furthermore, you can
Document
Document
Protected
Document
(Encrypted)
Workgroup
server
Permission
list
Information
rights server
UserID
authentication
Decryption
key
Figure 8.11
Information rights management. Each document is encrypted and the decryption keys
are stored on a special server. Users must be connected to the network to exchange
authorization identifers and obtain the decryption key. The document is unlocked but
users are allowed to perform only the specifed tasks.
specify a list of permissions that limit what each user can do with the document.
For example, you could prevent the use of cut-and-paste, or disable printing to
make it more diffcult for authorized users to make copies or spread the document.
The power of information rights management is that the permissions are en-
forced regardless of how the document was obtained. The main drawback is that
you need a server to identify the users and provide the decryption keys. Microsoft
provides a trial service to enable you to test the system, but ultimately, you would
have to pay to install a Microsoft server to implement their system. To activate
the Microsoft system, choose File/Permission within an Offce document. You
will need a Microsoft Live account or your own IRM server. With independent
servers, companies can create detailed customized permission schemes. The IRM
scheme also enables you to set an expiration date for content. After the specifed
date users will not be allowed to decrypt the fle.
Workfow
Workfow or business process management (BPM) is the concept that some
actions have to be performed in a sequence. In particular, documents need to be
delivered to and reviewed by various people in sequence. For a simple example,
consider a purchase order. You need to purchase some equipment or software for
your job. You begin by flling out a requisition form detailing the equipment and
how it will be used. You forward the document to your supervisor, who might ask
for some revisions or additional data. Once approved, it is forwarded to another
Figure 8.12
Business process management. A large purchase requires approvals and evaluations
by several groups. Each event triggers tasks by others. An automated system can
track the process and send messages to those involved as the process moves forward.
Purchase $59,000
Power $6,000
Network $7,000
Room $12,000
Estimate $84,000
Purchase
Purchase Request
Manager review
Cost analysis
1
2
Management
approval
3
manager or perhaps to the MIS department. With additional revisions and chang-
es, the document is forwarded to purchasing and accounting. These departments
create purchase orders and obtain responses and information from vendors. Some
of the vendor information might be forwarded to you for verifcation or approval.
Ultimately, the item is ordered, delivered, and installed. Remember that this was
a simple purchase example. Team projects can be considerably more complex
with more interactions and more approvals.
In the past, organizations created workfow systems using paper documents and
internal procedures. Each document contained sign-off lines to indicate how the
document was supposed to be circulated and to highlight the approvals. As shown
in Figure 8.12, electronic workfow systems are being developed that provide sim-
ilar features for electronic documents. Of course, you could just e-mail the docu-
ment to everyone for approval, but standard e-mail messages do not provide the
desired sequencing or the ability to track the current location of requests. Work-
fow systems work with e-mail but add the desired sequencing that you provide by
flling out a sequenced list of addresses. Each person receives the message with
attached documents. When fnished with the document, the message is forwarded
to the next person on the list. Ideally, the workfow system would provide feed-
back indicators to everyone on the list, so that you could check on the progress of
individual items or get a list of items that are likely to show up on your desk in
the next couple of days. Microsoft has embedded workfow features into Offce
and SharePoint, but more sophisticated controls are available through specialized
vendors such as Legato. Chapter 6 explains that workfow processes are also sup-
ported within ERP packages, because transactions involve approvals and over-
sight. Some systems support complex rules with conditions and multiple routing.
Some even provide programming languages to handle complex conditions and
interactions with other systems.
Group Decision Support Systems
How do groups make decisions? How can dozens of people present their ideas
and questions? How can you keep track of the various statements and problems
raised? Note taking at meetings is the traditional answer to these questions, but in
some cases, there is a better answer. Specialized tools such as a group decision
support system (GDSS) were created to answer these questions. Most versions
of a GDSS use a special meeting room, where each participant is seated at a net-
worked computer. A facilitator operates the support tools and keeps the discussion
moving in the right direction. Before the meeting, the primary decision maker
meets with the facilitator to establish the objective of the meeting. They set up
sample questions and design the overall strategy.
Typical meetings begin with a brainstorming session, where participants are
asked to think of ideas, problems, and potential solutions. They type each of these
into categories on their computers. The basic ideas and suggestions are stored in
a database and shared with the group through the networked computers. In some
ways, the sessions resemble online forumswhere people respond to comments
and ideas.
In terms of discussion and comments, the facilitator can choose individual
items and project them on a screen for the entire group to analyze. Participants can
write comments or criticisms of any idea at any time. This system is particularly
helpful if many participants come up with ideas and comments at the same time.
The computer enables everyone to enter comments at the same time, which is
faster than waiting for each person to fnish speaking.
Another feature of using the computer for the entry of ideas and comments is
that they can be anonymous. Although each comment is numbered, they are not
traced back to the original author, so people are free to criticize their supervisors
ideas. Anonymity reduces embarrassment and encourages people to submit riskier
ideas.
At various points, the facilitator can call for participants to vote on some of
the ideas and concepts. Depending on the software package, there can be sev-
eral ways to vote. In addition to traditional one-vote methods, there are several
schemes where you place weights on your choices. The votes are done on the
computer and results appear instantly. Because it is so easy to vote, the GDSS
encourages the group to take several votes. This approach makes it easier to drop
undesirable alternatives early in the discussion.
One useful feature of conducting the meeting over a computer network is that
all of the comments, criticisms, and votes are recorded. They can all be printed at
the end of the session. Managers can review all of the comments and add them to
their reports.
In theory, a meeting could be conducted entirely on a computer network, sav-
ing costs and travel time if the participants are located in different cities. Also, if
it is designed properly, a GDSS can give each participant access to the corporate
data while he or she is in the meeting. If a question arises about various facts, the
computer can fnd the answer without waiting for a second meeting.
Perhaps the greatest drawback to a GDSS is that it requires participants to type
in their ideas, comments, and criticisms. Most people are used to meetings based
on oral discussions. Even if they have adequate typing skills, a GDSS can inhibit
some managers.
Along the same lines, in a traditional meeting, only one person speaks at a time,
and everyone concentrates on the same issues at the same time. With a GDSS,
your focus is continually drawn to the many different comments and discussions
taking place at the same time. People who type rapidly and fit from topic to topic
will fnd that they can dominate the discussions.
In terms of costs, maintaining a separate meeting room with its own network
and several computers can be expensive. Unless the facility is used on a regular
basis, the computers will be idle a great deal of the time. When you factor in the
costs for network software, the GDSS software, and other utilities, the costs mul-
tiply. One way to minimize this problem is to lease the facilities that have been
established by a couple of universities and some companies (e.g., IBM).
The use of a GDSS also requires a trained facilitatorsomeone who can lead
discussions, help users, and control the GDSS software on the network. Hiring an
in-house specialist can be very expensive if there are only a few meetings a year.
Again, using facilities from an outside agency can reduce this cost, but it means
that someone outside your company is watching and controlling your meeting. Al-
though most facilitators are scrupulously honest, there might be some topics that
you do not want to discuss with nonemployees.
One way to overcome these limitations is to alter the approach to meetings. In-
stead of requiring everyone to get together at the same time in one room, meetings
could be held via network discussion groups. Each participant could read messag-
es, add comments, and vote on issues electronically at any time from any location.
Again, the Internet offers possibilities to provide these facilities, but it could be
a few years before organizations and managers can accept the changes required.
Knowledge Management
How does an organization remember past events? When an event
arises similar to one you dealt with in the past, do the employees use the knowl-
edge gained from the prior experience? What happens if several key managers
leave your company? These questions are critical to reducing costs, maintaining
continuity, and improving the organization. Every organization operates on some
type of business processes. Some frms are highly structured and spell out de-
tailed steps to follow in almost any situation. Other companies evolve implicit
procedures over time. Many organizations make the same diffcult decisions every
month or every year. Diffcult decisions can require the participation of dozens
of employees and analysis of gigabytes of data. It would be nice if the organiza-
tion could keep the knowledge gained from every decision and apply it to similar
RealityBytes:TheU.S.PatentOffce
Many people have criticized the U.S. Patent and Trademark Offce (USPTO) in the
past few years. The offce has issued several dubious patents that many people argue
should not have been granted because other systems already existed to do the same
thing. Patents are only supposed to be granted for new, non-obvious inventions.
Once a patent is granted, the owner gains a huge beneft in any court fght. Anyone
challenging it has the burden to prove that the prior work existed and was signifcant
enough to overturn the patent. The patent examiners are also overwhelmedfacing
a backlog of more than 600,000 applications in 2006. Patenting software applica-
tions takes four years from the date of submission to resolution. The area of patents
for computer code is also troublesome because much of the early work was never
published since patents for software have existed for only 20 years, there are few da-
tabases, and the examiners have little experience with computer code. Several people
and organizations have proposed projects that could help patent investigators. The
overall goal is to create a knowledge management system that would create a giant
library of existing practiceslargely by collecting code and tagging it with search-
able notes. Manny Schecter, an associate general counsel for IBM observed that the
[they] need a tool that will enable sifting through the code in such a way thats use-
ful to the patent examiner. We should be able to have it done this year. John Doll,
commissioner of patents, noted that theres a lot within the open source community
thats valid prior art, but because of how its stored, its not accessible to examiners.
We have a hope that if we have a standardized system, we can fnd it in the future.
Others have suggested using a Wiki approach to enable the development community
to contribute knowledge, but Doll noted that the problem we have is we have a stat-
ute and were restrained from opening the examination for anyone other than a pat-
ent offce examiner. However, he also suggested that it might be possible to allow
outside comments between the time the application has been made public and before
the patent is awarded. In the meantime, the USPTO is likely to continue granting pat-
ents for work that essentially copies things done 30 years ago. The complications are
the major reason software was not originally considered patentable, and why Europe
does not grant software patents.
Adapted from Eric Hellweg, The Patent Offce's Fix, Technology Review, January
13, 2006.
problems in the future. In the past, maintaining organizational knowledge was a
key management factor in retaining and promoting key employees. But in medi-
um-size and large organizations, turnover, distance, and the challenge of fnding
experts can make it diffcult to maintain and share the knowledge. So, some com-
panies have attempted to create knowledge management (KM) systems.
A KM is designed to store any type of data needed to convey the context of the
decision and the discussion involved in making the decision. While the system
might contain rules, it is primarily a giant database of easily accessible data for
experts. KM systems are designed to organize information and to assist people
in collaborative projects and research. The system can be relatively unstructured
and often consists of many individual cases. Decision makers can search the sys-
tem for cases with features similar to their current issues. The cases are cross-
referenced so that a decision maker might research one aspect of a case and fnd a
related issue. The links make it easy to explore the related issues, tying together a
variety of concepts and identifying related problems or consequences.
Because the feld is relatively new, the defnition of KM is somewhat nebulous
and many software vendors promote tools as useful for KM. One of the diffcul-
ties is that the decisions and knowledge required can be different for each orga-
nization. And organizations may approach decisions differently, so it is unlikely
that a single tool will be useful to every company. Instead, each company needs
to evaluate specifc decisions to determine whether it will be useful to explicitly
collect the information and process knowledge that was involved in making the
decision.
One of the biggest challenges with KM systems is creating an organizational
environment that encourages decision makers to store their knowledge in the sys-
tem. Initially, the system will have little data, and the early decision makers will
have to spend a great deal of time organizing their discussions and creating the
fles and links necessary to make the system valuable in the future. Companies
need to give managers enough time to consolidate their information and provide
incentives to encourage them to help build the new system.
Organizational Memory
Organizational memory represents the knowledge held within the company and
the processes and procedures used to perform the primary tasks. Imagine what
would happen to an organization if all of the managers retired on the same day
and all of the procedure books and documents were destroyed. All of the business
processes would have to be reinvented. Sometimes reinventing new processes can
result in more effcient systems (known as process reengineering). But it takes
time to re-create all of the knowledge.
Now, think from a more positive perspective. How can information technology
be used to improve organizational memory? Perhaps some events occur only rare-
ly and no one remembers them. Or, perhaps employees try to re-create the wheel
for every new event. If you can store the organizational memory in a searchable
database and make it easy to use, employees can turn to it to obtain fast, accurate
answers and even solutions to current problems.
One of the challenges to KM lies in getting everyone to use the system. Many
times it is hard enough to solve a current problem. Most workers will complain
about the lack of time to enter all of the issues into the computer. For that reason,
KM has to become part of the groupware project. Employees need to store the in-
formation so that they can share it with teammates. Organizing it for future use is
almost a secondary consideration. By using the system to gain immediate advan-
tages (collaboration and instant access to data), employees will have an incentive
to store the informationcreating the organizational memory in the process.
Service Processes
Service frms probably have the most to gain from the use of a KM system. Pro-
duction systems typically already have engineers and support frms testing new
ideas and tracking problems. This production knowledge is often saved explicitly
or discussed as a professional discipline. Service frms such as lawyers, doctors,
and fnancial institutions, have less specifc processes. Often the service is depen-
dent on one or two people within the frm.
As shown in Figure 8.13, documenting the processes makes it possible to im-
prove the quality of the service. For example, a major premise of the ISO 9001
(and related) quality standard is that all processes need to be documented. You
also need to assign measures that enable you to identify problem areas and mea-
sure your progress when you make changes. Theoretically, you could write all of
the documentation and assessment on paper forms and stash them in a fle cabi-
net. A more useful approach is to store all of the documents and assessment mea-
sures and comments in a KM system. The electronic system makes them available
throughout the company so that everyone can fnd the appropriate process, can see
the reasons and interactions for the process, and can suggest improvements.
Employee Client
North 235 312
South 334 416
East 116 315
West 645 879
Processes and
workow
Steps:
Client Interview
Project denition
Data collection
Forecasts
Assessment Measures
Quality
Productivity
Process evaluation and continual improvement
Figure 8.13
Service process improvement. A major step in quality improvement is to document
each step and provide an assessment measure. Ongoing evaluation of the processes
and measures shows you what areas need improvement and measures your progress.
Microsoft SharePoint
Where can you fnd a system that provides these groupware ca-
pabilities? The benefts of the tools are relatively strong, but will it be expen-
sive and require huge amounts of training to get workers to use these tools? In
some cases, the answer is yes. Depending on your needs, you can buy specialized
products in most of the groupware categories. On the other hand, Microsoft has
integrated groupware tools with Offce using the SharePoint product. As shown in
Figure 8.14, SharePoint is a set of utilities that reside on a departmental server de-
signed largely to share documents over the Web. Currently, Microsoft offers two
version of SharePoint: (1) SharePoint Server is a free add-on with Offce and
Windows Server, and (2) the enterprise version which must be purchased sepa-
rately. The discussion and capabilities described here are based on the free ver-
sion. Although the tools may not be perfect, they are relatively inexpensive and
easy to use. Consequently, you can install them and get employees using them
fairly quickly. In terms of the cost, keep in mind that you will probably have to
install a departmental server.
Most of the SharePoint tools use standard Web protocols, so authorized users
can use them even when they are traveling. A special Web site has to be created
to support the tools. Generally, this Web site should be run on a company server,
using standard security precautions. If you want higher security, you can run the
Web site as a secure site and encrypt the data transmissions. Small companies
might consider obtaining the service from their ISPs. Participants are required to
log in to the system, and security controls can specify detailed access rights.
Communication and Scheduling
It is always amazing to learn how diffcult it is to contact everyone on a team.
People work on many projects, at different times, in different locations. Conse-
quently, even simple information can be hard to share. Basic announcements are
useful for these situations. Announcements are short messages that are displayed
Figure 8.14
Microsoft SharePoint overview. SharePoint Server provides several useful tools to
support teams and collaboration. It requires running a server, but it also integrates
well with Offce.
Communication
Contacts E-mail
Chat Discussion
Survey Lists
Project Management
Tasks Schedule
Calendar
Collaboration
Libraries Shared documents
Tracked changes Version control
Wiki
Workfow
Routing
Alerts
Technology Toolbox: Collaborating with SharePoint
Problem: You need to collaborate on a project with several workers.
Tools: Several groupware and collaboration systems have been created in the last de-
cade, but SharePoint is probably the easiest to confgure and use since it is integrated
with the standard Offce suite.
The drawback to SharePoint is that you really need a departmental server running
the latest version of Windows Server software. You also have to ensure that everyone
has an account either on the server or on Active Directory. These issues are relatively
minor in a business setting, but they can be a hurdle in an educational environment.
Once you have the server confgured with the SharePoint extensions, you use the
administration tools to create a new team site and add members. You can specify the
role of each participant (browser, contributor, administrator). The system makes it
easy to e-mail each person to notify them of the site. When you confgure the server,
be sure to setup the e-mail server so participants can set alerts for changes.
You can confgure the home page of the team website with the Modify options.
For example, you can choose which items (Web parts) will be displayed and where
they are located. The main page is essentially a portal for the team that organizes all
of the work. Each person can also customize the home page with additional options.
For example, you can have the page display all of the contacts or provide links to
messages. Each day when you log on, you can check for urgent messages or examine
the progress on the task list. Programmers can create additional Web parts and cus-
tomize the data displayed on the site. Microsoft includes some pre-built parts such as
weather or stock feeds from MSN. By customizing the home page for each project,
workers can focus on the most important tasks and monitor progress. Additional fea-
tures and the standard tools are only a couple of clicks away for any team site.
Quick Quiz:
1. Why would you want to confgure the home page differently for different proj-
ects?
2. Is there a limit to the number of people you would want to place on a SharePoint
team?
3. Research the two products and list the additional features provided by Share-
Point Portal.
to everyonegenerally they are displayed prominently on the frst page. Lists can
be created to display timetables for tasks, contact names, or any other category
needed by the group.
The power of sharing lists through a Web site is that members of the team can
see the lists at any time. As authorized members make changes, everyone has ac-
cess to the current data. Moreover, the lists can be organized and searched by vari-
ous categories, such as deadline, project sponsor, or participant. Some standard
lists exist in SharePoint, such as contacts and schedules. Additional lists can be
created at any time. SharePoint also maintains e-mail contact lists for teammates
and for any other core contacts that you need to share. This address book is easily
accessible by everyone, and a single click launches your e-mail editor. If you need
more powerful communication capabilities, including the ability to conduct online
meetings, you will want to add the Microsoft Communicator server. If you have
a messenger service (e.g., Live Messenger or Communicator) installed on your
computer, while you are logged into the SharePoint system, it will show you a list
of other team members and tell you if they are online. You can click on the name
for each person online and send a quick message.
As shown in Figure 8.15, SharePoint includes a discussion system. It runs with-
in the standard system without relying on external tools such as Communicator.
Figure 8.15
Discussion. You can create multiple discussion threads so all comments on a topic are
grouped together. The system maintains the group memory so teammates can check
on the reasons for various decisions.
Most people are familiar with Web-based discussion groups. Discussion adminis-
trators establish topics and can specify the roles of the other team members, such
as the ability to read or reply to comments. Discussion groups are useful on team
projects to discuss issues that arise. The strength of the computer-assisted discus-
sion is that everyone has access to the comments, and the entire record is avail-
able if questions arise later. It also makes it easy to search for specifc problems.
Discussions can be created on any topic. Common business uses include overall
comments on scheduling, sharing research information, and discussing problems
that arise. You can also attach a Word document or spreadsheet to the original
message. Team members can make changes to the document and write comments
in the discussion list to explain the changes. However, the attachments can be a
little diffcult to fnd, so use this approach when you want to emphasize the dis-
cussion, not the document.
As shown in Figure 8.16, SharePoint also maintains a calendar that is tied to
the task list. You can use the calendar as a simple project management system.
The calendar shows scheduled events. You can defne recurring events such as
regularly-scheduled meetings that always meet at a certain time. The process of
creating events is similar to that used in Microsoft Outlook. In fact, you can in-
tegrate the SharePoint calendar with your Outlook calendar. You might want to
download the SharePoint events onto a laptop so you can check the calendar while
you travel.
The calendar is also tied to a task list that makes it easy to assign and monitor
tasks by employees. By scheduling tasks and assigning them to various workers,
you can easily see which tasks are on schedule and which are falling behind. You
can use the calendar and the task completion list to see if some employees need
additional help. The system does not offer all of the features of a true project man-
agement system, but it displays the basic schedules and is easy to use.
Surveys are useful for some business applications. In particular, they come in
handy when designing new systems. Generally they are used to obtain a quick
Event Calendar Recurring Event
Figure 8.16
Project events. The internal calendar integrates with the event list. You can create
tasks and assign them to team members. Recurring events are useful for regularly
scheduled meetings.
perspective on individual opinions. Paper surveys are a pain. Web-based surveys
are easy to change, easy to fll out, and can instantly report the data. As shown in
Figure 8.17, SharePoint includes a basic survey-building tool. You simply write
the questions by selecting a format, entering the question, and identifying the pos-
sible responses. When you post the survey to the SharePoint site, the other mem-
bers of the team enter their selections. The results are immediately available. One
advantage of the system is that the entire process is done through Web forms. Note
that the surveys are not available to people unless they are registered in the group,
so the technique does not work as well for public surveys. However, SharePoint
does support anonymous access, and the confguration system includes templates
for building Web sites. So you could use a SharePoint site to interact with public
responses, but it takes some effort to confgure. On the other hand, professional
survey design software is relatively inexpensive and has more features for creat-
ing and analyzing surveys.
Collaboration
Until recently, most organizations shared fles through shared directories on LANs
or via e-mail. Document libraries are simply Web-based folders that hold a related
collection of documentssuch as all work on a particular project. The fles are
accessed across the Web, so they are accessible to team members anywhere in the
world. Furthermore, the group leader can establish a template so that all docu-
ments have the same look.
As shown in Figure 8.18, once the site is set up, accessing the documents is
easythrough the familiar File and Open commands. Generally, you will create a
Figure 8.17
Create a survey. You add individual questions to the survey. They can have a variety
of formats. After flling in the template for each question, users can enter data over
the Web, and you can instantly see the responses on another Web page.
link to the directory in My Network Places so that you can fnd the documents
with one click. In a team environment, it is important to store your fles in a docu-
ment libraryinstead of on your personal machine. That way, everyone in the
group can read and contribute to the work. Once the documents are stored in the
shared library, some other powerful tools and options can be used to coordinate
the team, as described in the next sections. Note that direct access to the docu-
ments generally requires that each team member have the most recent version of
Offce installed. You will usually want to protect the document and track changes
so that everyone can see the changes that were made by others.
SharePoint can also handle version control. Moreover, version control systems
support check in and check out of documents, so that only one person can edit
a document at a timeminimizing the problem of needing to determine which
change to keep. It also tracks who made the changes and which team member is
currently using the document. If you want complete version control including the
ability to automatically track changes, you will also need Visual Source Safe, Mi-
crosofts version control software. As shown in Figure 8.19, you confgure version
control by setting the properties for the document library. You access the confgu-
ration options within the document library: Settings/Document Library Settings/
General Settings/Versioning Settings. Other options enable you to specify a de-
fault template for new documents to encourage teammates to apply a consistent
style so documents look the same. Furthermore, you can require that documents
be approved by the owner of the group site before they are made available to
everyone.
Figure 8.18
Shared documents. You can use the upload link to transfer a fle to the server for
sharing. You can also use the New Document link to add a Word document directly to
the folder. You can open and save any existing document directly over the Web.
Once version control is confgured, the process is automatic. When a team
member edits and saves a document, SharePoint automatically creates a new ver-
sion. From the document library, you can choose the drop-down action list to se-
lect versioning. As shown in Figure 8.20, you will see a list of the versions along
with dates and optional comments. From this version list, you can open prior ver-
sions to copy desired sections. Or you can delete an entire version, including the
latestcausing the system to revert to the prior version. However, in general, you
should avoid deleting versionsyou never know when the changes or data might
be useful. Instead, simply copy the older version, make the desired changes, and
save it as the new version.
Sharing documents presents additional problems, particularly concurrent
changes. What happens if two people alter a document at the same time? When
the versions are saved, which changes should be kept? One solution to this prob-
lem is to use Check-out/Check-in to lock the document. When you check out a
document, no one else will be able to edit it until you make your changes and
check it back in. If you plan to make major changes to a document and need to en-
sure that no one else alters it until you are fnished, you should use the check-out
option. Other users will be notifed that you have the document open and will be
allowed to read it but not make any changes.
Figure 8.19
Confgure library options. Use the server options to set up version control. Additional
options are used to specify default templates so all contributed documents follow a
standard style. You can require that fles be approved before they are made available
to the team.
Workfow
Although it is not as sophisticated as specialized business process management
systems, SharePoint provides a couple of tools to help with workfow. These tools
integrate the features of communication and collaboration. Routing documents
and messages is a useful workfow tool. For example, your manager probably
wants to review recommendations that you make, purchasing managers are re-
sponsible for approving the purchase of major items, and important documents
have to be approved by the legal department. Projects tend to have discrete steps.
Sometimes the steps are as simple as obtaining an approval and comments from
someone before proceeding.
As shown in Figure 8.21, it is straightforward to establish a workfow process
in SharePoint. Simply select the document from the library and choose the Work-
fow option. In this case, workfow means that the document and your message
will be e-mailed to a list of team members in order. The actions offered to each
person depend on the type of workfow that you choose. The two main types are
shown in the fgure. To access the other two, you need to change the site proper-
ties. The four types are: (1) Approval, (2) Collect Feedback, (3) Collect Signa-
tures, and (4) Disposition Approval. The Approval method sends the document
to each person and gives them options to (a) approve, (b) reject, (c) forward the
message to someone else to handle the approval, or (d) request changes. If one of
the recipients rejects the document, the process is interrupted and it is returned to
the sender. The Feedback option asks each person to enter comments. When the
last person has responded, all comments are returned to the originator. Collecting
signatures requires that the company have a digital signature system in place, and
it is integrated into and requires the use of Offce software (such as Word), not just
Figure 8.20
Version control. Choosing the version option from the document library screen shows
you the prior versions, who edited them, and the comments that were added. You can
open prior versions to copy sections or simply revert to the prior version.
Technology Toolbox: Meeting Space and IM
Problem: How can you meet with a few people online to discuss a document?
Tools: Several online meeting systems have been created to help you conduct a
meeting for a small group. Microsoft Meeting Space is a free tool that is built into
Windows Vista. You will also want to use Instant Messenger (IM) or Communicator.
More sophisticated systems are available as stand-alone packages or as Web servic-
es, but they can be expensive. IM and Communicator support both voice and video
connections as long as you have microphones and webcams.
The easiest way to start a meeting is to fnd the participants and connect through
IM frst. Otherwise, you can start a meeting and e-mail them an invitation. IM needs
to be confgured by each person with an e-mail account. Microsoft servers help users
fnd each other. Use Communicator if you need more privacy and security.
Setting up a meeting is as easy as having the host start Meeting Space, and enter
a meeting title and password. If the IM connections have already been made, simply
click the users name in the display list. Otherwise, click the link to invite people and
follow the instructions to save and e-mail the invitation fle.
Once the meeting connections have been established, you can start a program
and click the Share option so that others can work on the document or drawing. Ev-
eryone sees the changes, but only one person has control at a time. The fle and the
software remain on the computer of the person who shared it so it works even if only
one person has the software installed.
You can also share documents or PowerPoint fles as notes. Everyone can see
them, but only one person can change them at a time. Remember that you need rela-
tively fast connections (upstream and down) to transfer voice and video. If the con-
nection becomes overloaded, you can stop the video transmission or switch to tele-
phones.
Quick Quiz:
1. What problems are you likely to encounter if more than four or fve participants
try to use online meeting software?
2. Why is voice communication important in an online meeting?
3. Is a video connection critical to online meetings?

the Web browser. The Disposition option is a request sent to a list of people asking
if expired content can be deleted.
More complicated workfow rules can be created with InfoPath and Microsoft
Project. All of these tools integrate with some minor programming. For example,
a list can be created to describe the state of a document (draft, approved, fnal).
Then various conditions and triggers can be applied to specify conditions for each
state and how the document must be handled. For instance, two specifc people
must approve a document before it can leave the draft state and move to the next
steps. Or, a document must have three completed and approved fgures before it
is considered to be complete. These rules require some effort to set up, but once
established, the system enforces the basic procedures of the business. Moreover,
team members can check on the progress of a project to see what steps remain, or
identify which team member is holding up a particular document.
Another powerful workfow technique is to add alerts to the shared items. As
shown in Figure 8.22, you can set an alert so that you are notifed when a document
is changed. For example, you might need to obtain your supervisors approval be-
fore continuing with a document. Save the document in the shared folder, notify
the person that it is available, and then attach an alert to the document. When the
document is revised, you will be notifed by e-mail automatically. You can avoid
bothering your manager, or wondering if he or she has gotten to the document yet.
Note that you should inform everyone about these capabilities, partly so that they
can take advantage of them, but also to minimize privacy issues. You can attach
an alert to almost any object in SharePoint, including discussions.
SharePoint Confguration
Because SharePoint has so many options, it takes some time to learn to use Share-
Point effectively. Once a site is confgured, the basic options are generally acces-
sible through tabs or drop-down menus. Figure 8.23 shows some of the options
confgured for this particular site. Many of the options can be added or removed
by editing the site properties. Additional templates provide other functionality. For
example, the Site Directory template indexes all of the subsites and workspaces so
Figure 8.21
Workfow. SharePoint supports basic workfow options that will cause documents
to be e-mailed sequentially to a list of team members. The four primary options are:
approval, collect feedback, collect signatures, and disposition approval.
Figure 8.22
Change alert. Almost all shared documents and lists support alerts. Whenever a
document or folder is altered, an e-mail message will be sent to you. You can use
alerts to track the progress of documents as they are revised by various people.
Figure 8.23
Document options. As you customize the site, additional options and actions become
available in the tabs and drop-down menus. Users should explore the various options
to understand the power and value of each choice.
Reality Bytes: Teamwork Tools
Several online tools have arisen to help people work as groups. Many people are fa-
miliar with the basic fle-sharing tools of Google and Microsoft. But more advanced
tools exist to support brainstorming, editing, and reviewing.
Tool Site
Associations/Brainstorming
Mind Maps www.mindmapping.com
Bubbl www.bubbl.us
Mind Meister www.mindmeister.com
Spinscape www.spinscape.com
Shared Workspaces and Drawing
Show Document www.showdocument.com
Twiddla www.twiddla.com
Document Creation and Review
Writeboard www.writeboard.com
Google Docs docs.google.com
Zoho www.zoho.com
Microsoft 365 www.microsoft.com/en-us/ofce365/
online-software.aspx
Review Basics wwww.reviewbasics.com
Web Site Creation
Wikidot www.wikidot.com
Wikispaces www.wikispaces.com
Project Planning
Wrike www.wrike.com
WizeHive www.wizehive.com
Basecamp www.basecamp.com
Meetings
Adobe https://acrobat.com/welcome.html
Microsoft Live Meeting www.livemeetingplace.com
Cisco Webex www.webex.com
Go to Meeting www.gotomeeting.com
Log Me In www.logmein.com
Adapted from Logan Kugler, Online Collaboration on the Cheap: 20 Free and Low-
Cost Tools, Computerworld, June 3, 2010.
users can select sites by hierarchy, department, or through a search. You also use
the site properties to control security and access permissions.
It is also possible to substantially alter a SharePoint site. In addition to basic
templates and color schemes, you can use the SharePoint Designer to integrate
additional functionality into any site and to develop new templates and options.
SharePoint is built on top of Microsoft ASP .NET, which is a substantial Web
server programming language. Using Visual Studio, programmers can develop
custom modules and integrate them into SharePoint, or retrieve SharePoint data
and perform analyses. The process requires a trained programmer, but it means
that almost any type of confguration is possible. For example, you can use forms-
based authentication to enable external business partners to contribute to your
SharePoint site. SharePoint also provides Web Parts that are modules to integrate
some specifc complex actions into SharePoint. For example, the Excel Web Part
uses SharePoint (.NET) code to programmatically build new Excel worksheets.
Users can enter data on forms and your code can transfer that data to a custom
worksheet, which can then be added to a document library. These techniques will
take time to develop, but they can be used to automate relatively complex tasks.
Peer-to-Peer Collaboration
What if you do not have a server and want to collaborate? Share-
Point has many powerful features, but it requires setting up a special server to
hold the documents and handle the Web communications. Similarly, the open-
source collaboration tools rely on setting up an account with a third-party Web
site and adding everyone as users. A central server has some signifcant strengths;
such as version control, automatic backup, searches, and being easy to fnd. How-
ever, sometimes you do not need that much power, and do not want to spend the
money and time creating a central system. Also, you might want to use a tool that
supports immediate interactionwhere everyone can make changes and work on
a document simultaneously. However, remember that in any peer-to-peer situa-
tion, you are constrained by the limitations of NAT. If a person you want to con-
nect with is outside of your local network, you will need a server at least to fnd
the other computer and establish the connection. Microsoft has released several
specialty servers (Communicator Server, One Note Server, and so on), specifcally
designed to create these connections. If your company does not run these servers,
you will need to pay a third-party (e.g., Skype or Microsoft Live) to establish the
connection.
Sharing Documents
You can always resort to using e-mail (or USB drives) to share documents if you
do not have access to a server. But sometimes you want to share fles directly. As
shown in Figure 8.24, Microsoft provides a second sharing tool that has features
similar to those in SharePoint server. Confusingly, in 2010, the tool was renamed
as SharePoint Workspace. The main sharing session is still named Groove. All
participants need a copy of SharePoint Workspace installed on their computers.
Participants also need to be able to connect their computers. If everyone on the
team works for the same company, you should be able to select people from the
company network. Otherwise, you will need a server to coordinate users across
networks.
The system works by creating a shared workspace on each computer. A com-
puter can handle multiple workspacesfor different projects or teams, but creat-
ing too many will slow down communications. One user initiates a new work-
space and invites other people to participate in the session. At that point, Groove
creates a folder on each of the machines to store shared documents. Any document
created on one computer is replicated to the folders on all of the other machines.
Users who are traveling and not connected to the network can still work on all of
the fles. When the computer is reconnected to the network fles on all machines
are synchronized by exchanging all of the changed data. Similar to SharePoint,
Groove also supports shared calendars, a discussion system, and a meeting tool.
It also provides alerts to notify you when a fle has been changed. Groove lacks
some features, such as version control. However, it includes a SharePoint Files
tool that makes it easy to publish the documents to a SharePoint document library.
One possibility is to use Groove to work on preliminary versions of documents
with a few people on your team, and then publish major versions to the SharePoint
server to make them available to a larger team.
Groove also supports several tools for direct communication. For example, it
contains a discussion tool that tracks comments and replies. A shared calendar can
store tasks and meeting dates. More interestingly, Groove supports an interactive
notepad and sketchpad. These tools are shared in real time with participants who
are currently connected. For example, with the sketchpad, multiple people can be
drawing items at the same timeand all changes will be displayed on everyones
screen.
If you are familiar with the Microsoft One Note application, you should know
that it also supports a couple versions of sharing. One Note is a unique appli-
cation that is designed to provide features similar to a paper notebook. You can
Workspace
Replica
Synchronized
when connected
New Workspace
Files
Discussion
Calendar
Workspace replica
Figure 8.24
Microsoft Groove. Groove creates workspaces on each team members computer.
The fles are replicated onto each machine and synchronized when the computers are
connected. Groove also supports discussions, alerts, calendars, and meeting notes.
create notes by typing, using a pen on a tablet PC, drawing or importing fgures,
adding almost any type of fle or Web site, or adding audio or video notes. You
create pages on which you can put almost any content, and then you organize the
pages to make them useful. One Note also makes it easy for you to search your
typewritten notes. The Shared Notebook feature makes it relatively easy to share
a collection of pages with team members. The key step is that you must create a
master copy and place it on a folder that is accessible to all team members. They
can download and work on the notebook, and their changes will be synchronized
with the master copy. If you have a small offce, you can use a shared folder on a
desktop to hold the master copy. In larger companies, you should consider using
SharePoint to hold the master copy. It is easier for users to fnd the documents and
SharePoint maintains version control.
Immediate Interaction: Microsoft SharedView
Immediate interaction is a second type of peer-to-peer sharing that is more excit-
ing. Simple communication tools, such as instant messenger, count in this catego-
ry, but more powerful tools exist. For example, Microsoft SharePoint Workspace
supports live connections for notepad and sketchpad. Microsoft One Note also
supports Live Sharing. Another Microsoft product for interaction is SharedView,
which can be downloaded for free. All of these tools enable multiple people to
connect to a computer and work on a document at the same time. Changes made
by one person are immediately displayed on everyones computer. Of course, it
can get confusing if everyone tries to type or draw at exactly the same time, so
people usually take turns. The point is that everyone can see the changes imme-
diately. This process makes it easier to spot mistakes, fnd new directions, or so-
lidify ideas.
Shared Document
Each person sees the same
document or application.
Even if they do not have the
software installed on their
machines. Any person can be
given control to type text or
control the application.
Figure 8.25
Meeting Space. The host starts a meeting and invites others to join. The host starts
an application on his computer and shares it with the others. Everyone sees the same
document and can be given control to make changes and run the application.
These tools work the same way, so this discussion will focus on SharedView
because it is free and is relatively powerful. When you use these tools, you will
also probably want to make a connection with a voice or video communication
tool such as Messenger or Skype so that you can talk to the other people while you
are making changes.
As shown in Figure 8.25, once the computers are connected through the meet-
ing software, you can share applications, or even your desktop. As a host, you
start SharedView and log in. Then you start a new session, which generates a ses-
sion name and password. The tool also creates a simple invitation with a link to
your session that you can e-mail to other people. The link routes them through
Microsofts SharedView server to make the connection.
The entire purpose of SharedView is to enable multiple people to work on the
same document at the same time. The nice part is that you can share any applica-
tioneven if it is not installed on the other computers. You can even share the
desktop, so someone else can run your entire computer. Okay, that choice is a little
scary, so be careful who you grant permission to run your desktop. It is useful if
your project involves multiple applications or if you need help from an expert. Try
sharing a drawing application such as Word, PowerPoint, or Paint. Figure 8.26
shows a simple example using Microsoft Word. Note that track changes is enabled
to make it easier to see what changes are being made. Each person can draw, add
text, or bring in images or other fles. Everyone gets to see the changes as they are
made. Remember that there is only one copy of the documentthe fle is stored
Figure 8.26
Microsoft SharedView. Up to 15 people can work on a single copy of a document,
application, or desktop. Only one copy exists, but everyone sees the same screen.
You are sharing the screen, keyboard, and mouse.
on the hosting computer. The other computers are working with a simple screen
copy. You can always e-mail the fnal fle to everyone. SharedView also lets you
share any saved fle as a handout which can be retrieved immediately by the other
participants.
SharedView works with a relatively small number of people, but if you get too
many people working at the same time everyone will interfere with each other.
However, note that the system limits control to one person at a time. The person in
control has the ability to make changes to the document. The host can select who
is in control, but everyone sees the resulting changes. This tool can actually be
useful even if everyone is in the same room. Although only one person can enter
text or make changes at one time, it is easy to switch among the participants to let
people take turns.
Sharing with Multi-Touch Panels
What could you accomplish if fve people work on the same
screen at the same time? Several companies are developing computers that
use an entirely new form of interaction. It is known as a multi-touch panel. Es-
sentially, several video cameras beneath the display observe touches and gestures
by multiple fngers, which includes multiple people, at the same time. Professor
Jeff Han astounded people with his presentation in February 2006 at an annual
technology conference. Search for it on YouTube. The applications were mainly
graphical and geographical, but the ideas offer profoundly new ways of interact-
ing with the computer. Basically, users touch the screen and manipulate objects
using various motions. From a team perspective, the interesting part is that several
people can work on the same display at the same time. As shown in Figure 8.27,
Figure 8.27
Microsoft Surface. Multi-touch screens create a new interface for interacting with
objects on the screen. They also enable several people to work on objects at the same
time. (Photo courtesy of Microsoft).
in 2007, Microsoft announced a coffee-table-size computer called Surface based
on a similar technology. Essentially, users can sit around the display table, and
work on a project together. The system can even recognize (some) objects when
they are placed onto the display table. The technology might remain expensive
for a few years, because it requires fast computation speeds to make it look good.
Other vendors, such as PQ Labs, are now offering glass overlays that can be at-
tached to a standard large panel LCD monitor to enable multi-touch interaction.
Because the technology is new, few people have had the opportunity to work
with it. The bad news is that it will take a few years to develop good business
and data applications. The good news is that you could invent a new way of hav-
ing people work together, build the system, and sell it for tons of money. The
initial applications are likely to be graphically oriented, because they are easier
for people to understand and to have several people contribute at the same time.
But, the basic groupware tools described in this chapter could be modifed to work
with this hardware. People could work on different elements at the same time. For
example, one person could work on fnancial data in a spreadsheet while another
works on a presentation slide. Data from the spreadsheet could be dragged over
to the presentation when it is complete. Or, perhaps two people could search for
information at the same time and combine the results to do a quick comparison.
One drawback that will slow the development of software is that all of the oper-
ating systems (Windows) are built to work with only one mouse and one keyboard
at a time. The application focus controls where typed data is entered, and Win-
dows and other applications currently support only a single focus point. Eventu-
ally, this limitation can be overcome, but it will take time for developers to change
the old applications.
File Formats and Standards
How do you collaborate and share data if everyone uses differ-
ent software? To put it another way, why is this chapter so reliant on Micro-
soft technologies? The answer to the second question largely comes down to the
fact that Microsoft Offce has about a 95 percent market share (Business Week,
July 3, 2006) and at least 80 percent in businesses (Forrester, 2009: http://www.
networkworld.com/news/2009/060409-forrester-microsoft-offce-in-no.html).
The underlying issue is that if you want to share documents, everyone needs to
be able to read the fles. Microsoft has largely relied on proprietary fle formats.
Even though some programs exist to convert fles into other formats, the conver-
sions are never perfect; and they are likely to lose the sharing properties in the
process. To collaborate with shared documents, all participants need to be able to
read and write the fles in the exact same format. Historically, this compatibility
could be accomplished only if everyone used the same softwareand even the
same versions.
When programmers develop software, they need to design a way to store the
associated data. Historically, most companies have developed proprietary binary
fle formats that are optimized for their specifc purposes. Proprietary formats had
the side beneft of encouraging everyone to adopt the same software to avoid the
conversion hassles. The problem of transferring binary fles is bad enough when
everyone uses the same basic hardware. Binary fles store data in the underlying
format of the processor on which they were created. These formats vary for differ-
ent types of computer processors. The World Wide Web was actually designed to
solve this particular problem. Physicists wanted a way to share their research, so
Web pages are defned as simple text fles that can be read by any type of comput-
er, along with the markup language (HTML) to display pages consistently. How-
ever, HTML is not powerful enough to handle complex documents, so the world
is still searching for a method to share word processing documents.
The International Organization for Standardization (ISO) approved the Open
Document Format (ODF) in 2006 as a world-wide standard for storing common
offce documents. Conceivably, if enough vendors adopt this format, it will be eas-
ier for competitors to offer new tools, particularly collaboration tools, that work
with documents created from any software. Support for most of the ODF stan-
dard is built into Microsoft Offce (use File/Save As and pick the Open format).
But these formats simply enable people to read documents with other software.
Standardized support for collaboration notes and security are weak. Microsoft has
made its fle formats more accessible, but it is not clear that other companies will
adopt the features.
Cloud-Based Collaboration
Another way to share documents and collaborate is to use one of the online brows-
er-based systems. As browsers gain more power, it has become possible to create
relatively complex applications that run on the browsers alone and store the data
on a Web server. The two leading sites are Google Docs and Microsoft Live. Both
offer free access with a limited amount of storage or you can purchase accounts
with more storage capacity and other options. For example, Google Docs offers
business accounts for $50 per person per year. These accounts could be used by
one person to create common documents (word processing, spreadsheet, or slide
show). However, the real strength comes with the fact that the documents are on a
cloud server. With the documents in a central location, multiple people can access
them. By relying on Web browsers, it is not necessary for users to install software,
and the fles can be reached using almost any device from anywhere in the world.
Reality Bytes: How Do Workers Communicate?
In 2009, Forrester Research conducted a study of 2,001 American workers at me-
dium-to-large organizations in terms of their technology usage. The results showed
that several predicted trends have not yet taken root in mainstream corporations.
Only one-third used laptops for work and one in nine had a smartphone. Almost 20
percent shared a PC with a coworker. In terms of communication, e-mail was the
most frequently used methodwith 60 percent of the workers saying they checked
e-mail at least hourly. Almost three-fourths said they never use instant messaging at
work. Most could simply talk to coworkers or use the phone. Many people blamed
lack of IT support for newer tools. Around 70 to 80 percent of the workers never
use groupware tools, including Web conferencing, Microsoft SharePoint, social net-
working, or videoconferencing. Of those who did use these types of systems, less
than 20 percent said they were experts and about the same number said they were
very satisfed with the tools. At the same time, 45 percent of them said they spend
three or more hours a week looking for important information.
Adapted from Eric Lai, 8 Ways the American Information Worker Remains a Lud-
dite, Computerworld, October 12, 2009.
Figure 8.28 shows an example of a simple spreadsheet created in Google Docs.
The tools provide basic functionality and are built using commands you are al-
ready familiar with. The Google system relies on menu commands, which you
can browse to see what features are available. Partly because of limitations with
browsers, functionality is somewhat limited compared to the PC-based Offce pro-
grams. For example, there is no simple copy-and-paste operation to move a chart
from a spreadsheet into a document. Instead, it has to be saved as an image fle.
On the other hand, sharing a document is straightforward. Choose the Sharing
option in the document list and you can make the document private, public, or
share it with a specifc list of users. You can send a link to the document to people
who need to review or edit it. Google also supports up to 50 people editing the
same document at the same time (10 for presentations). Changes made by each
person are updated and pushed out to the other copies in real time. The names of
each collaborator are listed at the top of the page and you can open a chat window
to share comments.
Microsoft Live also has a set of online editing tools (www.live.com or www.
microsoft.com/live). Live started with a focus on communication tools (Hotmail
Figure 8.28
Spreadsheet in Google Docs. Spreadsheets and documents support standard features
including layout, functions, and charts. Usability and features are somewhat limited
because of the reliance on browser capabilities but it is easy to share access to the
documents.
and Messenger), so the frst page still emphasizes them. Click the Offce link to
get to the main documents site (offce.live.com). Figure 8.29 shows that the tools
use a ribbon-based editing systembut there are fewer options than in the PC-
based tools. However, there is a button to download the fle and edit it with the
PC-based software if you are working with a PC that has it installed. So it is easy
to move back and forth between your local computer and the server-based copies.
Microsoft Live also makes it straightforward to share the online fles with the
public or with a select group of friends. The system also keeps versions and au-
tomatically saves your work. With the 2010 online version, Microsoft introduced
simultaneous editing. Each user must be given edit permissions and open the fle
(spreadsheet) using the Web browser. Changes made by one person are pushed out
and displayed to the other people editing the document. You can test the process
with yourself by opening two separate browser windows and editing the same
document twice.
The primary features between Google Docs and Offce Live are similar, but the
Microsoft layout and commands are closer to the PC-based versions so it might be
Figure 8.29
Spreadsheet in Offce Live. Spreadsheets and documents support standard features
and can be downloaded and edited in the PC-based tools. They also can be shared
with other users.
easier to user. However, as a manager, you should be familiar with and have ac-
counts on both systems in case someone wants you to work with a specifc set of
tools. Both options are useful if you want to carry around a tablet computer. Most
of the early tablets do not have the ability to run common PC-based software,
but do support standard browsers. If you want to rely on a tablet, you need to use
cloud-based computing to create and store fles.
Summary
Working together and sharing data are crucial in todays companies. MIS can help
teams work better with tools designed to integrate data across an organization.
Managers need to know how to use a variety of tools, from data sharing over net-
works, to dynamic linking, to groupware products.
Teamwork begins with communication, and information technology provides
many ways to communicate, from voice mail, e-mail, text messaging, Web pages,
and blogs, to videoconferencing. As a communicator, one of your challenges is to
choose the appropriate communication channel for the specifc type of message.
Teamwork also requires scheduling and task coordination. Electronic calendars
and project management tools help everyone identify bottlenecks and keep proj-
ects on schedule. Meetings take big blocks of time from all managers, but they are
often critical to communication and the success of a project. Electronic meeting
systems make it possible to meet regardless of location. They also provide records
of the meetings that can be reviewed or searched later.
Collaboration requires sharing documents and knowledge. Shared document
features enable you to track changes made by all teammates. They also provide
version control so you can recover elements deleted from earlier versions. Work-
fow software helps you control the sequence of events and when people will re-
ceive each document. Information rights management provides tools to limit the
distribution of documents and control what people are allowed to do with each
document. Ultimately, an important goal of teamwork systems is to collect all of
the information used to make decisions and hold it as the organizations memory.
This knowledge management system can then be used for future decisions and
problem solving.
Microsoft SharePoint and Groove provide several tools to integrate the el-
ements of the Offce suite as well as provide communication and collaboration
tools. SharePoint makes it easy to share documents across the Web. It also main-
tains contact lists, events, and task items. All of the features are shared with team-
mates. Discussions can be created quickly on general topics or on specifc docu-
ments. Surveys can record votes on critical issues. When you need more interac-
tive tools, you can add communication and meeting tools as well as tools such as
SharedView that enable multiple people to work on documents at the same time.
Cloud-based tools including Google Docs and Microsoft Live make it easy to
share documents with multiple peopleusing only Web browsers. The tools work
for basic documents, spreadsheets, and presentations. The cloud servers make the
fles accessible to almost anyone and maintain backups and even version control.
Both tools also support simultaneous editing of documents so multiple people can
make changes at the same time.
Key Words
Web Site References
Teamwork Tools
Microsoft Live Meeting www.microsoft.com/livemeeting
Webex (Cisco meeting) www.webex.com
Version control comparison better-scm.berlios.de/comparison/
comparison.html
Microsoft SharePoint www.microsoft.com/sharepoint
KMWorld www.kmworld.com
Knowledge Board www.knowledgeboard.com
KMNews www.kmnews.com
KMResource www.kmresource.com
A Managers View
Teamwork is an increasingly important aspect of management. Integration
of business units so they work together is another important issue. Effec-
tively managed, the techniques can cut costs, improve quality, and improve
response time. Communication and collaboration are key elements to the
success of any team. Groupware tools support messaging and sharing docu-
ments. Business process management tools help automate collaborative pro-
cesses by adding triggers and rules. Ultimately, storing all information on
decisions contributes to the organizational memory, and the knowledge man-
agement system can help the company make better decisions.
business process management (BPM)
check in
check out
concurrent changes
full duplex
group decision sup-
port system (GDSS)
Groupware
Information rights man-
agement (IRM)
knowledge management (KM)
protect document
replicated
synchronized
telepresence
track changes
Version control
Videoconference
Workfow
Review Questions
1. What communication channels are available to teams? What are the strengths
and weaknesses of each method?
2. Do current communication and groupware tools reduce the need for business
meetings? Explain.
3. How does network address translation present problems with collaboration?
How are these problems solved?
4. Why is it helpful to use collaboration tools when sharing documents instead
of just e-mail?
5. Why is version control important to teams?
6. How is information rights management different from typical computer
access controls?
7. What is workfow processing and how is it supported by collaboration tools?
8. What are common features of a group decision support system?
9. What features are important in a knowledge management system?
10. What are the major features of Microsoft SharePoint?
11. How does peer-to-peer collaboration differ from Microsoft SharePoint?
12. Why are fle formats so important in collaboration?
13. What are the benefts to using cloud servers for collaboration?
14. What benefts would large multi-touch panels provide for small groups?
15. Briefy explain the main capabilities of the following tools:
a. Microsoft SharePoint server
b. Microsoft SharePoint workspace/Groove
c. Microsoft SharedView
d. Google Docs
e. WebEx
f. Skype
Exercises
1. Can smart phones be used for video conferences? Can they be used with
phones on other systems (e.g., AT&T v. Verizon)? Why or why not?
2. Research and compare the meeting capabilities and costs of at least two
products (e.g., Microsoft Live Meeting, Cisco WebEx, Skype). What types of
meetings (e.g., size) is each best suited for.
3. Research the costs of installing several Microsoft collaboration servers,
including Communicator, Groove, and SharePoint.
4. For each of the following business situations, identify the collaboration
method you would use and briefy explain why.
a. A salesperson wants to demonstrate a complex software package to a
customer group in a different city. The customers can meet in one room.
b. A corporate vice-president wants to give a two-hour presentation of
new products to a local audience of at least 100 people, plus at least 300
company employees in multiple cities, and make it available to authorized
reporters.
c. A design team consisting of a group of four artists and a programmer, each
in a different city, need to create a new design for a large Web site.
d. Students in a club need to write a short document and create a slideshow
for an upcoming event.
e. A group of students and faculty in an investment club need to share data
and research with each other and keep records so students can follow the
work next year. Outside directors also want to be able to see selected data
and reports.
5. Most organizations have a process for submitting and approving a business
travel request. If possible, interview someone from a company or the
university about the steps involved in creating and processing a travel
request. Then describe the steps to create that process in SharePoint server.
6. Describe a business situation in which you would need to implement
information rights management controls. Describe the specifc rights you
would assign.
7. How is Skype similar to but different from Microsoft Messenger?
Technology Toolbox
8. As a team project, if you have access to a SharePoint server, have one person
create a small document that describes a computer that he or she might
wish to purchase. Set change subscriptions on the document for each team
member. Then over the next few days, have members make changes to the
document, with suggestions for different components. As team members are
notifed of changes, they should check the suggestions and either accept or
reject them by adding additional comments.
9. Use a SharePoint server to discuss a recent movie. Include links to other
review sites and if possible upload image fles or movie clips. Create a
calendar to assign tasks to each team member. The fnal result should be
a formal review of the movie. Be sure to include your instructor as a team
member.
10. With at least two people, choose an industry and assign a frm in the industry
to each person. Set up a Groove workspace. Have each person retrieve
sales and proft data for the assigned frm. Use the workspace to build a
spreadsheet that combines the data from each frm to draw charts displaying
revenue and profts for each frm.
11. Set up Windows Meeting Space and create a meeting with at least three
people. Together, build a short PowerPoint slide show to come up with a new
marketing campaign for an automobile manufacturer (pick one).
12. Compare Meeting Space to One Note/Live Sharing. If you do not have
access to the software, use the Microsoft documentation to evaluate the
features of the two products. Which tool would you prefer to use for a small
team project and why? Test both tools if you have access to them.
Teamwork
13. As a group project, assume that each person in the group is a manager of a
different department. Each person creates a spreadsheet to list the salespeople
in his or her department (4 to 10), their hours worked, total sales, and
commissions. Compute the totals for each column. Once the individual
spreadsheets have been created and stored on separate computers, the
group will create a composite spreadsheet that brings in the data from each
individual spreadsheet. Compute the corporate total and draw pie charts for
each column. If possible, use dynamic linking across the network to capture
the data from the individual spreadsheets.
14. Using SharePoint, have one person write a short document. Use workfow
to send it to each of the other team members for feedback. Submit the fnal
document with the feedback to the instructor.
15. Using Groove, (or SharePoint if Groove is not available) confgure a
workspace for a project to evaluate cell phones. Use the tools to discuss the
features and try to agree on a common cell phone that could be purchased by
a company for each of the team members. Hand in the fnal document and
recommendation, with the discussion comments.
16. Using SharedView (with PowerPoint) or SharePoint workspace/Groove to
create a listing of the name and major of each person in the group. Create
a small diagram as a map that shows your current location compared to the
campus library.
17. Connect as many people as possible using a single instant messenger or chat
room system. As a topic, discuss a current movie. What problems do you
encounter? Should you limit these types of meetings to a certain number of
people? What would be the maximum?
18. If your team has to work from home on a project for the entire semester that
involves drawings, documents, and discussions or meetings, which tools
would you choose and why?
19. If you have access to a SharePoint server, set up a team site to discuss
production issues. Generate queries to show production delays over time.
Put the data into a spreadsheet and graph it. Create a discussion group for the
spreadsheet and discuss possible reasons for any issues.
20. (team) If you have the Groove client software, set up a workspace and build
a spreadsheet to display and chart U.S. retail sales of bicycles for at least
six years. Each person should keep notes on which sites were searched
and where data was found. Comment on any differences in data found and
produce a best-estimate chart.
21. Describe the hardware and software Rolling Thunder should install so that all
employees can coordinate schedules and conduct 15-minute meetings every
day.
22. Choose a system that would help Rolling Thunder marketing managers
communicate with retail stores to show new designs and products with the
goal of reducing the number of physical visits by salespeople.
23. Identify documents that Rolling Thunder should protect with an information
rights management system and briefy describe the desired rights.
24. What information would you recommend placing in a knowledge
management system for Rolling Thunder Bicycles?
25. Explain how version control would be useful for designing the bicycles at
Rolling Thunder. Find software that could provide version control.
Additional Reading
Han, Jeff, http://www.youtube.com/watch?v=zwGAKUForhM. The multi-
touch video that astounded the world.
Rosencrance, Linda, Sidebar: A Collaborative Difference, Computerworld,
January 26, 2004. [Using blogs as collaboration tools.]
Schlender, Brent, Microsoft: The beast is back, Fortune, June 11, 2001, 75-86.
[Outline of Microsofts new ideas for integrating data.]
Verton, Dan, Exostars Collaboration Platform Takes Security to New Level,
Computerworld, June 7, 2004. [External collaboration tool with powerful
security controls.]
Weiss, Todd R., Groove Touts Performance, Usability In Updated Virtual
Offce, Computerworld, July 8, 2004. [Alternative desktop collaboration
software.]
Cases: Package Delivery
Te Industry
The package delivery industry has changed dramatically over the past few years.
In fact, today, it should probably be called the business logistics industry. Several
large mergers have altered the industry. The competition among the remaining
frms is intensifying. The primary players today are FedEx Corp., United Par-
cel Service (UPS), the United States Postal Service (USPS), and Deutsche Post
(which owns DHL and Airborne).
To understand the changes, you need to understand how businesses operate to-
day. The most important issue is that large companies operate in many locations,
within the United States and around the world. This means that companies need
to move supplies, assembled components, and fnished products. Parts are moved
from suppliers to assembly plants, and fnished products are shipped to distribu-
tion centers and retail stores. Few frms today want to move their own products
on their own trucksit just does not pay. And smaller frms could never afford to
own their own feets. Instead, the large frms have focused on effciency, time, and
price to be able to provide a complete level of delivery services around the world.
0
10
20
30
40
50
60
70
80
90
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
B
i
l
l
i
o
n

$
Annual Revenue
USPS
Deutsche Post
UPS
FedEx
-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
R
a
t
i
o
USPS
Deutsche Post
UPS
FedEx
Today, that encompasses everything from freight handling (large shipments) to
overnight packages (express delivery).
UPS was at the top of the revenue list in 2003 slightly over $33 billion. DHL
was second with $24 billion, closely followed by FedEx Corp. with $22.5 billion
(Whiting 2003). However, USPS topped all of them with operating revenue of
over $68 billion in 2003; but that includes money from its monopoly position to
deliver letters to households (annual reports). The basic rankings had not changed
much by 2011 Deutsche Post and USPS had similar sales numbers at the top of the
list, followed by UPS and FedEx. Of course, the big difference is that USPS was
experiencing billions of dollars a year in losses.
Services and Information
In many ways, UPS was the leader in driving changes in the logistics industry. The
company was initially formed to deliver merchandise from manufacturers to retail
stores. This third-party service made it possible for manufacturers to increase their
sales without having to build their own distribution centers. Even smaller compa-
nies now had direct access to sales across the United States. And all companies
could use the system to ship short orders in a matter of days.
The industry has changed considerably over time as the frms have introduced
new services and combined others. FedEx introduced the concept of next-day de-
livery. At a time when many people doubted there was demand for such a ser-
vice, the company owned and then perfected the market. When computers became
cheap enough, FedEx introduced package trackingemphasizing that informa-
tion was just as important as the package (FedEx Web site) and (Ulfelder 2002).
The other carriers struggled to provide the same service, but eventually UPS,
USPS, and DHL; as well as a host of other frms moved into next-day shipping. At
the same time, FedEx began to move into ground and freight deliveries, shipping
larger items across the country and delivering packages to stores and individuals
in a few days. FedEx grew largely by acquiring other frms. DHL has an interest-
ing historylargely driven by one man: Larry L. Hillblom, who died in a plane
crash in 1995. His focus was air delivery to islands and nations in the South Pa-
cifc, and for years, DHL dominated that international market.
Overall, the international market is a challenge for any shipping company, or
even any company that wishes to ship packages around the world. Each nation
has different customs requirementsin terms of forms, taxes, and even items that
are banned. To make it more complex, most nations have their own defnitions of
items and terms. Complicating it even more, the rules tend to change over time.
Now, if you are a manufacturer shipping only a single item to a single nation on
a regular basis, you will eventually learn the procedures. But if you want to ship
different items to multiple countries, how can you gain the knowledge you need?
Ultimately, the major shipping companies (not including USPS), have created
huge online knowledge systems that automate most of the steps in shipping over-
seas. You simply fll out the declarations properly and their systems produce the
necessary forms. UPS and FedEx have special facilities to notify customs agents
when packages are arriving and produce detailed information on request. In 2003,
the United States Department of Homeland Security (Bureau of Customs and
Border Protection) proposed a regulation that all carriers have to notify offcials
of inbound cargos several hours before they arrive. The notice would have to be
electronic so that the bureau could run computerized analyses on shipping data to
spot potential terrorist threats or illegal shipments. DHL, UPS, and FedEx gener-
ally already meet this requirement. The DHL system is based on Open Harbors
global trade-management system that has over 8 million trade rules (Chabrow and
Bacheldor 2003). Ultimately, it is likely that other major nations will require the
same advance notice, and all of the carriers and software vendors are building
systems to meet these demands. The key point is that shippers now offer much
more advanced services. They are not simply ferrying packagesthey are taking
responsibility for the entire process.
Many companies today use these shipping frms as the frst line of their logistics
service. Even internal shipments between factories can be handled and tracked by
UPS or FedEx. Particularly with the integration of freight, package, and overnight
deliveries, companies can fexibly select and monitor their shipments. UPS and
FedEx have divisions and special software to help companies use it as a complete
transportation system. They can schedule deliveries for different times and days.
They can even have UPS and FedEx integrate shipments. For example, if you buy
a computer from Dell, the monitor might have been shipped from Asia, unloaded
in a UPS warehouse, and then shipped directly to youwithout frst having to go
to a Dell warehouse. And it would be shipped to arrive at the same time as the rest
of the computer, which was shipped from the Dell assembly plant.
In 2006 and 2007, all carriers were hammered by rising fuel prices, both for
trucks and airplanes. Faced with increasing costs, IT needs became lower pri-
oritiesparticularly since no astounding new technology appeared. Rob Carver,
CIO, noted that for budget money it is not a stellar IT year, and that although the
FedEx security team likes features on Windows Vista, they see no need to switch
desktop operating systems when Windows XP is working fne (Preston 2007).
Competition
The major frms have consolidated to the point where they now offer essentially
the same types of services. The declining costs of technology have made it pos-
sible for all of the frms to compete with information systemsfor example, they
all offer package tracking. The rivalry is ferceparticularly between FedEx and
UPS. Both compete on the basis of global reach, quality, and price. Currently,
DHL/Deutsche Post is not in the same league within the United States, but domi-
nates in Europe and Asia. So, all frms have to focus on effciency, and effciency
often requires automationreducing the number of workers where possible, and
cutting times as much as possible. All of them have benefted from the increased
level of shipping in the past decade. Will that trend continue? Or will it level off,
forcing frms to compete harder to steal market share from each other? Informa-
tion technology has played an enormous role in many aspects of the competition.
From providing information for package tracking, to the knowledge bases for in-
ternational shipping, to route scheduling, automation, and cost cutting.
Teamwork
These companies are huge, and they operate around the world. Moving packages
and information quickly and effciently in this environment requires tremendous
teamwork. A large portion of the information technology is devoted to ensuring
that everyone down the line knows exactly what is coming. When customers enter
package data on the main Web site, the local driver knows what to pick up. Sched-
ulers communicate in real time with drivers to alter routes as needed. The pack-
age data is provided to the central hub, which optimizes the plane loadings and
fights. Data is sent to customs as well as to the local destination hubs, so drivers
know the loads and routes for the next day. All through the process, customers are
treated as team members and have access to the information. Customers include
the shipper as well as the recipient. Firms that ship products can be notifed of any
problems before they become a disaster.
Additional Reading
Chabrow, Eric and Beth Bacheldor, Rules Could Force Shippers To Modernize,
Information Week, July 28, 2003.
FedEx Corporate History, http://fedex.com/us/about/today/history/.
Preston, Rob, Down To Business: Whos In For Innovation? No, Whos Really
In? Information Week, May 12, 2007.
Ulfelder, Steve, Dateline: A Timeline Of Technology Advances At Fedex And
UPS, Computerworld, September 30, 2002.
Whiting, Rick Logistics & Transportation: Ship To Sure: Tracking In Real
Time, Information Week, September 22, 2003.
Case: FedEx
The history of FedEx (ticker: FDX) is outlined on its Web site. Federal Express
began with a paper that Frederick W. Smith wrote in 1965 while at Yale Uni-
versity. After serving in the military, Smith purchased Arkansas Aviation Sales
in Little Rock in 1971. It offcially began operations on April 17, 1973, with 14
small aircraft and delivered 186 packages to 25 cities. Deregulation of air cargo
in 1977 enabled Federal Express to fy larger jets and the company grew rapidly.
In 2004, the company had a daily lift capacity of more than 26.5 million pounds
and the feet traveled 500,000 miles a day. The endpoint delivery vans logged 2.5
million miles a day (company Web site). The company reached $1 billion in sales
in 1983a mere 10 years after start-up.
Mergers
In 1989, Federal Express purchased the Flying Tigers networklargely oriented
to fights in the Pacifc region, and giving Federal Express expertise in interna-
tional shipments. In 1995, Federal Express acquired Evergreen International Air-
lineslargely because the acquisition gave it rights to deliver in China. In 1998,
Federal Express expanded into ground deliveries by purchasing RPS (Roadway
Package System, renamed Caliber System in 1996). The purchase of Caliber
Systems also gave Federal Express Viking Freight. Federal Express expanded its
freight system in 2001 by purchasing American Freightways. In 1999, FedEx ac-
quired Caribbean Transportation Services, focusing on shipments to Puerto Rico.
FedEx now has a full spectrum delivery system, from air shipping to large freight
transfers to package delivery.
In 2000, FedEx acquired TowerGroup and World Tariff. This acquisition pro-
vided important knowledge and experience in handling international shipments.
It provides customs clearance and electronic fling for all of the other divisions
of FedEx, enabling FedEx to provide end-to-end transportation around the world.
The more surprising merger happened in February 2004 when FedEx acquired
Kinkos Offce stores. For several years, FedEx had a relationship with Kinkos,
so that customers could drop off packages at the local stores. All 1,200 stores in
10 countries are digitally connected, making it possible to print documents na-
tionally. The company evolved from neighborhood print shops into local business
centers for mobile professionals and small businesses. The company also handles
commercial print jobs.
Operations
The essence of package delivery seems easywhen you look at one package:
pick up a package, transfer it to a plane, fy it to the nearest airport, put it on a
delivery truck, and drop it off at the destination. The problem is that drivers have
to pick up many packages, all of them have to be sorted, planes have to be loaded
properly and fown to airports where you have landing rights, then packages have
to be unloaded, re-sorted into delivery trucks, and the drivers have to get to mul-
tiple destinations. Now, you have to do all of that fast and at the lowest possible
cost. And the mix of packages and destinations can change every day.
Scheduling and optimizing package handling and deliveries is a key element in
the operations. In 1978, FedEx put voice radios in the trucks to assist and direct
drivers. Only two years later, in 1980, the company launched a revolutionary pro-
prietary wireless data network called Digitally Assisted Dispatch System (DADS).
Dispatchers could send text messages to drivers. By reducing chatter and enabling
drivers to check the messages when they needed, the system improved effciency
30 percent in the frst day (Ulfelder 2002). In 1986, the company adopted the
wireless handheld SuperTrackers that capture data with bar code readers. The sys-
tem is the foundation for tracking packages at every step, and uploads data to the
Customer Oriented Service and Management Operating System (COSMOS).
Tracking the package at every step in the process is critical to the operations.
It provides information for the operations at the next stops, it provides data to the
customers, and it makes it almost impossible to lose a package. On average, every
FedEx package is scanned 23 times. Once a package is picked up and scanned,
the data is transferred to the central computer. At the shipping hub, the computer
routes the package to the proper conveyor belt. The package is loaded into an ig-
loo container that fts the sloping sides of the plane. The container has a scan that
represents all of the items held inside. While being routed to the proper container,
each package is weighed and the billing corrected if necessary. The weights are
also used to balance the loading of the plane. As the 160 planes land each night at
the FedEx hub in Memphis, workers unload them within 20 minutes. Packages in
the Memphis hub travel through a giant convey system to be routed to new planes
or trucks to the nearest of 22 airports. The system scans 1,000 bar codes per min-
ute. The hub alone generates about 900 MB every two and a half hours (Brewin
and Rosencrance 2001). The main FedEx hub in Memphis is a key element in the
operations effciency. FedEx can move 325,000 letters an hour through the hub,
along with 125,000 small packages per hour. The average transit time is seven
minutes (Brewin April 2004).
In 2003, FedEx rolled out a completely new handheld system. Dropping most
of the proprietary technology, the PowerPad was developed by Motorola to run
the Microsoft Pocket PC operating system. More important, the system uses the
GPRS cellular wireless service from AT&T. It is designed to completely elimi-
nate paperwork, saving couriers 10 seconds at each stop. Which adds up to $20
million in annual savings when you factor in the 3.5 million packages delivered
a day (Brewin November 2002). GPRS has data rates between 20 and 40 kbps,
which exceed the 19 kbps on the older FedEx network (Brewin March 2002). The
PowerPad reduces errors by automatically checking Zip codes. It can print labels
through a label printer in the truck that is connected by Bluetooth radio. It even
checks the weather at the destination so that the driver can notify the shipper of
any potential delays. The PowerPad contains a third radio transmitter so that it can
connect via WiFi to the networks at the FedEx hub.
FedEx runs several fully automated sorting hubs. Packages are delivered at one
end, then passed on conveyor belts through weighing stations and sorted by ZIP
code to be delivered at the other side of the DC onto waiting trucks. Ken Spangler,
CIO at FedEx Ground, notes that automation is required to handle the 7,500 pack-
ages an hour. He also noted that Over the last fve years, we have been on a mis-
sion to get faster and faster. We re-engineer and speed up our lanes an average of
twice a year, and sometimes more frequently than that. Over 80,000 ZIP-code-to-
ZIP-code lanes are at least a day faster now (King 2011). An interesting twist to
the FedEx DC is that vendors are required to provide a standard interface, instead
of proprietary connections. That way FedEx can combine systems from different
vendors to achieve the best results at the best prices. FedEx developers create the
overall software and can control the performance.
Services
Providing global shipping is considerably more complex than simply transfer-
ring a package. Each nation has hundreds or thousands of different rules on tar-
iffs and procedures for importing items. Increasingly, FedEx customers were ask-
ing for information on how to ship globally. FedEx began a project in 1997 and
launched it in August 2000. Originally, the Web-based system arranged shipments
between the United States, Canada, the United Kingdom, Hong Kong, and Puerto
Rico. It was then expanded to 20 nations. The system runs on an Oracle data-
base and provides the rules and appropriate import and export forms needed. It
also lets shippers know about restrictions, embargoes, and licensing requirements.
In February 2001, the Global Trade Manager system began computing govern-
ment fees, duties, value-added tax (VAT), and other charges so that shippers can
quickly estimate the total costs. Robert B. Carter, head of the division, notes that
Global Trade Manager makes it easier to ship internationally, which can be an
intimidating experience. It allows customers to do import/export documentation
in a straightforward fashion.We are giving them [clients] access to technology
they wouldnt otherwise have (Rosencrance 2002).
In 2003, FedEx partnered with Cap Gemini Ernst & Young to offer detailed
logistics services to companies of any size. Many large companies already out-
source freight transportation (67 percent), and 42 percent outsource management
of their distribution centers. But most frms still maintain control over the rest of
the decision aspects of logistics and purchasing (Bacheldor 2003). The new offer-
ing from FedEx and Cap Gemini integrates software for transportation-manage-
ment, fulfllment management, supply chain event management, and business-to-
business hub management. With the integrated tools, customers can examine and
track shipments and warehouse or distribution center handling at the stockkeeping
unit (SKU) or even serial number level. They will be able to see the movement of
all products and obtain reports on all aspects of shipping, even though the entire
process will actually be handled by FedEx. Douglas Witt, president of FedEx Sup-
ply Chain Services, notes that the system is important because lack of visibility is
a major issue for a lot of customers (Bacheldor 2003).
With the global and domestic presence of FedEx, U.S. government police agen-
cies often turn to the company for information about deliveries and customers.
Prior to 2001, FedEx generally declined to cooperate without court ordersfear-
ing retribution against employees and concern about customer privacy. Since the
terrorist attacks of 2001, FedEx now encourages its employees to watch for po-
tentially dangerous activity and it works with the FBI and other agencies. FedEx
provides access to its international customer databases, including credit-card and
payment data. The company also has the legal right to open and inspect any cus-
tomer package at willa power not even granted to police. Fred Smith, the found-
er, ordered the cooperation and sees it as an important service to society. Plus, the
cooperation helps prevent delays in package deliveries. FedEx also created a 10-
man internal police force that is recognized by the state of Tennesseegiving the
company access to police databases and the only private seat on the FBI terrorism
task force (Block 2005).
Strategy
Battling head-to-head with UPS and forging entirely new ways of doing business
have forced FedEx to focus on strategy. Technology plays a key role in its plans.
But Smith is understandably cautious. He shies away from huge projects designed
to alter everything. Instead, he says, We try to do a lot of work on the front end,
divide things into bite-size pieces, do things in a more evolutionary way (Foley
February 2, 2004). Exactly how Kinkos fts into that strategy is not completely
clear yet, but it will probably take time for the value to be realized. FedEx CFO
Alan Graf notes that it diversifes our business. Kinkos is the back offce for
hundreds of thousands of businesses (Foley February 9 2002). Fred Smith notes
only that it will enable both companies to take advantage of growth opportuni-
ties in the fast-moving digital economy. Jim McCluskey, a FedEx spokesman,
adds that it will enable FedEx to push information electronically for its clients
(Brewin January 2004). If nothing else, the stores will raise the visibility of Fe-
dEx even higherproviding a convenient location for small business services and
drop-off points for FedEx shipments. With the stagnant economy, a shift to digital
documents, and a tight battle with UPS, the market for overnight delivery has lev-
eled. FedEx is looking to the ground and international shipping to carry its growth
forward. Furthermore, the company has to focus on costs to improve profts (Fol-
ey January 2004).
In an interesting strategic arrangement, FedEx has two agreements with the
USPS that run through August 2008. FedEx uses its planes to help carry Priority,
Express, and First-Class Mail as well as the USPS international delivery air ser-
vice. The second agreement enables FedEx to place 5,000 drop boxes at U.S. Post
Offces in 340 metropolitan areas (2006 Annual Report).
Faced with the economic slowdown and rising fuel prices in 2006 and 2007,
FedEx made few changes in strategy. However, in 2007 the company announced
that it would spend $3.5 billion in fscal 2008 to build new shipping hubs in China
(Machalaba 2007). Most of the expansion in FedEx has been related to services
in China and India (2006 Annual Report). On the Kinkos side, FedEx has a deal
with Adobe to include a button in Acrobat Reader that sends PDF fles to any
selected Kinkos store for printing. Users can print documents at an offce near a
desired U.S. destination and have them delivered for a fee. Initially, 133,000 cus-
tomers had signed up for the service (Golden 2007).
Teamwork
One of the challenges with growing through mergers is integrating all of the infor-
mation systems. To maintain the existing effciencies, FedEx chose to run the new
divisions independently. But that means managers do not have a single view of
customer data, and customers have to check multiple locations to fnd their ship-
ping information. In 2004, FedEx began development of its Customer Fusion da-
tabase. The package will pull data from all of the independent systems and make
it available from a single point. The product is a computer messaging system that
can talk to all of the individual applications. Larry Tieman, chief architect for the
project, notes that the message bus allows us to loosely couple the business ap-
plication, apply the proper business rules, and build a unifed application layer
(Sullivan 2004).
The company also created an integrated system for international managers in
2001. The international strategic information system (ISIS) uses a Web-based
front-end query tool to integrate data from internal legacy systems for market-
ing, sales, and inventory. It then consolidates information for local agents into
an Oracle database. The agents can then query the database and identify patterns
such as the amount of shipping by industry. A key aspect of the system is to pro-
vide business intelligence data to local agents and FedEx managers. By sharing
the same data and analysis, FedEx can better help the agents meet revenue targets
(Dash 2001).
Because FedEx develops most of its own applications, it is concerned about
software development processes. As it explores options to outsource development
and hire teams from off-shore development frms, the company needs a method
to test software code. In 2006, FedEx developed a program with the University
of Memphis to fnd new methods of testing software, including multistage system
testing, risk-based testing procedures, and an expert system for systems testing
(Krill 2006).
FedEx encourages teamwork learning throughout the IT organization. The
company has a job swap program, where employees can request to trade jobs for
a certain length of time. The swap can include temporary assignments in other
countries. The technique provides employees with a broader picture of the busi-
ness, a chance to see other problems, and to share approaches to solving them
(Pratt 2007).
Questions
1. How does FedEx use information technology to improve teamwork and
deliver packages more effciently?
2. How can FedEx integrate data from its diverse companies to improve
teamwork, without replacing all of the individual systems?
3. Why is wireless technology so important to FedEx?
Additional Reading
Bacheldor, Beth, Better Visibility Along The Transport Chain, Information
Block, Robert, In Terrorism Fight, Government Finds a Surprising Ally:
FedEx, The Wall Street Journal, May 26, 2005.
Brewin, Bob, FedEx Expands Net Reach to Mobile Data, Computerworld,
March 25, 2002.
Brewin, Bob, Fedex: New Courier System Will Save $20M Annually,
Brewin, Bob, Fedex To Use Kinkos Stores To Offer E-Services To Enterprise
Customers, Computerworld, January 2, 2004.
Brewin, Bob, Sidebar: FedEx vs. UPS: The Technology Arms Race,
Brewin, Bob and Linda Rosencrance, Follow That Package! Computerworld,
March 19, 2001.
Dash, Julekha, FedEx Launches Global Business-Intelligence Application,
FedEx Web site: http://fedex.com/us/about/today/history/
Foley, John, FedEx Success Doesnt Come With Big IT Budget Increase,
Information Week, January 12, 2004.
Foley, John, Redefning What You Are, Information Week, February 2, 2004.
Foley, John, Retail Plus: FedEx Buys Kinkos, Information Week, February 9,
2004.
Golden, Mark, Adobe Links Acrobat, Reader To FedEx Kinkos Stores, The
King, Julia, Extreme Automation: FedEx Ground Hubs Speed Deliveries,
Krill, Paul, FedEx Seeks Improved Software Testing, Computerworld,
December 20, 2006.
Machalaba, Daniel, FedEx to Boost Spending As Slowdown Hits Sector, The
Pratt, Mary K., IT Job Swapping at FedEx for Fun and Proft, Computerworld,
September 10, 2007.
Rosencrance, Linda, FedEx Corp., Computerworld, March 11, 2002.
Sullivan, Laurie, FedEx Achieves Fusion, Information Week, June 30, 2004.
Ulfelder, Steve, Signed, Sealed and Delivered, Computerworld, September 30,
2002.
Case: United Parcel Service
The UPS (ticker: UPS) Web site has details on its history and growth. UPS began
in Seattle in 1907 as a messenger service, with most deliveries made on foot by
teenage boys. In 1913, the fedgling company acquired its frst car: a Model T
Ford. Merging with a couple of competitors, the company renamed itself Mer-
chants Parcel Delivery and began focusing on delivering merchandise from the
retail stores. The company also was the primary carrier of special delivery pack-
ages in Seattle for the U.S. Post Offce. In 1919, the company expanded to Oak-
land, California, and adopted its current name: United Parcel Service. After World
War II, retail shopping changed, as customers drove to malls and carried their own
packages. UPS fought to provide common-carrier service in new areas, including
Chicago. Expanding into competition with the U.S. Post Offce raised regulatory
issues, and the company fought many legal battles to win the right to deliver pack-
ages. In 1953, UPS began shipping cargo on existing airline fights. Yet package
delivery had to remain inside a state to avoid regulation by the Interstate Com-
merce Commission (ICC). Fighting for three decades, the ICC fnally granted
UPS permission to deliver packages between Montana and Utahin 1975. By
1985, the company was shipping packages by air to all 48 contiguous states and,
in 1988, started its own airline. In 1993, to track the huge number of drivers and
deliveries, and to keep up with FedEx, UPS introduced the handheld Delivery In-
formation Acquisition Device (DIAD). In 1999, UPS joined the New York Stock
Exchange.
Operations
Fighting competitors from day one, UPS pioneered effciency in deliveryem-
phasizing the need to hold down costs and prices. Although UPS headquarters
are in Atlanta, it runs its operations through its main Worldport hub in Louisville,
Kentucky. The 4 million-square-foot facility handles over 100 planes a night,
pushing 600,000 packages a day through the hub. Because of a limited labor pool
and the desire to hold wages down, the facility emphasizes automation. It runs
on almost 14,000 computer devices, including 30 terabytes of online storage and
5,500 miles of fber-optic cable. It uses 122 miles of conveyor belts to move pack-
ages and sort them by bar code scanners as they fy by (Brewin April 2004). Most
of the individual packages (93 percent) are identifed by special bar codes. When
shippers fll out the shipping labels on the UPS Web site, the system notifes UPS
about incoming packages. Twice a day, this incoming information is used to create
a sort plan that confgures the hub to match the expected packages. The software
Flexible Lineup Editor reorganizes the loading and unloading positions to opti-
mize the package transfers. Packages are only touched twice by humans: once to
be loaded onto the conveyor belts and once to be taken off and put into shipping
containers (Brewin April 2004).
UPS also pays close attention to its drivers and their routes. Drivers are often
monitored in terms of performance, and managers share tips on how to improve.
In 2003, the company added a new software routing tool to plan the routes. Since
most customers provide pickup data early in the day, the computer can plan the
optimal route, speeding up the drivers while they make more stops and travel few-
er miles. UPS CIO Ken Lacy estimates the system will reduce feet miles in the
United States by more than 100 million miles a year, saving about 14 million gal-
lons of fuel (Brewin September 26, 2003).
UPS had thought about creating a massive routing system since at least the
mid-1990s. The catch is that many people thought the problem was too big to
solve. In computer science terms, the problem is called NP-complete because
there is no perfect solution to analyze the entire delivery problem other than to ex-
amine every possible choice. And, with trillions upon trillions of options, no ma-
chine could fnd the perfect solution. Operations research specialists studied the
problem and found ways to narrow it down to just examining feasible solutions
that are close to optimal. With research and faster machines, they managed to re-
duce the time to fnd a solution from 90 days down to three hours. By optimizing
its primary routes, UPS was able to cut a day off its guaranteed delivery times for
some routes, such as reducing shipping time from Los Angeles to New York from
four days to three (Bacheldor 2004).
In 2003, UPS began redesigning its handheld system to the DIAD IV terminal.
The terminal contains a color screen and runs on Windows CE. In addition to the
bar code scanner, it includes a global positioning system (GPS) receiver. It uses
Bluetooth wireless to connect to a label printer in the van. When parked at the
hub, the system connects to the local area network via a WiFi wireless connection.
On the road, it uses a built-in cellular modem, so that drivers can connect to the
network and provide immediate information to customers. Ultimately, the system
will be able to use its Bluetooth radio to connect directly to customer computers.
The GPS system will be particularly valuable in Europe, where drivers handle
less frequent trips. In the United States, many companies require daily deliveries
and pickups to fxed locations. But the GPS system will also help route drivers
for last-minute pickup requests (Brewin September 26, 2003). In 2005, after two
years of development, the DIAD IV was rolled out to all 70,000 UPS drivers. The
company spent $22 million to develop the new devices based on Symbol technol-
ogy and said it would spend $127 million to deploy them globally (Weiss 2005).
At the same time, the company reported that it was experiencing problems with its
package-fow technology. The analytical system was designed to work with GPS
systems to analyze routes and determine the most fuel-effcient directions and
loading instructions. Full implementation was projected to be delayed by a couple
of years and the system was estimated to save $50 to 100 million in fuel costs
each yeardown from the $700 million originally projected (Rosencrance 2005).
By 2006, UPS was delivering an average of 15 million packages a day around
the world. But UPS still focuses closely on keeping up with FedEx. Using the
Internet, UPS is working to collect package data earlier and applying it to loading
trucks. When customers fll out the online form to generate a package label, the
information is sent to one of the UPS hubs. A computer system sets up a delivery
route and identifes where each package should be loaded in the truck, with earlier
deliveries stored in the front of the truck. UPS reports that drivers have gained
enough time to make an additional seven to nine stops per shifton top of the
100 pickups and deliveries they handled before the tool. Integrated with the new
DIAD IV, the system sounds an alarm if a driver forgets a package or stops at the
wrong address (Dade 2006). The technology also enabled UPS to introduce a new
service in 2007delivery intercept. For a $10 fee, senders can use the Web site to
stop the delivery of a package or reroute it to a new destination. UPS suggests the
service might be useful if a sender realizes at the last minute that the wrong data
or package was sent (press release March 26, 2007).
Capital expenditures at UPS were relatively stable over time. In 2006, the com-
pany spent $720 million on buildings and facilities, $1.15 billion on aircraft and
parts, $831 million on vehicles, and $384 million on information technology. In
total, the company spends about $1 billion a year on IT costs, including labor
In 2010, the UPS Web site was processing more than 26 million package track-
ing requests a day. But technology is also heavily used in operations to set the
most effcient package loading and to defne routes for delivery trucks. The hand-
held technology introduced in 2008 includes telematics, or information about the
truck including GPS and automotive sensors that include engine RPM, braking,
and speed. The goal is to fne-tune routing as well as track maintenance (2010 An-
nual Report). The new system took 10 years to design and build and focused on
seemingly simple things such as loading trucks more effciently. The automated
system puts tags on boxes so loaders know exactly which order to load them on
the truck. Developers also had to fnd ways to handle poorly-entered data (Mitch-
ell 2008).
Integrated Services
In 2001, UPS purchased the entire chain of 4,000 MailBoxes Etc. stores and re-
named them The UPS Store. The chain provides UPS with local contact points
for individual and small business pickups. But most of the stores are still run as
franchises. The stores are eliminating the need for paper shipping forms. Custom-
ers simply enter the data into a kiosk PC (or have the clerk enter it), and it prints
out the bar-coded label. In 2003, UPS took the interesting step of testing public-
access Wi-Fi in 66 of the retail stores in Chicago. Customers would pay either
an hourly or monthly fee to connect their own laptops to the Internet through the
WiFi connection. The demand for this service is still uncertain (Brewin September
11, 2003).
In 2002, UPS introduced a Web-based shipping solution to large customers
with offces in multiple locations. The system enables employees to ship a pack-
age by simply entering the details in a Web site. The system automatically bills
the company and provides a comprehensive report to managers. Best Buy is using
the system to make it easy for employees in its stores to ship packages to other
Best Buy locations or to send them directly to customers. Jim Hay, general trans-
portation manager of small package solutions for Best Buy, observes that because
employees do not have to handwrite the labels, this saves 15 minutes per user per
package. The biggest area of savings is in time and (in fewer) lost and misrouted
packages due to poor handwriting or a bad address (Rosencrance March 2002).
UPS began offering a useful service in early 2003 for exporters and importers.
The company launched an electronic payment system called UPS Exchange Col-
lect. Basically, UPS collects the money from the recipient/customer. Sellers are
free to expand to new markets with unknown customers with minimal risk. UPS
drivers will only deliver the product when they receive payment. When the ship-
per sends an item, the recipient is notifed electronically. UPS then collects the
money and delivers the package. The money is electronically sent to the seller.
The system provides huge gains over the current system. Currently, shippers often
wait 90 days for paymentand still run risks that the buyer will not pay. UPS will
transfer the funds within 10 days (Rosencrance 2003).
UPS also provides expertise in shipping products internationallyin terms of
handling forms and getting packages through customs. By identifying and con-
trolling every package that UPS carries, the companys Worldport hub qualifes
as a controlled building for the customs bureau. More important, the product in-
formation database is accessible to customs agents, who can scan it for potential
problems. Any suspect packages are instantly rerouted within the Worldport hub
to the customs offce for detailed inspection (Rosencrance March 11, 2003).
In 2004, UPS introduced new tools for customers to use the system as a com-
plete logistics solution. The Quantum View Manage software makes it easy for
customers to identify all shipping information. In particular, customers can see
all packages as a group, without needing to enter every tracking number. They
can see exactly where packages are located, along with any reasons for delays.
The system can even trigger alerts if problems arise and notify customer service
automatically, or use automated delivery notices to send electronic invoices (Ros-
encrance 2004).
UPS demonstrated the true power of its integrated logistics system in 2004 by
offering a new service to Toshiba. Toshiba outsources all laptop repairs to UPS
Supply Chain Solutions. Broken laptops are sent by UPS to its Louisville hub
where UPS engineers diagnose the problem, fx the laptop, and return it to its
owner. Mark Simons, general manager of Toshibas digital products division,
notes that diagnosing and repairing computers is actually the easy part. Instead,
moving a unit around and getting replacement parts consumes most of the time.
The actual service only takes about an hour (James 2004). UPS has been helping
Toshiba with repairs by performing an initial inspection since 1999. UPS has also
been servicing Lexmark and Hewlett-Packard printers since 1996.
Questions
1. Does it make sense for UPS to move into the computer repair business?
2. How does the Quantum View system provide new revenue or profts for
UPS?
3 .How does UPS use the Internet to reduce costs?
Additional Reading
Bacheldor, Beth, Breakthrough, Information Week, February 9, 2004.
Brewin, Bob, UPS Delivers Public-Access Wi-Fi To Its Retail Outlets,
Computerworld, September 11, 2003.
Brewin, Bob, UPS Invests $30M In IT To Speed Package Delivery,
Brewin, Bob, IT Drives the UPS Machine, Computerworld, April 19, 2004.
Dade, Corey, Moving Ahead, The Wall Street Journal, July 24, 2006.
James, Geoffrey, The Next Delivery? Computer Repairs by UPS, Business 2.0,
July 2004.
Mitchell, Robert L., Project Delivers Savings for UPS, Computerworld, April
21, 2008.
Rosencrance, Linda, UPS Launches CampusShip Shipping Tool,
Rosencrance, Linda, UPS Uses Technology to Speed International Shipments,
Rosencrance, Linda, New UPS Service Speeds Payments Between Exporters
and Importers, Computerworld, March 11, 2003.
Rosencrance, Linda, UPS Launches Quantum View Manage, Computerworld,
February 4, 2004.
Rosencrance, Linda, Planning System Isnt Fully Delivering at UPS,
UPS Web site: http://www.ups.com/content/corp/about/history/index.html.
Weiss, Todd R. UPS Deploys its Fourth-Generation Handheld Delivery
Terminals, Computerworld, May 9, 2005.
Case: DHL/Deutsche Post
DHL has an interesting history. The company was started and driven by Larry L.
Hillblom who loved to fyparticularly among the Pacifc islands. Because of
this love, DHL focused largely on international shipments and became a leader
in Southeast Asia. After Hillblom died in a plane crash in 1995, a huge number
of East Asian and Pacifc Island women came forward and claimed that they had
had children with Hillblom, and that he promised to support them. After hundreds
of lawyers and several court battles, four children were shown to be his heirs by
DNA testing, but several others received settlement money as well (Metropolitan
News-Enterprise 2003). In 2003, DHL was purchased by Deutsche Post. With its
strength in Europe, Deutsche Post and DHL provide a strong international solution
to shipping (company Web site). In 2004, DHL announced a major commitment to
expanding its operations in the United States. In addition to its international ties,
the company wants to focus on competing with UPS and FedEx through highly
visible trucks and a national marketing campaign (Press Release 2004).
DHL has one of the largest customs-clearance systems of any of the shipping
companies. With over 8 million trade rules, the WorldWide Clearance System
(WCS) helps millions of packages clear customs throughout Asia and Europe. The
process is largely automatic for the shipper. Most items are shipped without in-
tervention. If a problem arises, the system fres a message back to the shipper to
handle it. As a result, the customer does not have to constantly monitor the pro-
cess (Rosencrance 2003).
By the time the recession hit in 2008, DHL had mostly pulled out of local de-
livery service in the United States. The company is still big in Europe and Asia
and it still delivers packages to and from the U.S., but it no longer runs head-to-
head against FedEx, UPS, and the USPS in delivering within the United States.
Teamwork
DHL does not have the huge amount of money to spend on IT that UPS and Fe-
dEx do. Consequently, it is rarely a leader. But it has made some smart moves in
communications networks, and by waiting for prices to drop, it now has the ability
to compete with the giants. Alan Boehme, director of business planning in 1998,
noted that the nice thing is Internet technology and the change it enablesin-
expensive communication on an open, standards-based networkis like Davids
rock. Im not going to say that DHL is going to kill the giant, but Internet technol-
ogy is the great equalizer for any business (Sliwa 1998). Actually, even before
the Internet took off, DHL made some smart investments. In 1988, the company
built a TCP UNIX-based network to connect its offces around the world. Their
package tracking data is routed on the internal frame-relay based network. It uses
object-oriented technology to integrate data regardless of whether it came from
the Web site or the automated phone system. The worldwide network also con-
nects programmers and developers so they can share data electronically and work
together even when they are scattered around the globe (Sliwa 1998).
In 2004, DHL expanded its operations to provide radio-frequency identifca-
tion (RFID) technology and support to its customers. By tagging all packages, the
company helps suppliers meet RFID requirements throughout Europe. As world-
wide standards are developed, the company wants to expand the program glob-
ally. Clems Beckmann, managing director of corporate development at Deutsche
Post, notes that Our goal is to become the core logistics supplier for RFID ser-
vices worldwide (Sullivan 2004). In 2007, DHL slowed down the move to RFID
in the U.S. noting a lack of demand for the service and the $1 billion cost. The
company is a leader in Europe, providing RFID tags and data tracking for Metro
AG. The company believes it can add value to shipments of cold perishables such
as seafood, and is trying to convince customers to adopt it (Weier 2007).
One problem that DHL experienced in 2001 was that its pricing was too diffcult
for customers. It could take days for a sales representative to quote a price, and as
long as two to three weeks to get fnal approval from management. Lacking the re-
sources to develop the system themselves, DHL turned to Metreo Inc to automate
the process. Going live in 2002, the system provides sales representatives with
real-time customer data and pricing information based on the customers ship-
ment. Tied to the corporate network, the salesperson can suggest an optimal price
and receive approval almost immediately, sealing the deal before the customer can
be coaxed away by FedEx or UPS (Rosencrance 2002).
In 2002, DHL realized that it needed a new data center to consolidate its op-
erations. Built in Scottsdale, Arizona, the system connects centers in Kuala Lum-
pur, Malaysia and London. DHLs CIO, Steve J. Bandrowczak, noted that DHL
wanted to create a seamless single offering to our customers, like a single invoice
or a single Web interface. In order to enable that global logistics business vision,
we had to consolidate our IT infrastructure, which included the data center, our
network backbone and our key global applications (Rosencrance 2004).
Merging Airbone with DHL in 2003 was a challenge for the IT departments.
Airbornes system were designed to handle high data volumes rapidly and reli-
ably. DHL, headquartered in Belgium, focused on ground transportation and the
complexities of international shipping. DHL also heavily used Indias Infosys to
integrate IT systems, ultimately resulting in layoffs on the Airborne development
team. The Airborne applications, centered in Seattle, were largely based on legacy
IBM CICS and DB2 systems. The DHL Americas systems ran in Scottsdale on
Hewlett-Packard servers running Unix and Oracle databases. In the frst pass, in-
tegration required selecting key applications from each platform to keep the sys-
tems running, but ultimately, operations were moved to Scottsdale. In 2006, the
company began searching for a new integrated system (Whiting 2006).
In terms of development of applications, because of its global nature, DHL
turned to offshore development early on. By hiring programmers in the UK, In-
dia, and Asia, it is able to develop critical applications throughout a 24-hour day.
Colum Joyce, global e-business strategy manager based in Brussels, observed that
for us, large-scale development is not a hothouse environment, its an everyday
reality (Gilhooly 2001). Using the time differences around the world, develop-
ment effectively continues around the clocksomething you cannot do very well
if everyone is located in the same time zone. Plus, the lower turnover rates and
low salaries for skilled overseas programmers help hold costs down. Finally, the
multilingual and multicultural programmers can tailor the applications to specifc
nations.
DHL has generally been behind the curve in adopting technology. For example,
it lags UPS and FedEx in providing wireless tracking devices to its drivers. Only
in late 2006 did the company provide Web-based tools to customers to enable
them to track all of their shipments (Rosencrance 2006).
Similar to many companies, DHL faces challenges in hiring and retaining tal-
ented IT workers. The companys 1,200 U.S. IT workers had a turnover rate of
about 9 percent a year. After surveying workers, CIO Steve Bandrowczak was
told that they wanted greater recognition of their work. So he created the carrot
a day philosophy where workers will be rewarded with items from verbal praise
by a manager at a staff meeting, to an e-mail from a senior IT executive, to gift
certifcates (Hoffman 2005). At the end of 2005, Bandrowczak was named CIO of
Lenovo and in early 2006, DHL hired Maryann Goebel, formerly from GM, as the
CIO for the Americas and Pacifc region.
Deutsche Post faced many challenges as it grew through acquisitions. Each
company that it acquired had different forms of accounting and different ways of
measuring progress. To standardize the accounting and reporting, the company
created the Common Reporting System (CREST). Instead of focusing on cutting
costs, the CREST system emphasized sharing data in a standardized format. It is
based on the SAP platform and uses an Oracle DBMS to handle the data require-
ments. The Web-based application is used across 200 countries and provides con-
sistent basic fnancial data to managers in each area and at corporate headquarters
(http://www.cwhonors.org/viewCaseStudy.asp?NominationID=245).
Questions
1. How do RFID tags help Deutsche Post offer new services to customers?
2. What benefts does Deutsche Post gain by using offshore programmers?
3. How does the Deutsche Post information system make it easy for small
businesses to ship internationally? Why is this important?
4. How can the Internet enable DHL to compete equally with FedEx and UPS?
Additional Reading
Deutsche Post Web site: http://www.dhl.com/publish/g0/en/about/history.high.
html
Gilhooly, Kym, The Staff That Never Sleeps, Computerworld, June 25, 2001.
Hoffman, Thomas, DHL Program Seeks to Hold on to IT Staff,
Metropolitan News-Enterprise, Ninth Circuit Rejects Judicial Bias Claim in
Dispute Between Lawyers for Heirs of DHL Founder, February 7, 2003.
Press Release, DHL Brings Competition And Choice to U.S. Express Delivery
Business With New Campaign, June 14, 2004. http://www.dhl-usa.com/
about/pr/PRDetail.asp?nav=&seq=664
Rosencrance, Linda, DHL Picks Metreo Software to Improve Effciency,
Rosencrance, Linda, Brief: DHL expands Worldwide Clearance System to
Europe, Computerworld, July 23, 2003.
Rosencrance, Linda, DHL Reaps Massive Savings in Data Center
Consolidation, Computerworld, March 15, 2004.
Rosencrance, Linda, Brief: DHL Launches Suite of Web-based Tools,
Sliwa, Carol, Look Out, Goliath, David Has A Rock, Computerworld, June 15,
1998.
Sullivan, Laurie, Logistics Providers Ready RFID Services, Information Week,
June 28, 2004.
Weier, Mary Hayes, DHL Backs Off Plans For RFID On Every Package By
2015, Information Week, March 27, 2007.
Whiting, Rick, DHL-Airborne Didnt Take the Easy Road To Post-Merger IT
Integration, Information Week, June 26, 2006.
Case: United States Postal Service
The U.S. Postal Service (USPS) has been around for centuries. Sometimes hailed,
sometimes maligned, it has struggled to turn a proft. Technically, it is a govern-
ment agency, but it is not funded by Congress. In fact, the USPS provides free
stamps to members of Congress. In exchange, the USPS has a federally mandated
monopoly on delivering mail to individual mail boxes. It is also required to de-
liver frst-class mail to any U.S. address for a fxed price. Letters sent to an ad-
dress down the street cost the same as letters delivered to a rural town in Alaska.
Rate changes must be submitted to and approved by a political Postal board of
governors. With the advent of low-cost long-distance phone calls and e-mail, the
demand for frst-class letters has dropped for several years. In 2005 and 2006,
the quantity of junk mail carried surpassed the delivery of frst-class items (Hall
2007). In response, the Postal Service has continued to increase the price of frst-
class stampsdriving demand even lower. Most of the mail carried consists of
advertising pieces charged at varying rates below frst-class postage, but even
those rates were increased by as much as 40 percent in 2007. Those rates have
increased to the point where some advertisers are testing hand-delivered fyers
(Chabrow 2004 and usps.gov).
In 2010 and 2011, the USPS is still struggling to adapt to a digital world. De-
mand for mail continues to drop, so revenue is declining, and losses are increas-
ing. To compensate, the USPS continues to increase prices and is contemplating
reducing services by dropping Saturday mail delivery. In 2011, the agency sus-
pended its employer contribution to the Federal Employee Retirement System be-
cause it did not have enough money. The Postal Service was $8 billion in the red
for 2010 and risked running out of money before the end of 2011. The agency had
already reduced staffng by 110,000 and cut costs by $12 billion, but it was still
not enough to keep it alive (Wall Street Journal 2011). From 2001 to 2010, frst-
class mail volume dropped from 103.7 billion to 78.2 billion pieces (www.usps.
com).
Technology
The USPS relies heavily on outside contractors to provide information technology
and other infrastructure. For example, for years, the USPS has relied on Lockheed
Martin to provide mail handling systems. Several mail-processing centers are run
by outside contractors. In October 2004, fearing problems with the bankruptcy fl-
ing of MCI, USPS signed a long-term contract with Lockheed Martin to provide
telecommunications services as well. At the end of 2006, after some apparently
bad experiences, USPS cancelled the contract and returned to MCI as its primary
provider (Marsan 2006).
For the most part, the USPS lacks the delivery tools of the giants FedEx and
UPS. Workers still sort and deliver individual pieces of mail on fxed delivery
routes. But, the task of the workers is quite different because they almost always
have to stop at every mailbox every day. Hence, the potential gains from tracking
and scheduling are minor. However, in 2003, the organization spent $20 million
to develop logistics software to aggregate ZIP code data and map out how to load
packages on trucks more effciently (McDougall 2003).
The mail system was the frst to support person-to-person messages across the
U.S. and eventually the globe. Consequently, it was also a primary method of
propagating scams. The Postal Services Postal Inspection Service was created in
1909 to investigate various levels of fraud and scams. Various federal laws were
enacted over the years to make specifc scams a federal crime. You might think the
Inspection Service would have no role in todays Web-based e-mail scams. How-
ever, any activity, such as mailing a check, that uses the mail system falls under its
jurisdiction. If someone orders an item via the Web and it is delivered through the
Postal Service, its inspectors can take chargeas happens often in pornography
and drug cases. The Service has been involved in tracking several major cyber-
scams and attacks (Greenemeier 2007).
With 700,000 employees, the USPS needed a way to train its employees on
procedures and rules. For years, it used an internally-developed system. In 2003,
USPS purchased a commercial testing learning-management system from Thinq
Learning Solutions. Robert Otto, VP of IT, was primarily impressed with the data-
capturing capabilities of the tools that validate an employee has completed the
course. He said that we had a hole in the administrative end of managing our
training (George 2003).
Like any other organization, the USPS needs an accounting system to track
expenses and receipts. In 2004, the Postal Service began a process to replace its
decades-old custom-built applications with off-the-shelf software. In particular,
it chose to install Oracles ERP system beginning with the general ledger. The
USPS had already installed Oracle.s Web-based time-and-attendance application
in 2001. A Retek Inc. retail accounting software package collects sales data from
the 40,000 Postal Service branches and posts them to the Oracle general ledger.
Using the Oracle system, the Postal Service was able to reduce its accounting cler-
ical staff by 1,000 positions, reducing expenses between $300 million to $1.5 bil-
lion annually in 80 districts. The organization planned to implement an SAP HRM
application to handle benefts and promotions data in 2007 (Greenemeier 2005).
In October 2009, the government accounting organization (GAO) evaluated the
technology for the Post Offce Intelligent Mail Full Service Program. This pro-
gram was created in 2003 to use barcodes to track mail. The basic conclusion was
that the program was only partially fnished and that the rest of the project was
unlikely to be completed (GAO-10-145).
Some people have tried to search for ways to keep the traditional Post Offce
alive, but the task seems grim. The Postal Regulatory Commission notes that the
Postal Service wants to close thousands of post offces. John Callan, a mailing
industry consultant organized conference for June 2011 to help fnd a future for
the Postal Servie. He noted that I think we need to stage a serious conversation
about the future, so that the folks on the Hill can understand whats going on and
those of us in the postal industry can learn how we would do something differ-
ently (OKeefe 2011). Other people and government oversight agencies are even
bleaker. The agency relies on frst-class mail to fund its operations, but frst-class
mail volume is plummeting. Two complicating factors are that the union contract
forbids layoffs and Congress will not allow closing a Post Offce for economic
reasons. Through privatization, European postal services have used these tactics
to survive, but they are not yet an option in the U.S. (Leonard 2011).
A few people still fnd the Postal Service valuableparticularly for package
shipping. The prices are competitive and often lower than the alternatives (FedEx
and UPS). The Postal Service also offers electronic postage. It is now relatively
easy to purchase postage online and print an electronic stamp on an envelope
directly from a PC printer (Khechfe 2010, but note he works for a company that
creates electronic postage).
Questions
1. How does the USPS use its Web site to sell services?
2. When was the last time you mailed a letter via frst-class mail?
3. What happens if USPS goes under?
Additional Reading
Chabrow, Eric, Will Feds Pay For Post Offces Funeral Using E-Payments?
GAO, Information Technology: U.S. Postal Service Needs to Strengthen System
Acquisition and Management Capabilities to Improve Its Intelligent Mail
Full Service Program, October 2009.
George, Tischelle, USPS Dumps Its E-Learning App, Information Week,
January 6, 2003.
Gilhooly, Kym, The Staff That Never Sleeps, Computerworld, June 25, 2001.
Greenemeier, Larry, Postal Service Emerging From Holiday Deluge With Big
Plans For 2005, Information Week, December 17, 2004.
Greenemeier, Larry, U.S. Postal Service Cracks Down On Cyberscams,
Information Week, June 4, 2007.
Hall, Mark, Neither Snow, nor Rain, nor Heat ..., Computerworld, May 14,
2007.
Khechfe, Amine, Postage Technologyand U.S. Postal ServiceReally
Delivers for Parcel Shippers, December 20, 2010. http://www.
parcelindustry.com/ME2/dirmod.asp?sid=23C6283BD51B46348B616C079E
EB2E21&nm=Miscellaneous&type=Publishing&mod=Publications%3A%3
AArticle&mid=8F3A7027421841978F18BE895F87F791&tier=4&id=947F7
EE1BD3848BF9CBC3C43D7D5D5D2.
Leonard, Devin, The U.S. Postal Service Nears Collapse, Business Week, May
26, 2011.
Marsan, Carolyn Duffy, USPS Quietly Cancels 18-year Network Outsourcing
Deal, Computerworld, December 19, 2006.
McDougall, Paul, Load Em Up, Information Week, September 29, 2003.
OKeefe, Ed, Google Execs, Tech Experts Focus on Future of Postal Service,
The Washington Post, May 4, 2011.
Wall Street Journal, Post Offce Suspends Retirement Contributions, June 22,
2011.
adopted by rivals?
sectors?
3
Part
Decisions and Strategies
How does information technology help teams share information?
Can computers make decisions? How can information systems
make your company better than the competition?
Business decisions can be complex. Complexity can arise from
several areas, including the use of huge amounts of data, diff-
cult mathematical formulations, uncertain relationships, detailed
linkages to multiple business units, and physical or procedural
constraints. Middle-level managers in all functional areas face
complex problems. Various models have been created to help you
analyze these problems and evaluate alternative answers. Infor-
mation technology provides several tools to help managers col-
lect data, evaluate models, evaluate output, and make decisions.
Ongoing research into artifcial intelligence has led to addi-
tional tools to tackle specifc problems. Expert systems, robot-
ics, and neural networks are sophisticated tools to solve complex
problems.
Strategic analyses represent some of the most diffcult deci-
sions a manager can face. Strategy represents fundamental chang-
es in the operations of the business. Information systems are used
to search for useful changes. Information systems have also been
useful in creating a competitive advantage.
Chapter 9: Business Decisions
Chapter 10: Strategic Analysis
Chapter 11: Entrepreneurship
How do information systems help managers make better deci-
sions?
576
How do businesses make decisions?
How do you make a good decision? Why do people make bad decisions?
How do you fnd and retrieve data to analyze it?
How can you quickly examine data and view subtotals without hundreds of queries?
How does a decision support system help you analyze data?
How do you visualize data that depends on location?
Is it possible to automate the analysis of data?
Can information technology be more intelligent?
How do you create an expert system?
Can machines be made even smarter?
What would it take to convince you that a machine is intelligent?
What are the differences between DSS, ES, and AI systems?
How can more intelligent systems beneft e-business?
How can cloud computing be used to analyze data?
Chapter Outline
Business Decisions
9
Chapter
Introduction, 577
It is Hard to Make Good Decisions, 578
Human Biases, 581
Models, 584
Data Warehouse, 589
Online Analytical Processing (OLAP), 590
Decision Support System, 594
Marketing Forecasts, 594
Human Resources Management, 596
Geographical Information Systems, 597
Maps and Location Data, 599
Example, 601
Data Mining, 602
Expert Systems, 606
Specialized Problems, 608
Diagnostic Problems, 608
Speedy Decisions, 609
Consistency, 610
Training, 611
Building Expert Systems, 612
Knowledge Base, 613
Rules, 613
Creating an ES, 614
Limitations of Expert Systems, 615
Management Issues of Expert Systems,
616
Specialized Tools, 616
Pattern Recognition and Neural
Networks, 617
Machine Vision, 619
Language Comprehension and
Translation, 620
Robotics and Motion, 621
Machine Intelligence, 623
DSS, ES, and AI, 623
The Importance of Intelligent Systems in
e-Business, 625
Agents, 625
Support and Problem-Solving
Applications, 627
Summary, 628
Key Words, 630
Exercises, 631
Cases: Financial Services Industry, 636
577 Chapter 9: Business Decisions
Citigroup
Introduction
How do businesses make decisions? Figure 9.1 shows that as a man-
ager, you will have access to huge amounts of data. How do you analyze it to
understand what it means? How can information systems help you make better
decisions? These questions are diffcult to answer, but there is a much bigger un-
derlying question. Why would companies need to hire you as a manager? Ex-
ecutives already have access to databases containing integrated data for the entire
company. ERP systems can provide detailed data on any aspect of the business.
EIS systems can show charts, summaries, and detailed data. Now, what if com-
puter systems can be built to analyze the data and make decisions? What job will
you have?
Analyzing data and making decisions depends somewhat on the discipline you
choose and your background in statistics and operations management. A key as-
pect in many introductory business courses is to teach you the basic models used
within the discipline. It is unlikely that you are an expert in any particular disci-
pline at this point, and you might not be aware of the variety and power of the
statistical tools available to analyze data. Nonetheless, you need to learn how in-
formation technology can be applied to many business decisions. This chapter fo-
cuses on the tools available and illustrates them with relatively simple decisions.
Even if you do not fully understand the discipline-specifc models, you need to
learn what tools are available.
Can computers make decisions? How can information systems help manag-
ers make decisions? Some business problems are straightforward. In these cases,
developers simply create a set of rules or procedures that the computer can be
programmed to follow. As long as the business behaves in a predictable manner,
the rules apply and the computer can handle the details. However, many business
problems are less structured and cannot be solved so easily. In addition, problems
often involve data that is not well defned. For example, it is straightforward to
create a computer system to handle inventory because the computer can easily
keep track of item numbers and quantity sold. Consider the more diffcult prob-
How do you use information technology to make better decisions? Citigroup is one
of the largest banks in the world. And the world part is important: the company op-
erates in 101 countries. Running a global bank requires making thousands of deci-
sionsfrom basic questions about approving loans to structuring mega-deals with
huge corporations. Banks obviously adopted information technology earlyto han-
dle basic transactions. After all, money today is really just numbers in a computer.
Banks, including Citigroup, have been slower to adopt technology to make deci-
sions; but that reluctance has been changing in the last few years.
Citigroup is faced with competitive pressures as well as disruptions from eco-
nomic fuctuations. It has attempted to reduce risk by spreading into multiple busi-
ness areas, including insurance (Travelers), brokerage (Salomon Smith Barney), and
investment banking. The company is one of the largest issuers of credit cards in the
United States. With stagnant growth of the U.S. credit industry, Citigroup has ex-
panded into South America and even China. Evaluating customers to minimize risk
is a key aspect of the credit card industry.
lem faced by a manager who has to decide where to locate a new plant. Some at-
tributes are measurable, such as distance from suppliers, cost of land, and taxes.
Other features are diffcult to quantify: quality of the labor force, attitudes of gov-
ernment offcials, and long-run political stability of the area.
Many problems involve nonnumeric data and complex interrelationships
among the various factors. Without computers, businesses often call in experts or
hire consultants to help solve these problems. Special software programs called
expert systems (ESs) provide many of these features. From the beginning, re-
searchers and computer designers have known that humans perform some tasks
much better than computers can. These differences led researchers to investigate
how people solve problems and investigate how humans think. The research into
techniques that might make computers think more like humans is known as ar-
tifcial intelligence (AI). There is some question as to whether it will ever be
possible to build machines that can think the same way humans do. Nonetheless,
the research has led to some useful tools that perform more complex analysis and
can solve diffcult problems. These tools attempt to mimic the processes used by
humans in a simpler form that can be processed by a computer system.
The answer to the big question is yes, computers can make decisions in some
situations. In other cases, the computer is a tool that helps you analyze the data.
The level of support provided depends on the type of problem and on your skills
as an analyst. As a manager, it is your responsibility to identify decisions that can
be handled by machine systems and to recognize when these systems do not work.
It is Hard to Make Good Decisions
How do you make a good decision? Why do people make bad
decisions? Most businesses have evolved over time. In many cases, their busi-
ness processes have been built and patched in response to various changes in the
industry. The frms that made better decisions and changes survived, while the
others failed. But the existing process might not be the most effcient. Consider
the apparently simple process of farming. Farmers feed the animals and then sell
Figure 9.1
Tactical decisions often require complicated analysis. Problems utilize forecasts,
optimization, and in-depth analysis. Information systems provide support through
data, modeling, and presentation tools. Managers use information system tools to
build, evaluate, and maintain various models.
Data
Sales and Operations
f x
x
( ) exp =

(
|
\
|
.
|
1
2
1
2
2

0
20
40
60
80
100
1st Qtr 2nd Qtr 3rd Qtr 4th Qtr
Actual Forecast
Models
Analysis and Output
Decisions
them. The hard part is that the animals can be fed and housed many different
wayseach with different costs. Should the animals be fed high-protein food that
costs more and grows bigger animals faster, or should they be fed simple diets and
take more time to mature? In the 1970s and 1980s, experts created software that
analyzed these questions from the standpoint of minimizing the cost of feeding the
animals. Using optimization methods, they were able to substantially reduce the
production costs. But some experts have found it is possible to do even better by
focusing on profts across the entire industry chain. The same types of problems
apply to any business process, because most companies solve individual problems
frst because they are easier. As tools become better, it is important to expand the
perspective and look at broader decision problems.
Even if you do have a system for making a better decision, you need to con-
vince managers to use it. Many managers distrust new technologies and different
answers, because they see an element of risk. A few companies have established
a culture that focuses on continual improvement and growth. In these companies,
managers are encouraged to explore new ideas and replace the existing processes.
Trends
Through the 1970s, computers were largely used to assist with transaction
processing. Support for making decisions was generally limited to the ba-
sic reports produced from the data. Computers were too expensive and pro-
gramming too diffcult to be used by every manager. As personal computers
became commonplace through the 1980s, managers began transferring data
from the corporate central computers to their personal machines. Spread-
sheets made it easier to analyze data, evaluate models, and create charts.
In the 1990s, networks, improved spreadsheets, and better ties to databases
made it possible to build more complex, interactive models and create fore-
casts.
Along with technology, improvements were made to modeling and ana-
lytical tools. Scientifc advancements made it possible to add more intelli-
gence to software tools. Data mining systems use statistical tools to semiau-
tomatically evaluate data, searching for important relationships and cluster-
ing or classifying groups of data. Expert systems evolved from early work
on artifcial intelligence. Focusing on narrow domains, these tools encode
the rules of an expert to analyze data and suggest solutions. Today, thou-
sands of expert systems are used to improve decisions and provide quick
results 24 hours a day.
The study of human brains yielded clues that led to the development of
neural networks. Today, neural networks are widely used in pattern match-
ing applications. Humans are good at pattern recognition, and neural net-
works dramatically improve the ability of machines to perform these tasks.
Writers and researchers have long wondered whether machines can be-
come intelligent. No one is close to an answer yet, but the new technolo-
gies mean that todays systems are more intelligent and can handle more
complex problems than machines a few years ago. As the range of solvable
problems increases, managers need to understand the capabilities and limita-
tions of each method.
Reality Bytes: Googles Self-Driving Cars
Following the successes of the DARPA Grand Challenges of the late 2000s, some
companies have expanded research and development of self-driving cars. Interest-
ingly, Google has been one of the leaders. Perhaps driven by ties to Stanford (which
one the second Grand Challenge), and by the importance of mapping in Google. In
2010, Google told the world that it had been using off-the-shelf components to test
run self-driving cars a total of 150,000 mileswith almost no human intervention.
However, a little know aspect of Googles methodology is that humans frst drive the
cars around the test arealargely to provide a more specifc map of the route, stop-
lights, school zones, and so on. Chris Urmson, technical lead for the Google project,
and a leader from the Carnegie Mellon team that won the 2007 DARPA Urban Chal-
lenge, notes that There are things that right now are a challenge for us. For instance,
if most of the world stayed the same but the lanes are shiftedso the physical road
didnt move, but for whatever reason, the department of transportation decided we
should drive a half lane to the leftthat would probably confuse the car today. The
challenge lies in unexpected eventsgetting the software to recognize and react to
them using some level of common sense. Ultimately, the tradeoff is an interesting
question in ethics: Does a machine have to be perfect? The National Highway Traffc
Safety Administration (NHTSA) noted that in 2008 the U.S. experienced 5.8 million
car accidents, with 1.6 million of those resulting in personal injuries and 34,000 in
deaths. If machine-drivers can cut that rate in half, would we still blame the develop-
ers for the remaining accidents?
Adapted from Nick Chambers, Hands-Off Training: Google's Self-Driving Car
Holds Tantalizing Promise, but Major Roadblocks Remain, Scientifc American,
May 23, 2011.
Company A
Company B
Stock Price
130
125
120
115
110
105
100
95
90
1 2 3 4 5 6 7 8 9 10 11 12
Month
Figure 9.2
Sample decision. Do you invest your money in Company A or Company B? Be
careful, it is a trick question.
Human Biases
Assume you have money to invest in the stock market. Someone shows you two
companies. As shown in Figure 9.2, Company As share prices have risen by 2
percent per month for the last year. The others share price was fat for fve months
but has increased by 3 percent per month since then. Which stock do you buy? But
wait a minute. How can you possibly decide based on the little information you
have? It sounds silly, but people make these decisions with minimal data and no
logical analysis every single day. When people make decisions this way, the re-
sults are going to be inconsistent and dangerous. But it is so much easier to make
a snap decision, and so much harder to do the research and complex analyses to
make the right decision.
Consider a true example: designing a new automobile. The automobile in-
dustry, epitomized by GM in the early 1980s, presents an interesting example of
making decisions. GM demonstrated from the 1940s forward that people buy cars
based on design and appealnot simply a collection of features. In particular,
GM demonstrated that customers can be heavily persuaded by style and advertis-
ing. Assuming you have money, what kind of car would you buy? Sporty, luxuri-
ous, fashy, utilitarian, big, small? What color? How many doors? Now ask a few
friends or relatives what they would buy. Will all of the answers be the same? Not
likely. Now think about the problem from the perspective of an automobile manu-
facturer such as General Motors. What features are car buyers going to demand in
two or three years? This classic marketing problem is diffcult to solve. For years,
GM used its multiple divisions to create separate identities that appealed to dif-
ferent segments of the population. Designers within each division focused on the
preferences and lifestyles of their specifc target. Most of that structure fell apart
in the early-1980s with the introduction of a completely new line of cars based on
the A-body. At that time, the GM divisions introduced a new car model from four
main divisions. As shown in Figure 9.3, all four cars (Oldsmobile Cutlass Ciera,
Pontiac 6000, Chevrolet Celebrity, and Buick Century) were virtually identical.
Buick Century Oldsmobile Cutlass Ciera
Chevrolet Celebrity Pontiac 6000
Figure 9.3
GM 1984 models. Note the body similarities. Even the base color offerings were the
same. How could GM throw away its division strategy and offer virtually identical
models? Source: Wikipedia.
Figure 9.4
Biases in decision making. Without models, people tend to rely on simplistic rules
of thumb and fall prey to a variety of common mistakes. These errors can be
minimized with training and experience in a discipline. They can also be minimized
by having computer systems perform much of the initial analysis.
Acquisition/Input
Bias Description Example
Data availability Ease with which specifc
instances can be recalled afects
judgmentsoffrequency.
People overestimate
the risk of dying due to
homicides compared to
heartdisease.
Illusory correlation Belief that two variables are
relatedwhentheyarenot.
Ask any conspiracy buf
aboutthedeathofJFK.
Data presentation Orderefects. First (or last) items in a list
aregivenmoreimportance.
Processing
Inconsistency Difculty in being consistent
forsimilardecisions.
Judgments involving
selection, such as
personnel.
Conservatism Failure to completely use new
information.
Resistancetochange.
Stress Stress causes people to make
hastydecisions.
Panicjudgmentsandquick
fxes.
Social pressure Social pressures cause people
to alter their decisions and
decision-makingprocesses.
Majority opinion can
unduly infuence everyone
else:mobrule.
Output
Scale efects Te scale on which responses
are recorded can afect
responses.
Ask a group of people to
rate how they feel on a
scalefrom1to10.Aska
similar group to use a scale
from1to1,000.
Wishful thinking Preference for an outcome
afectstheassessment.
People sometimes place
a higher probability on
events that they want to
happen.
Feedback
Learning from
irrelevant outcomes
People gain unrealistic
expectations when they see
incompleteorinaccuratedata.
In personnel selection
you see how good your
selection is for candidates
youaccepted.Youdonot
receive data on candidates
yourejected.
Success/failure
attributions
Tendency to attribute success to
onesskillandfailuretochance.
Only taking credit for the
successesinyourjob.
The available colors were even the sameparticularly maroon. In effect, GM was
assuming that millions of customers all wanted the same car.
Designing cars, or any product, is a diffcult decision problem. In good situ-
ations, like GM, you have tons of data available. You have sales data, data on
competitors, surveys, and focus groups. But, the results are meaningless if you
cherry-pick data to match your preconceived ideas. Oh look, here are 5,000
people who want larger engines. If you search hard enough, you can fnd
data to match any opinion you want; but it will not accurately represent the opin-
ions of the population.
In response to these problems, Barabba and Zaltman, two marketing research-
ers working with GM, analyzed decision making at General Motors and no-
ticed that several common problems arose. In summary, they found that people
are weak at making decisions. For example, people place too much emphasis on
recent events, they tend to discard data that does not ft their prior beliefs, they
follow rules of thumb instead of statistical analysis, and they choose outcomes
based on wishful thinking. As shown in Figure 9.4, all of these problems and more
infuenced the decisions of designers at GM. In particular, they found that the de-
signers tended to discuss ideas with their bosses in an attempt to identify manage-
ment preferences that would help get a particular design approved. So cars were
designed to the preferences of a few managers, instead of to the needs of custom-
ers. Despite attempts to improve, the fasco eventually forced GM to eliminate
the Oldsmobile division. The books written by Barabba and Zaltman discuss even
more examples and human biases in decision making.
Before you think that businesses (and GM) have solved the design problem, go
look at current designs. In 2008, the Wall Street Journal examined cars pro-
duced by GM and Ford and found similar overlap in styles within each company.
And GMs overall market share continued to decline at least through 2009, when
Model Building
Understand the Process
Models force us to defne objects and specify relationships. Model-
ing is a frst step in improving the business processes.
Optimization
Models are used to search for the best solutions: Minimizing costs,
improving effciency, increasing profts, and so on.
Prediction
Model parameters can be estimated from prior data. Sample data is
used to forecast future changes based on the model.
Simulation
Models are used to examine what might happen if we make changes
to the process or to examine relationships in more detail.
Figure 9.5
The four primary reasons for building and using models. Descriptive, graphical, and
mathematical models can be used for each of these purposes. However, mathematical
models tend to be emphasized for optimization and simulation.
the company was forced to fle for bankruptcy protectionwhen the company
sold off or eliminated several other divisions. The main point to remember is that
making decisions without a good model and process leads to poor decisions. Sure,
you might get lucky for a while (like investors in the 1990s), but ultimately you
need a solid decision-making process.
Models
Models are key aspects to any decision; they are simplifcations designed to help
you understand and analyze a problem. Many of the models you will use in busi-
ness decisions were created by academics. You will be introduced to many of
these models in other business courses. As a manager, you are responsible for
knowing that hundreds of models are available to help you make decisions, and
for knowing which model best applies to the problem you are facing. Understand-
ing and evaluating models is an important aspect of a business education.
Models often use drawings and pictures to represent the various objects. How-
ever, at heart they typically use mathematical equations to represent the process
and the various relationships. For example, an operations engineer would model
a machine as a mathematical formula that converts raw materials and labor into
products. Using equations for each step of the production process, the engineer
could search for ways to reorganize production to make it more effcient or to
improve quality.
Models are used to help managers make decisions. Most businesses are far too
complex for any single person to understand all of the details. Consequently, a
variety of models may be created to present a simplifed view of the business. In
particular, one of the original purposes of accounting was to create a standardized
model of the fnancial aspects of business. Another common model of business is
the practice of dividing the company into functional areas. For example, a man-
ager with experience in the fnance department of one company can usually apply
knowledge and problem-solving skills to fnance departments in other companies
and even other industries. The basic goals are summarized in Figure 9.5. Models
help you simplify the world. They help you search for similarities in different
situations. Models also enable managers to predict how changes might affect the
business. As the decision maker, it is up to you to determine which models to use
and to make sure they actually apply to the situation. Once you have selected the
appropriate model, you apply whatever data you have, evaluate the results, and
make the decision.
Prediction and Optimization
An important use of models is for prediction. If a model is reasonably accurate,
it can be used to predict future outcomes. For instance, when you buy a car, you
might want to know how much it will be worth in three years when you want to
sell it. It is possible to estimate how the price of used cars changes over time.
Prediction frst requires that you have a model that describes the situation. Then
data is collected and statistical techniques are used to estimate the parameters
of the model for the specifc problem. Next you fll in values for any parameters
that you already know, and the model provides a prediction. Prediction techniques
such as regression and time series forecasting are used to examine the data, identi-
fy trends, and predict possible future changes. To use statistics effectively requires
a model of the underlying system. For instance, to use regression methods you
frst identify the dependent variable and a set of possible independent variables.
These choices come from the underlying model.
Figure 9.6 illustrates how a spreadsheet can be used to display the results of
a forecast. Although dedicated statistical packages contain more options, spread-
sheets contain several tools to help with basic statistical regression and forecast-
ing. Once you load the Data Analysis toolpak in Excel, you can run basic regres-
sions, T-Tests, or ANOVA to identify how a set of independent X variable affect
the dependent Y variable. You can use the trend line function to highlight averages
on charts. If you are familiar with statistical functions, you can use the built-in
functions to make comparisons and evaluate data. You can also purchase add-ins
to perform more-sophisticated analysis of your data. When you need even more
power, you can use the import and export tools in the high-end statistical software
packages to extract and analyze your data. You might have to hire a statistician to
help determine which tools you need and to create the initial models. But once the
model and system are confgured, you can run the system to analyze your results
on a regular basis.
Optimization
Optimization evaluates a model in terms of the inputs or control parameters
and searches for the best solution. Optimization requires a detailed mathematical
model. Several tools such as linear programming are used to fnd optimal values.
Some optimization models have resulted in substantial savings in cost or increases
in proft. Tasks that are repeated hundreds or thousands of times can often beneft
through optimization modeling.
0
5
10
15
20
25
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2
Time/quarters
O
u
t
p
u
t
Moving Average
Trend/Forecast
Economic/
regression
Forecast
Figure 9.6
Prediction model. Several statistical techniques exist for analyzing data and making
forecasts. Two common methods are regression and moving averages. Both methods
require substantial amounts of data. Choosing between the two requires some
expertise in statistical analysis, but many times we display both methods to show a
range of possible outcomes.
Figure 9.7 shows a simple optimization model with two input variables and
a single output. Optimization typically requires a mathematical model that de-
scribes the complex relationships between the variables. You will often use sta-
tistical techniques to estimate the underlying model parameters, and then turn to
optimization tools to fnd the best operating point. Linear programming is an op-
timization tool that has demonstrated considerable success in solving some rela-
tively complex problems. As the name implies, basic tools assume that all of the
relationships are linear, but modern tools can support a limited amount of nonlin-
earity, including quadratic terms.
The Solver tool that is sold with Excel is a limited version of a commercial
optimization product, but even the basic version is capable of solving moderately-
large problems. If your model grows too large, you can upgrade to the full version
or purchase software from other companies. One of the strengths of the linear
programming (and Solver) approach is that it is designed for constrained optimi-
zation problemswhich are common in business. Operations research and man-
agement classes explain these problems in more detail. The problems are a little
tricky to set up, so you should consult an expert if you lack experience with these
problems. The key is to identify the output goal (perhaps proft maximization or
cost minimization), and then list all of the input variables that affect that result.
Once you have defned the optimization equation, you need to specify all of the
constraints that affect the problem. For example, a production plant has a maxi-
mum amount that can be produced; and people work a limited number of hours
a week producing a defned amount of output per person and machine. Once you
understand optimization, you will recognize that it can be used to solve a vari-
ety of problems, including complex pricing decisions, worker scheduling, product
mix, and input selection.
20
15
10
5
0
1
2
3
4
5
6
7
8
9
10
1

2

3

4

5
input levels
O
u
t
p
u
t
Control variables
Figure 9.7
Optimization model. Optimization requires a mathematical model. Output is defned
in terms of various input variables. The maximum point can be found by altering the
values of the input variables. Special software, such as linear programming tools, can
help evaluate complex models.
Simulation or What-If Scenarios
Simulation is a modeling technique with many uses. Once a model is created,
managers use simulation to examine how the item being studied will respond to
changes. With simulation, various options can be tested on the model to examine
what might happen. For example, engineers always build models of airplanes and
engines before they try to build the real thing. The models are much cheaper. In
fact, most engineers today start with mathematical computer models because they
are cheaper to create than physical models and can contain more detail. Moreover,
they can perform experiments on models that would not be safe to perform in real
life. For example, an engineer could stress a model of an airplane until it broke up.
It would be dangerous and expensive to try such an experiment on a real plane.
Similarly, a business model could examine what would happen if prices were in-
creased by 20 percent, without worrying about losing real money. Of course, the
model is only useful and valuable if it accurately refects real conditions.
Most simulation models are mathematical instead of descriptive models, be-
cause they are easy to evaluate. Mathematical models contain parameters, or
variables that can be controlled by the managers. For instance, you might use a
spreadsheet to create an accounting model of an income statement and balance
Reality Bytes: Make Decisions by Putting Down the Cell Phone and Going for
a Walk
If you think choosing what to eat for dinner is tough, stay away from combinatorial
auctions. Consider an airline that wants to bid on landing slots at an airport. The slots
can be bid on individually or in bundle of combinations. Data includes the passenger
loads, weather, and connecting fights. Angelika Dimoka, director of the Center for
Neural Decision Making at Temple University, uses these types of complex puzzles
to study the ability of humans to make decisions. She used functional MRI (fMRI)
to examine brain patterns while volunteers worked on these puzzles. At the start,
increasing the amount of data caused activity in the dorsolateral prefrontal cortex
region of the brain to increase. This region is responsible for decision making and
control of emotions. However, as the amount of data increased, activity suddenly
dropped as the humans reached overload. At that point, frustration and anxiety soar
as the emotions take over. Other research into human aspects of decision making
show that unconscious systems guide human decisions and that decisions that re-
quire creativity beneft from patienceallowing the brain to think about the problem
passively. But interrupting the brain with increasing amounts trivia makes it more
diffcult for the brain to operate. It is one of the reasons pilots and frst responders
train for emergencies and maintain checklists of critical tasks. Studies, such as those
by Shenna Iyengar of Columbia University, show that providing more choices and
more data can cause people to give up and simply opt out. For instance, employees
who faced retirement plans with more options tended to opt out more often than
those with fewer choices. The brain needs time to evaluate data, integrate the ele-
ments, and determine which items to ignore. So, collect the data, use analytical tools
and models that were developed for the problems, try to identify the most important
factors, then take a break and let the brain evaluate the choices for a while.
Adapted from Sharon Begley, I Cant Think! Newsweek, March 7, 2011.
sheet. When you create the spreadsheet, production quantity and price of your
products are controllable parameters that affect the income and profts of the frm.
You could use the model to investigate decisions, like the effect on profts if you
increase production. Costs will increase, but so will revenue from sales. The net
result depends on the specifc details of the frm and the model. Spreadsheets are
often used to analyze small models and graph the results. More sophisticated sim-
ulation packages can support more complex analysis and will automatically create
graphs and pictures to show interrelationships.
More complex models provide more opportunities for simulation. In part be-
cause they are diffcult to solve and understand. In a production system, a more
detailed model might enable you to investigate alternatives such as increased
overtime, hiring another shift, building additional plants, or subcontracting the
work to another frm.
Object-oriented simulation tools developed in the last few years make it easy
to create many simulations. As shown in Figure 9.8, you can place icons on the
screen to represent various business objects. Behind each of these objects, you
specify their behavior. For example, you would need to estimate the average num-
ber of customers that arrive in an hour and how long they are willing to wait
for service. For physical processes like order entry, inventory, and shipping, you
specify the number of transactions that can be handled in a given time. When all
of the details have been entered, the system runs simulations and tracks statistics.
You can change the parameters such as adding clerks to see what happens if you
change the operations.
Customer
Order Entry
Custom Manufacturing
Production
Inventory &
Purchasing
Shipping
Purchase
Order
Purchase
Order
Routing &
Scheduling
Invoice
Parts List
Shipping
Schedule
Figure 9.8
Object-oriented simulation. A simple example of custom manufacturing. Each object
(parts list, purchase order, etc.) and each process are defned in detail by the modeler.
The simulation system generates orders, makes shipments, and orders inventory
according to programmed rules. The simulator collects a variety of statistics that can
be displayed graphically or in summary reports.
Data Warehouse
How do you fnd and retrieve data to analyze it? ERP and other
transaction systems can provide enough data to bury you. But the transaction sys-
tems are designed to store data, not to search and analyze it. Relational databases
turn out to be relatively slow when you need to analyze several gigabytes or even
terabytes of data. If you do not have an ERP system, you have even greater prob-
lems trying to integrate and clean data from all of your systems. As shown in
Figure 9.9, the answer is to create a separate data warehouse that extracts and
stores the data in a clean, easy-to-analyze format. The process shown in Figure 9.9
is known as extraction, transformation, and loading (ETL). Larger database
management systems have specifc tools and data storage methods to create data
warehouses. Some companies also create specifc data marts that are basically
copies of a small portion of the data warehouse designed to feed a specifc ap-
plication. For instance, a fnancial data mart might be used by the accounting and
fnance department just to monitor investments and bank accounts.
Finding and cleaning the data is the most time-consuming step in most data
analysis projects. Data is often stored in multiple systems and proprietary formats.
Even if the data is stored in a DBMS, someone has to verify the data content and
format to ensure it matches exactly with the other data. Even a number as simple
as sales could be defned differently by various departments. Some departments
might count sales when a customer signs a contract, but others might wait until
the contract has been approved by top management. Sometimes organizations are
not even aware that these differences existuntil they decide to start integrating
the data.
OLTP Database
3NF tables
Operations
data
Predened
reports
Data warehouse
Star conguration
Daily data
transfer
Interactive
data analysis
Flat les
Figure 9.9
Data warehouse. A data warehouse is commonly used as a method to provide data
to decision makers without interfering with the transaction-processing operations.
Selected data items are regularly pulled from the transaction data fles and stored in a
central location. DSS tools query the data warehouse for analysis and reporting.
Documenting the data is critical because managers have to understand what
each item represents and they need to be able to fnd specifc items. Metadata is
used to describe the source data, identify the transformation and integration steps,
and defne the way the data warehouse is organized. A data warehouse represents
a subset of the total data in the company. In most cases, it is a static copy that is
refreshed on a daily or hourly basis. This type of system is relatively easy to use;
managers do not have to learn data access commands (SQL or QBE). However, it
is less fexible than using a database management system. Decision makers will
be unable to get additional data or to compare the data in some previously unex-
pected way. The success of a data warehouse depends on how well the managers
needs have been anticipated.
Online Analytical Processing (OLAP)
How can you quickly examine data and view subtotals without
writing hundreds of queries? Retrieving data and examining totals from
different perspectives is an important part of making decisions. When the prob-
lem is unstructured and there is no existing mathematical model, it often helps to
look at subtotals of data. Sometimes it is useful to browse through the data and
examine various subtotals. Which customers bought the most? Click: there are
the totals. Which employees sold the most? In December? Click: see the totals
Figure 9.10
Multidimensional cube for item sales. Managers are interested in various
combinations of the dimensions. For example, total item sales of dog items in the last
quarter. OLAP tools rapidly provide answers to questions involving any perspective
of this cube.
CA
MI
NY
TX
Race
Road
MTB
Full S
Hybrid
Jan Feb Mar Apr May
Time
Sale Month
Customer
Location
1420 1253 1184 1098 1578
437 579 683 873 745
1011 1257 985 874 1256
880 750 935 684 993
or view a chart. This process is a major function of online analytical processing
(OLAP). Most OLAP tools rely on a data warehouse to provide consistent data
and fast access.
The ability to browse through the data is a useful feature of OLAP. Most deci-
sion makers want to see subtotals of data. A key method of organizing OLAP data
is to identify a key fact (often sales) and then collect attributes that presumably
affect the fact. As shown in Figure 9.10, OLAP tools depict this data as a mul-
tidimensional cube. Managers use specifc tools to examine various sections of
the data. To illustrate the process, consider a simple example from the Rolling
Thunder Bicycle company database. Managers are interested in sales of bikes.
In particular, they want to look at sales by date, by the model type (race, road,
mountain, full suspension, and hybrid), and by the location of the customer. The
fact they want to measure is the value or amount of the items sold, which is the
price times the quantity. The OLAP tools enable managers to examine any ques-
tion that involves the dimensions of the cube. For instance, they can quickly ex-
amine totals by state, city, month, or category. They can look at subtotals for the
different categories of products or details within individual states. The tools can
provide detail items that can be pictured as a slice of the cube, or they can provide
subtotals of any section. Each attribute represents a potential subtotal. In terms
Reality Bytes: Face Recognition
The general public is easily misled by science fction and Hollywood. It is easy to
think that computers can do anything, or that breakthroughs are only a few months
away. But, computers still struggle at some tasks that humans perform easily. Most
of these tasks fall into the category of pattern recognition, and face recognition is
just the latest example. Many photo sites use facial recognition technology to help
users identify new photos. For example, Googles Picasa automatically tries to tag
new photos by matching faces to existing photos, which simplifes organization and
makes it easier to share them. But the system often fails. One user, Stacey Schlitten-
hard, noted that All babies kind of look alikethey all have little round faces. If I
label one baby as my son, it will label almost every baby as my son. She also noted
that Picasa once labeled a lollipop as her friend. Yi Ma, an associate professor of
electrical and computer engineering at the University of Illinois, and a researcher at
Microsoft Research in China, notes that I dont think, currently, any facial recogni-
tion system is good enough for security purposesnot even close, actually. He also
believes a reliable system is at least a decade away. On the other hand, in June 2011,
when Vancouver fans rioted after losing the last round of the NHL championship,
several photos of people were published on the Internet. Within days, people on the
Web had identifed most of the rioters and two people who were caught passionately
kissing in the middle of the street. Similarly, someone posted a video of a woman
arguing with a subway conductor in New York and she was quickly identifed by
people who watched the video. In that sense, computerized facial recognition is not
the threat to privacythe prevalence of cameras and the ability to share them on the
Web is removing anonymity in everyday life.
Adapted from John D. Sutter, Why Face Recognition Isn't Scary Yet, CNN On-
line, July 9, 2010; and Brian Stelter, Upending Anonymity, These Days the Web
Unmasks Everyone, The New York Times, June 20, 2011.
of database queries, think of an OLAP cube as a collection of GROUP BY state-
ments. However, the browsing tools are interactive and show managers the results
instantly. For example, you can use checkboxes to select specifc states. You can
begin looking at total sales of all bikes, and then look at the detail by model type.
With a few clicks, you can choose a specifc slice of the cube, such as one state,
and then look at data for a single month or model type, or rotate the cube and look
at totals by month.
Although most DBMS vendors (including Microsoft, Oracle, and IBM) pro-
vide OLAP cube browsers, Microsoft provides the PivotTable interface that works
with almost any DBMS or spreadsheet. A PivotTable is an interactive interface to
a multidimensional cube. A PivotTable is created on the users machinemost us-
ers will build pivot tables inside of Microsoft Excel. This tool has several options
and provides a great deal of fexibility for the user.
A PivotTable report for Rolling Thunder Bicycles is shown in Figure 9.11. By
clicking on a row or column dimension, managers can see detail or subtotals. They
can also select specifc items to include in the subtotals. Managers even have the
fexibility to drag the dimensions aroundto move them from columns to rows,
or to change the order of the summations. The four windows in the lower-right-
hand corner specify where each item will be displayed, and you can drag-and-
drop the columns across those windows. Additional options provide other statis-
tics, such as averages.
Figure 9.11
Microsoft PivotTable report. Pivot tools make it easy for managers to examine cube
data from any perspective, to select subsets of the data, to perform calculations, and
to create charts.
The Pivot Chart graphics option makes it easy to create chartsusing the Ex-
cel interface that is familiar to most business managers. As shown in Figure 9.12,
the system uses the same data as the PivotTable, but creates interactive charts to
display the data visually. Once the connection to the data is made, you select the
basic chart type and pick the attributes to be displayed. You, or your manager, can
change the chart type easily, and even drag data columns to look at new relation-
ships. The options are similar to those in the standard Excel charting package. But
remember that the system is interactive, and you can quickly look at any relation-
ship of interest. The data connection is also dynamic, so that as the data changes
in the underlying database, you can refresh the chart or PivotTable to get up-to-
date values without altering the spreadsheet or chart design.
Several vendors provide interactive cube browsers. Microsoft and Oracle have
versions built into their DBMS reporting systems so that you can embed them
on a Web reportgiving managers Web access to these tools. The PivotTables
and Charts built into Excel are convenient and accessible to most organizations
without spending additional money. Even if your data is stored in other Excel
spreadsheets, you can build PivotTables and Charts that extract the data from oth-
er spreadsheets and provide the same interactive data browsing. Additionally, you
can use the statistics, functions, and programming capabilities of Excel to analyze
the extracted data.
Figure 9.12
Microsoft Pivot Chart. A Pivot Chart uses the same type of data as a PivotTable, but
displays the data graphically. It is easy to change the type of chart or to change the
data displayed by clicking options or dragging attributes to different sections.
Decision Support System
How does a decision support system help you analyze data?
Once you have the data, you often need to analyze it statistically, or by applying
discipline-specifc models. A decision support system (DSS) consists of three
basic components to help you analyze data: (1) data retrieval, (2) model evalua-
tion, and (3) visualization of the results. Today, a data warehouse or OLAP cube
is often used as the data source. The model is often developed by experts (usually
consultants) and evaluated in a spreadsheet. The visualization component gener-
ally consists of charts, but more sophisticated time lines and schedules are used
for complex problems.
To understand the value of a DSS, it is easiest to work with a couple of ex-
amples. Thousands of examples exist, but they often require detailed knowledge
from a specialized discipline. On the other hand, most managers will need some
familiarity with marketing and with human resources management. The examples
are relatively small and you will cover more complex models in other business
classes. The principles are the same, and you can often use the same tools. Along
the same lines, if you want to apply these models to real business problems, you
will have to collect more data and add more features to the models.
Marketing Forecasts
Marketing departments are responsible for market research, sales forecasting,
management of the sales staff, advertising, and promotion. In some frms they also
process orders and manage the design of new products and features. Processing
orders is essentially a transaction-processing task. The others involve tactical or
strategic questions that are more complex, so we will focus on those tasks.
An enormous amount of data is available for market research. Figure 9.13 pres-
ents some of the common data available for marketing purposes. Internally, the
marketing department maintains records of sales and basic customer attributes.
With some frms, there can be a longer distance between the frm and the fnal
customer. For instance, manufacturers typically sell products to wholesalers, who
place the products in individual stores, where they reach the fnal customer. In
these cases it is more diffcult to identify customer needs and forecast sales. There
Figure 9.13
Common marketing data sources. There are three primary sources of marketing
data: internal collections, specialty research companies, and government agencies.
Detailed data is available on the industry, customers, regions, and competitors.
Internal Purchase Government
Sales
Warrantycards
Customerservicelines
Coupons
Surveys
Focusgroups
Scannerdata
Competitivemarketanalysis
Mailingandphonelists
Subscriberlists
Ratingservices(e.g.,Arbitron)
Shipping,especiallyforeign
Census
Income
Demographics
Regional data
Legalregistration
Drivers license
Marriage
Housing/construction
will be delays in receiving sales data because the retailers and wholesalers typi-
cally place bulk orders. Furthermore, it is more diffcult to identify customer pref-
erences because their purchases are fltered through other companies. Marketing
departments also have access to data that is collected by other frms. In a manufac-
turing environment, marketers might get raw sales data from the wholesalers and
retailers. On the retail side, with the pervasiveness of checkout scanners, it is now
possible to buy daily and hourly sales records from thousands of stores in various
cities. This data contains sales of your products as well as rivals products.
Marketing is often asked to forecast sales. Several different methods can be
used, but a straightforward approach is to begin with statistical forecasts. Consid-
er a simple example shown in Figure 9.14 for a fctional store that sells consumer
products nationwide. The sales estimate is based on economics where sales are
dependent on time and on consumers income. As consumer income increases,
they will be more likely to purchase the companys merchandise. Gross domestic
product (GDP) is often used as a proxy for consumer income.
Notice the seasonal peaks in sales for each fourth quarter. It is important to cap-
ture this holiday sales effect. Since national GDP and household income have this
same effect, you can build a model based on the relationship of your sales to GDP
and time. The process is described in Figure 9.15, and all of the steps can be per-
formed in a spreadsheet. Regression provides the estimate of the coeffcients that
describe the relationship between sales and GDP and time. The R-squared value
is over 90 percent and all of the coeffcients have high t-values, so it is a strong
relationship. If you look closely at the forecast values for the fourth quarters, you
will notice some errors that indicate the model is missing some elements.
GDP
Sales
Forecast
GDP and Sales
3000
2500
2000
1500
1000
500
0
140
120
100
80
60
40
20
0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48
Figure 9.14
Sales forecast. Note the seasonal peaks in the fourth quarter. The points beyond
quarter 40 are forecasts based on time and the relationship to income (GDP). This
forecast requires that GDP predictions be made for each future quarter, but these
values can often be obtained from government forecasts.
Forecasting the GDP is a little tricky, so most people just use government fore-
casts. But be careful to get nonseasonally adjusted values (which are diffcult to
fnd), so they show the quarterly cycle. Unfortunately, the U.S. Bureau of Eco-
nomic Analysis has been reporting only seasonally adjusted data since 2004. You
can also forecast the quarterly values yourself. The easy way to preserve the cycle
is to forecast the quarters independently (all frst quarters, all second quarters, and
so on). Then plug these values in for the forecast, multiply by the estimated coef-
fcients, and graph the result. It might not be quite as accurate as a full time-series
estimation technique, but any business student can make it work, and it is better
than guessing or wishful thinking.
An important HRM task in any organization is the need to allocate raises. Using
a merit pay system, each employee is evaluated on the basis of factors related to
his or her job. Typically, each manager is given a fxed amount of money to allo-
cate among the employees. The goal is to distribute the money relative to the per-
formance appraisals, provide suffcient incentives to retain employees, and meet
equal employment opportunity guidelines. Many of these goals are conficting,
especially with a fnite amount of money available. To set the actual raises, man-
agers need to examine the raw data. On the other hand, a graph makes it easier to
compare the various goals.
A few specialized software packages can help you determine merit raises.
However, as shown in Figure 9.16, it is possible to create a small system using
a spreadsheet. A spreadsheet that can display a graph alongside the data tables
is particularly useful. Assume that the company wishes to give a certain portion
of the raise based on the average performance ratings. The amount of money per
point (currently $100) can be changed. Each person can be given an additional
market adjustment raise. The total departmental raises cannot exceed the allocated
total ($10,000).
Figure 9.15
Forecasting process. All of these steps can be performed by a spreadsheet. To forecast
the quarterly GDP values, simply split them into four columns, one for each quarter.
Then use the spreadsheets linear forecast tool to extend the columns.
Data: Quarterly Sales and GDP for 16 years.
Model: Sales = b0 + b1 Time + b2 GDP
Analysis: Estimate model coeffcients with regression
Coefcients Std. Error T-Stat
Intercept -48.887 13.631 -3.586
Time -0.941 0.294 -3.197
GDP 0.067 0.011 6.326
Output: Compute Sales prediction. Graph forecast.
The goal is to fll in the market adjustment column so that the raises match the
performance appraisals. As illustrated by the graph in Figure 9.17, the manager
can evaluate both absolute dollar raise or the percentage increase. The total de-
partmental raises should be equal to $10,000. By displaying the graph next to the
last columns in the spreadsheet, it is possible to watch the changes as you enter
the data. This immediate feedback makes it easier to set the raises you prefer. Use
of some type of DSS analytical system is helpful for identifying and minimizing
illegal discrimination in salaries.
As shown in Figure 9.16, you can use the Offce 2010 conditional formatting
feature to create a simple chart to show where each persons salary falls within the
defned range for the job category. The additional chart is somewhat diffcult to
read, but it does provide useful information. Also, if the system has access to data
across the organization, it is possible to statistically analyze the raises assigned by
each manager to ensure raises are given in a non-discriminatory fashion. Such a
system will probably be unlikely to catch small deviations, but if systematic dif-
ferences across managers or individuals appear, HR managers can investigate fur-
ther. The main point of this example is that managers can visually see the effects
of their decisions.
Many, many other problems can be analyzed with decision support systems.
The key to a DSS is that you need a model to analyze. Typically, an expert will
create the model and build a DSS that loads current data. Managers still need to
understand the model and the results, because they are the ones who ultimately
analyze the data and make the decisions.
Geographical Information Systems
How do you visualize data that depends on location? Many as-
pects of business can beneft by modeling problems as geographical relationships.
Merit Pay raise pool 10000
perf. pct salary range (000) current merit market total New
Name R1 R2 R3 perf low high avg salary 100 adjust. raise raise pct salary Pct Range
Caulkins 9 7 6 73% 28.4 37.5 36.4 35.8 733 800 1533 4.3% 37.3 0.98
Jihong 3 6 7 53% 15.4 18.9 16.3 17.9 533 100 633 3.5% 18.5 0.90
Louganis 8 7 7 73% 26.7 30.2 28.9 29.5 733 850 1583 5.4% 31.1 1.25
Naber 9 8 8 83% 19.5 23.2 21.4 19.8 833 1030 1863 9.4% 21.7 0.58
Spitz 3 4 3 33% 17.3 22.4 18.4 17.5 333 600 933 5.3% 18.4 0.22
Weissmuller 5 4 6 50% 32.5 60.4 45.2 53.2 500 2955 3455 6.5% 56.7 0.87
Department 6.2 6.0 6.2 23.3 32.1 23.8 21.7 3665 6335 10000 5.7% 30.6 0.83
Corporate 5.0 6.0 5.0 124.3 124.3 18.9 18.9 Remain 0
Figure 9.16
Merit pay analysis. With a merit system, salary increases should be related to
performance evaluations (denoted r1, r2, r3). Managers are typically given a fxed
pool of money to distribute among the employees. Employee raises should be based
on merit evaluations, current salary, the salary range for the job. Market adjustments
are often paid to attract workers in high-demand felds. A spreadsheet can be used
to model the effects of various policies. In this example, the manager has allocated
$100 for each merit percentage point. The rest of the money will be given as market
adjustments. The effects of the adjustments can be seen in the graph displayed in the
next fgure.
For instance, to choose the site of retail outlets, you need to know the location and
travel patterns of your potential customers as well as the locations of your compet-
itors. Manufacturing can be made more effcient if you know the locations of raw
materials, suppliers, and workers. Similarly, locations of distribution warehouses
need to be chosen based on retail outlets, manufacturing facilities, and transpor-
tation routes. Thousands of other geographical considerations exist in business,
such as monitoring pollution discharges, routing and tracking delivery vehicles,
classifying areas for risk of crimes and fre, following weather patterns, or tracing
migration paths of fsh for commercial fshing. Geographic information systems
(GISs) have been designed to identify and display relationships between business
data and locations. Arc Info and Microsofts MapPoint are two of many com-
mercial GIS packages available.
A GIS begins with the capability of drawing a map of the area in which you are
interested. It might be a world or national map that displays political boundaries.
It might be a regional map that emphasizes the various transportation routes or
utility lines. It might be a local map that displays roads or even buildings. An oil
exploration company might use a map that displays three-dimensional features
of a small area. A shipping company could use ocean maps that display three-
dimensional images of the ocean passageways. The level of detail depends on the
problem you wish to solve.
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
80.0%
90.0%
0
500
1000
1500
2000
2500
3000
3500
4000
Raises
raise raise pct performance
Figure 9.17
Performance evaluation. Using a separate y-axis for the two types of data and
overlaying line plots on the bar chart makes this graph easier to read. If this graph
is dynamically linked to the salary table, the manager can make salary changes and
instantly compare the raises to the performance ratings.
Maps and Location Data
Most maps today are built from digitized map data. Each item is stored as a geo-
graphic point or line segment; defned by latitude, longitude, and altitude. The
maps are drawn mathematically from a collection of these points. Consequently,
the maps can be drawn to any scale and the tools make it easy to zoom in or out
at will. Most U.S. digital maps are based on data that the Bureau of the Census
created for the 1990 national census, known as TIGER. The Bureau of the Census
has every road and house number entered into a giant database. Because of pri-
vacy reasons, they will not sell house locations, but you can get the range of street
numbers for each city block. The U.S. Department of Defense has digital data
available for many areas, including international locations, and often includes el-
evation data. The U.S. Geological Survey topographical maps are also being con-
verted to digital systems. However, keep in mind that the systems being mapped
are constantly changing, so even digital maps often contain missing, incomplete,
or inaccurate dataas the United States learned when it accidentally blew up the
Chinese embassy in Belgrade because the CIA maps were out of date. Due to the
popularity of online mapping systems, several private companies work to add new
data and fx errors.
Several companies have integrated satellite photos into their mapping systems.
Through a variety of programs, the U.S. government has accumulated and re-
leased low-resolution photos of most areas of the planet. A couple of private com-
panies supplement those photos with images captured from vans driving through
major cities. Computers can match the digital map data with the satellite photos
and the drive-by photos to provide realistic views of many cities.
Once you have the base maps, the objective is to overlay additional data on the
maps. For example, you can obtain census data that displays average consumer
characteristics such as income, house price, and number of autos within each geo-
graphic area. The GIS could be used to plot different colors for each income level.
Next you can overlay the locations of your retail stores. If you are selling a high-
price item such as a Cadillac, you want to locate the stores in areas of higher
income.
Although you can buy base geographical data, how do you know the location
of your retail stores? Or how do you plot the locations of delivery vehicles, or
police cars, or trains? The easiest answer today is to use the global positioning
system (GPS), which is a set of satellites maintained by the U.S. government.
A portable receiver tuned to the satellites will identify your location in latitude,
longitude, and elevation (if it can reach four satellites) within 20 feet. Several
handheld units are available for a few hundred dollars. If you work for the De-
partment of Defense, you can get receivers that will identify your location within
a few millimeters, but you need appropriate security clearances to obtain these
receivers. Currently, the government is broadcasting the higher-resolution signal
to civilian receivers, but sometimes blocks it during emergencies. Civilian models
that combine signals from U.S. and Russian satellites provide even better resolu-
tion, but cost about $8,000. Europe has been trying to raise the funds to build its
own GPS system (Galileo) and China has announced that it intends to create its
own system as well. There is probably not enough money, or orbital space, for that
many satellites; but competition is good and should encourage the U.S. to keep
the system open.
As a model, the GIS makes it easier to spot relationships among items. Visual
presentations are generally easy to understand and are persuasive. A GIS can be an
Technology Toolbox: Browsing Data with a PivotTable
Problem: You and your manager need to analyze sales based on several attributes.
Tools: OLAP cube browsers are designed to make it easy for you to examine slices
of the cube and examine a fact (sales) based on several dimensions. The big DBMS
vendors sell cube browsers with their packages. However, you can also use the Mi-
crosoft PivotTable that is built into Excel.
The Rolling Thunder Bicycle data provides a good example for a cube. The fact to
be evaluated is the SalePrice. The more interesting dimensions are OrderDate, Mod-
elType, and SaleState. The date presents a common problem: dates are hierarchies.
You might want to examine sales by year, or you might want to drill down and see
sales by quarter or month. Some cube browsers make it easy to create this hierarchy.
Excel does not create it automatically, but you can use a query to convert the date
into the different dimensions. The qryPivotAll query shows you how to compute the
various year, month, and quarter felds.
In an Excel worksheet, use the Data/PivotTable option to begin. Select the Ex-
ternal Data Source option because the data comes from a DBMS. Follow the basic
steps to get data from a Microsoft Access Database and select your copy of the Roll-
ing Thunder database. In the Query Wizard, fnd the qryPivotAll query, and select
all of its columns by moving them to the right side box. Follow the Next and Finish
prompts to return the data to the spreadsheet.
The structure of the PivotTable contains a space for row variables, column vari-
ables, page variables, and the main fact. Drag Year and YearMonth from the feld list
onto the table column location. Drag the ModelType and SaleState felds onto the
row location. Finally, drag the SalePrice onto the main body of the table. You can
place the remaining felds in the page location if you want to make them easy to fnd
later.
The fun part is playing with the cube. Select a Year heading and click the minus
button to collapse the detail. Do the same with the ModelType or SaleState column.
Watch as the totals are automatically updated. Drag the ModelType or SaleState feld
heading to swap the order. Collapse or expand the feld to summarize or drill down
into the data. You can just as easily expand only one year or one state.
Quick Quiz:
1. How is the cube browser better than writing queries?
2. How would you display quarterly instead of monthly data?
3. How many dimensions can you reasonably include in the cube? How would you
handle additional dimensions?
effective means to convince management that neighborhoods have changed and
that you need to move your retail outlets. A GIS can also be used for simulations
to examine alternatives. For example, a GIS oriented to roadmaps can compute
the time it would take to travel by different routes, helping you establish a distri-
bution pattern for delivery trucks.
Example
Consider the problem faced by a manager in a small retail chain that has stores
located in 10 Florida cities. It sells a combination of hard goods (such as cleaning
supplies, snack items, and drapery rods) and soft goods (mostly clothing). For the
most part, proft margins for soft goods are higher than for hard goods. However,
total sales of hard goods seem to be better than those of soft goodsexcept in
certain stores. The manager has been unable to fnd a reason for the difference, but
a friend who has lived in Florida longer suggested that there might be some geo-
graphical relationship. The basic numbers are presented in Figure 9.18.
Because there are only 10 cities, it might be possible to identify patterns in the
data without using a GIS. However, an actual frm might have several hundred or
a few thousand stores to evaluate. In this case, it is much more diffcult to identify
relationships by examining the raw data. It is better to use a GIS to plot the data.
Different colors can be used to highlight large increases in sales. By overlaying
this data with the population and income data, it is easier to spot patterns. Notice
in Figure 9.19 that there is a correlation between population and total sales. Also,
notice that sales in the northern cities are concentrated more in hard goods than
in the southern cities. Each of the radar charts in the north points to the northwest
(left), while those in the south point to the northeast (right). Once you see this pat-
tern, it becomes clear; yet it is diffcult to see the pattern within the raw data, or
through any other chart.
City 2000 pop 2009
pop
2000
per-capita
income
2009
per-capita
income
2000 hard
good sales
(000)
2000 soft
good sales
(000)
2009 hard
good sales
(000)
2009 soft
good sales
(000)
Clewiston 8,549 7,107 15,466 15,487 452.0 562.5 367.6 525.4
Fort Myers 59,491 64,674 20,256 30,077 535.2 652.9 928.2 1010.3
Gainesville 101,724 116,616 19,428 24,270 365.2 281.7 550.5 459.4
Jacksonville 734,961 813,518 19,275 24,828 990.2 849.1 1321.7 1109.3
Miami 300,691 433,136 18,812 23,169 721.7 833.4 967.1 1280.6
Ocala 55,878 55,568 15,130 20,748 359.0 321.7 486.2 407.3
Orlando 217,889 235,860 20,729 23,936 425.7 509.2 691.5 803.5
Perry 8,045 6,669 14,144 19,295 300.1 267.2 452.9 291.0
Tallahassee 155,218 172,574 20,185 27,845 595.4 489.7 843.8 611.7
Tampa 335,458 343,890 19,062 25,851 767.4 851.0 953.4 1009.1
Figure 9.18
Geographic sales data. We suspect that sales of hard and soft goods are related to
population and income. We also want to know whether there are regional patterns to
the sales.
Most GIS tools use simpler methods of displaying data. Figure 9.20 shows a
common approach where geographic regions (states) are shown in darker colors
for higher values. The chart was created by creating a query in Access to compute
sales by state for the year 2008. Microsoft Map Point was then used to draw the
map and shade the values by state. Map Point also includes demographic data
from the U.S. Census Bureau at relatively detailed levels that can be added to
the chart. More sophisticated, and more expensive, tools such as ESRIs ArcInfo
can display multiple layers of data using overlays. Google maps (and Microsofts
Bing) provide some simple GIS tools using their online systems. The beneft is
that the tools are free, but the options available are limited. Typically, you can
display only push-pins to highlight specifc locations. However, they do support
some programmatic tools so you can automatically create charts from your Web
server.
Data Mining
Is it possible to automate the analysis of data? Many decision mak-
ers, particularly researchers, are buried in data. Transaction-processing systems,
process control, automated research tools, even Web sites all generate thousands,
Figure 9.19
Geographic-based data. It is diffcult to display this much data without overwhelming
managers. Notice that the sales (radar) graphs use size and shape to highlight total
sales, and the changing sales mix. Income is color-coded in a smaller graph. Notice
in the sales graphs that the northern counties experienced a greater increase in sales
in hard goods compared to the southern counties.
Tampa
Miami
Fort Myers
Jacksonville
Tallahassee
Gainesville
Ocala
Orlando
Clewiston
Perry
17,000
15,800
14,600
13,400
12,200-
1990 2000
20,700
19,400
18,100
16,800
15,500-
per capita income
Red
3. 2
Yellow
2. 3
Blue
1. 9
Green
2. 3
Red
1. 7
Yellow
1. 1
Blue
1. 0
Green
1. 1
Red
2. 1
Yellow
1. 7
Blue
1. 1
Green
1. 4
Red
5. 0
Yellow
4. 2
Blue
3. 2
Green
3. 7
Red
1. 8
Yellow
1. 5
Blue
1. 2
Green
1. 4
Red
2. 6
Yellow
3. 0
Blue
1. 9
Green
1. 6
Red
3. 6
Yellow
3. 8
Blue
3. 2
Green
2. 9
Red
3. 5
Yellow
3. 8
Blue
2. 5
Green
2. 0
Red
1. 4
Yellow
2. 0
Blue
2. 1
Green
1. 7
Red
3. 7
Yellow
4. 8
Blue
3. 2
Green
2. 7
2000
Hard
Goods
2000
Soft
Goods
1990
Hard
Goods
1990
Soft
Goods
millions, or even billions of pieces of data a day. Sure, you can retrieve the data
into a data warehouse, run queries, or put it into a spreadsheet, but it is still dif-
fcult for humans to comprehend raw data.
Data mining consists of a variety of tools and techniques to automatically re-
trieve and search data for information. Originally, the term was derogatory be-
cause it represents an undirected search for relationships. Statistically, results
obtained from nonscientifc searches can be spurious and not repeatable. On the
other hand, the tools can fnd minute comparisons that are not ordinarily found
through traditional statistical methods. As the food of data increases, more com-
panies are turning to automated and semiautomated tools to help search databases
and make it easier to visualize patterns. Additional models and research can then
be used to investigate and validate the relationships.
Data mining tools exist for numerical and other types of data. The numerical
tools tend to be based on statistical theory. The others sometimes use statistical
theory but are often highly specialized. For example, event analyses look for pat-
terns in data based on timing or sequencesuch as the path viewers take through
a Web site. This chapter focuses on the more generic statistical tools because they
are used in many different companies and tasks. The specialty tools typically re-
quire a detailed background in a specialty discipline, so you can learn more about
them when you study specifc topics.
Figure 9.20
Geographic-based data. Tools such as Microsoft Map Point often display location-
based data by shading geographic regions based on the relative levels of the data.
This map shows sales by state for Rolling Thunder Bicycles in 2008.
As shown in Figure 9.21, a common theme in the numerical data mining tools
is the search for items that infuence a given fact variable. Much like the data
cube, you might choose the Sales variable as the fact to be investigated. In sta-
tistics terms, it becomes the dependent variable. You then want the system to ex-
amine various attributes or independent variables to see how they infuence the
dependent variable. For instance, you might have data on customer location, in-
come, product category, time and month purchased, marketing budget, and a host
of other variables that could conceivably affect the level of sales. Remember this
goal as you read the rest of this section. Even if you have limited experience with
statistics, remember that most of the tools are trying to identify which independent
variables strongly affect the dependent variable.
Correlation is a key statistical tool that is leveraged in data mining. A data min-
ing system can compute the cross correlation for all dimensions. High correla-
tions provide a useful indicator of how one dimension (variable) affects another.
Correlation typically represents direct effects between pairs of variables. Multiple
regression is an extension of correlation where multiple dimensions (independent
variables) are used to predict values for a dependent variable. The regression co-
effcients have long been used by statisticians to measure the importance of each
attribute. Special data mining tools extend these concepts into nonlinear relation-
ships, automatically searching for relationships between variables. For example,
marketers could mine the sales data to determine which variables had the stron-
gest impact on sales revenue, including price, quality, advertising, packaging, or
collections of product attributes. Regression tools provide numerical estimates of
the coeffcients for direct and indirect effects, along with measures of the accuracy
of the estimate (standard errors).
Clustering is another data mining technique. It tries to fnd groups of items
that have similar attribute values. For example, a car manufacturer might fnd that
younger buyers are attracted to one car model, while older buyers tend to pur-
Sales
Location
Dependent Variable
Fact
Independent Variables
Dimensions/Attributes
Age
Income
Time
Month
Category
Direct eects
Indirect eects
Figure 9.21
Common data mining task. Determine how independent variables affect the
dependent fact variable. Models can have direct or indirect effects. Basic linear
models are easy to estimate, non-linear ones are harder to estimate and more diffcult
to understand.
chase a different model. These concepts are relatively easy to understand in two or
three dimensions. The tools can search in higher dimensions, looking for group-
ings across dozens of attributes. The results are more diffcult to interpret and act
on, but can provide useful information in dealing with complex datasets. Figure
9.22 shows an example that compares cars on three dimensions: acceleration time,
weight, and miles per gallon. The results show that there were two distinct groups
of cars at that point in time. Heavy cars with low acceleration times (faster), and
low miles per gallon versus light, slow cars that achieved more miles per gallon.
Clustering is a useful tool when you are frst looking at data or when you need
to reduce the number of dimensions. In the car example, you can reduce the analy-
sis to looking at those two groups of cars. Similarly, you might learn that you have
three or four types of primary customers, enabling you to focus your analysis and
decisions on how representatives from each group will react to changes.
One of the classic data mining tools is market basket analysis (or association
rules). Market basket analysis was designed to address the question, What items
do people tend to purchase together? Figure 9.23 shows the classic example that
was purportedly identifed by a convenience store. Managers found that on week-
ends, people often purchased both beer and diapers. This raises the immediate
issue of how you can use the results. You might stock the items near each other
to encourage people to buy both. Or, you might place the items at opposite ends
of an aisleforcing people to walk past the high-impulse items such as chips.
Or, perhaps you discount the diapers to attract more customers and make up the
additional proft by not discounting the beer. The tool is generic and is used by
Figure 9.22
Clustering. Sometimes you want to know which items are similar to others. You
might face distinct groups of customers, or your products might be perceived
in certain clusters. This example compares cars in terms of three dimensions
(acceleration time, weight, and gas mileage) to reveal two distinct groups.
companies such as Amazon.com and Netfix to show you books that were bought
by other customers. If you are interested in the frst book, you might also like
the books that other people purchased along with that book. Tivo, the television
recording system, uses a similar process to identify programs that you might be
interested in watching. To illustrate the hazards of data mining, many people have
reported interesting twists with Tivo. For instance, if you watch one movie for
children, the system will begin to offer more shows geared toward children. Mar-
ket basket or association analysis is relatively easy to computeparticularly for
pairs of items. Several tools exist to perform the computations for you, including a
couple of open source tools such as Weka.
Do not let the name fool you. Market basket analysis has many more uses than
the example shows. More properly known as association analysis, it works well
with categorical data that does not need to be measured numerically. It basically
estimates the probability that any two (or more) items will occur together. These
items could be purchases, or they could be diseases, events such as Web clicks,
votes, and so on.
Several data mining tools have been built in the past few years. They are rela-
tively easy to use, but you might want to review your statistics to help understand
the results. One catch with the tools is that some of them can take time to run
even with fast machines, some complex analyses can take hours. For example,
market basket analysis is relatively fast with pairs of items, but requires expo-
nentially more computations and time if you attempt to compare more than two
items at a time. A bigger problem is that the results might be meaningless or not
reproducible. They might simply have arisen because of some random occurrence
in the data. The bottom line is that you have to carefully evaluate each piece of
information to make sure it is relevant and important.
Expert Systems
Can information technology be more intelligent? Can it ana-
lyze data and evaluate rules? Imagine your life as a top-notch manager.
Coworkers perceive you as an expert and value your advice and problem-solving
skills. You are constantly answering questions and there are always more prob-
lems than you can handle. You are using decision support systems and integrated
information technology to perform your job better and more effciently, but it is
Figure 9.23
Market basket analysis. What items do people buy together? Data mining tools can
examine each purchase to identify relationships. You can use this information to
increase cross selling.
Technology Toolbox: Forecasting a Trend
Problem: You need a basic forecast of sales data.
Tools: Many statistical tools and packages exist to build models and estimate trends.
The simplest technique is a linear regression. Excel has tools to calculate and display
the results.
Open the Rolling Thunder database and create a query that computes the total
sales by year and month. The Format function converts the date: Format(OrderDate,
yyyy-mm). Save this query within Access, and close the database. To import the
data into Excel, open a new worksheet and choose the Data/Import/New Database
Query option. This approach will build a dynamic link to the query.
Create a line chart of the data. Right-click the plotted line and select the option to
add a trend line. You should stick with the linear trend, but you might want to test the
other choices. Select the options to display the equation on the chart.
To see the actual forecast values, select the entire column of data for SumOf-
SalePrice, but do not include the title row. Scroll down to the last row in the series
and fnd the small square handle on the lower-right-side corner. Drag this handle to
extend the series for six months. Add the appropriate months in the month column.
Edit the chart and edit the series selections to include the new rows. Notice that the
new points fall exactly on the trend line.
Excel includes more powerful statistical tools, including one that performs multi-
ple regression when you want to examine the effect of many variables. Although this
dataset contains only one variable, you can still use it to test the tool. First, insert a
new column that numbers the months as 1, 2, 3, and so on. Ensure the Analysis Tool-
Pak is loaded and select Tools/Data Analysis and pick the Regression tool. As the Y
range, select the sales column including the title row but excluding the six rows you
forecast earlier. Select the month column as the X range. Be sure to check the Labels
box, and then run the regression. This tool provides the standard regression coeff-
cients, probabilities, and diagnostics.
Quick Quiz:
1. Why is a linear forecast usually safer than nonlinear?
2. Why do you need to create a new column with month numbers for regression
instead of using the formatted year-month column?
3. What happens to the trend line r-squared value on the chart when you add the
new forecast rows to the chart?
not enough. Can technology help you with more complex decisions and problem
solving? From another perspective, when you encounter new problems with dif-
ferent, complex models, it would be helpful to have an expert assist you with ap-
plying and understanding the models. Yet experts or consultants are expensive and
not always available. Can you somehow capture the knowledge and methods of
experts and use technology to make this knowledge available to workers through-
out the company?
Specialized Problems
Expert systems have proven useful for many problems. The goal of an expert sys-
tem is to enable novices to achieve results similar to those of an expert. The users
need to understand the basic problem, learn the terminology, and be able to an-
swer questions. For example, a typical patient would not be able to use a medical
expert system because the questions and terms would not make any sense.
Think of an expert system as a consultant in a box. The consultant can solve
only certain specifc problems. For example, perhaps a retail store manager needs
to estimate buying patterns for the next few months. The manager might call a
marketing consultant to survey buyers and statistically search for patterns. The
consultant will ask questions to determine the basic objectives and identify prob-
lems. Similarly, a production manager might be having problems with a certain
machine. The manager might call a support line or a repair technician. The advice
in this situation will be quite different from the marketing example, because the
topics (or domains) of the two problems are different. It would be diffcult to cre-
ate one computer program that could help you with both types of problems. On
the other hand, there are similarities in the approach to the two problems. Com-
puterized expert systems are designed to solve narrow, specialized problems. Each
problem can be relatively complex, but it must be reasonably well defned. Many
business problems fall into this category, and expert systems can be built for each
problem.
Diagnostic Problems
Several problems in the world can be classifed as diagnostic situations. These
problems arise when the decision maker is presented with a set of symptoms and
is asked to fnd the cause of the problem, as well as solutions. Consider a frm that
uses a complex machine. If the machine breaks down, production stops until it
is fxed. In addition, maintenance tasks have to be performed every day to keep
the machine running. The company hires an engineer to perform these tasks. The
engineer also knows which adjustments to make if various symptoms appear. This
system has been working well, and the company wishes to expand to other loca-
tions with a franchise system. The problem is that there is only one engineer, and
it would be too expensive to have a highly trained engineer at each location.
One possible solution would be to set up a phone link between the franchises
and the engineer. One person at each franchise would be trained in the basics of
the machine. If problems arise, the person could call the engineer. The engineer
would ask specifc questions, such as What do the gauges show? The answers
will lead the engineer to ask other questions. Eventually, the engineer makes rec-
ommendations based on the answers.
Of course, if there are many franchises, the engineer will be too busy to solve
all of the problems. Also, if the businesses are located in different countries, the
time differences may not allow everyone enough access to the engineer. A bet-
ter solution is to create a computerized expert system. All the experts questions,
recommendations, and rules can be entered into a computer system that is distrib-
uted to each franchise. If there is a problem, the on-site person turns to the expert
system. The system asks the same questions that the engineer would and arrives at
the same recommendations.
As shown in Figure 9.24, expert systems also have the ability to explain their
recommendations. In more complex examples, while running the ES, the user can
ask it to explain why it asked a particular question or why it arrived at some con-
clusion. The ES traces through the answers it was given and explains its reason-
ing. This ability helps the user gain confdence in the decisions, allows mistakes to
be corrected, and helps the users remember the answer for future reference.
The business world offers many examples of diagnostic situations, such as
identifying causes of defects, fnding the source of delays, and keeping complex
equipment running. The common characteristic is that you are faced with a set of
symptoms, and you need to fnd the cause.
Speedy Decisions
Other situations can beneft from the use of expert systems. Even if a problem is
not exceedingly complex, you could use an expert system to provide faster re-
sponses or to provide more consistent recommendations. Several advantages can
be gained from making decisions faster than your competitors do. If you can iden-
tify a trend in stock prices before anyone else, you can make a higher proft. If you
can answer customer questions faster, they will be more likely to shop with you in
the future. If you can provide a loan to a customer sooner than anyone else, you
will do more business.
Figure 9.24
Expert System example from ExSys. This sample expert system acts as a
knowledgeable dog lover and asks questions about how you view characteristics of
dogs. Based on your responses it makes a recommendation of which breed might be
good for you. The Web site contains many other examples.
Transaction-processing systems keep much of the basic data that you need to
make decisions. Decision support systems help you analyze that raw data. Both of
these tools enable you to make decisions faster than trying to make the decision
without any computers. However, it still takes time for a human to analyze all of
the information.
Consider the case of a bank loan. In order to get a loan, you go to the bank
and fll out a loan application form. You tell the loan offcer why you want the
loan and provide basic data on income and expenses. Depending on the amount of
money involved, the banker will probably check your credit history, get appraisals
on any collateral, and perhaps get approval by a review offcer or loan committee.
All of these actions take time.
Now, consider the steps involved with a computerized process. First, you need
to tell the bank that you want a loan. Instead of driving to the bank, you could use
the telephone. With a push-button phone, you enter information directly into the
banks computer. The computer would give you a choice of loan types (car, boat,
personal, etc.), and you push a button to select one. You enter the amount of mon-
ey you want to borrow. The next step is to check your credit history. Your income,
expenses, and credit record are available to the bank from national credit reporting
agencies. The bank might also have its own database. The banks computer could
be connected to credit agency computers to collect additional data on your credit
history.
To make the fnal decision, the bank needs a set of rules. These rules take into
account the size of the loan, the value of the collateral, as well as your income,
expenses, credit history, and existing loans. When the bank has determined the
proper rules, the computer performs the analyses. If the bankers trust the rules, the
computer could make the fnal decision. For example, there would be no need for
a loan offcer to be involved in simple decisions, such as making small car loans
to customers with large savings accounts. With an expert system, a bank can cut
the loan-approval period down to a few minutes on the phone.
Many other decisions need to be made rapidly. The frst step in all of these cas-
es is to make sure that the transaction-processing system provides the necessary
raw data. The second step is to create a set of rules for making the decision. The
diffculty lies in fnding these rules. For some problems, there are well-defned
rules that can be trusted. For other problems, the rules may not exist. In this case,
the company will probably still need a human to make the fnal decision.
Consistency
The example of the bank loan demonstrates another advantage of expert systems.
Business decisions are subject to a wide variety of nondiscrimination laws. An
expert system can be used to provide consistent decisions. The rules followed by
the ES can be set up to avoid illegal discrimination. Businesses also have credit
ratings, which are often determined by Credit Clearing House (CCH). CCH
uses an expert system to make the easy decisions, which speeds up the process
by allowing humans to focus on the more complicated cases. It also leads to con-
sistent application of the rules.
Consider the loan example. If each loan offcer makes individual decisions, it is
hard to determine whether they are consistent with corporate policy. Each individ-
ual decision would have to be checked to make sure it was nondiscriminatory. On
the other hand, a committee could spend several weeks creating a set of lending
rules that can be verifed to be sure they are legal and ethical. As long as the bank
employees follow the recommendations of the ES, the outcome should not be dis-
criminatory. Because there should be few cases where the loan offcer overrules
the ES, managers will have more time to examine each of these circumstances.
Many business decisions need to be performed consistently to avoid bias and
to treat people equally. Loans, pricing, raises, and promotions are some examples.
However, there can be problems with using a computer system to enforce stan-
dards. The main diffculty lies in creating a set of rules that accurately describe
the decisions and standards. For example, it might be useful to have a set of rules
regarding raises and promotions, but think about what happens if an employees
job does not ft the basic rules. Organizations continually change, which means
the rules have to be monitored and changed regularly.
Training
Training employees is closely associated with problems of consistency. All orga-
nizations must train employees. If the tasks are complex and the decisions highly
unstructured, it can take years for employees to learn the rules and gain the experi-
ence needed to deal with problems. Two features of expert systems help employ-
Reality Bytes: Dont Bet Against the Machine
Online poker sites are hot. Many people, particularly college students, see it as a
simple diversion or a way to pick up a couple of bucks. But have you considered
how easy it is for someone to cheat at online poker? Some of the better gaming sites
try to detect obvious cheats, but it can be hard to catch a group of players who work
together to feece an outsider. Players also face a new threat: poker botsautomated
programs that play poker. Some people are aboveboard and build them to search for
new ideas in AI. There is even an international competition for poker bots. Others
build poker bots and put them online without telling opponents. But how good are
the poker bots? Dr. Jonathan Schaeffer, a professor of computer science at the Uni-
versity of Alberta, has spent over 14 years developing a poker bot, and he competes
aboveboard in public tournaments. In 2003, his program went head-to-head against a
top poker player. For the frst 4,000 hands, the computer was leading, but the poker
player eventually analyzed the programs approach and developed a new strategy
to beat it. But how many poker games involve top-level players? Dr. Schaeffer ob-
served that when you look at the low-limit tables, I have not doubt that there are
computer programs that can win at that level consistently. And dont forget that these
programs can play all the time, without getting tired. And they can play at multiple
tables. In 2005, the Golden Palace in Las Vegas sponsored a poker bot competi-
tion. Poker ProBot created by Hilton Givens of Lafayette, IN won after nearly 5,000
hands, earning the $100,000 prize. He also had the opportunity to match his program
against human pro player Phil Laak. Phil trounced the machine in 399 hands, and
also beat the University of Alertas PokiX bot in 290 hands. But most experts believe
that just as in the chess world, the machines will eventually improve to beat humans.
In 2011, the U.S. government shut down most of the online poker sites doing busi-
ness America--primarily because they allegedly violated the U.S.laws against online
gambling. Ultimately, online poker is a fool's game.
Adapted from Shawn P. Roarke, Bots Now Battle Humans for Poker Supremacy,
Fox Sports, July 20, 2005.
ees learn. First, employees learn what questions need to be asked. In particular,
after using the system for a while, certain groups of questions will occur together.
Second, most expert systems have provisions for explaining their answers (and
the motivation for each question). At any point, an employee can ask the expert
system why it asked a certain question or why it reached a conclusion.
Building Expert Systems
How do you create an expert system? At frst glance, you would sus-
pect that expert systems are hard to create. However, except for one step, which
is hard, tools exist to make the job easier. The area that causes the most problems
when you are creating expert systems is fnding a cooperative expert who fully
understands and can explain the problem. Some problems are so complex that it
is diffcult to explain the reasoning process. Sometimes the expert may rely on
vague descriptions and minor nuances that cannot be written down. Even though
expert systems can deal with these types of problems, it might take too long to
determine the entire process. Also, if you transfer the experts knowledge to a
computer, the expert might worry about losing his or her job.
Most expert systems are built as a knowledge base that is processed or ana-
lyzed by an inference engine. A knowledge base consists of basic data and a set
of rules. In most situations, an inference engine applies new observations to the
knowledge base and analyzes the rules to reach a conclusion.
The basic steps to create an expert system are (1) analyze the situation and
identify needed data and possible outcomes, (2) determine relationships between
data and rules that are followed in making the decision, (3) enter the data and rules
into an expert system shell, and (4) design questions and responses. A knowledge
engineer is often hired to organize the data, help devise the rules, and enter the
criteria into the expert system shell, or supervise programmers as they create an
expert system.
First, compute the monthly income before taxes.
Next, compute the monthly payment of the loan.
If the payment is greater than 5% of income:
Compute total of other loans payments.
Compute payments as percent of monthly income.
If this percent is less than 25%:
If the new loan is less than 10%, make loan.
Else:
If total monthly expenses are less than 40% of income,
make the loan.
Else:
If less than 50% and has been a customer for more than
5 years or if less than 60% and has been a customer
for 10 years and has lived at the same address for 5 years,
make the loan.
Figure 9.25
Sample rules for the bank loan. A portion of the business rules that are used to
determine whether a person should get a loan.
Knowledge Base
A knowledge base is more than a simple database. It consists of data but also
contains rules, logic, and links among data elements. In most cases, it contains
less structured and more descriptive data. For example, an ES for medicine might
have a list of symptoms that contains items like high temperature and intense
muscle pain. This knowledge base is the reason why the problem must be nar-
row in scope. Even narrow, well-defned problems can require large amounts of
information and thousands of rules or relationships. The real challenge in building
expert systems is to devise the knowledge base with its associated rules.
Three basic types of expert systems are in use today. They are defned by how
the knowledge base is organized: by rules, frames, or cases.
Rules
The heart of a rule-based ES is a set of logical rules. These rules are often com-
plicated. Consider some of the rules that might be needed for an ES to evaluate
bank loans, as shown in Figure 9.25. This example has been simplifed to keep it
short. There will usually be hundreds of rules or conditions to cover a wide variety
of situations. Rules are often presented as If Then else statements. They
can include Boolean conjunctions such as AND, OR, NOT. Figure 9.26 presents a
portion of a decision tree that visually displays the rules.
The diffculty with any ES lies in determining these rules. Some of them will
be easy. Others will be complex. Most of them will come from the expert. Unfor-
tunately, most people do not usually express their thoughts in the form of these
rules. Although a person might follow rules of this sort, they can be diffcult to ex-
press. It is even more diffcult to remember all the rules at one time. For instance,
Payments
< 10%
monthly income?
Other loans
total < 30%
monthly income?
Credit
History
Job
Stability Approve
the loan
Deny
the loan
No
Yes
Good
Yes
No
Bad
So-so
Good Poor
Figure 9.26
Decision tree for sample bank loan expert system. Parts of a knowledge base are
often expressed as a decision tree. Each answer to a question leads to additional
questions and eventually to a decision. Notice that questions sometimes require
numeric answers but can also rely on subjective comments.
say you have lived in the same place for fve years and a new person moves into
the neighborhood. She asks you to describe the best ways to get to school, the
mall, and the grocery store. Then she asks you for the best shortcuts if one of the
roads is closed. This problem is relatively simple, but can you sit down right now
and provide a complete list of all the rules?
Creating an ES
More commonly today, an ES is built from an expert system shell. This program
provides a way to collect data, enter rules, talk to users, present results, and evalu-
ate the rules. To create an ES, you must know what data you need and all of the
rules. Once you express this knowledge base in the format used by the shells
inference engine, the shell takes care of the other problems. Many ES shells are
available on a wide variety of computers; Jess and Clips are two common systems
that are available free or at relatively low cost.
To understand how to create an ES, consider the bank loan example. A typical
dialogue with the user (the loan clerk) appears in Figure 9.27. Notice that the ES
begins by asking some basic information-gathering questions. The responses of
the user are underlined. Once the basic data is collected, the ES performs some
computations and follows the built-in rules. Notice that the ES follows the rule
that asks for the other loan payments. However, the loan clerk does not know
about this rule, so he or she asks for clarifcation. This ability to ask questions is a
powerful feature of expert systems.
Figure 9.27
Bank loan sample screen. An expert system carries on a dialogue with the user. The
ES asks questions and uses the answers to ask additional questions. The user can
ask the ES to explain a decision or a question. Hence the ES can be used for training
purposes.
Welcome to the Loan Evaluation System.
What is the purpose of the loan? car
How much money will be loaned? 15,000
For how many years? 5
The current interest rate is 10%.
The payment will be $297.02 per month.
What is the annual income? 24,000
What is the total monthly payments of other loans?
Why?
Because the payment is more than 10% of the monthly income.
What is the total monthly payments of other loans?
50.00
The loan should be approved, because there is only a 2% chance of default.
Once you have collected all of the rules involved in the problem, you enter
them into the ES shell. The shell lets you type in the questions you want to ask
the user. You defne the calculations and tell the shell how to look up any other
information you need (e.g., the interest rates for auto loans). You then enter the
conditions that tell the shell what questions to ask next. If there are many rules
with complex interactions, it is more diffcult to enter the rules into the shell. One
advantage of ES shells is that you generally have to enter only the basic rules and
data. As the user enters the data, the shell performs the calculations and follows
the rules. The shell also automatically answers the user questions. You do not have
to be a computer programmer to create an ES with a shell. With training, many
users can create their own expert systems using a shell. However, there are many
dangers inherent in ES development, so it helps to have someone evaluate and test
the resulting system.
Limitations of Expert Systems
Expert systems are useful tools that can be applied to several specialized prob-
lems. However, several important drawbacks arise in their design and use. First,
they can be created only for specifc, narrowly defned problems. Some complex
problems contain too many rules with too many interactions. It quickly becomes
impossible to express all of the interrelationships. For example, it is currently im-
possible to create a medical diagnostic system that covers all possible diseases.
However, smaller systems are in use that help determine drug dosages and other
treatments such as radiation levels for cancer patients.
Another problem that users and designers have encountered is that it can be
diffcult to modify the knowledge base in an expert system. As the environment or
Reality Bytes: One Smart Machine is Better than Thousands of People
Many people have touted the potential benefts of crowd-sourcing: breaking a prob-
lem into small pieces and using the Internet to assign each piece to thousands of
workers. Amazon developed the Mechanical Turk to facilitate these types of tasks.
Anyone can sign up to complete tasks and receive wagesoften only pennies a
chore. The goal is to handle tasks that are easy for humans but diffcult for machines.
The problem is that people being paid low wages are usually not very good and
machines are getting smarter. The Web site Yelp ran a comparison test. First, re-
searchers developed a multiple choice test to try to identify the best Turkers who
could correctly categorize a business (restaurant, clothing, etc.), phone number, and
address; based on its Web site. The initial test was given to 4,660 applicants, only
79 of them passed. This select group of the best were then given the problem of
classifying business information where three people saw the same site. If a major-
ity (2-3) agreed, the site was then classifed. The researchers then applied a Nave
Bayes classifer data mining algorithm on the same set of problems. The algorithm
was trained using 12 million examples submitted earlier by Yelp users. The computer
algorithm easily beat the humanscorrectly classifying the sites a third more often
than the people. Does that mean machines are smarter than humans? Or just better
than poorly-paid humans with low motivation?
Adapted from Christopher Mims, AI Defeats the Hivemind, Technology Review,
December 20, 2010.
problem changes, the expert system needs to be updated. The changes are relative-
ly easy to make if they affect only a few rules. However, many expert systems use
hundreds of interrelated rules. It is not always clear which rules need to be altered,
and changes to one rule can affect many of the others. In essence, as the situa-
tion changes, the company is forced to completely redesign the expert system. In
fast-changing industries, it would cost too much to continually redesign an expert
system. In the lending example, a policy change based on monthly income would
be relatively easy to implement. On the other hand, some changes in policy would
force a complete redesign of the expert system. For instance, a bank might decide
to grant loans to almost everyone but charge riskier people higher interest rates.
Probably the greatest diffculty in creating an expert system is determining the
logic rules or frames that will lead to the proper conclusions. It requires fnding an
expert who understands the process and can express the rules in a form that can be
used by the expert system.
Management Issues of Expert Systems
Creating and building an expert system involve many of the same issues encoun-
tered in building any other information system. For instance, the problem must be
well defned, the designers must communicate with the users, and management
and fnancial controls must be in place to evaluate and control the project.
However, expert systems raise additional management issues. Two issues are
particularly important: (1) if an expert transfers knowledge to an expert system,
is there still a need for the expert; and (2) what happens when the expert system
encounters an exception that it was not designed to solve?
The answer to the frst question depends on the individual situation. In cases
where the problem is relatively stable over time, it is possible to transfer expert
knowledge to softwareenabling the frm to reduce the number of experts need-
ed. If this action results in layoffs, the experts will need additional incentives to
cooperate with the development of the system. In other cases, the frm will con-
tinue to need the services of the experts, to make changes to the ES and to solve
new problems. Before starting an ES project, managers need to determine which
situation applies and negotiate appropriately with the experts.
The second problem can be more diffcult to identify. Consider what happens
when workers rely on an expert system to make decisions, and management then
cuts costs by hiring less-skilled workers. The new workers do not understand the
system or the proceduresthey simply follow decisions made by the rules in the
ES. If an exception arises, the ES may not know how to respond or it may respond
inappropriately. A customer then would be left to deal with an underskilled worker
who does not understand the process and cannot resolve the problem.
Specialized Tools
Can machines be made even smarter? What technologies can
be used to help managers? Research in artifcial intelligence (AI) exam-
ined how humans are different from computers. This research led to tools that can
be used for certain types of problems. Some of the ideas come from the early days
of computers, but it has taken until now for machines to be developed that are fast
enough to handle the sophisticated tasks. Ideas in AI have come from many disci-
plines, from biology to psychology to computer science and engineering.
Humans are noticeably better than computers in six broad areas: pattern rec-
ognition, performing multiple tasks at one time, movement, speech recognition,
vision, and language comprehension. Some of these concepts are related, but they
all represent features that would make machines much more useful. Even with
current technological improvements, most observers agree that it will be several
years before these features are available.
Pattern Recognition and Neural Networks
One of the early issues in AI research was the question of how human brains
worked. Some people suggested that to make intelligent computers, the computers
would have to work the same way as the human brain does. An important conclu-
sion from this research is that humans are good at pattern recognition.
Humans use pattern recognition thousands of times a day. It enables people
to recognize coworkers, to spot trends in data, to relate todays problems to last
years changes. Many problems in business could beneft from machines that can
reliably recognize patterns. For example, what characteristics do good borrow-
ers have in common? How will changes in the economy affect next years sales?
How are sales affected by management styles of the sales managers?
Pattern recognition is used by people to solve problems. It is one of the reasons
teachers use cases to teach students to solve business problems. If you notice that
a problem is similar to a case you have seen before, you can use your prior knowl-
edge to solve the problem. Imagine how useful it would be if an expert system
could recognize patterns automatically.
One current technique that is used to spot patterns is the use of neural net-
works. Initial study indicated that the brain is a collection of cells called neurons
that have many connections to each other. Each of these cells is relatively simple,
but there are approximately 100 million of them. In some respects, a neuron re-
sembles a simple computer. It can be at rest (off), or it can fre a message (on). A
neuron responds to other cells (input) to send messages to other neurons (output).
A collection of these cells is called a neural network. Human neural cells are
actually more complicated, but researchers have focused on this simplifed form.
A common current example is a bank that uses a neural network to spot credit
card fraud. In some cases, Mellon Banks neural network identifed fraudulent
patterns even before the human investigators spotted them. It is faster and more
accurate than an earlier expert system. The original expert system looked at a lim-
ited number of variables and indicated 1,000 suspects a day, which was far more
than actually existed and too many for the investigators to keep up with. The new
neural network system examines more variables, lists fewer false suspects, and
adjusts its methods on its own.
A fnance manager might use a form of pattern recognition to search for pat-
terns in the fnancial markets to forecast future movements. Of course, with thou-
sands of other people searching for patterns, the patterns would not last very long.
Similarly, a banker might use pattern recognition to classify loan prospects.
Neural networks can be built with software. Also, computer chips are avail-
able today that function as neural networks. Neural networks can be measured in
two ways by (1) the number of neurons and (2) the number of interconnections
between the individual cells. It is fairly easy to increase the number of cells, but
the number of possible interconnections increases very rapidly. For instance, if
there are four cells, there are six possible connections. With 10 cells, there are 45
connections. With 1,000 cells, there are half a million connections. In general, if
there are N cells, there are N(N - 1)/2 possible connections. For many purposes,
not every connection is needed, but with millions of cells, a neural network would
incorporate a large number of connections. Most existing networks use only a few
thousand cells.
Figure 9.28 presents a version of how a neural network converts an array of
input sensors into a hidden layer and then stores patterns on an output layer. One
useful feature of the neural network approach is that it is fairly good at identifying
patterns even if some of the inputs are missing.
What can neural networks do that cannot be done with traditional computers?
The basic answer is nothing. However, they provide a new way of thinking
about problems. More important, with hardware specifcally designed to process
neural networks, some diffcult problems can be solved faster than with traditional
computers. The primary objective of neural networks is the ability to store and
recognize patterns. A well-designed network is capable of identifying patterns
(such as faces or sounds) even if some of the data is missing or altered.
Another advantage that researchers hope to achieve with neural networks is the
ability to simplify training of the computer. The discussion of expert systems not-
ed that changes in the business often mean that knowledge engineers have to rede-
sign the entire expert system. A neural network has a limited ability to learn by
examining past data. Feeding it proper examples establishes the interconnection
weights that enable the network to identify patterns. In theory, neural networks
Output Cells
Sensory Input Cells
Hidden Layer
Some of the connections
3
-2
7
4
Input weights
Incomplete
pattern/missing inputs.
6
Figure 9.28
Neural net for pattern matching. Input cells convert data to binary form. The required
hidden layer recodes the inputs into a new internal representation. The connections
represent outputs from the lower layers. When total input levels exceed some value,
the receiving cell fres. Any cell can be connected to many other cells. Input weights
are determined by training. The output cells are triggered when total input levels
from the connections exceed some threshold. Note that a pattern can be recognized
even if some input cells are wrong or not fring.
have the ability to learn on their own. In practice, the learning stage is the most
diffcult component of building a neural network. Most times the designer has to
understand the problem and provide hints to the network, along with good sample
data. In many ways, training a neural network uses basic properties of statistics
related to data sampling and regression.
Machine Vision
Machine vision has many uses in manufacturing environments. Machines are
used in optical character recognition, welding and assembly, and quality control.
Mechanical sensors have several advantages over humans. They do not suffer
from fatigue, they can examine a broader spectrum of light (including ultravio-
let and infrared), and they can quickly focus at many different levels (including
microscopic).
On the other hand, traditional computer systems are literal in their vision. It is
hard for computers to compare objects of different sizes or to match mirror im-
ages. It is hard for machines to determine whether differences between objects are
minor and should be ignored or if they are major distinguishing features.
Say you are shown a picture of your instructor, and someone adds or subtracts
features to it, such as bigger eyebrows, longer hair, or glasses. In most cases, you
would still recognize the face. Computers would have diffculty with that problem
because they see pictures as a collection of dots (or lines). How does the computer
know which changes are important and which are minor?
Machine vision systems are improving rapidly but still have a way to go to
become commonplace. For example, companies are working on applications in
facial recognition and facial expressions, body tracking (so you can use your hand
as a computer pointer), visual tracking of handwriting for use in computer tablets,
product inspections for defects, and shape identifcation.
Reality Bytes: Make the Trains Run on Time
Scheduling trains, planes, or healthcare appointments might appear to be an easy
problem. Just fgure out the capacity, compute the time needed to get from point A to
point B, and then route everything. But, the problem is much more complex. To start
with, a complete routing scheme is a diffcult problem even in computer science,
with no simple solution. But, it gets considerably worse when things go wrong
such as a train delayed by weather. A slowdown at one point can interfere with many
other schedules. However, a university research project called Arrival resulted in
new scheduling software that effciently routes trains, even when disruptions hap-
pen. The system can handle almost instantaneous rescheduling as disruptions arise.
It adjusts timetables, platform allocations, and staff scheduling. The system has been
tested in the Netherlands, Germany, and Switzerland. In Berlin, the average wait
time was cut in half from four minutes to two. In a test in Italy, the system created a
25 percent reduction in train delays in Palermo and Genoa.
Adapted from Daily Mail Reporter, Brussels: We Can Make the Trains Run on
Time, July 16, 2010.
Language Comprehension and Translation
Related to voice recognition is the issue of language comprehension, or the ability
of the computer to actually understand what we are saying. Technically the two
topics are separate, since it might be possible to have a machine understand what
we type onto a keyboard. Language comprehension exists when the machine actu-
ally understands what we mean. One test of comprehension would be the ability
of the computer to carry on a conversation. In fact, Alan Turing, a British pioneer
in the computer feld, suggested the Turing test for computer intelligence. In this
test, a human judge communicates with a machine and another person in a sepa-
rate room. If the judge cannot determine which user is the machine and which is
a person, the machine should be considered to be intelligent. Some people have
tested this concept (using specifc topics). Other people have noted that perhaps
you do not have to be intelligent to carry on a conversation.
The Loebner Prize is a Turing test that runs every year to evaluate how close
programs are coming to meeting the challenge. Several of the challengers have
set up Web sites (e.g., www.alicebot.org, and www.jabberwacky.com) so you
can play with these conversational bots. Why do you care? First, you can buy
this technology to use it in business applicationssuch as building Web sites that
answer basic questions from customers. Second, as a customer, you might want to
learn to recognize when you are dealing with a machine instead of a human.
Language comprehension would be useful because it would make it easier for
humans to use computers. Instead of needing to learn a language such as SQL
to access data, imagine being able to get answers to questions asked in English
(or some other natural language). Of course, any natural language has its lim-
itations. The greatest danger with language comprehension is that the machine
will interpret your question incorrectly and give you the right answer to the
wrong question. Figure 9.29 provides a simple illustration of the complexities
of language comprehension. The frst example involves the use of punctuation. A
misinterpretation of the command can result in deleting the wrong fle. Similarly,
See what happens when you give a computer the frst set of instructions, but
it does not hear the commas correctly and thinks you said the second line:
(1) Copy the red, fle the blue, delete the yellow mark.
(2) Copy the red fle, the blue delete, the yellow mark.
Consider the following sentence, which can be interpreted by humans, but
would not make much sense to a computer that tries to interpret it literally.
I saw the Grand Canyon fying to New York.
Or the phrase epitomized in a grammar book:
The panda enters a bar, eats, shoots, and leaves.
Figure 9.29
There are inherent problems with voice recognition. Punctuation and implicit
meaning are two diffcult areas. Even communication between people has frequent
misinterpretations.
interpretation of a natural language involves understanding some basic concepts,
such as the fact that the Grand Canyon cannot fy.
Ultimately, translating languages requires an understanding of the underlying
meaning of a sentence and paragraph. Early translation systems can convert a
word from one language into another. For example, check out babel.altavista.com,
or any of the other online translators. Some high-end translation systems have
been developed that do a better job by at least recognizing common phrases and
idioms. Yet even these systems need to be supplemented by human translators.
IBM raised the bar for natural language comprehension in 2011. The company
created the Watson system designed to understand language, using statistics to
help it understand questions and provide answers. As shown in Figure 9.30, to
show off the system, it played several games of Jeopardy against two of the best
human opponents. The Watson system won easily. The use of statistical analysis
is still debated in computing circles. In language, it provides the beneft of letting
the system decide not just an answer but how confdent it is in the answer. Just as
in real life, if a statement is not clearly understood, the computer can ask for clari-
fcation. IBM has suggested that the system can applied to many types of business
and research problems, including medical research.
Robotics and Motion
Modern manufacturing relies heavily on robots, and the capabilities of robots con-
tinually increase. Most existing robots are specialized machines that perform a
limited number of tasks, such as welding or painting. In many frms, there is little
need for a general-purpose robot that can do everything. However, one area that
Figure 9.30
A practice Jeopardy match with IBMs Watson and two of the best human players. In
three days of televised matches, Watson handily beat the humans.
Source: http://www.youtube.com/watch?v=12rNbGf2Wwo
remains troublesome is the ability of machines to move. Making a machine that
can navigate through an unknown or crowded space is especially diffcult. Some
work is being done in this area. Liability is a major problem when robots attempt
to move among people.
In 2000, Honda built a humanoid robot (Asimo) in Japan. The Asimo robot has
two legs and arms. Its most impressive feature is the ability to walk like a human,
including up and down stairs. It can also shake hands and hand objects to people.
The multi-million-dollar project is the latest step of a 16-year evolution.
In March 2004, DARPA held the frst Grand Challenge contest for automated
vehicles in the Mojave Desert. The challenge was to create an automated vehicle
that could drive itself across 189 miles of desert hitting about 1,000 GPS way-
points in about 10 hours. The Carnegie Mellon $3 million Red Team vehicle (a
modifed Hummer) made it 7.4 miles. In fall 2005, the contest was run a second
timeand fve vehicles fnished the 132-mile course. A huge gain in a year and a
half, and the winning Stanford vehicle was created in less than a year! DARPA an-
nounced a new contest for November 2007, where the vehicles will have to drive
through city traffc. Congress has mandated that 30 percent of Army vehicles be
automated by 2015, and DARPA is attempting to stimulate innovation with these
prizes.
As computer processors decline in size and price, it becomes easier to build
intelligent mobile systems, making it possible to build robots and automated ve-
hicles. Major automobile manufacturers are working on slightly less automated
systems to provide assistance to drivers. For example, crash-warning sensors can
automatically apply brakes for an inattentive driver. Slide-control systems to pre-
vent spinouts on sharp turns have been installed in luxury vehicles for several
years.
Reality Bytes: Dealer Services Corp
Dealer Services Corp. provides fnancing to about 10,000 car dealers. The company
essentially writes separate loans for every vehicle, providing individual conditions
and payback schedules. The system generated huge amounts of data. Technically-
oriented analysts within the company developed their own spreadsheets to analyze
data, but they often had inconsistent or bad data. The company purchased WebFocus
as a tool to provide centralized access to data and business intelligence. The system
largely focused on interactive reporting and charting. When the recession hit in 2008
and 2009, the company managers were able to see results in individual markets and
by dealers to see that inventory was not selling. CIO Chris Brady noted that these
key insights warned the company early that the recession was going to cause severe
problems. So the company tightened lending standards and increased fnancial re-
serves. It was also able to provide advice to dealers, such as stop buying SUVs, they
arent selling. Brady also noted that We defnitely reduced our losses from bad
loans and didnt start to see a negative effect until the very end of 2008.
Adapted from Elizabeth Horwitt, Self-Service BI Catches On, Computerworld,
December 13, 2010.
Machine Intelligence
What would it take to convince you that a machine is intelli-
gent? The Turing test has been proposed as one method. Many other tests have
been proposed in the past. At one time, people suggested that a machine that could
win at chess would be intelligent. Todays chess-playing computers have beaten
even the top human players. Another test proposed was the ability to solve math-
ematical problems, in particular, the ability to write mathematical proofs. An early
AI program created in the 1950s could do that. Today, for a few hundred dollars,
you can buy programs or even small calculators that manipulate mathematical
symbols to solve equations.
Some people have suggested that intelligence involves creativity. Creativity is
probably as hard to measure as intelligence. Even so, examples of computer cre-
ativity abound. A few years ago, a programmer developed a system that created
music. The interesting feature of the program was that it allowed people to call
on the phone and vote on the music. The computer used this feedback to change
its next composition. Not only was the computer creative, but it was learning and
adapting, albeit in a limited context. Today, you can buy software that creates mu-
sic or plays background to your solo.
Although business applications to much of this current research is somewhat
limited, there are two main reasons for staying abreast of the capabilities. First,
anything that makes the computer easier to use will make it more useful, and these
techniques continue to improve. Second, you need to understand the current limi-
tations to avoid costly mistakes.
DSS, ES, and AI
What are the differences between DSS, ES, and AI systems? The
differences among decision support systems, expert systems, and artifcial intel-
ligence can be confusing at frst. Take a simple problem and see how a computer
system based on each method might operate. A common fnancial problem is to
determine how much money to lend to customers. Any frm that grants terms to
customersnot just fnancial institutionsmust make this decision. Figure 9.31
discusses the differences among a DSS, ES, and AI approach to the inventory
problem.
Reality Bytes: Guessing for Holiday Sales
Holiday shopping is critical for many U.S. businesses. At the high end, hobby, toy
and game stores make 34 percent of their sales in November and December. Even
for other retailers, the season is critical. Almost 20 percent of all retail sales value
arises in the last two months of the year. Department stores hit 25 percent. The one
sector that is lower than then one-sixth time frame is car dealers. The problem is that
retailers need to order products and have them on the shelves well before the season
begins. Consequently, retailers need good forecasts of both the expected sales level
and details on which products are going to be hot.
Adapted from Phil Izzo, Number of the Week: Outsize Importance of Holidays for
Retailers, The Wall Street Journal, December 24, 2010.
In a relatively simple system, the computer would retrieve data about the cus-
tomer and the prior loans to that customer. Historically, loan offcers used basic
data and personal factors to make the lending decision. In some instances, these
rules of thumb led to problemswith bad decisions and sometimes discrimina-
tion. The DSS could also be used to monitor existing loans and payments. As part
of a transaction-processing system, it can notify managers when customers con-
tinually make late payments and help identify problem loans.
To improve consistency and reduce the decision time, many frms have moved
to expert systems to help evaluate loans. Statistical analysis of prior loans is used
to establish a set of rules that are coded into the ES. In some cases, the ES can
then be operated with push-button phones or over the Internet. In straightforward
cases, the ES can make the fnal decision and approve the loan. In more diffcult
situations, the preliminary results and data can be forwarded to a human loan of-
fcer to factor in personal judgment and factors not considered by the ES.
Of course, the value of the ES depends heavily on the accuracy of the under-
lying rules (and the supplied data). These rules might change over time or as
economic conditions change. A neural network can be used to examine the prior
loans automatically to identify the factors that predict successful and unsuccessful
loans. Once these factors are identifed, they can be coded into the ES to automate
the decision process. In this situation, the AI/neural network takes the place of (or
supplements) the decisions of the human expert.
Decision Support System Expert System AI/Nerual Network
Loan Ofcer
Data:
Income
Existing loans
Credit report
Model:
Lend in all but the worst
cases.Monitorforlateand
missingpayments.
Output:
Name Loan #Late Amt
Brown 25,000 5 1,250
Jones 62,500 1 135
Smith 83,000 3 2,435
ES Rules
What is the monthly
income? 3,000
What are the total monthly
payments on other loans?
450
How long have they had
the current job? 5 years

Should grant the loan since
there is a 5% chance of
default.
Determine Rules
Data/Training Cases
Loan 1 datapaid
Loan 2 data 5 late
Loan 3 data bankrupt
Loan 4 data 1 late
Neural Network Weights
Evaluate new data, make
recommendations.
Figure 9.31
Comparison of techniques for a loan. A DSS can display background data for a loan
offcer and can also monitor customer payments. An ES could help managers decide
if they should make the loan by evaluating more complex rules. An AI such as a
neural network can analyze past loans and determine the rules that should be used to
grant or deny future loans.
Te Importance of Intelligent Systems in e-Business
How can more intelligent systems beneft e-business? Disinterme-
diation is a primary aspect of e-business. Businesses can interact directly with
customers, with less need for middle levels such as retail stores. However, these
middle levels often existed because they provided more explanations and support
to customers. If you remove that level, how are you going to deal with thousands
or millions of customers? If you have to hire hundreds of workers to answer cus-
tomer questions, you will lose most of the potential benefts of disintermediation.
One of the solutions to this problem is to implement more intelligent systems that
can provide automated support to customers.
In many ways, the Internet adds complexity to the daily lives of customers and
managers. The Internet provides access to huge amounts of dataand it is grow-
ing constantly. The growth adds more data, but it also means that the availability
and use of information is constantly changing
Agents
A recent application of AI techniques has arisen in the context of the Internet. A
key issue of the Internet is searching for data. Although the Internet dramatically
improves communication, there are problems with maintaining the interpreta-
tion of the information from various systems. Originally, most data on the Web
was stored as standard pages of text using HTML. Search engines would simply
scan these pages and build searchable indexes.
Increasingly the Internet is being used to store and transmit objects composed
of data, pictures, spreadsheets, sounds, and video. From a pure transmission
standpoint, any object can be decomposed into raw data bits and sent between
computers. Where we run into problems is searching for the objects. Consider a
simple example where you want to fnd a new printer, so you search the Internet
for prices. Today, many vendors store the product descriptions and prices in a da-
tabase, and then build the HTML page on demand when you go to the site. Since
the page is not static, the search engines do not index it.
One solution to this problem is to create software agents. Agents are object-
oriented programs designed for networks that are written to perform specifc tasks
in response to user requests. The concept of object orientation is important be-
cause it means that agents know how to exchange object attributes, and they have
the ability to activate object functions in other agents. The tasks could be simple,
such as fnding all fles on a network that refer to a specifc topic. One key feature
of agents is that they are designed to communicate with each other. As long as
your agent knows the abilities or functions of another agent, they can exchange
messages and commands. General Magic was a pioneering company that cre-
ated a standard programming language for agents. With this language, agents can
transfer themselves and run on other computers. Agents also have a degree of
intelligence. They can be given relatively general commands, which the agents
reinterpret and apply to each situation they encounter.
Consider an example illustrated by Figure 9.32. You have been working hard
and decide to take a vacation. You want to go to a beach but do not have much
money to spend. You are looking for a place where you can swim, scuba dive, and
meet people at night. But you also want the place to have some secluded beaches
where you can get away from the crowds and relax. You could call a travel agent
and buy a package deal, but every agent you call just laughs and says that next
time you should call three months ahead of time instead of only three days ahead.
You suspect that a beach resort probably has last-minute cancellations and you
could get in, but how do you fnd out? Thousands of possibilities exit. If all of the
resort computers had automatic reservation agents, the task would be fairly easy.
You would start an agent on your computer and tell it the features you want. Your
agent sends messages to all of the automated resort agents looking for open spots
at places that matched your features. When your agent fnds something close, it
brings back details and pictures to display on your screen. When you decide on a
resort, the agent automatically makes the reservations.
You might argue that some of the existing travel search sites come close to this
example. On the surface, yes, sites such as PriceLine offer a few of the elements.
However, the technology and methods beneath the surface are completely differ-
ent and largely rely on humans to set prices. The travel industry is likely to be
one of the frst to incorporate these technologieslargely because many of the
reservation systems are already linked together at some level. At some point, the
negotiating agent needs access to the data. Today, these decisions are made at cen-
tral search engine sites that have access to the data on millions of sites. In a true
agent-based system, the search engine would not need direct access to the data,
but would communicate with the agents located at each site.
Notice three important features of software agents. First, the agents need to
know how to communicate. It is not as simple as transmitting raw data. They must
understand the data and respond to questions. Second, imagine the amount of net-
work traffc involved. In the vacation search example, your agent might have to
Vacation
Resorts
Software agent
Resort
Databases
Locate &
book trip.
Figure 9.32
Software agents. A personal software agent might be used to book a vacation. It
would take your initial preferences and communicate with other agents to fnd
sites that matched your preferences. It might also be able to negotiate prices with
competing resorts.
contact thousands of other computers. Now picture what happens when a thou-
sand other people do the same thing! Third, all of the agents are independent. You,
as well as other computer owners, are free to create or modify your own agent. As
long as there are standard methods for agents to exchange attributes and activate
functions, they can be modifed and improved. For instance, you might program
your agent to weight the vacation spots according to some system, or you might
teach it to begin its search in specifc locations.
Programmers have begun to incorporate expert system and other AI capabili-
ties into these agents. By adding a set of rules, the agent becomes more than just
a simple search mechanism. The more complex the rules, the more intelligent it
becomes, which means you have to do less work. In fact, software agents have the
potential to dramatically increase the research in AI. Currently, because of limited
standards and the diffculty of creating them, there are few examples of useful
agents. As increasing numbers of people use agents and begin demanding more
intelligence, it will become proftable for researchers to work harder at building
reliable, intelligent software.
Support and Problem-Solving Applications
Increasingly, your customers want personalized attention to help in both selecting
products and solving problems. Yet it is expensive to provide individual personal
support to every customer. Instead, frms are developing expert systems and other
intelligent applications to help customers with a more personalized touch. For ex-
ample, look at Amazon.coms recommendation system. It began with books but
has been expanded to most of their products. As you purchase items at Amazon.
com, the system gives you a list of similar products that you might be interested
in. For instance, if you purchase several science fction books, it will suggest new
releases of similar books. The system can increase sales because it helps show
customers items that they might not have found otherwise.
More complex products can beneft from more sophisticated expert systems
that help analyze customer needs and help confgure the correct components. For
example, a computer vendor could build a system that asks questions to help iden-
tify the applications that a customer will run. It could then suggest specifc en-
hancements such as adding RAM or a second disk drive to improve performance.
Similarly, many frms are building expert systems to help customers with prob-
lems. If a customer has a problem installing a new product, he or she can turn to
the Web site. The system asks questions to identify the problem and then make
suggestions. The advantage of the expert system is that it is available 24 hours a
day, can solve most of the easy problems, and is less embarrassing to customers
who might think their questions are too silly to ask a human troubleshooter.
Intelligent systems can also be useful for B2B and other forms of e-business.
The systems might analyze past purchases and suggest new products, or automati-
cally analyze sales patterns and help managers develop new products and close
out unproftable lines. They can be used to develop automated ordering systems
that predict customer demands, schedule production, and generate automated
sales orders and payments.
It can be diffcult to develop these applications, but frms that build powerful
systems will attract customers and increase the level of sales to each customer. Ul-
timately, these systems could be the primary reason people switch to buying items
over the Internet.
Cloud Computing
How can cloud computing be used to analyze data? Cloud comput-
ing offers some options for analytical toolsbut the process is only beginning.
Some decision tools require relatively large amounts of computation and storage
capabilities. In some cases, these tools could be run on large parallel processing
supercomputers. Some research institutions provide online access to supercom-
putersbut they are traditionally used for scientifc research. For more typical
business research, it should be possible to use multiple processor systems leased
from cloud computing providers. Companies such as Amazon provide access
to multiple processors based solely on the amount of time used. Consequently,
researchers can develop algorithms, test procedures, and fne-tune the analysis.
Then, when the main data sets are loaded, additional processors can be leased for
a short time to process the main results. With minimal fxed costs, cloud comput-
ing makes it possible to tackle large problems at relatively low costs.
It is likely that companies will offer data analysis as package services through
online systems. Currently, several market research frms will conduct studies, for-
mat data, and run analyses for other companies. As more data becomes acces-
sible through online systems, it will become possible for cloud-based companies
to offer direct access to analytical services. For instance, if you want to offer ser-
vices similar to Amazon or Netfix to show customers similar items purchased
by customers, a company could automatically read your sales data, perform the
forecasts, and display the results using links across the Web. You would not have
to develop the algorithms or purchase high-speed computers to handle the com-
putations. On the other hand, you might have less control over customization or
modifcation of the algorithms. Leased systems might not provide a competitive
advantage, but they might make it easy for your company to add desirable features
to your Web site at relatively low cost.
Summary
Managers make many different decisions. Every business discipline builds models
to help people analyze problems and make decisions. Without models and tools,
people rarely make good decisions. Many tools have been created to help you ana-
lyze data and make decisions. In many cases, you will want to build a data ware-
house to retrieve and organize the data. You can build a DSS, building a model in
a spreadsheet or more advanced statistical analysis tool. A good DSS extracts the
needed data from the database, evaluates the model, and displays the results in a
form that helps managers visualize the problem and quickly choose a solution.
The biggest diffculty with DSS tools is that you generally need to be trained to
create the models and understand the results.
More sophisticated data mining tools can be employed to semiautomatically
search for correlations and other relationships within the data. Common regres-
sion, clustering, and classifcation tools use statistical measures to identify the
importance of various attributes and to build forecast equations. Market basket
analysis identifes items that customers purchase together. The information pro-
vided by this analysis is a powerful tool for cross selling related products. Data
mining tools are powerful techniques, but the results are not always useful, and
you must carefully evaluate the implications of the results to ensure that they are
realistic and repeatable.
Expert systems provide a different level of decision automation. They analyze
data based on rules defned by an expert. They can handle complex and missing
data. They are particularly useful at helping novices reach a better decision. They
are excellent tools for specialized, narrowly defned problems. The biggest con-
cern arises when someone tries to build an ES for a problem that is too large, too
variable, or unstructured. The ES is not likely to work, and it will cost money and
time to build. Worse, the proposed decisions might be nonsensical.
Scientists are continually working on ways to automate even more decisions.
Neural networks are incredible tools for analyzing pattern data. They have been
successfully applied to problems ranging from speech recognition to lending anal-
ysis. You can purchase software that will quickly build a neural network to analyze
data. One drawback to neural networks is that the relationships tend to be highly
nonlinear and diffcult to interpret. So, you might reach an answer, but you might
not be able to explain the relationships. Other tools that are increasingly useful in
business applications are machine vision, robotics, and language comprehension.
Automated tools can help you provide better service to customers and suppli-
ers, particularly when the intelligent systems can be reached online. If you can
help a customer solve a problem 24 hours a day without having to pay hundreds of
humans, you can gain happier customers with little additional cost. Several com-
panies are working on building intelligent agents that will enable customers and
vendors to interact automatically by following rules that you specify.
A Managers View
It is hard to make good decisions. You need fast access to huge amounts
of data, the ability to evaluate various models, and a way to visualize the
problem and the solution. Various levels of tools are available to help. The
tools provide different types of intelligence and support. As a manager, you
need to understand the context of the problem and know which tools can be
applied to solve a problem. You also need to be enough of an expert to rec-
ognize when a system provides useless or bad answers so that you can avoid
disasters.
Key Words
Web Site References
Expert Systems and AI Tools
A.L.I.C.E.conversation Alice.pandorabots.com
CLIPS (Started by NASA) clipsrules.sourceforge.net
ExSys (Commercial) www.exsys.com
International NeuralNetwork Society www.inns.org
Jess ( Java) www.jessrules.com
Mathworks neural network toolbox www.mathworks.com/products/
neuralnet
Machine Vision
CalTech www.vision.caltech.edu
Carnegie Red Team www.cs.cmu.edu/~red/Red
British Machine Vision www.bmva.org
Review Questions
1. Why is it so important to build models and analyze data using a scientifc
process?
2. What is the purpose of a data warehouse?
3. How is an OLAP cube browser better than using queries?
4. What is the role of a manager in a DSS?
5. How is a GIS used to answer business questions?
6. How can data mining help you make better decisions and what are some of
the primary techniques?
7. What is an expert system and what are the characteristics of the problems
that it is designed to solve?
8. What tools are available to help businesses construct expert systems and
what features do they provide?
agent
artifcial intelligence (AI)
data marts
data mining
data warehouse
decision support system (DSS)
decision tree
expert system
extraction, transforma-
tion, and loading (ETL)
geographic information system (GIS)
global positioning system (GPS)
knowledge base
knowledge engineer
market basket analysis
metadata
models
natural language
neural network
online analytical processing (OLAP)
optimization
parameters
prediction
rules
simulation
Turing test
9. What types of problems are best suited for a neural network?
10. How would a fully autonomous vehicle alter business?
11. What are the differences between a DSS, ES, and AI features?
12. Why are intelligent systems so important for e-business?
13. How could cloud-based computing and the Internet be used to add
intelligence to business applications and decisions?
Exercises
1. Work through at least two of the examples on the Exsys Web site. What
features do the two examples have in common? Give an example (unrelated
to any on the Web site) of a business problem or something from your life
that could beneft from an expert system.
2. An HR manager wants to develop an expert system to evaluate potential
employees applying to work for a job in your department. Assuming the job
can be performed by a business-school intern, list some of the questions you
would ask potential employees. Create a decision tree to evaluate the basic
questions. Try to generate three possible outcomes: acceptable, unacceptable,
and personal interview to decide. Your goal is to reduce the number of people
needing a personal interview.
3. Interview an expert in some area and create an initial set of rules that you
could use for an expert system. If you cannot fnd a cooperative expert, try
researching one of the following topics in your library: fruit tree propagation
and pruning (what trees are needed for cross-pollination, what varieties
grow best in each region, what fertilizers are needed, when they should be
pruned); requirements or qualifcations for public assistance or some other
governmental program (check government documents); legal requirements to
determine whether a contract is in effect (check books on business law).
4. Describe how you could use data mining tools to help you fnd a new vehicle
to buy. How well do the car-buying sites such as Edmunds perform these
tasks for you?
5. Obtain an expert system (e.g., Jess and CLIPS are free). Create a set of rules
to evaluate a simple request for a car loan. If you do not have access to the
tool, at least build the decision tree.
6. Identify a problem that would be well suited for a neural network. Explain
how the system would be trained (e.g., what existing data can be used?).
Explain why you think the problem needs a neural network and what benefts
can be gained.
7. For the following problems identify those that would be best suited for an
expert system, decision support system, or a more advanced AI system.
Explain why.
a. A fnancial manager in a bank wants to fnd the best investments for her
clients.
b. A venture capital investor wants a system to do a frst-pass evaluation of
new business plans that are submitted via a Web site.
c. A plant manager wants to be able to forecast when equipment will need
special maintenance. Currently, the engineers use a fxed schedule, but
some products and operations seem to cause more problems than others.
d. A fast-food restaurant chain wants to help managers determine the
appropriate number of employees to schedule for each day in the year.
e. A cell-phone provider wants an application on its Web site to help
customers select a cell phone that best meets their needs.
f. A company has to select a city to host a new factory which happens about
once a decade.
g. A marketing department wants an application to help them set the best
prices for new products in different markets.
8. Examine existing travel sites, including aggregators such as Kayak.com.
What features could be added to increase the intelligence level to assist
consumers?
9. Use a spreadsheet to create the example from the Human Resources
Management example. Fill in the market adjustment column so that raises
match the performance appraisals. Remember, total raises cannot exceed
$10,000.
DeptStor.mdb
You are a midlevel manager for
a small department store. You have
collected a large amount of data on
sales for 2012. Your transaction sys-
tem kept track of every sale (order)
by customer. Most customers paid
by credit card or check, so you have
complete customer data. Walk-in
customers who paid cash are given
a separate customer number, so you
still have the sales data.
You are trying to determine staff-
ing levels for each department. You
know that the store becomes much
busier during the end-of-the-year
holiday season. For summer months,
you have thought about combining staff from the departments. From conversa-
tions with experienced workers, you have determined that there is a maximum
number of customers that can be handled by one person in a department. These
numbers are expressed as monthly averages in the table.
You are thinking about combining workers from some of the departments to
save on staffngespecially over the spring and summer months. However, work-
ing multiple departments makes the sales staff less effcient. There are two consid-
erations in combining staff members. First, if any of the departments are reduced
to a staff of zero, sales in that department will drop by 10 percent for that month.
Second, total staffng should be kept at the level defned by the monthly averages.
If average staffng (total across all departments) falls below the total suggested,
then sales in all departments will fall by 2 percent for each tenth of a percentage
point below the suggested average.
Department Customers/month
ClothingChildren 180
ClothingMen 150
ClothingWomen 180
Electronics 200
Furniture 150
Household 250
Linen 300
Shoes 300
Sports 400
Tools 340
10. Using the database and a spreadsheet, determine how many workers we
need in each department for each month. Present a plan for combining
departments if it can save the company money. Assume that sales members
cost an average of $1,000 a month. Two queries have already been created
by the MIS department and are stored in the database: SalesbyMonth and
SalesCountbyMonth. The frst totals the dollar value; the second counts the
number of transactions.
11. Write a report to upper management designating the appropriate sales staff
levels for each department by month. Include data and graphs to support your
position. (Hint: Use a spreadsheet that lets you enter various staffng levels in
each department in each month, and then calculate any sales declines.)
12. Create a PivotTable for the company that enables managers to evaluate sales
by employee by department by month.
Technology Toolbox
13. Create the PivotTable report for Rolling Thunder Bicycles. Briefy
summarize any patterns or problems you identify.
14. Using the Rolling Thunder Bicycles query, create a PivotChart and compare
sales of the different models over time. Identify any patterns that you see.
15. Research an alternate cube browser (such as SQL Server or Oracle) and
explain how it is different from the Excel PivotTable. If you have access to
the tool, build a small example.
16. Compute the average number of days it takes to build a bicycle (ShipDate
OrderDate) for each month. Import the data into Excel and forecast the
trend. First, forecast it based on all of the data. Second, forecast it for three
time periods: (a) the early years, (b) the middle years, and (c) the most recent
years. Look at the initial chart to estimate the breaks between these three sets,
or just divide it into three equal-size groups if you do not see any good break
points. Comment on any differences or problems.
17. Using federal data (start at www.fedstats.gov), compute a regression
analysis of Rolling Thunder sales by state by year compared with at least
population and income.
18. Choose a company and identify the primary dimensions that you would
create in an OLAP cube if you were a manager at that company. Hint: Focus
on your major area.
Teamwork
19. Have each person fnd and describe a problem that could beneft from a GIS.
Make sure it needs a GIS, not just a mapping system. Combine the results
and compare the types of problems to identify similarities.
20. Build a decision tree to test whether a specifc student next year can take this
class.
21. Many classes fll up early and the dean keeps a wait list for adding students
at the last minute. Build a decision tree to evaluate each student on the list to
determine priority for adding that specifc person to the class.
22. Each person should identify a problem that could be solved with a rules-
based expert system. Combine the results and compare the types of problems
to identify similarities.
23. Using Rolling Thunder Bicycles, have each person forecast the sales by
one model type for six months. Combine the individual model results and
compare this value to the forecast based only on total sales.
24. Select an economic data series such as personal income (check www.
fedstats.gov). Place members into one of three subgroups. Have each group
forecast the series using a different methodology. Compare the results. If you
have suffcient data, leave out the most recent data, and then forecast those
values and compare the forecasts to the actual.
25. Choose a publicly-traded company and collect basic quarterly fnancial data
for the company for at least 10 years. It is easiest if you assign specifc years
to each person. Put the data into a spreadsheet or simple database. Create a
PivotTable and Pivot Chart to examine the data. Create a report containing
some of the charts and tables and any conclusions you can make.
26. Have each person fnd a problem that could beneft from a neural network.
Describe how the system would be trained. Combine the individual
comments and identify any commonalities.

27. Identify shipments where receipts do not match the original order. Provide a
count and value (and percentages) by supplier/manufacturer.
28. Analyze sales and discounts by employee and by model type. Are some
employees providing higher discounts than others? Are we discounting some
models too much or not enough?
29. The company wants to create an online ordering system. Create a decision
tree to help novices select the appropriate bicycle and components. If
necessary, consult with a friend or relative who can be considered a bicycle
expert.
30. Use queries to extract sales data by model type and month. Use a spreadsheet
to forecast the sales of each quantity of model type by week for the next year.
Hint: Use Format([OrderDate],yyyy-mm) to get the month.
31. What pattern-matching types of decisions arise at Rolling Thunder that could
beneft from the use of neural networks?
32. What aspects of customer service might be automated with expert systems?
What are the potential advantages and disadvantages?
Additional Reading
Barabba, Vincent and Gerald Zaltman, Hearing the Voice of the Market, Harvard
Business Press: Cambridge, MA, 1991. [Overcoming design biases at GM.]
Franses , Philip Hans, Time Series Models for Business and Economic
Forecasting, 1998, Cambridge University Press: Port Chester, NY. [An
introduction to forecasting.]
Dangermond, Jack and Adena Schutzberg, Engineering, geographic information
systems, and databases: a new frontier, Journal of Computing in Civil
Engineering, July 1998 pp. 121-122. [Uses of GIS.]
Krill, Paul, Analytics redraw CRM lines, Computerworld, December 3, 2001.
[A summary of vendors developing products to analyze customer data.]
Neil, Stephanie, Blue Cross dissects data to improve care, PC Week, February
8, 1999. [BCBS uses data warehouse and online application processing DSS
to improve service.]
Wilkinson, Stephanie, PC apps help to take a byte out of crime, PC Week,
February 23, 1998. [GIS and imaging tools help police predict and solve
crimes.]
Cases: Financial Services Industry
Te Industry
The fnancial industry is interesting. It is huge and everyone interacts with the tril-
lions of dollars a day handled by the global fnancial system. It is also complex,
with many different types of frms involved, and constantly creating new fnancial
tools. At one point, the fnancial industry was driven by banks. Today, banks still
play important roles, but where do people put their money? In the stock market
or other investments. So brokerage frms have a strong role. How do consumers
pay for things? With debit cards and credit cards. So the card-processing sector,
led by Visa and MasterCard, plays a huge role in the industry. As the national laws
have changed, mostly relaxed, over the past few years, the industry has become
even more complex. Banks can once again sell stocks and investments, as well as
insurance. And brokerage frms and insurance companies can perform banking
functions.
0
20
40
60
80
100
120
140
160
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
B
i
l
l
i
o
n

$
Annual Revenue
Citigroup
Bank America
Capital One
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
R
a
t
i
o
Citigroup
Bank America
Capital One
Making Money
How do brokerage frms make money? They used to make a proft on transaction
commissions, as much as several percentage points per tradeamounting to hun-
dreds or thousands of dollars every time a client bought or sold a stock. With the
advent of online trading, pushing the discount brokerage frms, commissions fell
to fxed rates of $10 to $20 per trade, a tiny fraction of what they were and under
10 cents a share. The big frms tried to convince customers that their higher fees
were worthwhile because they also provided investment advice. But with tainted
advice, and a huge amount of data available free to customers online, the full-
price frms suffered.
How do banks make money? Banks make money in two ways: interest rate
spreads and fees. Interest rate spreads are the difference between the rates the
banks charge borrowers and the amount the bank has to pay to obtain the money.
Banks select different lending markets (consumer versus business) and have di-
verse ways of obtaining funds (deposits, loans from other banks or the Federal
Reserve). Likewise, fees vary depending on the type of bank. Consumer-oriented
banks receive fees from customers (such as checking account fees). Business or
investment banks charge fees for more sophisticated services.
How do credit card companies make money? Here the answer is trickier, be-
cause several types of companies are involved. The transaction-processing organi-
zations (Visa and MasterCard) receive a fee for every transaction handled through
the clearing system. However, banks ultimately issue the cards and are responsible
for the money. A consumer bank issues a card to customers and is responsible for
paying vendors for legitimate transactions. The consumer bank makes money on
consumer fees (usually low today) and interest charges to customers on outstand-
ing balances. It also uses the cards as a marketing tool to attract customers and
encourage them to purchase other services. The merchant bank is responsible for
ensuring that the merchant is legitimate. It makes money by taking a percentage
cut of the merchants credit card (or debit card) sales. In 2011, the U.S. govern-
ment joined Europe and imposed caps on merchant fees for debit cards (not credit
cards). So banks began cancelling debit card promotional programs and scram-
bled to fnd other ways to charge fees.
Brokerage Firms
More than three-fourths of Americans liquid assets ($12 trillion) are in stocks,
bonds, and money market funds [Revell 2002). How safe are these assets? Do you
trust your broker? Do you trust the advice from brokers? In April 2003, forced by
a lawsuit led by New York attorney general Eliot Spitzer, 10 of the largest Wall
Street investment banks paid fnes of $1.4 billion (Nocera 2004). The money was
a settlement for providing misleading investment analysis. The brokerage frms
usually wrote glowing reports of companiesurging people to buy stock in spe-
cifc companies. Many of the reports were written because the frms were vying
for investment banker contracts with the same companies. In exchange for glow-
ing reports, the companies provided more business (and fees) to the brokerage
frms. James Freeman, a former research director, notes that analysts went from
having investment banking deals pushed on them in the 1980s to becoming the
greedy pigs at the center of it (Nocera 2004).
Credit Cards
At one time, the credit card industry was neatly divided into three main segments:
the upper end controlled by American Express, the middle tier targeted by Citi-
corp, and the subprime market pursued by Providian. As a fourth group, MBNA
targeted a wide swath of customers with its affliate cards. Most card companies
gained market share by blindly sending 5 billion solicitations a year to U.S. con-
sumers (Gross 2002). Capital One grew by using targeted marketing and thou-
sands of marketing tests with different groups of potential customers. Nonethe-
less, with competition at a peak for high-end customers, profts were squeezed.
Companies were increasingly tempted to go after the subprime marketwhere
they could charge higher interest rates in exchange for greater risk. But as the
economy turned down and interest rates increased, this strategy results in greater
losses when customers do not pay their bills or even fle for bankruptcy (Dugas
2002). Providian lost 90 percent of its stock market value from mid-2001 to mid-
2002. Capital One also faced greater losses, but not as large as many other compa-
nies. The 2003 settlement with Visa and MasterCard ultimately affected profts for
the industry. Wal-Mart had sued the two credit card companies to break up their
power to force merchants to accept higher costs for debit cards. The win by the
merchants reduces the power of the card companies and banks, and should ulti-
mately reduce fees paid by merchants (www.cardweb.com).
Banks
Citibank (or the parent company Citigroup) is one of the largest banks in the
world. In 2003, it earned $17.9 billion in profts, the most ever by a single bank
(Economist 2004). In 2010, net income was $10.6 billion (2010 Annual Report).
Over the years, it has suffered several problems along with other banks, including
from the huge losses on international loans to South America in the 1980s to the
investment banking scandals in the 1990s. The company agreed to pay $400 mil-
lion as part of the $1.4 billion investment banking settlement in 2003, the largest
payment by any of the companies involved. Also like other banks, the company
grew through acquisitions of other banks (Primerica in 1988), insurance compa-
nies (Travelers in 1993), brokerage frms (Shearson and Salomon in 1998), and
credit cards (Sears in 2004) (Rosenberg 2003). In fact, the last decade of banking
in the United States can be best characterized by the mergers. A handful of banks
control a huge number of deposits (Citibank, Bank of America, J.P. Morgan Chase,
Wells Fargo, and Wachovia) along with a few large regional banks (Stone 2003).
Ultimately, one limiting factor to mergers is that federal law currently prohibits
a single U.S. bank from holding more than 10 percent of the consumer deposits.
Banks have also been one of the big benefciaries of the loose federal monetary
policy of the early 2000smaking it easy to borrow money at almost no cost.
The entire banking industry was rattled in the 2008 recession and housing mar-
ket crash. All of the major banks in the U.S. and many banks in other countries
were bailed out by governments. A few banks were allowed to fail, and a few oth-
ers merged with larger banks. The banks most involved in shaky mortgage loans
were the hardest hit and the frst to go. Brokerage frms were also forced to merge
in order to survive. For example, WaMu (savings bank) was eventually sold to JP
Morgan Chase, and Merrill-Lynch (brokerage) was picked up by Bank of Ameri-
ca. The crisis revamped banking rules and brokerage oversight, but the process is
ongoing and will take many more years to stabilize (Sorkin 2009).
Decisions
All aspects of the fnancial industry heavily utilize information technology. All
frms face diffcult daily questions on how to balance risks and profts. Overall,
banks have profted because they have become more adept at identifying poten-
tial bad loans and diversifying risk. The entire industry has become more inter-
connected through new fnancial devices to share risk and create new products.
The use of credit-default swaps and derivatives has been credited with protecting
banks in the downturn of the early 2000s (Economist 2004). Derivatives spread
risks across the entire industry and multiple frms, which is good. But, it means
that all companies, investors, and regulators have to carefully monitor a wide
range of fnancial indicators and evaluate large-scale effects. How can anyone
possibly handle this much data to make informed decisions?
Additional Reading
Dugas, Christine, Bankruptcy Filings Set Reocrd in 2002, USA Today,
February 14, 2003.
Gross, Daniel, Capital One, Whats in Your Wallet? Slate, October 25, 2002.
Just Deserts? The Economist, April 15, 2004.
Nocera, Joseph, Wall Street On the Run, Fortune, June 1, 2004.
Revell, Janice, How Much Do Brokers Have to Hide? Fortune, April 28, 2002.
Rosenberg, Yuval, Sandy Hands Over the Keys to Citi, Fortune, July 16, 2003.
Stone, Amey, Will Citi Upstage BofAs Big Deal? BusinessWeek, October 29,
2003.
Sorkin, Andrew Ross, Too Big to Fail: The Inside Story of How Wall Street and
Washington Fought to Save the Financial Systemand Themselves, Viking,
2009.
http://www.cardweb.com
Case: Citigroup
Charles Prince III, general counsel of Citigroup explained the overall challenges
facing management: Youve got fve or six or seven businessescredit cards to
mortgages to personal loans to investment banking to commercial insurance. They
relate in important ways, but theyre different. And theyre all over the place. No
one has ever had a company as broad in geographic scope [101 countries], as
broad in product set, and as deep in size (Loomis 2001). Sandy Weill gained con-
trol of Citigroup (ticker: C), the parent company, in 2000 after a merger between
his company (Travelers) and Citibank. In 2001, most of the top-level management
was reorganized, but the structure is a little non-traditional. Robert Willumstad is
head of consumer business. Michael Carpenter runs the corporate division includ-
ing the Salomon Smith Barney brokerage and Citis commercial bank. Thomas
Jones runs the investment management and private bank. Bictor Menezes heads
the emerging markets section. Charles Prince and Jay Fishman, technically co-
chief operating offcers, also report directly to Weill. Prince handles operations
and administration, Fishman risk and fnance. Neither has any direct authority
over the other business leaders. One of Fishmans jobs is to synthesize the overall
risk picture of the business (into a 54-page monthly book), so that everyone can
see the current issues and risks. The lines in many cases are somewhat blurry.
Prince, for example, has a responsibility to promote cross selling of products. Be-
cause most of the company growth has come from acquisitions, few people down
in the hierarchy know much about the other divisions and options. Chief fnancial
offcer Todd Thomson also has oversight of cross selling. Yet, neither Prince nor
Thomson has direct authority over the businesses that do the cross selling. Weill
points out that this company is too big to micromanage, but its not so big that
you cant know whats going on (Loomis 2001). Overall, the combination of a
huge bank and brokerage frm creates signifcant advantages over the competitors.
When necessary, Citigroup can swing an investment underwriting deal by offer-
ing a Citibank loansomething that is much harder to do at brokerage frms like
Merrill Lynch.
One controversial step Weill has taken is to minimize the use of the Internet.
Instead of trying to run it separately, he moved the projects into the subdivision
units and asked them to merge them into their own operationswhile trying to re-
duce costs. Similarly, the Salomon Smith Barney retail brokerage is a full-service
company and does not offer low-cost Internet trading. On the other hand, Deryck
Maughan, CEO of E-Citi worked hard to establish an electronic foreign-exchange
market called Atriax in partnership with three other large banks. We are saying
that we would rather disintermediate ourselves in partnership with others and grab
a large share of the new market than sit around in some pre-Information Age fac-
tory (Loomis 2001). In October 2003, Weill remained chairman but made Prince
CEO and Willumstad president (Stone 2003).
American fnancial institutions were hammered by the recession of the early
1990s, even Citibank came close to going under. Yet with the recession of the
early 2000s, most banks, including Citigroup, increased profts. Average return
on bank assets jumped from 8 percent in the 1990s to 16 percent in the 2000s.
One reason is that the 2001 recession was not as deep, and banks are not as tied to
corporations as before, supplying 40 percent of the funds today versus 50 percent
a decade earlier. Banks also profted from the drop in interest rates engineered
by the Fed. But banks are still the biggest source of funds for companies. The
big change: a substantial improvement in risk management at American banks.
Doug Woodham of Moodys observes there has been a step change in risk man-
agement (The Economist 2004). For example, mortgages have been turned into
securities that are sold on a national marketspreading the risk away from the
issuing bank. Credit-default derivatives have accomplished the same task for cor-
porate loans, with the market reaching $350 billion in 2003.
The Hispanic population represents a growing marketparticularly for Ci-
tibank with its international reach and reputation. In 2004, Citibank created a
checkless checking account specifcally for Hispanic customers. Many workers
from South America come to the United States to make money, which they send
back to their families. Yet many of them do not realize the importance of banks
and ATMs in the U.S. economy. Most have diffculty opening checking accounts.
The Citibank account relies on a debit card and Citibanks international collection
of ATMs. Workers can deposit money in the account, and family members can
withdraw it almost anywhere in the world through a local ATM (Wentz 2004).
Citibank is one of the largest issuers of credit cards in the United States. Yet
that market is stagnant with intense competition and little opportunity for proft.
Some frms have tried to expand profts by going after the subprime market, but
Citigroup has resisted. Instead, it is aiming to expand its reach globally. Leverag-
ing its experience in emerging markets, it began offering Visa cards to Chinese
consumers in February 2004. The charges can be paid using either dollars or Chi-
nese yuan. Charles Prince and other dignitaries opened the service with parades
through the streets of Shanghai. In 2003, only 25 million of the 1.3 billion Chinese
citizens held credit cards. Some experts expect the number to easily triple in 10
years. Although the untapped market seems to offer incredible benefts, the risks
are enormous. There is only one credit bureau in the entire country, minimal mar-
ket research, and high bank fraud. Similar expansions of credit in Hong Kong and
Korea resulted in huge losses from personal bankruptcies. Citigroup is mitigating
the risk by performing their own background checks, only issuing cards to clients
older than 21, and requiring an annual income of at least $6,000 (restricting the
market to only 10 percent of Shanghais 16 million residents). Citibank has also
devised a mathematical scoring system to asses each individuals credit risk. The
McKinsey consulting frms notes that the current default rate in China is 1.5 per-
cent compared with 5 percent in the United States. An interesting twist to the risk
problem is that the customers tend to pay off their card in full each month, so the
bank has to make its money on fees instead of interest charges. (Baglole 2004).
To train brokers and agents, Citigroup needed a better way to communicate. In
2004 and 2005, the company installed IBMs Digital Media for Banking technol-
ogy that enables it to stream live and archived content to 300,000 PC desktops
across the companys offce and branches (Mearian 2004). Communication with
external organizations is more complicated. A key feature in the modern bank-
ing system is communications from the banks to the credit bureaus. On a regular
basis, Citigroup sends a tape of customer data and payment histories to various
credit bureaus. On May 2, 2005, Citigroup sent a tape to an Experian facility in
Texas via UPS. However, the tape never made it to the company and UPS was
unable to track the package. The unencrypted tape contained data on 3.9 million
customers including Social Security numbers and payment history. After the inci-
dent, Citigroup decided it would be safer to begin encrypting the data on the tapes
(McMillan 2005).
In early 2007, Citigroup laid off 17,000 workers in an effort to cut costs by
$10 billion. The company also reorganized IT operations, largely by consolidating
data centers. The company said that simplifcation and standardization of Citis
information technology platform will be critical to increase effciency and drive
lower costs as well as decrease time to market. The company also planned to
move 9,500 back-offce positions to lower-cost locations, including off shore (Vi-
jayan 2007). Analysts observed that Citi had little choice because between 2000
and 2005 its revenue increased by 8 percent while costs soared by 15 percent.
Citigroup planned to close half of its 42 data centers. Much of the savings will
come by replacing server clusters with grid computing based on thousands of
smaller computers networked together (Crossman 2007). Another part of the plan
involved reducing its credit card platforms from 12 down to 2, and reducing mort-
gage-origination software systems from 5 to 1 (McDougall 2007). However, at the
same time, Citigroup announced that it was upgrading 325,000 employee desktop
PCs to Windows Vista (Lai 2007).
The housing market collapse of 2007-2008-ongoing caused huge problems for
most banks. The main problem was that banks had made loans to weak customers
who could not pay them back. When housing prices collapsed (many cut in half),
borrowers decided to default on their loans and walk away. Banks who made the
loans had resold the loans as packages to everyone else. The entire banking indus-
try was hit with a liquidity crisis. Citigroup alone posted losses of $27.7 billion in
2008 and $1.6 billion in 2009. With a huge infusion of cash by the federal govern-
ment, Citigroup returned to proftability ($10.6 billion) in 2010. Homeowners did
not really gain anything. Banks were also required to increase their capital hold-
ings to reduce the risk of future meltdowns.
To reduce costs, Citigroup cut another 52,000 jobs at the end of 2008. Many of
the job cuts came through selling subsidiaries, including the sale of Citigroups
India-based computer management center to Wipro (Thibodeau 2008).
Largely due to the fnancial crisis, banks largely made few changes in opera-
tions or technology. On the other hand, security issues continued. Citigroup was
hit by hackers who stole card numbers from over 360,000 accounts (McMillan
2011). Hackers did not get into the main database, but were able to guess account
numbers and log into the customer Web site. Sensitive data such as Social Secu-
rity Numbers, expiration dates, and CVV codes were not compromised. Still, the
thieves racked up $2.7 million in charges on 3,400 accounts. The bank covered
those losses, and issued new cards to all of the affected customers.
Many banks have jumped on the smartphone platform and offer apps to help
customers monitor their accounts and transfer money. Citigroup encountered a
problem with their early version. The software stored account information on a
fled on the iPhone, and apparently this data was not encrypted. If someone lost a
phone, it would have been relatively easy for someone to read the fle to get the
account information. Citi releaed an upgrade that removed the fle (Keizer 2010).
In another interesting security problem, Citigroup reported that a former vice
president (Gary Foster) allegedly stole $19.2 million from corporate accounts that
were in his responsibility (Bray and Rothfeld 2011). Basically, he is alleged to
have set up false contracts and just wired the money to his own account. The
transactions were uncovered in an audit. Shannon Bell, a Citi spokeswoman stated
that We are outraged by the actions of this former employee. Citi informed law
enforcement immediately upon discovery of the suspicious transactions and we
are cooperating fully to ensure Mr. Foster is prosecuted to the full extent of the
law.
Questions
1. How does Citigroup use models and information systems to make decisions
in the credit card market?
2. How is the Chinese and Southeast Asian market for credit (especially in
Korea) different from that in the United States?
3. How could Citigroup make better use of the Internet?
4. How can models, expert systems, and other information tools help Citigroup
manage such a large organization?
Additional Reading
Baglole, Joel, Citibank Takes Risk by Issuing Cards in China, The Wall Street
Journal, March 10, 2004.
Bray, Chad and Michael Rothfeld, Former Citi Executive Charged in $19
Million Theft, The Wall Street Journal, June 27, 2011.
Crossman, Penny, Wall Street-Style Power, Information Week, April 14, 2007.
The Economist, Just Deserts? April 15, 2004.
Keizer, Gregg, Citi Confrms Critical Bug in iPhone Mobile Banking App,
Computerworld, July 26, 2010.
Lai, Eric, Ballmer: Citigroup to Upgrade 500,000 PCs to Vista in Next Year,
Loomis, Carol J., Sandy Weills Monster, Fortune, April 1, 2001.
McDougall, Paul, Citigroup Cost Cutting Puts a Bulls-Eye On IT, Information
McMillan, Robert, Citigroup to Encrypt Data sent to Credit Bureaus,
McMillan, Robert, Citigroup Hackers Made $2.7 Million, Computerworld,
June 24, 2011.
Mearian, Lucas, Citigroup Rolls out Streaming Media to Corporate Desktops,
Stone, Amey, Will Citi Upstage BofAs Big Deal? BusinessWeek, October 29,
2003.
Thibodeau, Patrick, Wipro Says Buying Citigroup Unit will Boost IT
Infrastructure Outsourcing in India, Computerworld, December 23, 2008.
Vijayan, Jaikumar, Citigroup to Lay Off 17,000, Overhaul IT Operations,
Wentz, Laurel, Banks Tailor Efforts to Homesick Hispanics, Advertising Age,
April 5, 2004.
Case: Capital One
As a newcomer to the credit card industry, Capital One (ticker: COF) needed an
edge. The company created the edge by building the Information Based Strategy
(IBS). The system is highly focused on testing and evaluating various options be-
fore offering them to the public. The three-step approach: (1) create a new product
and fnd a target population, (2) create a test by changing the variables and seeing
how the group members react, and (3) use the test results to further divide the
segment, and then test specifc campaigns against those segments. By specifcally
targeting rates, fees, and options to each market, Capital One was able to obtain
high response rates, reducing the costs of acquiring a customer. The company
conducted 45,000 tests in 2000 alone (120 per day). The company extended the
process to its Web site, tracking visitor activity, tailoring options to each specifc
customer, and using the background data to buy ads on other sites with the appro-
priate demographics. For example, if sports enthusiasts responded well to certain
features, those features would be advertised on sporting sites. With this process,
the company opened 2 million new accounts online (Cohen 2001).
The company uses the test results and marketing to tailor interest rates and fees
to individual customers. In effect, it has created 100,000 different segments or
product combinations. The process extends to the customer call center. Capital
One uses Cisco Systems Global Service Logistics (GSL) system to route calls.
The system retrieves information about the customer and routes the call to the
most appropriate representative. For example, a customer who routinely pays off
the monthly charges might be routed to a representative to sell a platinum card
with a higher balance.
In a different twist, Capital One uses a similar process to hire and promote em-
ployees. It records specifc data on every hire, including scores on a timed math
test and a behavioral test. During promotions, the characteristics of successful em-
ployees are listed for each job. These characteristics, and the initial test scores, are
used to refne the hiring process for each type of job. The process fts employees
to jobs and reduces hiring costs (Cohen 2001).
For several years, Capital One had enviable growth rates with revenue grow-
ing from $95 million in 1995 to $4.97 billion in 2001, from 6 million customers
to 33 million (Cohen 2001). Yet because of the extremely competitive nature of
the industry, Capital One had to extend into the subprime market to capture more
customers. With the economic downturn, the subprime market crashed in 2002.
Capital One suffered along with the others in the industry. For the frst time, man-
agers also revealed how dependent Capital One was on the subprime market. In
2002, 40 percent of its cards were in the hands of subprime customers (Albergotti
2003), far above what most investors had believed. The market punished Capital
Ones stock. Yet, ultimately, Capital Ones system worked. The companys risk
management techniques gave it one of the lowest levels of bad loansuntil 2003.
The company wrote off bad debt, but earnings eventually rose and the company
survived. The company has reduced its exposure in the subprime market. Federal
oversight also forced companies to reevaluate the subprime marketregulators
have been attempting to limit fees and penalty charges (Smith 2002). But, the
strategic shift leaves Capital One facing more competition in the higher quality
markets. Its proft margins have shrunk accordinglydown to 16 percent in 2004,
half their earlier levels (Byrnes 2004).
Searching for new markets, Capital One is widening its search into more tra-
ditional banking areas. Richard D. Fairbank, the CEO, knows that growth in the
card market will be slow. A lot of the different fnancial markets are evolving, but
the most evolved is the credit card business. This is pretty close to the endgame.
Weve really got to work for a living these days in the credit card market. His
new target: auto loans, installment loans, and even international loans. In particu-
lar, he wants to target small businesses. Capital Ones lending to small businesses
has grown from $400 million in 1999 to $3.3 billion in 2003. At the end of 2003,
the company managed $46.3 billion in credit card loans, $8.5 billion in autos, $5.4
billion in installment loans, and $7.6 billion in international loans (Kuykendall
2004).
In the meantime, the company is cutting expenses, eliminating as many as
2,500 of its 9,000 positions in its Richmond, Virginia, headquarters. Fairbanks
told employees he needed to cut expenses by 20 percent. Our businesses need
to improve their cost positions to compete in the future against leading players in
the fnancial-services industry (Hazard 2004). Capital One is also interested in
using the Internet to reduce costs. Processing payments online instead of paper
checks reduces costs by $1 a year per customer. Persuading customers to accept
electronic statements instead of mail cuts costs an additional $5 per customer per
year. Rick Long, director of U.S. card operations at Capital One, notes that our
ROI [return on investment] models are built on lowering costs [Wade 2004). He
was emphasizing that Capital One is primarily interested in using the Internet only
where it is cost effective. The company is also concerned that customers have to
be enticed to use the Internet instead of coerced. If we drive them to the site, they
may not stay [Wade 2004). Most customers who are interested in using the Inter-
net, enroll within the frst six months.
To broaden its reach and fnancial offerings, Capital One embarked on a cam-
paign of buying retail banks. The company purchased North Fork Bancoprt as
well as Hibernia Corp. (Wei 2007) Capital One IT managers were so impressed
with Hibernias deposit-transaction system that they migrated their banking sys-
tem onto its mainframe package. In its other 9 recent acquisitions, Capital One
migrated most operations onto its existing platforms. In particular, HRM, fnance,
and other operations are transferred to Capital Ones PeopleSoft ERP system (Mc-
Gee 2006).
Like other institutions, Capital One is trying to reduce expenses. In 2007, the
company announced job cuts of 2,000 employees in an effort to reduce costs by
$700 million (Wei 2007). The IT department is working to cut expenses by reduc-
ing the number of servers it runs. Using server virtualization, a hardware-based
server can be split into multiple virtual servers, with each one running different
operating systems and different applications. One goal is to reduce costs by mov-
ing from 1,600 down to 1,100 servers. Le Congdon, VP of corporate technology
also said that the technology buys us time in circumstances where a specifc ap-
plication is ready to move over and we need to shut down that physical location,
or where we need to move of aging hardware or software. It also gives us better
management tools and controls across the environment. If we can go from 50 per-
cent to 85 percent utilization on the servers, thats free money (Thibodeau 2006).
Capital One was also hurt by the fnancial crash. But, because the frm focused
on higher-end customers (compared to Providian), it was able to survive the li-
quidity crash. Rob Alexander, CIO, pointed out that the company was still in-
vesting in technology to handle customer transactions. He also emphasized the
importance of business analytics at Capital One to ensure the company was mak-
ing the best decisions possible (Fonseca 2008). Alexander also noted that the IT
department maintains an advanced technology group to look at cutting-edge tech-
nologies and test them to see how they might improve operations or deliver new
business opportunities in the near future.
In 2011, Capital One purchased the U.S. online-banking business of INGa
Dutch bank that pioneered some online banking. The deal will make Capital One
the ffth largest U.S. bank measured by the amount of deposits (Grocer 2011).
Questions
1. How does Capital One use decision support systems to reduce risk and
increase sales?
2. With its advanced systems, how did Capital One lose so much money in
2003?
3. How can Capital One use technology to reduce its costs?
Additional Reading
Albergotti, Reed, Fewer Credit-Card Solicitations Are in the Mail, The Wall
Street Journal, November 5, 2003.
Byrnes, Nanette, Coming of Age At Capital One, BusinessWeek, February 16,
2004.
Cohen, Jackie, Growth Formula, Computerworld, July 2, 2001.
Fonseca, Brian, The Grill: Capital Ones CIO Discusses Business Strategies for
Tough Financial Times, Computerworld, November 3, 2008.
Grocer, Stephen, Deal Profle: Capital One Wins ING Direct USA, The Wall
Hazard, Carol, Capital One Shifts Focus; Firm to Cut Richmond, Va., Jobs to
Reduce Costs, Richmond Times, April 29, 2004.
Kuykendall, Lavonne, Amex, Capital One Talk About the Future, The
American Banker, June 7, 2004.
McGee, Marianne Kolbasuk, Capital One: A Serial Acquirer, Information
Smith, Geoffrey, The Bill Comes Due For Capital One, BusinessWeek,
November 4, 2002.
Thibodeau, Patrick, Capital One Cuts Servers With Virtualization Push,
Wade, Will, Coming Full Circle on E-Strategies, The American Banker, June
15, 2004.
Wei, Linling, Capital One to Cut About 2,000 Jobs, The Wall Street Journal,
June 28, 2007.
Case: Providian/Washington Mutual
The credit card industry has changed in the decades since the 1960s when Bank of
America fooded households with millions of cards. Today, trillions of dollars of
transactions are paid with credit cards, with 1.2 billion in use in the United States
alone. The average cardholder has 2.7 bank cards, 3.8 retail cards, and 1.1 debit
cards for a total of 7.6 cards per person. Despite the huge variety, the top 10 issu-
ers handled about 78 percent of the total value at the end of 2002. About 24 per-
cent of all consumer retail transactions are paid with credit and debit cards (www.
cardweb.com).
A huge percentage of the U.S. population already holds several credit cards, and
banks are constantly competing to give more cards to high-income customers. To
most observers, the market appears to be saturated. Customers will switch banks
if one offers a better rate or different benefts. Providian Financial (PVN), a San
Francisco-based company, looked at this market and decided to fnd a new niche:
the subprime market. According to cardweb, approximately 25 percent of the pop-
ulation has weak credit ratings. The objective was to fnd low-income people with
poor or no credit, offer them a credit card, and then charge huge fees and high
interest rates. Shailesh Mehta, who eventually became CEO, developed a complex
mathematical model that allowed the company to identify the subprime custom-
ers who would be most likely to use the cards, but not default on the loans. One
executive characterizes the process as we found the best of the bad. Since other
banks refused to serve this market segment, Providian had solid growth rates and,
despite the 24 percent and higher interest rates, was able to claim that it was pro-
viding a useful service to the customers. Eventually, other banks (notably Capital
One and Household International) jumped into the market. By 2000, 20 percent of
the cardholders were classifed as subprime (cardweb), and Providian was the ffth
largest card issuer in the nation (Koudsi 2002). In 2000, the company was ordered
to pay $300 million in restitution for misleading, unfair, and deceptive business
practices. To keep growing, the company had to fnd a new market. It failed in an
experiment to go after the platinum market. So, it went after increasingly risky
customers. By 2001, when the economy faltered, the subprime market felt the
impact frst, and the company saw its default rates jumpto a huge 12.7 percent.
Amid claims of misleading accounting, the stock price plummeted 90 percent.
Cardweb notes that 1.3 million cardholders declared bankruptcy in 2001which
generally erases unsecured credit card debt. Other banks focusing on the subprime
market faced similar problems. Bernhard Nann of Fair Isaac, which evaluates
consumer creditworthiness, notes that data used in traditional behavioral models
is not quite as powerful as we would like. Wed like to go and expand the universe
of data that can be looked at to produce better predictions. Thats particularly im-
portant in the subprime area. In particular, the company wants to include utility
and rent payments in its calculations (Punch 2003).
With the economy improving in 2003 and 2004, Providian was able to sell off
some of its loans and write off the really bad ones. The company also refocused its
efforts to go after the middle market, people with a FICO score between 600 and
720 (Albergotti, 2003). Although weak, the company was able to report a proft
in late 2003 (Business Week 2003). By 2004, some experts saw an improvement
in the credit card industry profts. Many homeowners had already refnanced their
mortgagesand paid down credit card debt. But, that leaves them free to increase
the borrowinghence increasing profts (Stovall 2004).
Yet, with increasing competition for the platinum and midmarket segments,
banks continue to search for ways to make money on credit cards. With low bal-
ances and low interest rates, and few new customers, banks turned to fees. Income
from late fees and penalties was predicted to reach $13 billion for 2004, with total
fees likely to account for 39 percent of revenue (Simon 2004). Providian joined
in the parade and increased its fees annually. However, the company also created
a Real Rewards program, where cardholders accumulate points for cash rebates
that can also be used to offset late fees.
In 2004, Providian settled the accounting and insider-trading lawsuits for $65
million (Kuykendall 2004).
In 2005, Providian was purchased by Washington Mutual Bank WaMu (Hal-
linan 2005). WaMu was formed in 1889 in Washington State. It runs consumer
and small-business retail banks from the West Coast to the Midwest. It also has
an aggressive home-lending program. In the past few years, it has become known
as a bank for taking risks and targeting subprime loansboth in credit cards and
mortgage loans. WaMu was pressured by the crashing housing market in 2006
and 2007. Because of its aggressive subprime loans, the company was one of the
frst to offer new loans to customers to try and move them into safer loans as rates
increased, in an effort to keep large number of customers from defaulting on hous-
ing loans (Carrns 2007 and Hagerty and Carrns 2007).
WaMu aggressively uses technology to reduce labor costs in its banks. In most
of its branches, tellers assist customers with transactions that are handled elec-
tronically. Effectively, tellers work at advanced ATM stations that enable them
to handle tasks only slightly more complex than a standard ATM machine. If a
customer processes a transaction involving cash, the money is dispensed by the
machine.
The real estate crash and fnancial liquidity crunch were tough on WaMu. Es-
sentially, the government forced the sale of WaMu to JP Morgan Chase, who
picked up the assets for relatively low cost. Shareholders of WaMu stock pretty
much lost all of their value. For a couple of years after the crash, top-level manag-
ers tried to argue that the government should not have forced the sale of WaMu,
and that government actions might have contributed to the failure of the company.
But, in the end, most observers simply blamed the over-aggressive lending prac-
tices of the company. When the real estate market crashed, an increasing number
of loans turned bad, and even moderately risky customers walked away from their
debt. These were the customers targeted by WaMu, and it lacked the capital and
stable profts to balance these risks. JP Morgan had to write off at least $31 billion
in bad loans (Sidel, Enrich, and Fitzpatrick 2008).
Questions
1. Why is the subprime market so risky and how does Providian use
information technology to minimize the risks?
2. How are the platinum and midmarket accounts different from the subprime
market, and can Providian use the same models?
3. How can Providian use information technology to increase its revenue from
fees?
Additional Reading
Albergotti, Reed, Fewer Credit-Card Solicitations Are in the Mail, The Wall
Street Journal, November 5, 2003.
BusinessWeek, UBS Financial Cuts Providian to Neutral, October 30, 2003.
Carrns, Ann, Washington Mutual Earnings Slide on Home-Loan Losses, The
http://www.cardweb.com/cardlearn/stat.html
Hagerty, James R. and Ann Carrns, WaMu Leads in Risky Type of Lending,
The Wall Street Journal, April 17, 2007.
Hallinan, Joseph T, WaMu Has Deal To Buy Providian For $6.45 Billion, The
Koudsi, Suzanne, Sleazy Credit, Fortune, February 19, 2002.
Kuykendall, Lavonne, Providian Says It Can Move On After Settling Class
Action for $65M, The American Banker, June 9, 2004.
Punch, Linda, Subprime Cards, Act II, Credit Card Management, November
2003.
Sidel, Robin, David Enrich, and Dan Fitzpatrick, WaMu Is Seized, Sold Off to
J.P. morgan, In Largest Failure in U.S. Banking History, The Wall Street
Journal, September 26, 2008.
Simon, Ruth, Credit-Card Fees Are Rising Faster; As Economics of Industry
Change, Issuers Seek To Boost Income From Penalties; $35 Late Charge,
The Wall Street Journal, May 31, 2004.
Stovall, Sam, A Rising Rating for Consumer Finance, BusinessWeek, February
24, 2004.
Case: Merrill Lynch
For decades, Merrill Lynch (ticker: MER) was the premier brokerage frm. It led
the industry in sales and innovations. The frm helped create hundreds of modern
companies with its access to capital and the ability to sell stocks. In the 1970s
and 1980s, its innovations in money market funds rocked the fnancial world and
resulted in substantial changes to the federal banking laws. It accomplished many
of these innovations through technologyand the recognition that money is re-
ally data. It expanded its reach overseas, and became the largest retail broker-
age and the largest underwriter of stocks and bonds in the world. Through the
stock market bubble of the 1990s, Merrill prosperedat least in terms of market
share. They helped fnance several start-ups and retail stock brokerage revenues
grew by $3 billion from 1996 to 1998. But the growth carried equally huge costs:
only $100 million of that money made it to the bottom line. By 1998, Merrills
proft margins were 10 percentage points lower than its competitors. At the same
time, the growth of online account management and discount brokerage frms had
chipped away at the retail end of Merrills operations. By 2001, even the directors
knew that Merrill was in trouble and might not survive. Being the largest was not
very useful if the company was not making money (Rynecki 2004). In 2001, the
board made the surprising move of appointing Stan ONeal, the CFO to become
the new CEO of Merrill. Recognizing the importance of profts, and seeing limits
on the ability to increase revenue, he began focusing on cutting costs. Ultimately,
he eliminated 24,000 jobs, including 20 percent of investment banking and ana-
lyst positions. He closed 300 feld offces and completely pulled out of Australia,
Canada, New Zealand, and South Africa. He reduced the number of stocks traded
directly by the frm by 75 percent. He replaced almost all of the management from
the top downlargely replacing them with younger staffers looking to make their
marks. He reportedly often states that ruthless isnt always that bad (Rynecki
2004). The result: in 2003, the frm earned a record $4 billion in proft, and pre-
tax margins reached 28 percentvastly exceeding the high-growth years of the
1990s.
Despite the proftability, the frm still has to worry about revenue. In 2004, with
the slow economy and shaky fnancial markets, Merrill ranked last in sales growth
among the Newsday Top 100with a three-year sales decline of 14.8 percent
(Murray 2004). Merrill Lynch is refocusing its retail brokerage operationson
wealthy customers. The company pays higher interest rates to customers with bal-
ances over $10 million (McGeehan 2004). Merrill also jumped into the credit card
market in 2004. The company realized that customers withdrew $3 billion from
their accounts in 2003 to pay off credit card debt. Co-branded with MBNA, the
card is designed to provide one-stop shopping for its customers. Merrill is target-
ing its wealthier customersin an effort to provide more services, but also to keep
control over a larger percentage of their money. The company is offering a variety
of awards to entice customers to use them for large transactions (Lieber 2004).
Despite the renewed emphasis on cost, Merrill still has to handle business, and
that requires technology. In 2003, the company began the frst steps of a $1 billion
upgrade to their broker workstation systems. The new Client 360 system is de-
signed to be a total wealth-management toollargely focused on customer rela-
tionship management. It is designed to provide a complete view of customer data
to the broker. John Killeen, the chief technology offcer, notes that the tool suite
around wealth management is pretty well established across the industry, still what
will make this unique is that we are facilitating the relationships for our fnan-
cial advisers with their clients. Its a single-screen representation of the most
important and most prevalent questions that a client may ask a fnancial adviser
when they have them on the phone. It will talk to balances; it will talk to progress
towards plans, and any important notices that affect that client. And within one
or two clicks, we can drill down to greater and greater detail. It represents a huge
productivity gain for our fnancial advisers (Pallay 2004). Most of the front-end
system being developed is installed by Thomson Financial, which is overseeing
the integration of work from 400 vendors. The back-end system remains Merrills
proprietary system, largely running on Microsoft software. The company built a
framework that separates the back-end and middle tier systems from the front end.
This approach enables the company to alter various components on either end
without having to rebuild the entire system. Despite the complexity of the project,
Killeens focus has been on the end-use applications: I think the lynchpin around
the tools will be client datawho has the greater understanding of the total client
relationship, what is important to the client at the different phases of their fnan-
cial life cycle, and how do we put all of that together so that our advisers are best
positioned to work with their client?
Technology has been driving several changes in market trading for several
years. For a while, the stock exchanges frowned on automated trading schemes,
but by 2007, they had adopted them wholeheartedly. Brokers, including Merrill
Lynch, have developed complex models of trading strategies. Some look for ar-
bitrage opportunities across markets, others look for timing gaps. These systems
can generate hundreds of trade orders per second, and generally hold positions for
only a few seconds or minutes (Martin 2007). Economically, these systems serve
to remove bumps and reduce imperfections in the market. In the process, the large
brokers make profts not available to the average investors, but the profts pay for
the technology.
In terms of customer-based technology, Merrill Lynch is reportedly the largest
customer of Salesforce.com with licenses for 25,000 users (Weier 2007). Sales-
force provides customer relationship management software via an online system.
The company offers its Wealth Management Edition for Merrill Lynch and other
fnancial institutions. In terms of processing transactions, federal regulations be-
gan changing the rules in July 2007 with the implementation of Reg NMS. The
regulation requires brokers to automatically and instantly route orders to the mar-
ket that displays the best price (Horowitz 2007). Only the largest brokerage frms
have been able to afford the technology to maintain constant connections and
price searches with multiple markets. With many of the exchanges automated, the
brokerage computers need to be fast and secure to handle the number of searches
required for each transaction. Tim Cox, VP of market structure strategy at Merrill
Lynch observed that The big are going to get bigger. Reg NMS over the course of
a couple of months will make technology a real differentiator (Horowitz 2007).
Merrill Lynch only barely survived the fnancial crisis of 2008. The company
had made huge bets on the housing marketand held those bets on its balance
sheet. Its competitor Lehman Brothers was allowed to fail by federal regula-
torspartly to send a message to the rest of the industry. Merrill Lynch would
probably have been close behind, but Bank of America took a $50 billion gamble
and bought out the company (Farrell 2010). Ultimately, the question of wheth-
er Merrill Lynch was worth billions of dollars to Bank of America remains to
be seen. Can the division generate enough revenue to justify its continued exis-
tence? After all, it is still a high-priced brokerage frm. Perhaps it can help Bank of
America in high-end deals for IPOs and other investment banking tasks.
Questions
1. How can Merrill Lynch survive as a full-service broker, particularly with the
regulatory changes in investment research?
2. How can Merrill Lynch use information technology to attract and keep the
high-end investors that it wants?
3. How can the company use technology to reduce its costs and improve its
proft margins?
Additional Reading
Farrell, Greg, Crash of the Titans: Greed, Hubris, The Fall of Merrill Lynch and
the Near-Collapse of Bank of America, Crown Business 2010.
Horowitz, Jed, How New Rule May Tip Brokerage Scales, The Wall Street
Lieber, Ron, A New Contender for Your Wallet; Merrill Jumps into Card War
With No-Fee Rewards Offer, But Does It Beat Platinum? The Wall Street
Journal, April 20, 2004.
Martin, Richard, Wall Streets Quest To Process Data At The Speed Of Light,
Information Week, April 21, 2007.
McGeehan, Patrick, Schwab Tells Less-Affuent Customers They Have To Pay
More Than Richer Clients To Make Trades, The New York Times, May 26,
2004.
Murray, Christian, Top 100; After a Low, Merrill Looking Up, Newsday, June
14, 2004.
Pallay, Jessica, An Interview with Merrill Lynchs John Killeen, Wall Street &
Technology, May 2004.
Rynecki, David, Putting the Muscle Back in the Bull, Fortune, April 5, 2004.
Weier, Mary Hayes, New Deals Show Software-As-A-Service More Than A
Niche, Information Week, February 28, 2007.
adopted by rivals?
sectors?
653
How can you use information technology to improve your organization and make it
better than your competitors?
How competitive is your world?
What are the main factors affecting a frms competitive advantage? Where do you
begin looking for an edge?
How can you use IT to gain a competitive advantage? Where do you begin your
search?
How can IT support the operations of the frm to provide a competitive advantage?
Why is it so diffcult to convince management to make strategic changes? What are the
risks of strategic decisions?
Why did so many dot-com frms fail? Do their failures mean there is no viable Internet
strategy?
How do you convince an organization to change strategies?
Can cloud computing provide strategic advantages?
Chapter Outline
Strategic Analysis
10
Chapter
Introduction, 654
The Competitive Environment, 656
External Agents, 658
Buyers, 658
Suppliers, 660
Rivals, New Entrants, and Substitutes,
660
Government Regulations, 662
IS Techniques to Gain Competitive
Advantage, 662
Barriers to Entry, 662
Distribution Channels, 664
Switching Costs, 666
Lower Production Costs, 667
Product Differentiation and New
Products, 667
Quality Management, 669
The Value Chain, 670
The Search for Innovation, 671
Research, 672
Engineering and Design, 673
Manufacturing, 673
Logistics and Supply, 674
Marketing, 674
Sales and Order Management, 676
Service, 676
Management, 676
Costs and Dangers of Strategies, 678
High Capital Costs, 678
When the Competition Follows, 679
Changing Industry, 679
Sharing Data, 680
Government Intervention, 681
The Role of Economics, 682
Challenges of Strategy, 683
Sustainability, 685
Leader or Follower, 686
Change Agents, 688
Summary, 690
Key Words, 691
Exercises, 692
Cases: The Airline Industry, 696
654 Chapter 10: Strategic Analysis
Delta Air Lines
Introduction
How can you use information technology to improve your orga-
nization and make it better than your competitors? Can technol-
ogy make your company the best in the industry? These questions are still being
debated, but it is clear that in some cases information technology has provided the
ability for companies to dramatically change industries. Figure 10.1 shows that
technology can be used to build ties to customers and supplierseffectively mak-
ing it harder for competitors to enter your market. A key theory from economics
(extensively applied in marketing) states that a frm can make additional profts
only by blocking new frms from entering the market. And one of the few legal
ways to block frms is to differentiate your product so that customers will perceive
it as a separate item. Technology can provide differentiating factors including bet-
ter service, lower prices, stronger customer relationships, and new features for
products.
The reengineering provided by information technology creates ad-
ditional benefts. It enables you to alter the way an organization oper-
ates. However, it is critical that you remember that technology by itself
is not a magic wand. Now, for the important question: How can you use
information technology to fnd new opportunities and gain a competi-
tive advantage?
Strategy requires looking beyond the internal factors such as costs and imple-
mentation details. Managers need to look at the potential effects on customers,
How can you use information technology to make your company better than your
rivals? Delta is one of the original airlines in the United States. Only a handful of
the original remain, and most of them, including Delta, are firting with bankruptcy.
Many of the airline executives blame September 11, 2001, the SARS scare of 2002-
2003, and the oil price increases of 2003-2004 for their problems. But if you look at
revenue and profts over time, it is clear that the big airlines were getting their butts
kicked long before these events. In 2005, Delta, Northwest, and Continental were
in bankruptcy courtlargely to drop their pension plans and renegotiate labor and
other contracts. At one point, 50 percent of the U.S. air transportation capacity was
operating in bankruptcy. The discount airlines, led by Southwest and joined by Jet-
Blue and ATA, have been proftable over the last few yearsbecause they found a
way to attract customers and hold down costs.
Discounters and upstarts are not new to the airline industry. In a classic case,
American Airlines was able to drive away People Express in the early 1980s with its
Sabre reservation system. American pioneered yield managementcharging every-
one on the plane different pricestrying to get as much money from each person as
possible, meaning that businesses paid through the nose while tourists who book in
advance get cheap seats. The advantage generated from this system was refned with
frequent fier miles and kept American, Delta, and the others on top for 15 years.
But Southwest, with a radically different strategy, has slowly grown to challenge the
entire industry. Is there another rabbit in the hat for the big, traditional airlines? Can
Delta fnd a new way to use information technology to save the day?
suppliers, competitors, and other external agents. In some cases, the technology
can be relatively commonbut the trick is to look for additional benefts. For
instance, using social networks for marketing uses relatively standard technology,
but it requires a marketing staff that understands the products and the customers
supported by custom search engines and data analytics to determine the most im-
portant contacts and strongest effects. Using the technology changes the way the
company operates. In this example, instead of relying on traditional sales through
distributors and retail stores, companies can choose to interact directly with high-
profle customers who then affect decisions by other consumers. The real differ-
ence with strategy lies in its goal: to change the way the business operates, and
gain an advantage over the other frms in the industry.
Information systems can provide a competitive advantage through decreasing
costs, improving quality, establishing ties to consumers or suppliers, differentiat-
ing products or creating entirely new products. Computer systems can also create
barriers to entry through the need for greater technical skills or increased costs.
Designing strategic systems can be a dangerous task with many opportunities
to fail. One complication is that development costs are high. Some strategic sys-
tems use new technology, which carries higher costs and a greater risk of incom-
patibilities and other problems. It is also important to remember that attempts to
monopolize a market are illegal, and strategic systems can sometimes come close
to breaking the antitrust laws.
The most diffcult aspect of strategic systems is coming up with ideas that might
give you an advantage. Consequently, a related problem is convincing top manag-
ers that a potentially radically new, costly idea could actually work. One way to
get ideas is to see what frms in other industries have done. You never know when
some of the techniques and tricks used by other companies might be useful to you.
Figure 10.1
Strategies. Managers are increasingly being asked to fnd ways to give their frm an
advantage over the competition. Information systems can help identify areas that can
provide a competitive edge. Information systems can also directly provide services
and advantages that are not offered by your rivals.
Competition
Become the best rm in the
industry. Block the competitors by
keeping your customers happy.
Connections to suppliers
and customers.
Te Competitive Environment
How competitive is your world? One of the important trends facing most
businesses today is the increased level of competition. As indicated in Figure 10.2,
improved telecommunications and faster delivery services mean that local frms
face competition from regional, national, and international frms. Local frms have
to compete against national mail-order companies, which offer wide selections,
next-day delivery, and low prices. The Internet, home shopping channels, and toll-
free phone numbers make it easier for consumers to compare prices, putting pres-
sure on all frms.
Large national retailers and franchises put pressure on local stores. They also
compete against themselves for market territories. Their size gives them leverage
in dealing with manufacturers. By purchasing in large quantities, they can nego-
tiate lower prices. Their high volume also makes it easier for them to buy from
foreign producers.
Trends
Ideas and concepts for managing businesses are constantly changing. Many
current practices are often traced to Alfred Sloan, who drove the consoli-
dation and expansion of General Motors from 1920 to 1956. Management
techniques evolve over time and ideas come from many sources. Through
the 1950s, many companies focused on making production more effcient.
In the 1950s and 1960s, U.S. frms expanded into wider markets, both na-
tionally and internationally. In the 1970s, managers were preoccupied with
the economic changes brought on by oil price rises and consequent shocks
of high infation and high interest rates. The 1970s and 1980s also saw the
emergence of increased international competitionfor example, between
1960 and 1985, U.S. imports as a percentage of GDP increased from 5.6 per-
cent to 11.5 percent. The 1990s and 2000s saw increasing international com-
petition as well as consolidation among large frms. Many industries now
support only a handful of large frms.
Through the 1960s and 1970s, the use of MIS was largely governed by
its capabilities and the immediate needs of the organizations. The most com-
mon MIS objective was to save money and time by automating transaction-
processing tasks. The projects were evaluated on the basis of how much
money they could save. Eventually, managers came to realize that computer
systems have other advantages. A new technology might enable the frm
to provide better service to customers. The company that is the frst imple-
menter of a technology might fnd it easier to attract customers, giving it a
competitive advantage over the other frms. For example, the frst banks that
installed ATMs to provide 24-hour access gained an advantage over their
competitors.
Now, everyone uses information technology for transaction processing
and some decision-making. Still, new opportunities arise to be a leader, such
as connecting to customers through the Web, social networking, and mobile
devices. The key is to always look for ways to use new technologies and
identify the benefts as well as the costs to fnd the optimal timing.
Several international trends are creating increased competition. The interna-
tional search for lower manufacturing costs puts pressure on frms to cut their
costs. For instance, the Japanese have moved production to other Asian nations
to build television sets and automobiles. The demographics of an aging popula-
tion have led Japan from a labor surplus to a labor shortage, and Japan has moved
much production to lower-cost nations. Decreasing trade barriers throughout the
world also creates larger markets. As eastern European economies rebuild, as the
European Union takes shape, and as Chinese, Indian, and Mexican incomes in-
crease, consumers will be able to buy more products. Although the prospect of
these increased sales is enticing to U.S. manufacturers, many complications exist.
If a competitor becomes established frst, it will be a stronger and tougher com-
Figure 10.2
Competition. Even industries with one or two front-runners often have a pack of
hungry competitors trying to chase them down. Todays businesses face competitors
from around the world.
Reality Bytes: Business Trends
Business statistics indicate a clear trend toward the increased importance of service-
oriented frms. Service frms are well suited to certain strategic uses of information
systems. In particular, product differentiation, product quality, and new products are
typically useful strategies. In many service industries, information is the primary
product, so technology is especially valuable.
The fnancial industry provides several strategic examples, such as the Merrill
Lynch Cash Management Account, ATMs, or new fnancial instruments created by
brokers. Similarly, Federal Express uses tracking information to differentiate its ser-
vice from its rivals offerings. Likewise, the airlines used their reservation systems to
give them a competitive advantage in transportation services.
petitor in the United States. New frms will arise or expand in these international
markets, giving them a stronger base to increase sales in the United States, provid-
ing for increased competition.
The business world has defnitely become more price-competitive in the last
decade. Large frms competing on price continually squeeze suppliers to cut costs
and offer discounts. Think about the effect of Wal-Mart. How does Wal-Mart con-
tinue to decrease costs? Yes, through internal savings, but also through leaning on
suppliers. On the other side, the Internet and mobile computing have made it easy
for consumers to fnd price information and compare products instantly. It is easy
for consumers to use their cell phones to take a photo of a bar code and receive
instant price comparisons, quality ratings, and customer comments. Certainly re-
tailers need to consider the effects of these technologies, but manufacturers must
also evaluate the competition in terms of price and quality.
External Agents
What are the main factors affecting a frms competitive advan-
tage? Where do you begin looking for an edge? Competitive advan-
tage can be gained by establishing or changing relationships between the frm and
its external agents. External agents consist of suppliers, customers, rivals, poten-
tial new entrants, substitute products, and sometimes the government. Figure 10.3
portrays these relationships in Porters Five Forces model. From a systems per-
spective, each of these entities is outside the control of the frm. Yet they strongly
affect the company. Through improved ties to these agents, they become part of
your system, which can be used to improve the competitive position of the frm.
Buyers
Who are your customers? This famous question is used to highlight the issues in
an industry. The answer might seem obvious, but many frms have layers of cus-
tomers. To a retail outlet, customers are likely to be individual people. As shown
in Figure 10.4, a large manufacturer might have several levels of customers, rang-
ing from wholesale frms that buy in bulk, then sell to distributors, which deliv-
Threat of
New Entrants
Rivalry Among
Existing Competitors
Threat of Substitute
Products or Services
Bargaining Power
of Buyers
Bargaining Power
of Suppliers
Figure 10.3
Porters Five Forces model. Strategies often involve external agents: customers,
suppliers, competitors, the threat of new entrants, and substitute products.
er products to retailers, where the fnal customer purchases the product. Having
more intermediate levels between the manufacturer and the customer can make
it much harder to manage the frm. It also makes it more diffcult to identify the
needs of your customersparticularly when information from the various levels
conficts. For instance, an end consumer might want more features, while a retail
store might want simpler products and fewer models to reduce stocking problems.
A common strategic goal is to get closer to the customers. Information sys-
tems can be used to strengthen the ties among the customers, manufacturers, and
various intermediaries. For example, you could build electronic ordering systems,
with terminals in the retail stores to capture current sales levels. The systems
could also be used to send new product information to the customers, or collect
feedback on various attributes, or provide immediate answers to question from
retailers and customers.
The issue of buyers power is critical in Porters model. For example, if you
are a small company selling parts to General Motors, then you have little power
in that relationship. So, you will need to look at the supplier side for strategic
options.
Figure 10.4
Production chain. Modern companies have ties to hundreds or thousands of entities.
Sometimes a company will own several pieces of the production chain (vertical
integration). Sometimes the company might expand horizontally by building related
businesses. Each linkage requires communication and offers the possibility for
strategic gain.
parts
supplier
parts
supplier
parts
supplier
warehouse warehouse
tool
manufacturer
Manufacturer
workers
wholesaler
wholesaler
distributor
retail store
Consumers
Suppliers
Suppliers can provide individual parts, entire products, or even services (such as a
bank that lends money). Three major issues involving suppliers are price, quality,
and delivery schedules. Just as with customers, problems can arise when many
layers of suppliers exist. For instance, increased layers can result in longer delays
between ordering and delivery because the supplier has to contact its supplier,
who contacts its supplier.
Quality management is also more diffcult when there are several layers of sup-
pliers. A fundamental element of total quality management (TQM) states that
quality must be built into every process and item. Picture the problems that arise
if quality is measured only in terms of the output at the manufacturer. When a
defective product is found, there is no information about the cause. How can the
problem be corrected? Managers need to know where each component came from
and evaluate the quality as soon as possible. For instance, if there is a defective
product, you could check each component to determine its original manufacturer.
The manufacturer could be notifed of problems, and you could search other items
for similar defects. The manufacturer could use this data to identify problems with
individual production lines.
Information systems can be used to build electronic ties to suppliers. Com-
mon uses of these systems include placing orders, tracking shipments, monitor-
ing quality control, notifying partners of changes in plans, and making payments.
Electronic links provide faster responses, better record keeping, and fewer errors.
They also offer the potential strategic benefts described in the next section.
Rivals, New Entrants, and Substitutes
The goal of a strategic approach is to derive a competitive advantage over the
rivals, or other frms in the industry. There could be many competitors or just a
few larger rivals. The competition could take place in a small town, across a na-
tion, or worldwide. One of the frst steps in any strategic analysis is to identify the
primary competitors and to assess their strengths and weaknesses. Is the industry
rivalry intense with constant price movements and attempts to gain market share?
Or do companies rarely adjust prices and are largely content to service existing
customers?
One issue to remember about competition is that it never stops. Coming up
with one strategic idea is not good enough. For example, American Airlines
and United Airlines spent millions of dollars to build reservation systems as
strategic systems. Today, all major airlines have access to these systems, and each
airline must continually work to improve its system to provide new enticements
to customers. Similarly, automobile companies designed computerized diagnostic
systems to improve services offered by repair shops. Today, all of the manufactur-
ers have essentially the same systems. In some cases, rivals might offer improve-
ments over your ideas, which will put the originator at a disadvantage. However,
the frm that frst implements a new strategy can gain recognition and market
share. It is important to remember that companies must continually improve and
seek new opportunities.
A related issue is the concept of potential competitors or entrants in the busi-
ness. In some cases, you might identify the major rivals, implement a strategy,
and then immediately lose everything as new frms enter your business. Entrants
might build their frms from scratch, such as the way Burger King built new
stores in the same areas as McDonalds restaurants. Alternatively, other frms
may increase the sales of products that are similar to your products. Substitute
products are related economically by the degree to which consumers are willing to
use one product instead of the other. A classic example comes from the late 1970s,
when the U.S. economy faced high infation rates and banks were subject to limits
on the interest rates they could pay on deposits. Merrill Lynch, the stock broker-
age frm, introduced a service enabling customers to store their money in a wide
variety of fnancial instruments that paid signifcantly higher interest rates than
did checking accounts, and still write checks on the account. Many larger custom-
ers took their money away from banks and put it in these asset accounts. These
new accounts were perceived as close substitutes for traditional bank services, and
people transferred huge sums of money out of the banking system.
The key point is that you need to take a broad look at your frm and the indus-
try. Know who your competitors are and how they operate. Are other products or
services offered by other industries that might attract your customers? If you make
a change in the way you do business, fnd out how it will affect your rivals. Deter-
mine how changes will alter the industry. Will they provide an opening for frms
in other industries?
Barrier
To Entry
Firm Rival
Supplier
Supplier
Supplier
Consumer
Consumer
Consumer
Consumer
Consumer
Consumer
Consumer
Decreased
Costs
Improved
Quality
Ties To
Customers
Increased
Switching Costs
Wholesale
Ties To
Suppliers
Control Of
Distribution
Innovation and
Dierentiation
Figure 10.5
Methods to gain competitive advantage. Examining the production chain highlights
several useful techniques. Barriers to entry keep out potential competitors and
substitutes. Ties to suppliers can cut costs, improve quality, and lock out competitors.
Control over distribution provides stronger markets and keeps out competitors.
Building ties to customers builds loyalty, improves products and increases margins.
Creating switching costs keeps customers loyal.
Government Regulations
In any economy, government intervention has a strong infuence on the frm.
There are myriad government agencies, regulations, taxes, and reports. The situa-
tion multiplies for multinational frms that are subject to the regulations of many
nations. These agencies and regulations can have strong effects on the proftability
of a frm. Generally, an individual frm has no control over government regula-
tions, but sometimes suggestions can lead to modifcations. For instance, it is now
possible to submit some documents to government agencies in computer form. In
fact, some reports (such as 10K or 10Q fnancial reports) are required to be fled
electronically. Electronic forms can decrease your storage costs and make it easier
to fnd documents that have been stored for long periods of time.
IS Techniques to Gain Competitive Advantage
How can you use IT to gain a competitive advantage? Where do
you begin your search? These questions are diffcult to answer. Keep in
mind that your competitors are asking the same questions every day. Competitive
advantage may be achieved with many techniques in business. Information tech-
nology is one area that may provide several opportunities. In general, MIS tech-
niques may not be better than other methods. However, some frms have experi-
enced considerable success from using these techniques, so they are well worth
considering. Moreover, the rapid changes in technology often lead to competi-
tive advantages if your frm is the frst to fnd a creative use for the new technol-
ogy. The other side of the coin is that untested new technologies may not work as
planned. Hence, the pioneer is taking a risk: If the project fails, the development
costs may put the frm at a competitive disadvantage.
The fundamental mechanisms for gaining competitive advantage are barriers
to entry, switching costs, lower production costs, product differentiation, control
over distribution channels, innovation, and quality control. These techniques are
illustrated in Figure 10.5. The question we wish to examine is how information
systems can take advantage of these techniques.
Barriers to Entry
A fundamental concept of economics is that to make extra profts, you need some
mechanism to prevent other frms from entering the industry. Otherwise, as soon
Sources of Barriers to Entry
Economies of scale (size).
Economies of scope (breadth).
Product differentiation.
Capital requirements.
Cost disadvantages (independent of size).
Distribution channel access.
Government policy.
Figure 10.6
Several methods can build barriers to entry. Be careful. Many attempts to erect
barriers are considered illegal under antitrust legislation.
Reality Bytes: People Express Airlines (Classic Case)
In 1981, Donald Burrs People Express Airlines was the darling of the airline indus-
try and American management. In four years the fedgling airline grew to a $2 billion
company. People Express was cited in In Search of Excellence as an ideal American
business because of its fat organizational structure and compensation plan that based
reward on stock growth. All employees, whether customer representatives or pilots,
were viewed as equally valuable to the company. Growth seemed to be unlimited and
the airline could not process applications or reservations fast enough. Yet on January
18, 1985, People Express Airlines declared bankruptcy. Soon thereafter, the parts of
the empire that Burr constructed were auctioned off and the routes redistributed.
The basic philosophy driving People Express was to make air travel available to
everyone. At its peak, Peoples low fares brought thousands of students, the elderly,
and the middle class through Newark, New Jersey. The waits were horrendous and
the service was chaotic. Yet the $29 fare made the hassle worth it, particularly when
the other airlines were charging fve times as much. Peoples fares allowed the car-
rier to book and fy full planes.
As long as the fights were full, the profts were easy to calculate: determine the
price of the fuel and the equipment and employee cost per fight of the plane; deter-
mine a per fight fare that would provide a proft when the expenses were subtracted;
and repeat this formula across the fight pattern. Keep the fares so low that the fight
would always be booked. By developing the demand in this new market segment,
Burr felt that he had found a formula for success that could not be broken.
This approach looked promising until American Airlines used its Sabre reserva-
tion system to implement yield pricing. Through advance ticketing and other restric-
tions, American was able to discount seats that would have gone unsold because of
Peoples low fares. The fying public now had a choice. They could continue to fy
on People Express Airlines and deal with the chaos and the crowds, or they could
make reservations and fy on American Airlines with comfort. Besides, they could
fy directly and not go through Newark. The remainder of the seats were sold at full
price to business people who could not plan far enough ahead to make advance res-
ervations.
Peoples vision was a good one. It centered on cost cutting and motivating the
workforce. Overexpansion and the lack of a marketing focus contributed to the fail-
ure of People Express. However, a third major factor was the failure to integrate
technology into solving its business problems. Before its frst plane left the ground,
People Express managers decided not to duplicate American and United Airlines
sophisticated reservation systems. Another reason People Express shied away from
technological development was that the airline lacked the internal expertise to build
or even buy a reservation system. In 1983, the carrier contracted with NCR Corpo-
ration to build a system to handle yield management. After 18 months, the project
failed. According to Burr, the failure was due to poor communication on both sides
and a lack of management attention.
Adapted from Richard Pastore, 1990, Coffee, Tea and a Sales Pitch, Computerworld,
7/3/89, p. 1. and Clinton Wilder 1989, Dont Blame the System, Computerworld,
7/3/89, p. 42.
as your frm develops a strategy that pays higher returns, other frms will fock to
the industry and drive the prices and profts down. Figure 10.6 summarizes the
common barriers to entry. One way that information systems create barriers to
entry is from their cost. Consider what happens when you create a new informa-
tion system that provides additional services, like banks did with ATMs. Custom-
ers soon expect all frms to offer those services. If a new company wishes to enter
the industry, it will have to spend additional money to buy the computers and
create the systems to offer those services. The information system raises the cost
of entering the industry. A classic example of this situation was the introduction
of People Express Airlines in 1981. The CEO of People Express stated that
he knew the airline needed a reservation system to compete effectively with other
airlines, but after raising $100 million to start the airline, top management found
it impossible to raise the additional $100 million needed to create the reservation
system.
As shown in Figure 10.7, computer systems might also be used to create more
direct barriers to entry. For instance, as a manufacturer you could build a computer
system that is tied to retail stores. The stores would use the system to place orders
and to inquire about products, warranties, and delivery schedules. You might be
able to forbid the stores from using the system to connect to any other manufac-
turers. If the stores gain advantages from the new system, they will end up placing
more orders from you, and you will keep out potential competitors. However, you
will have to be careful not to violate antitrust regulations.
Distribution Channels
Controlling distribution channels is a method of gaining competitive advan-
tage that is similar to creating barriers to entry. The Japanese economy has long
been a classic example of controlling distribution channels, although the role of
information systems is minimal. In Japan, sales relationships are developed over
long periods of time, and companies have many interrelationships and ties. In par-
ticular, distribution of products from manufacturers to retailers is controlled by
a few large companies that are loosely organized into support groups (keiretsu).
If you want to sell products in Japan, you must build a relationship with one of
these companies. American executives have often complained about the problems
they experience in dealing with these distributors, which creates a barrier to sell-
ing U.S. products in Japan. Although there is disagreement on the cause of the
problems, the ability to control distribution channels can be an effective strategy
for maintaining market share and deterring rivals. The distributors gain power
through their close personal ties to the customers. For example, in Japan, most
new automobiles are sold by salespeople who call on customers at their homes.
Information systems can be used to control distribution channels. As a manu-
facturer, you could build a computer link to the retail stores. In addition to provid-
ing faster ordering and more information, you encourage the store to order direct-
ly from your company and avoid competitors. For example, Levi Strauss, the
jeans manufacturer, has installed such a system in some retail chains. Assume that
you work for a competitor and you call on the retail store to convince the buyers
to carry your products. Probably the frst question you will be asked is whether the
store can order your jeans through the Levi Strauss computer link. If the answer
is no, the store manager is going to be less willing to buy your products.
Now, imagine the confusion that can result for the retail manager who wishes
to sell similar products from three companies. What happens if each company
has its own private computer link? Does the manager need to have three different
computer terminals and learn three different systems?
Partly because of the loss of access to distribution channels and partly because
of the confusion resulting from having multiple systems, attempts are being made
to standardize some electronic relationships. An important component of electron-
ic data interchange (EDI) is to defne standards so that managers have to work
with only one system and everyone has reasonable access to that system. If EDI
does become standardized, there will be fewer opportunities to control distribu-
tion channels with information systems. However, businesses might still be able
Gaining a Competitive Advantage
Barriers to Entry
The additional costs of creating a sophisticated information system make it
harder for frms to enter the industry. Classic case: People Express.
Distribution Channels
Control over distribution prevents others from entering the industry. Case:
iTunes or Napster.
Switching Costs
Consumers are reluctant to switch to a competitor if they have to learn a new
system or transfer data. Classic Case: Baxter Healthcare.
Lower Production Costs
Using technology to become the least-cost producer gives an advantage over
the competition. Classic case: Wal-Mart.
Product Differentiation
Technology can add new features to a product or create entirely new products
that entice consumers. Classic cases: Federal Express and Merrill Lynch.
Quality Management
Monitoring production lines and analyzing data are important aspects of
quality control. Improving quality leads to more repeat sales. Classic case:
Digital Equipment Corp.
The Value Chain
Evaluating the entire production process identifes how value is added at
each step. Combining steps or acquiring additional stages of the value chain
can lead to greater profts. Case: Qwest.
Figure 10.7
Classic cases. Several classic cases illustrate some important methods of acquiring
a competitive advantage. Understanding these cases will help you identify potential
strategies in other situations. They will also help you communicate with IS
professionals.
to gain a competitive edge by providing better, more sophisticated electronic ser-
vices through the links. For example, expert systems might be used to provide
faster responses to retailer and consumer questions.
One of the interesting aspects of the Internet is its ability to alter traditional
distribution channels. In particular, the Internet is becoming the major distribution
system for digital data such as music, books, video, software, and news. Some
traditional organizations fear this change as a loss of control. For example, in the
U.S. music industry, a handful of frms have controlled the production and dis-
tribution of most music. In 1998, the frms attempted to stop the expansion of
digital music (e.g., MP3 format), but the courts did not support this interference.
Consequently, it is now relatively easy for anyone to create music in a commer-
cial format and distribute it cheaply over the Internet. The same industry-altering
effects are occurring within the book and news industries. As the number of mo-
bile devices (readers, tablets, and large-screen phones) increases, more people are
switching to digital content away from traditional paper.
Switching Costs
An interesting strategic capability of information systems is their ability to create
switching costs for your consumers. Consider the case of a brokerage frm that
creates a system that enables you to manage your accounts with your personal
computer. You buy and sell fnancial instruments and write checks against your
account. The computer automatically tracks your portfolio, notifes you of major
changes, and automatically sweeps uninvested cash into interest-bearing assets.
At the end of the year, it prints a complete summary of your transactions for tax
purposes.
Now, what happens if another broker or bank offers you the same capabilities?
Will you switch to a new frm? You might, but it depends on what other incentives
the company offers. If everything else is the same, most people would be reluctant
to change since they incur costs to switch. For example, you would have to learn
how to use the new system. Besides, you would have to reenter your investment
data and program new reports and graphs. If you are one of the frst frms to create
a new system, the deterrence of switching costs can be a powerful tool to maintain
Reality Bytes: Control over Distribution
Sony was one of the leaders in designing and selling e-book readers; even before
Amazon. Sony has its own digital book store tied into its reader. When the Apple
iPad was released, it presented a challenge to Sony and other e-Book readers. Al-
though the display quality is lower on the iPads, it is color, and it is backlit so people
can use it at night without a reading light. Sony (like Amazon and Barnes and No-
ble), created an iPhone/iPad app that enables people to buy books directly from the
Sony bookstore. In February 2011, Apple blocked the Sony app. Apple is now insist-
ing that content must be sold through the Apple store instead of inside an application.
The reason is not technicalit involves money. Apple takes a 30 percent commis-
sion on all sales through its store.
Adapted from Yukari Iwatani Kane and Stu Woo, Apple Rejects Sony E-Book
App, The Wall Street Journal, February 1, 2011.
your market share. Figure 10.6 summarizes the tools to create competitive advan-
tages as practiced by companies in the classic cases.
Lower Production Costs
In some cases, an effective strategy is to become the lowest-cost producer. If you
can consistently sell your product for lower prices than your competitors do, you
will have an important advantage. However, consumers need to believe that your
products are as good as the competitions.
Computer systems have long been used to decrease costs. Transaction-process-
ing and accounting systems decrease administrative costs. Robots and process-
control systems can be used to control manufacturing costs. Inventory systems are
used to track parts and reduce inventory ordering and holding costs. Marketing
systems might be used to create better target marketing, with the advantages of
improved response and lower marketing costs. Financial systems that control in-
vestments and cash fow also can result in decreased costs.
Product Diferentiation and New Products
Another strategic use of information systems is the ability to create new or differ-
ent products. If you can add features to your product so that consumers believe it
is different from the competition, you will be able to make more money. A classic
case of using technology to create a new product is portrayed by Merrill Lynch.
Reality Bytes: Merrill Lynch Cash Management Account (Classic Case)
Until the 1970s, banks and other fnancial institutions were treated differently by the
government than stock brokers such as Merrill Lynch. Financial institutions could
not sell stocks, and there were limits on interest rates that could be paid to deposi-
tors. Brokerage companies focused on investments in stocks. In this environment,
Merrill Lynch created its Cash Management Account (CMA). For a minimum sum
of $25,000, investors could open a new account with Merrill Lynch. The account was
similar to a bank account. The money could be placed in risk-free government bonds
or it could be used to purchase stocks and bonds. The money could be obtained with
minimal problems, including writing checks against the account. In short, the CMA
became a bank account for medium and large investors. The primary advantage to
the CMA over traditional bank accounts was that there were no government restric-
tions on the interest rates. As commercial interest rates rose in the late 1970s and
early 1980s, huge sums of money left the banking industry and were deposited in the
CMA.
Merrill Lynch used its information system to offer additional features, such as au-
tomatic transfers between accounts, overnight repurchases and sales of government
bonds, and automatic investments and sales of stocks. All the investment options
were controlled by individual investors. Banks could not offer these services because
of governmental restrictions, and other brokerage frms did not have the information
systems. This use of information technology gave an advantage to Merrill Lynch.
While Merrill Lynch was not known for other innovations, it is one of the largest
fnancial institutions in the United States with a balance sheet comparable to Citi-
corps. In 1995, the brokerage frm had 44,000 employees and operated in 31 coun-
tries. The 1994 proft amounted to 18.6 percent return on equity.
Another classic case of using information systems to modify a product for
competitive advantage came from Federal Expressan overnight package de-
livery company. Federal Express was the frst major delivery company to track
individual packages. The service places bar codes on every package and scans
them every time the package is moved. By storing this data in a central database,
Federal Express employees can tell customers exactly where any package is lo-
cated. Besides decreasing the number of lost packages, this system provides a new
service for customers. Nervous customers can use the information to determine
when a package will be delivered. The information system tracks the current loca-
tion of each package. When the system was created it provided a unique service to
customers. To consumers, Federal Express is offering not just package delivery
but also information on the location of the package. This product differentiation
will help attract customers and might allow the company to charge higher prices.
In some cases, information systems can be used to create entirely new products
or services. For example, many banks offer sweep accounts to customers who
place large sums of money in their bank accounts. Variations exit, but the purpose
of a sweep account is to automatically place money into higher-interest-bearing
assets. For instance, you might need cash available during the day to cover any
withdrawals. But if you do not make major withdrawals at night, the bank could
lend your money to someone for overnight use. The bank needs a sophisticated
information system to keep track of which customers are participating, monitor
what limits they have imposed, and automatically transfer the money to the bor-
rowers accounts. (As a side note, you might wonder who wants to borrow money
for just one night. Of the many possibilities, two major players are governments
and large banks. Some interesting international possibilities also arise by lending
across time zones.) Customers receive more interest, borrowers have access to
more funds, and banks make money on the transaction fees and interest rate dif-
ferentials. These accounts can be provided only by investing in new information
systems.
Reality Bytes: Postal Service Death Spiral?
In many ways, the U.S. Postal Service (USPS) appears to be in a death spiral. The
number of users (mail) declines, so the company increases prices to get more rev-
enue, which drives away more customers and the process repeats. In 2010, mail vol-
ume declined 3.5 percent from the prior year. However, 170.6 billion pieces is still a
lot of mail. For the years 2008, 2009, and 2010, the volume of frst class mail fell by
4.8%, 8.6% and 6.6% respectively. Typically, frst class mail produces at least half
of the revenue. The Post Offce eliminated 105,000 jobs and reduced costs by $9
billion, but still lost $8.5 billion for the year. The agency asked Congress for permis-
sion to eliminate Saturday deliveriesessentially reducing services. The demand for
physical shipments has not disappeared, but no one knows how far or how quickly
the demand will drop.
Adapted from The Wall Street Journal, Post Offce Has $8.5 Billion Loss, Novem-
ber 12, 2010.
Quality Management
Firms can gain a competitive advantage by offering higher-quality products.
Through the 1980s, surveys indicated that owners reported fewer problems with
automobiles manufactured by Japanese frms compared with those produced by
U.S. manufacturers. This difference in quality gave the Japanese frms a com-
petitive advantage. Similarly, Motorola is one of the leading proponents of total
quality management. The company is constantly encouraging its suppliers to work
at improving quality through the entire manufacturing process.
Information systems have a role in improving quality management. For start-
ers, they can be used to collect data about quality measures. If quality measures
come directly from production machines, there can be an overwhelming amount
of data. In other cases, quality measures might be collected electronically from
your suppliers. Collecting data seems like an obvious idea, but the huge amount of
data complicates the process. In many cases, manufacturers have trouble identify-
ing the original source when a component fails. Often, just knowing which sup-
pliers cause the most problems is a useful step in quality management. This data
can also help the supplier. Failure data can be used by the supplier to pinpoint the
source of problems. Since 1992, nations in the European Union (EU) have been
requiring frms to improve quality by complying with the statements in the ISO
9001:2008 (International Organization for Standardization) directive. ISO 9001
requires companies to measure quality at all stages of production. Any frm that
wishes to sell products or parts to frms in the EU must build an information sys-
tem to monitor quality and provide information to customers.
No machine is perfect. There is always an element of error in the output. The
diffcult part is to determine which errors imply that the machine needs to be read-
justed. Decision support systems can be used to improve quality. Statistical qual-
ity control (SQC) is an important tool. Several statistical calculations and graphs
are used to determine whether fuctuations are purely random or represent major
changes that need to be corrected.
Figure 10.8
Value chain. The value chain illustrates the essential operations in a business.
Every frm has operations for purchasing, production, shipping, marketing, and
customer service. These processes are supported by the organization of the frm,
human resources management, technology development, and procurement services.
Providing services desired by customers contributes to the proft margin of the frm.
Firm Infrastructure
Technology Development
Procurement
Inbound
Logistics
Operations Outbound
Logistics
Marketing
& Sales
Service
M
a
r
g
i
n
M
a
r
g
i
n
Expert systems also can be employed to control errors and locate the source of
the problems. Consider a production line that has 50 major machines. In addition,
several hundred parts are purchased from external suppliers. The fnal product has
thousands of parts and hundreds of assembly operations. Total quality manage-
ment requires that quality be monitored at each step of the process. A typical prob-
lem facing a machine operator is that a machine might stray off the baseline and
need to be corrected. The operator faces several questions, such as: Which adjust-
ment should be made? Should we overcorrect to compensate for the prior errors?
Was the problem caused by this machine, or did earlier operations contribute? If
corrections are made now, how will they affect other machines down the line? An
experienced operator might be able to answer some of these questions. On the
other hand, an expert system might be helpful at solving the more complex prob-
lems. Digital used expert systems to improve quality and cut the cost of installing
minicomputers. Digitals weak performance in the 1990s also illustrates the dif-
fculty in maintaining a competitive advantage as the market changes.
Te Value Chain
One method of searching for areas that might provide you with strategic benefts
is to examine the entire value chain of the industry. As shown in Figure 10.8, the
key feature of a value chain is to examine each step of production and determine
how value is added at each step. If some steps show larger increases in value than
others, they will be key points to target for strategic action. The second objective
of value chain analysis is to encourage decision makers to examine the bigger pic-
ture in the industry. In many cases, a frm can beneft by expanding its operations
beyond its traditional activities. For instance, an automobile manufacturer (Ford)
might buy a car rental agency (Hertz). Now the manufacturer can control a large
consumer of its products and control the sale of the used vehicles.
Manufacturing
Engineering
and Design
Marketing
Research
Customer
Service
Management
Sales and
Order
Management
Logistics/
Supply
Suppliers
Customers
Production
Figure 10.9
Process innovation. Production consists of the processes of supply logistics,
manufacturing, and sales management. These processes are directly supported by
design, engineering, and marketing. Research and customer service support all of the
processes; top management organizes and controls the frm. Technology can provide
innovations in all of these processes.
Te Search for Innovation
How can IT support the operations of the frm to provide a com-
petitive advantage? Industry and academic leaders are constantly searching
for ways to improve organizations and gain a competitive advantage. Illustrat-
ed by Figure 10.9, one method to organize the search is to examine the primary
processes of the frm: research, engineering and design, manufacturing, logistics
and supply, marketing, sales and order management, service, and general manage-
ment. Each of these processes has its own inputs, outputs, and objectives. Analyz-
ing them in detail enables managers to spot problems and to search for innovative
opportunities.
The following sections present general ideas for each of these processes that
have generated interest and some success. Most of them use technology to im-
prove the process or to help the processes work together better. Keep in mind that
in any frm, there can be many ways of improving processes. Relying on informa-
tion technology is not always the best answer.
Just coming up with a new corporate strategy is diffcult, but it is not enough.
As indicated by Figure 10.10, an effective strategic plan must also describe the
changes in the process, identify the new data needs, and describe how the infor-
mation system will be changed to support the new strategy. Business strategy re-
Figure 10.10
Strategy requires more than a good idea. Innovative strategies often carry high
risk and potentially high rewards. Unique ideas are hard to fnd, hard to evaluate,
and diffcult to sell to top management. And success can depend on careful
implementation, management support, and careful monitoring.
Strategy idea
Convince top management
Data
Competition
Costs
Forecast
Research/Plan
Implement
and Monitor
quires more study than can be provided in this book. Several books and courses
study the various techniques for analyzing businesses and searching for new busi-
ness strategies. The key is that the executive analysis should include an awareness
of the opportunities provided by information systems. Also, the information sys-
tem strategic plan needs to adapt to the organizations strategies.
It is easier to understand the strategic possibilities of IT by looking at the spe-
cifc organization functions. In most situations, you will examine the organization
in detail, and search for ways to apply technology to operations and support func-
tions. Figure 10.11 summarizes the capabilities of IT to support innovation. All
of these technologies are discussed in this book. The difference now is that you
are specifcally searching for new ways to improve an organization. All of these
methods have been used in other organizations to support the various areas. Some
organizations use many of these tools, others only a few. Some technologies are
well-established and would hardly count as strategic in some organizations today.
But, the point of the table is to summarize the possibilities to help ensure you do
not overlook them in your search for ideas.
Research
Research in frms varies enormously depending on the industry and the overall
corporate strategy. At a minimum, most frms at least have a product development
Area Information Technology Support
Research Analysis and modeling, project management, workgroup support,
databases,decisionsupport.
Engineering and Design CAD/CAM,testing,networks,workgroupsupport.
Manufacturing Mass customization, links from customers and suppliers, robotics,
qualitymonitoring,expertsystemsformaintenance,production
databases,businessintegrationtools.
Logistics and Supply Just-in-time linkages, forecasts, models, links for design, transaction
processing.
Marketing Frequentbuyerdatabases,targetmarketandmediaanalysis,survey
design and analysis, multimedia promotion design, links between
customersanddesignteams.
Sales and Orders Portable computers for salesperson contact, expert systems for order
customization and confguration, workgroup tools for customer
support.
Service Phone support systems, location monitoring and scheduling of
servicepeople,expertsystemdiagnostics,databases.
Management Enterprise information systems, links to service providers
(accountants,consultants,etc.),e-mail,bulletinboards,decision
supportsystems,personalproductivitytools,workgroupsupport.
Figure 10.11
The search for innovation. Information technology provides many opportunities
for improving the fundamental business processes. IT is used to improve
communication, decrease costs, reduce design times, monitor customers and rivals,
and improve customer service.
team that is constantly searching for new products or improvements in existing
products. Some companies, such as 3M, DuPont, Microsoft, and Intel, spend
considerable sums of money on basic research to create entirely new products. To
these frms, strategic advantage comes from being the leader in the industry with a
constant cycle of new products.
IT support for research takes the form of computer analysis and modeling,
statistical analysis of data, project management and budgeting, and workgroup
technologies that make it easy for researchers to collaborate and share informa-
tion with each other and with managers throughout the company. Data mining,
business intelligence, and neural network tools can also be useful for exploring
and analyzing research results. For example, many experiments in chemistry and
physical sciences can be monitored and summarized by automated equipment.
Engineering and Design
Engineering and design processes are responsible for converting theoretical re-
search into new products. Engineers establish manufacturing procedures, design
new equipment, and coordinate suppliers with production. In particular, the de-
sign process must optimize the production line to minimize costs and retain high
quality.
Support for engineering and design takes the form of CAD/CAM systems that
make it easy to create, modify, store, and share new designs. If these systems are
coupled to integrated design databases, engineers can more easily reuse prior re-
sults. Tying into production databases enables the engineers to model and test var-
ious aspects of their designs. Engineers can also be supported with expert systems
that help them analyze production aspects of their designs. As General Motors
engineers design new cars, software helps them improve the layout to simplify
production and to use existing components. Engineers are also being supported by
workgroup technologies that make it easy to share designs and receive input from
teams of workers throughout the company.
Manufacturing
Four key features are critical in production: costs, speed or timing, quality, and
fexibility. Competing through lower costs and higher quality are time-honored
means of gaining a competitive advantage. It might not be suffcient today. In-
creasingly, frms are turning to mass customization in an attempt to gain market
share. Twenty or thirty years ago, the large frms in an industry were content to
build huge plants, gain economies of scale, and aim at the mass market. This ap-
proach tended to leave niches open for competing frms. The problem with this
strategy is that it allows rival frms to gain a toehold, which they might use to
build market share and eventually compete directly against your primary market.
Todays frms are trying to shift production fast enough so that they can cover vir-
tually all of the niche markets.
Mass customization requires an IT system that links the sales system directly to
the production line and through to supply. It also involves heavy use of robotics
that are confgurable directly from one computer. Other uses of IT include expert
systems for maintenance and diagnostics. Japanese frms have long been propo-
nents of preventive maintenance. If you wait until a machine breaks, it is too late.
Expert systems can be used to schedule routine maintenance and spot problems
before they cause problems. IT systems are also heavily used to monitor quality
and suggest improvements.
Logistics and Supply
The implementation of just-in-time (JIT) inventory systems is largely credited to
Japanese manufacturers. Today they are used by manufacturers worldwide. Manu-
facturers attempt to cut costs by holding minimal inventories. Instead, inventories
are maintained by the suppliers, who deliver the products to the assembly line just
as they are needed. The system can work only if the suppliers and factories are
linked electronicallyoften there is only a one- or two-hour delay between order-
ing and delivery.
Suppliers are often involved in the design phase. Their knowledge is useful in
identifying supply availability, costs, and substitutability of components. Some-
times, it is diffcult to locate suppliers for equipment. Computer networks such as
IndustryNet help frms connect with potential suppliers and identify equipment,
parts, and prices.
Marketing
A well-known application of IT to improve marketing is the use of frequent-buyer
databases that identify major customers. More traditional point-of-sale transaction
systems can be leveraged by identifying preferences and rapidly spotting patterns
or trends. At the tactical level, expert systems are used to help analyze data and
perform statistical trend analysis. Geographic information systems are being used
by leading frms to identify patterns and possibilities for new sales. Information
systems can also be used to link frms more closely to external marketing frms for
research data, communication, and development of promotional materials.
Multimedia tools are being used by leading frms to develop initial ideas for
advertising and promotional campaigns. Companies such as General Motors are
also using video tools and computer dissemination of video to link customers and
marketing departments closer to the design team.
Social networking is an increasingly important marketing channel in many
frms. If nothing else, it is important to monitor the Web for comments about your
companyboth to gain information and to help stop bad comments as quickly as
possible. Finding ways to gather, evaluate, and leverage consumer information is
critical to success in marketing.
Reality Bytes: Artists Without Labels
Information technology has made huge leaps in the past few decades and companies
have spent billions of dollars on hardware, software, and personnel. But, do compa-
nies actually gain anything from all this technology? This question is challenging to
answer because of the complexity of companies and technology. One answer comes
from the MIT Center for Digital Business that worked with McKinsey & Co to sur-
vey 330 U.S. public companies. An important conclusion of the study was that com-
panies that used data and analytical tools to make decisions (instead of guesswork)
had four percent higher productivity and six percent higher profts than the average.
Another interesting result is that many frms are not using the technology as much as
they could be.
Adapted from Andrew McAfee and Erik Brynjolfsson, What Makes a Company
Good at IT? The Wall Street Journal, April 25, 2011.
Technology Toolbox: Locating Customers with a GIS
Problem: You need to identify locations and characteristics of your customers.
Tools: A geographic information system (GIS) helps you analyze and display re-
lationships based on location. ESRIs ArcInfo and Microsofts MapPoint are two
leading products for
displaying overlays
of geographical in-
formation. ArcInfo
is relatively expen-
sive, but is widely
used for large proj-
ects. MapPoint is
not included with
Offce, but is afford-
able and demon-
stration copies are
available. GIS soft-
ware is designed to
display basic maps,
showing boundar-
ies and major infra-
structure items such
as roads and rivers. You can choose the level of detail you want to see by zooming
in or out. The critical feature of a GIS is the ability to display geo-coded data. Data
is usually displayed using shading or different circles or other icons. You can also
overlay pie or bar charts in each location, but you need to keep the number of charts
relatively low to remain readable.
To further refne your strategies, you can add your own data, such as sales, facto-
ries, or competitor locations. Advanced systems also provide optimization routines
that help select distribution routes.
One of the greatest strengths of a GIS is the ability to persuade others. Even if
you already know the general relationships, a map is a powerful tool to convince
executives. Instead of reading through pages of data, they can quickly see the spatial
relationships on the map.
Quick Quiz:
1. How often does the U.s. Census Bureau update its data?
2. Why is location an important element in business decisions?
3. How many location-based pie charts do you think could be placed on a map?
Sales and Order Management
Sales and order management are often handled simply as an operations or transac-
tion-processing area. However, in the last 10 years, several frms have used tech-
nology to gain a competitive advantage by improving the way they handle sales
and orders. Frito-Lays use of handheld computers is a classic example. The sys-
tems enable managers to more closely track their own sales, sales of competitors,
and other external factors, because salespeople can enter data immediately. For
certain industries, the concept can be extended further to installing workstations
at the customer sites that tap into your central databases. Federal Express and
Baxter Healthcare both used this technology to gain a leadership position.
Leading frms are also using expert systems to assist customers in choosing the
products and options that best match their needs. These systems assist order takers
and improve sales by matching customer needs. Expert systems are similarly used
to improve confguration and shipping.
Workgroup technologies, e-mail, and expert systems all combine to give more
power to the frontline workers dealing directly with customers. Resolving prob-
lems and meeting customer needs faster improve customer satisfaction and cut
costs.
Service
Service industries and service-based processes (like accounting, MIS, and law)
have their own problems and opportunities. Technology is used to support ser-
vices with on-site, portable computers. These systems enable workers to have
complete access to information almost anywhere in the world. Leading companies
are building specialized databases to support their service workers, such as the
answer line databases that support General Electric and Whirlpool customer
service representatives.
Systems are built that monitor locations of service personnel, enabling frms
to identify the closest workers to problems and to fne-tune schedules through-
out the day. Complex products are increasingly being sold with internal diagnos-
tic systems that automatically notify service departments. Similarly, companies
are cutting costs and reducing repair time by building expert systems to diagnose
problems.
Management
One of the more dramatic IT support mechanisms for management is an execu-
tive information system. Giving top managers better access to data allows them
to identify and correct problems faster. More sophisticated models can be built to
examine alternativesespecially to analyze the potential reactions of rivals in a
competitive situation.
Larger frms are building electronic links to their strategic partners, for in-
stance, by providing electronic access to corporate data to accounting and legal
frms. These links enable the external partners to keep a closer eye on the frm,
speeding the identifcation of problems and assisting them in spotting broad pat-
terns and opportunities.
Executives are also increasingly turning to electronic conferencing tools and
workgroup software, even e-mail. Executives can cover more areas and deal with
more people with these systems than they can by phone or through face-to-face
contact. Some studies have shown that, in traditional conversations, managers
spend as much as 50 percent of the time on personal chitchat. Electronic systems
(although they might be less personal) tend to be more effcient. On the other
hand, some companies have been restricting employee access to electronic net-
works (especially the Internet) because they waste too much time on personal
communications.
Another approach taken by management is the move toward standardization:
the effort to make all jobs similar, routine, and interchangeable. By reducing jobs
to their most basic level, they become easier to control and easier to support or
replace with information technology. Franchises make good use of this concept.
At the same time, management jobs in some companies are being reformulated as
teams of knowledge workers. In the past, managers worked on fxed tasks within
the corporate hierarchy. Today, you are more likely to be hired for your specifc
skills and knowledge. As the needs of the company change, you will work with
different teams at solving problems and creating new products and services. Per-
sonal computers and client-server technologies are often used to support these
management teams. Instead of relying on one central computing facility, each
team has its own set of resources, which are shared over networks throughout the
company.
Strategy
Skills & Resources
Required
Organizational
Requirements Risks
Diferentiation Strongmarketing.
Productengineering.
Basicresearchskills.
Distributionchannel
acceptance and
cooperation.
Internal
coordination, R&D,
production, and
marketing.
Incentivesfor
innovation.
Resourcestoattract
creative and skilled
labor.
Competitors
imitate.
Customersdonot
acceptdiferences.
Costistoohigh.
Cost Leadership Continuedcapital
investment.
Processengineering.
Continuousquality
improvement.
Tightsupervisionof
laborandcosts.
Productsdesignedfor
low-costproduction.
Low-costdistribution.
Tightcostcontrol.
Frequent,detailed
costreports.
Highlystructured
organization.
Incentivesbased
onquantitative
measures.
Competitors
imitate.
Technology
changes.
Loseproduction
or distribution
advantage.
Customer-Supplier
Links
Infuencewith
partners.
Communiation
channels.
Standardsor
agreements.
Flexibility
to respond to
customers.
Serviceculture.
Abilitytoadaptto
emergencies.
Securitythreats.
Changingstandars.
Competitorscopy
withmorelinks.
Figure 10.12
Implementing strategy can be diffcult, costly, and time consuming. Firms generally
choose one primary strategy and then build the resources and shape the organization
to best support that strategy.
Costs and Dangers of Strategies
Why is it so diffcult to convince management to make strate-
gic changes? What are the risks of strategic decisions? Strategic
uses of information systems can be seductive. There are many interesting cases
in which companies have created innovative information systems. Inventing stra-
tegic alternatives requires a considerable amount of creativity. It is easy to get
caught up in the excitement of designing new approaches and to forget about the
risks. Evaluation of any project requires weighing the risks against the potential
gains. Although it is often diffcult to measure the potential gains and risks, it
is important to consider all consequences. By their nature, strategic changes can
alter the entire course of the frm. Figure 10.12 summarizes the skills, organiza-
tional effects, and risks involved with several strategies.
Robert Morison and Kirtland Mead (A Hard Look at Strategic Systems 1989)
pointed out that it is easy to misinterpret the various classic cases regarding stra-
tegic use of technology. For example, in many cases, the true strategy does not
lie in the computer system; instead, the gains came from changing the way the
business operates. For instance, the gains experienced by American Hospital
Supply (Baxter Healthcare) came about because they improved the way their
customers (hospitals) handled supplies and inventory. The computer system facili-
tated this change but was not necessarily responsible for it. In other words, rather
than search for a killer strategic computer system, it is wiser to identify ways to
improve the overall business, then ask how technology will support that change.
High Capital Costs
One of the important considerations in strategic analysis is the cost. Strategic
changes often involve implementing new technology before any of your competi-
tors. Yet new technology tends to carry high costs. Manufacturers of technology
may not have reached economies of scale, and they might have monopoly power
over prices through patent rights. Furthermore the IS teams will have less experi-
ence with the technology, so it will take longer to implement and may result in
Money for research
Money for IT
Figure 10.13
Dangers of strategy. When developing and choosing strategies, you must always
remember that innovations can be risky and often carry high capital costs. Although
it may be exciting to spend millions of dollars on technology, it can destroy the frm
if you do not have enough resources to support research and operations.
missteps and require additional testing. For instance, Morison and Mead (1989)
report that it took six years and $350 million before American Airlines Sa-
bre travel agency reservation system started paying off. As Figure 10.13 implies,
these costs might take away money from other projects.
It can be diffcult to estimate the cost of major projects, especially when they
involve new technologies. There are many examples of MIS projects going over
budget and beyond deadlines. Moreover, strategic projects often require major
capital outlays up front, but increased revenues do not appear until much later.
A big question with new technology is trying to decide when it should be im-
plemented. There is an inherent confict. If you offer an innovative service from
the technology before your competitors, you can gain a competitive advantage.
However, if you wait, the costs will be lower. In making this decision, you will
also have to guess what action your competitors will take.
When the Competition Follows
Another diffculty with strategic systems is that much of the advantage comes
from creating a service that is not offered by your rivals. Once you introduce the
service, your rivals will watch the customer response closely. If the customers
begin to switch to your frm, your rivals will undoubtedly create a system to offer
the same services. At that point, you lose most of the competitive advantage. Even
worse, you might end up with an escalating war of technology. Although the
competition is good for the customer, the increased expenditures can cause prob-
lems for the company if the ideas do not work as well as you expected. In this as-
pect, strategy is similar to chess: you must plan for multiple moves ahead of time.
The gains to technology occur from when you frst implement the strategy to
the point that your rivals follow. For example, almost all of the major overnight
delivery services now provide the ability to track shipments. If the system is
easy to create, you may not gain much. However, it is likely that customers who
switched to your frm will stay, so you can gain a larger share of the market.
On the other hand, if your strategic ideas do not pay off, your rivals will gain,
because you will likely lose most of the money invested in the project. Some frms
use this tactic to great advantage, by using a fast-follower strategy. They allow
smaller frms to take the risk and experiment with new technologies. If the project
succeeds, the large frm steps in with more money and more clout and creates its
own, improved version. About the only risk it takes is that the smaller frm might
become successful enough to grab a serious share of the market. But, if the large
frm is big enough, it can often buy out the smaller rival and integrate the technol-
ogy with minimal risk.
Changing Industry
An intriguing problem that can arise is that even if your strategic project suc-
ceeds, the company might lose because your project has changed the industry.
Consider an insurance company that sells software to companies to allow them
to track insurance deductions and payments to workers. The insurance company
decides that it can add a program to compute payroll, so the companies could drop
their existing payroll software. These features appear to give the company an edge
over its rivals in the insurance industry. The problem is that there are many more
companies that create payroll software, and one of them is an 800-pound gorilla
(ADP). It is relatively easy for these payroll companies to add insurance capabili-
ties to their existing software. The actions of the insurance company encourage
the payroll software frms to move into the insurance market. As illustrated in
Figure 10.14, the insurance company suddenly has hundreds of new competitors
and could lose customers.
Sharing Data
One common technique in strategic systems is to share your data with customers
and suppliers. Two questions arise from this situation. First, do you really want
suppliers and customers to have access to the data? Second, can you control their
access to protect other data? Security and control issues are examined in detail
in Chapter 5. The main point to think about here is what might happen as your
customers gain access to your data. The process of protecting the rest of your data
becomes more complex. Consider the situation of a supplier to General Motors.
To save costs and improve communications, GM wants you to connect your com-
puter to the GM factory computers. GM intends to use the links to place orders,
monitor quality data, and track shipments. Are you willing to give GM access
to your computer? Can you control the information that the large corporation is
allowed to see? Maybe when checking on their orders, GM will also be able to
determine how much you are producing for other companies. Or maybe GM will
gain access to your supplier lists and raw material prices. Even if the GM manag-
ers are ethical and do not reveal this data to anyone else, you still might worry.
What happens when you have to negotiate prices with GM the next time? If the
corporation has access to your data, you might be concerned that it could infu-
ence the negotiations. Figure 10.15 illustrates the need for security systems that
will enable you to control the access to your data.
Figure 10.14
Changing industry and government intervention. A complication with strategy is that
it might alter the industry. A frm in Industry 1 might use IT to attract customers from
a different industry. Because of this expansion, the frm gains new competitors (from
Industry 2). While competition is often benefcial, you must thoroughly analyze the
effect of the new competition before embarking on changing the industry. In a related
manner, sometimes changes in government regulations alter relationships between
industries, as in the telephone and cable TV markets.
Industry 1
(expands into
industry 2)
Industry 2
(new
competitor)
Customer
Customer
Customer
Customer
Customer
Customer
Customer
Customer
New technology
New services
Same technology
Government Intervention
You have to be careful when considering strategic maneuvers. Many potential
strategies violate antitrust laws. For example, many barriers to entry are illegal,
as is price discrimination. In fact, attempts to monopolize a market are forbidden
by the Sherman Antitrust Act. Price fxing and other forms of collusion are also
outlawed. Information system agreements between competitors could be scruti-
nized by the Justice Department or the Federal Trade Commission. More recently,
the European Union has become a major player in evaluating mergers and anti-
trust complaints. Even if a company is largely based in the U.S., major sales in the
European Union can make the company subject to European rules and procedures.
If government agents choose strict interpretations of the laws, it could compli-
cate many information system issues. For instance, frms might be discouraged
from forming consortiums that defne standards. Agreements to share disaster
backup facilities might be outlawed. Computer links to customers might be seen
as limiting competition. So far, the U.S. agencies have allowed all of these ac-
tivities to proceed without interference. However, there is always the chance that
other nations or different political administrations will view the issues differently.
From the 1980s through the 2000s, the government was relatively lenient
about antitrust issues, including those regarding information systems. However,
one interesting case arose with the airline reservation systems. For many years,
American Airlines and United Airlines had the leading reservation systems.
Other airlines could list fights on the systems, but they had to pay money for each
ticket sold through the system. A confict eventually arose because the airlines
Production Firm
Data to be shared. Data to be protected.
Hackers
or competitors
rejected
External
rewall
Internal
rewall
Web
servers
Internet
router
Figure 10.15
Security complications. Improving communication and sharing data are common
themes in using technology for competitive advantage. This need to share data with
outsiders makes it more diffcult to protect your internal data. Some data you will
want to share; other data must be kept secret. Security systems can provide some
protections, but the more outsiders who are involved, the more diffcult it is to
provide adequate security.
that created the system programmed it to list their fights frst. Flights from other
airlines were often listed on separate screens, so travel agents and customers were
encouraged to choose fights from the airline that built the system. Although this
mechanism did not directly violate antitrust laws, Congress decided to pass a new
law, making the practice illegal. Lawmakers decided that as long as the other air-
lines had to pay for access to the system, everyone should have an equal chance
at being listed frst. The point is that even though the initial action was not illegal,
Congress has the ability to pass new laws and remove the advantages, so you can-
not assume that the benefts will last.
Te Role of Economics
Why did so many dot-com frms fail? Do their failures mean
there is no viable Internet strategy? The main lesson from the failures
of the early dot-com frms in 2000 and 2001 is that no matter what anyone tries
to tell you, to succeed in business, you must make a proft on operations. The
second lesson is that it takes time to acquire loyal customerslonger if you want
to change the world. Many of the early e-commerce managers felt that to become
the dominant player, they had to be the frst and biggest frm. So their primary
strategy was to sell products below cost and spend huge amounts of money on
national advertising. The advertising was successful at attracting investors, whose
cash kept the frms alive for a year or so. But when the sales failed to generate
profts, there was no way to keep the companies running.
The advertising strategy also created an interesting domino effect in the ear-
ly industry. By pushing the importance of name recognition (and a good domain
name), many of the early frms were able to survive by attracting advertising mon-
ey from other frms. For example, Excite was a leading Web-portal frm. With its
easy name and relatively popular search engine, many people used the site on a
regular basis. Based on the number of people visiting the site (known as eyeballs),
Excite was able to sell advertising space on its pages to other Web frms. Over 80
percent of Excites revenue came from advertising. As the other frms in the indus-
try fell apart, they stopped their advertising spending and Excites revenue plum-
meted. Many other frms faced the same problem, and the chain reaction caused
hundreds of frms to fail.
Clearly, Web sales and Web advertising are here to stay. Many frms make mon-
ey from sales of products, services, and advertising. Yet, many frms still face dif-
Reality Bytes: Borders Battles for Life
Borders Group is the second largest bookstore chain in America, behind Barnes and
Noble; and not counting Amazon. The switch from print books to digital caught Bor-
ders fat-footed. The company was late to the digital marketplace. Borders purchased
a stake in Kobo Inc., a Canadian e-book retailer, and launched an online bookstore
in mid-2010. Albert Greco, a researcher of the book industry estimated that in 2009
Borders had a 10 percent share of the retail book market. Borders said that it hoped
to control about 17 percent of the e-book market by mid-2011. In early 2011, Bor-
ders began closing many of its retail stores.
Adapted from Jeffrey A. Trachtenberg, Borders Group Launches e-Book Store,
The Wall Street Journal, July 7, 2010.
fcult decisions about how to draw those lines. The newspaper industry is one that
is facing diffcult questions. In the early days, many magazines and newspapers
attempted to charge for access to their sites. Gradually, most of them dropped sub-
scriptions and relied on advertising revenue as well as sales of traditional paper-
based versions. A notable exception was The Wall Street Journal which focused
on creating quality content for a feeeven online. But, as print subscriptions
dwindled, newspapers began creating pay walls where people had to subscribe
to see content online. The situation is an interesting strategic problemwhich
could mean life or death for many companies. The challenge is that to sell online
advertising, the sites need to attract a large number of unique visitors. But, charg-
ing for content has the potential to drive away readers who will be inclined to fnd
free resources. So, a newspaper would lose money from advertising and not gain
much money through subscriptions. Yet, if print subscriptions and print advertis-
ing continue to dwindle, how can a newspaper cover its costs?
Television and radio stations have been able to survive on the basis of advertis-
ing revenue. Some Web sites (notably Google), became wealthy through advertis-
ing. But, is there room for thousands or millions of sites to succeed based solely
on advertising?
The strategic options in e-business are increasing as new tools are created,
wireless capabilities improve, and people begin to adopt connectivity as a way of
life. At this point, there is no single answer, which makes it even more important
that you carefully defne your goals, analyze the profts, evaluate your competi-
tion, and build a creative business plan.
Challenges of Strategy
How do you convince an organization to change strategies? This
question is probably the most diffcult question any upper-level manager will face.
Even when you join an organization in an entry-level management position, you
will quickly see the challenges presented to new ideas. Most people and almost
all organizations resist change. Organizations develop a method of operating that
works. It might not be the perfect answer, but it was developed over many years
and hardened through experience and competition. The method must have been
reasonably successful or the company would not have survived. But, is it the right
strategy for the future? And, if you think you have a new strategy, how will you
convince people to adopt it. Keep in mind that new strategies cost money and
an organization usually faces multiple options for strategies. How can you prove
your ideas have merit and will work? The challenge is even higher if the risk of
failure is a large loss or a loss in market share.
Information technology solutions can be even harder to sell to top manage-
ment. Few managers truly understand technology. Most have experienced prob-
lems with technology failuresparticularly leading edge technologies. Most large
companies have encountered the delays and frustrations of trying to develop and
implement new technologies. Managers have become cautious about being the
frst organization to try new technologies. Managers also know that the price of
new technology is often high and tends to decline rapidly. Being the frst person to
adopt new hardware usually means that you are paying more money than anyone
else. Consequently, many people choose to wait and see which technologies will
succeed and wait for the price to drop to levels that reduce the risk.
Technology Toolbox: Analyzing Businesses
Problem: You are presented with a business problem to solve.
Tools: You need an analytical methodology to evaluate the problem, focus on the
causes, and identify appropriate solutions. If the situation is based on business pro-
cesses, you can use the data fow diagram techniques to show how the frm is orga-
nized and how data is supposed to move. Because it is a graphical approach, it is
good for spotting the cause of problems and for communicating relationships to oth-
ers. You can also adapt database and object-oriented tools to help you understand de-
sign details for data-intensive problems. Ultimately, every manager needs to develop
a methodology for approaching business problems and cases. When you encounter a
new situation, where do you start?
Foundation Business Plan Expectations
Solve the right problem
Choose the right tools
Divide the system
Make a decision
Considertheconsequences
Detail the implementation
Problem description
Statefactsandproblems.
Identify most important problems
andcauses.
Plan
Describethenewsystem.
Detailhowtoimplementtheplan.
Provideacontingencyplan.
Advantages
Show how your plan will solve the
problems.
List additional advantages and
strategicefects.
Measurable goals
Financial implications
Efect on human resources
Strategic efects
Critical success factors
Potential risks
The important frst step is to solve the right problem. Most businesses encounter
multiple problems. Often, you see only the symptoms, and you need to trace back
to the cause. This work requires detective and diagnostic skillswhich are usually
acquired through experience with cases and real-world problems. Once you have
identifed the root problems, you have to choose the correct tools. There is an old
saying: When all you have is a hammer, everything looks like a nail. Make sure you
examine all of the available tools before pounding away.
A good starting point for developing a structure is the old debating approach:
problem description, plan, and advantages. Use the problem description to identi-
fy the most important problems and the cause of the problems. This section must
clearly communicate the need for a change. The plan itself must be detailed and as
specifc as possible. You should include implementation issues, costs, and alterna-
tives or contingency plans. Be sure to include a section that explains how your plan
is going to solve the problems you identifed.
Your analysis needs to identify any measurable goals and fnancial implications.
If the goal is to reduce costs or save time, quantify those numbers. Consider the ef-
fects on human resources. Will the technology reduce the number of employees?
Will you retrain employees, or need to hire more within the MIS department? What
are the critical success factors for the frm and the project? What are potential risks
of the project? Such as cost, alienation of employees or customres. Identifying these
risks up front makes it easier to fnd ways to mitigate them later.
Quick Quiz:
1. Why is practice so important in learning to diagnose business problems? Where
will you get this practice?
2. Where do you place the expectation elements in the business plan?
3. How is the problem description different for a business case compared with an
actual business problem?
Sustainability
Remember that to be successful a strategy needs to create a barrier to entry. If you
spend large amounts of money on new technology to gain market share, you need
to be able to sustain an increase in profts for a long enough time to recover your
costs. As shown in Figure 10.16, if you spend the money and your rivals imme-
diately follow, everyone ends up with higher costs and no additional profts. This
argument has been made many times over the years as a reason to avoid strategies
involving technology. It can be diffcult to determine if the argument is legitimate
or if it is merely an excuse to avoid change.
Nicholas Carr explored these ideas in an article in the Harvard Business Re-
view in April 2003, claiming that IT no longer has a strategic beneft. The thrust
of the argument is that technology prices have declined so far that IT no longer
provides a barrier to entry. Any frm can almost instantly copy anything you do.
Any short-term advantage you might gain from technology is not sustainable. In
fact, most information technologies today can be purchased as services from vari-
ous frms. Since the computing hardware and software are readily available to
any organization, it will be diffcult or impossible to use it to gain a competitive
advantage.
Many aspects of this argument are true. Rarely is the information technology
the source of a competitive advantage. Instead, the benefts arise from how an or-
ganization applies the technology to solve problems and improve the management
and decision making of the frm. The use and application of technology depends
heavily on the people and management of the organization. As a weak analogy,
you could give guitars to 1,000 people, but only a few of them would ever become
$
time
Technology cost
$
time
Technology cost
Additional prots
Additional prots
Technology creates
barriers or switching costs
so you gain prots.
Anyone can copy your
strategy and competitors
atract customers.
Figure 10.16
Sustainability. If technology does not enable you to keep customers and profts, rivals
will follow and you will not earn enough money to pay for the technology cost.
profcient musicians and only a couple would be able to form a successful band.
Hardware and software are not enough to provide a competitive advantage. You
need an organization that has the knowledge and willingness to change and use
the new technologies to alter the way it operates.
Examine the classic cases in detail and you will see that the gains arose from
the way the technology was applied. American Hospital Supply/Baxter Healthcare
used the technology to provide better information and lock in customers. Merrill-
Lynch and FedEx used the technology to create new products and services. Amer-
ican Airlines used its giant databases to provide knowledge to make better pricing
decisions. Even if you can copy the hardware and most of the software, can your
organization achieve the same benefts?
Consider groupware as a relatively new technology. Any organization can in-
stall SharePoint servers at minimal cost in terms of hardware and software. Yet,
even today, ten years after the introduction of the technology, how many have
done so? How many are effectively using the technology? And, how many have
altered their operations to take advantage of the communication and teamwork
features to reduce costs, fnd new solutions to problems, and add fexibility to
the organization? The technology is defnitely important, but ultimately it is the
management and workers in the frm that must adopt the technology to alter the
operations and gain the full benefts.
Leader or Follower
Another often-debated aspect of strategy is the importance of the frst-mover ad-
vantage. The concept comes from game theory, which is often used to study ri-
valry. In some situations, the organization that adopts a strategy or technology frst
will be seen as the leader and will acquire customers who will then be reluctant to
switch back. Even after a competitor adopts the same technology, the customers
will stay with the frst frm. The frst mover might also gain a reputation for being
innovative, which might attract additional customers. Ultimately, the frst-mover
Reality Bytes: USA Today: What is a Paper?
USA Today was one of the frst national newspaperscertainly the frst to target
general news, and to use color. At one point, it had the largest newspaper circulation
in America. By 2010, the company hit serious problems. The biggest two: Circula-
tion was plummeting and advertising revenue had collapsed. On fnally recognizing
the shift in the way people obtain news, the paper began downsizing, eliminating 130
jobs (9 percent of its employees), and shifting its focus away from the printed edi-
tion. The goal was to create an organization that could deliver news through the Web
site and react quickly to changing preferences. Over time, USA Today had become
stale. The environment changed to online, instant access to news, entertainment, and
sports; but USA Today had made few changes. The company was fnally beginning
to understand the need for change. But, the task is arduous. Craig A. Huber, an ana-
lyst at research frm Access 342 noted that I dont blame them for tinkering with
the model, but everybody has been trying to do that. They have to restructure. They
dont have a choice in this quickly evolving media landscape.
Adapted from Jeremy W. Peters, USA Today to Remake Itself to Stress Digital Op-
erations, The New York Times, August 27, 2010.
advantage depends on the customers and the nature of the innovation and the in-
dustry. If the technology increases the switching costs, then the value of being the
frst frm to adopt a technology is higher. With each strategic idea, you will face
arguments over the potential strength of being the frst adopter.
The issue of being a leader or a follower is critical with information technol-
ogybecause the costs will drop over time. As shown in Figure 10.17, if you
choose to wait, you lose the frst-mover advantage, but costs decrease the longer
you wait. Additionally, any technology strategy carries risks. By waiting, you can
let some other frm take the risk, while you wait to see if customers actually care
about the new services or options available. Remember that at any point in time,
you have a choice of several technologies. It is diffcult to guess which technology
will succeed and become a standard.
Many times you face this same decision as a consumer. Look at television and
cell phone technologies. If you adopted HDTV, 3D, or 4G cell phones early then
you paid a relatively high price for the technology. Plus, content and access were
limited because most people had not yet switched. If you can manage to wait a
year or two, prices inevitably drop (or weak technologies disappear), and you can
save money by not being the leading adopter. Fortunately, consumers do not have
to worry about making money or beating other consumers with these technolo-
gies, so the decision is quite a bit easier for them than it is for businesses.
time
$
Technology Costs
Leader Follower
Leader Follower
Lock in customers
Reputation
Lock in technologies
Experience
Lead for next step
Reduced risk standards
Lower cost
Supplier experience
Figure 10.17
Leader or follower. Technology costs decline over time, so leaders pay a higher
price for technology. In good situations, the leader gets to lock in customers, build
a reputation, and gain experience. It is also primed for the next step. The follower
reduces risk by picking the technologies that survive and paying a lower cost for
hardware and experienced suppliers.
Large frms can be reluctant to be leaders. Their size makes it diffcult to change
operating procedures, and tends to add bureaucracy that delays making decisions.
It is often easier to watch small competitors and let them adopt a technology frst.
Even if a small competitor is successful, the risk is minimal because the num-
ber of customers affected will be small. Once a technology is seen as useful and
the prices drop, the big frm can use its size to purchase the technology, fx any
perceived problems, and generate a marketing campaign to sell the new services
to the rest of the world. Because it is diffcult to patent the use of information
technology, this approach is generally legal; and most customers will never know
about the original small frms that tested the technologies. In some extreme cases,
large companies have spun off smaller frms to test experimental technologies.
In other words, if you cannot fnd a smaller frm to do the testing, you can create
your own small frm to test a technology. Companies with many small stores, such
as franchises, often adopt a similar approach by selecting one or two stores to test
new products and services on a small scale.
Choosing to be a leader or follower in information technology is one of the big-
gest strategic decisions an organization must make. The choice does not have to
be all-or-nothing. It can be applied differently to various divisions or even types
of technologies. Also, the decision needs to be revisited periodically to correct for
changes in technology, the business environment, and the competition.
Change Agents
Organizations need rules and methodologies to function. Basic day-to-day deci-
sions need to be handled at the customer and operations level. Without rules and
procedures, all decisions would flter back to the top levels of the organization and
it would take forever to get anything done. But the drawback to rules and proce-
dures is that everyone is reluctant to change them. The larger the change, the more
diffcult it will be to get it accepted and implemented. More importantly, organiza-
tional change does not happen by itself. Change agents are people or events that
cause an organization to reevaluate its rules and procedures. When a person is the
change agent, that person often becomes the champion for the new ideas. How-
ever, change agents are rarely appointed and tend to gain prominence by random
chance. In theory, you could appoint someone to be a change agentcontinually
searching for new ideas and technologies and bringing them to the attention of ev-
eryone else. But, it would likely be a frustrating job, and the person would likely
get blamed for a lot of problems. On the other hand, some people like to try new
products and new ideas, and other people naturally turn to them for advice and
recommendations. These people are often natural change agentssimply by pro-
viding information and experience to others.
Economic events can be natural change agents. Many of the classic cases were
driven by economic changes. These changes opened the way for new business op-
portunities and made it possible for frms to expand or move into new areas. For
example, Merrill-Lynch altered the entire world of banking by using technology
to offer a new type of account to customers. But, this account was wildly adopted
because of the unprecedented high interest rates and the government restrictions
imposed on commercial banks. Merrill-Lynch had the technology and the man-
agement insight to take advantage of these changes. Similarly, American Airlines
was driven to adopt yield pricing by the emergence of People Express as a new
competitorbecause of the relaxation of government regulation in the airline in-
dustry. American Airlines had the technology and data in place for several years,
but felt no need to adopt it until pressured by this external change agent.
As shown in Figure 10.18, most large organizations tend to resist change, caus-
ing them to adopt ideas and technologies in lumps. This approach might be useful
because it minimizes day-to-day disruptions. However, it tends to make the frm
a follower instead of a leader, and the changes that are eventually implemented
might be more disruptive than necessary. Yet, few frms have the fexibility to
make continuous changes. On the other hand, continual evaluation and adjustment
are hallmarks of managing for quality. In these organizations, change and inno-
vation are encouraged and rewarded. Each approach has its strengths and weak-
nesses, and again, the choice of how to identify and adopt change is a strategic
management decision. Just be sure that you consciously adopt the strategy, and
not fall into a choice by inaction.
Cloud Computing
Can cloud computing provide strategic advantages? In many
ways, cloud computing is an indicator that many aspects of information technol-
ogy have limited power as a strategic tool. One goal of cloud computing is to
reduce fxed costs and to make it easier for more frms to implement the same
technologies. When even small companies can implement the same technologies
as large companies, it is going to be diffcult to use it for a competitive advantage.
This statement does not mean that technology is unnecessary. On one level, it
means that frms have to at least experiment with every possible technology to see
how it might create a competitive advantage. On a broader level, frms can still
time
time
Change agent intervention
Change agent can be person
or event (economic) that
convices organization to
change its policies.
Continuous change where
everyone is encouraged to
contribute new ideas making
many small changes over time.
Figure 10.18
Change agent. Most organizations resist change and alter procedures in jumps when a
change agent forces them to. Some organizations encourage employees to contribute
new ideas and adopt change regularly, leading to continuous improvement through
small changes.
gain some benefts to technology if the workers and managers are better at apply-
ing it to solve problems in unique or more effcient methods.
These concepts echo many of the fundamental ideas of strategy. Simply adopt-
ing a specifc technology rarely provides a competitive advantage. Instead, frms
need to fnd a concept or new way of doing business. Technology provides the
means to create these new methods or concepts. And it requires management cre-
ativity to fnd and effciently apply these solutions.
On the other hand, powerful technology available at low costs and with a global
reach enables frms to experiment and innovate at low costs. Firms that emphasize
creativity and design could use the cloud computing technologies to continually
adapt and lead the way with new products and services. In particular, services can
be tailored to ever-changing regions. Yes, other frms could follow, but by then
the leader will have moved on to something new. In a sense, creativity becomes a
competitive tool that can be amplifed with information technology.
Summary
Information systems can provide benefts beyond traditional cost saving. Com-
petitive advantages can be gained by creating barriers to entry and gaining control
over distribution channels. Using information systems to build ties to suppliers
and customers can provide lower costs and better quality products. Computer sys-
tems also provide incentives for customers to remain with your company if they
incur costs of learning new systems and transferring data when switching to a
competitor. Information systems can also be used to differentiate your products
from the others in the marketplace. Similarly, innovative services offered with the
help of technology can entice customers and expand your market.
You can search for competitive advantages by examining Porters external
forces of rivals, customers, suppliers, substitute products, and new entrants. You
can also search for strategies in research, engineering, and design. In manufactur-
ing, you can look for ways to decrease costs and improve logistics. In marketing,
potential gains can be found in better understanding of customer wants, as well as
sales and order management. Services can be supported through better informa-
tion fows and workgroup products. Management can be helped with better data
and better decision tools. The government also strongly affects organizations and
strategy, but fnding IT interactions with the government is relatively diffcult.
Strategic systems face many risks. They tend to be expensive and diffcult to
create. Any gains created may disappear when competitors pick up the technol-
ogy and imitate your offerings. In addition, making strategic changes to your frm
might alter the industry, which might adversely affect your frm. And if these
problems are not enough to discourage you, remember that attempts to monopo-
lize a market are illegal, so you have to make sure that your plans do not violate
governmental regulations.
Key Words
Web Site References
Management Consulting
Accenture www.accenture.com
Bain & Company www.bain.com
Boston Consulting Group www.bcg.com
Booz Allen & Hamilton www.bah.com
Deloitte Consulting www.deloitte.com
McKinsey & Company www.mckinsey.com
Review Questions
1. If an industry has four large frms versus a thousand small frms, is it easier
or harder to create strategies?
2. Why is strategy so dependent on external agents?
3. What are the risks and costs of strategic implementations?
4. For a large manufacturing frm, who are the customers? How many different
types of customers can there be?
5. Why are barriers to entry important to gain a competitive advantage?
6. How does control over distribution channels give a frm a competitive
advantage?
A Managers View
With increased competition, every manager is being asked to identify ways
to improve the company and fnd an advantage over the rivals. Gaining a
competitive edge is not easy. Examining the entire value chain is a useful
place to start. Information systems can provide data and evaluate models to
help you identify strategic applications. Information systems can also pro-
vide direct advantages by creating a barrier to entry, gaining control over
distribution, cutting costs, improving quality, and improving ties between
suppliers and customers. Choosing an effective strategy is a critical task
in e-business. Creating a plan and successfully implementing it are critical
steps in strategy.
antitrust laws
barriers to entry
change agents
distribution channels
external agents
frst mover
Five Forces model
mass customization
product differentiation
rivals
statistical quality control (SQC)
switching costs
total quality management (TQM)
value chain
7. How can information systems be used to gain control over distribution
channels?
8. What are switching costs, and how can they give a company a competitive
advantage?
9. How can information systems be used to enhance product differentiation and
create new products?
10. What role is played by information systems in improving quality
management?
11. Within management, how can information technology provide a competitive
advantage?
12. How do antitrust laws make the role of IT more important in fnding a
strategy?
13. Is it possible for a cloud computing application to lead to a strategic
advantage?
14. What are change agents and what is their role in the adoption of new
technology?
15. How could social networks be used as a strategic tool by a manufacturer?
Exercises
1. Choose one large frm and read the annual report (www.sec.gov). Identify
the primary strategic directions and which of the main techniques it uses to
gain a competitive advantage.
2. How long can frms maintain an advantage using an information system?
Research one of the classic cases and fnd out how long it took for the
competitors to implement a similar information system (for example,
Merrill Lynch and its Cash Management Account, American Airlines
and the Sabre System, Levi-Strauss and its Levi-Link ordering system, or
Federal Express and its tracking system). Find out when the system was
implemented, identify the competitors, and fnd out when they implemented
similar systems. Did the original company continue to update its strategy?
Collect data on sales and profts for the frms to see whether there were major
changes.
3. Pick an industry. Find two frms in the industryone a technology leader, the
other a follower. Get the fnancial information on those frms for the last fve
years. Find analyst summaries of their operations. Compare the two frms.
Are there differences in fnances, operating methods, or customers?
4. Choose an area of management, such as marketing, manufacturing, or
logistics. Find an example of a frm that you believe is doing a good job
applying IT in this area. Briefy explain how this usage could provide
strategic benefts to the frm.
5. Choose one of the IS techniques to gain a competitive advantage. Identify
a frm (not one of the examples in the chapter) that is using that method.
Briefy describe the fnancial position of the frm and how it is using
information systems.
6. For each of the following frms, identify the primary strategy of the frm
(such as cost, quality, customer links, and so on).
a. Wal-Mart
b. Toyota
c. Apple
d. Kohler
e. Yelp
f. Pfzer
g. Starbucks
h. General Motors
i. Red Bull
7. Has the Internet changed the distribution of songs? In particular, is it easier
or harder for small labels and independent groups to get their music heard
and sold to consumers?
8. Examine the market for tablet computers starting with the Apple iPad.
Identify how the various companies introduced new features and how the
others responded. What did Apple gain by being the frst mover?
9. How could an exercise gym use information technology to gain a competitive
advantage?
Technology Toolbox
10. Using a GIS tool, compare the sales for Rolling Thunder Bicycles against
population and income.
11. Find at least two Web sites or news reports that use Google maps to highlight
geographic information.
12. Create a map with Microsoft MapPoint or Google maps and insert markers
for the headquarters of at least fve companies.
13. Choose a company case at the end of one of the chapters in this book.
Perform an initial business analysis of the company. Focus on identifying
the major problems faced by the company and the cause of those problems.
Identify the primary level of the problem (operations, tactics, or strategies).
14. Select a company in the Fortune 500 (or Global 1000). Get some basic
background information on the company and read its latest annual report.
Using Porters framework, identify its primary rivals and the level of rivalry.
If they exist, identify major customers and suppliers and discuss the sales
methods used by the organization. Discuss any potential competitors or
substitute products that might arise in the near future. Describe any barriers
to entry that might keep them out.
Teamwork
15. Each team member should read through at least two industry cases in the
chapters of this book. Identify whether the frm is a leader or a follower in
terms of strategy and technology. Compare each frms fnancial data to that
of the industry (for example, by sales and number of employees). Combine
the individual analyses and summarize them. Identify any patterns you
might see. For example, do the larger frms tend to be leaders or followers in
technology?
16. Choose a frm that provides reasonable amounts of management information
(such as a local frm or a well-documented public frm). Have each team
member choose one area (research, engineering, marketing, and so on).
Identify the strengths of the frm in the area. Create a short plan to improve
the companys use of IT within that area. Make sure the usage fts with the
overall strategy of the company.
17. Choose an industry. Assign each team member to investigate a level within
the production chain. Each person should identify the tasks that occur at
the specifed level along with the major frms. Identify the rivalry and any
dominant frms at each level. Identify the use of IT at each level and any ties
across levels. Combine the results and briefy discuss where on the chain you
would prefer to enter as a new frm.
18. Choose a large frm (perhaps an automobile manufacturer). Use the annual
report to identify the fnancial condition of the frm. Use the management
letter to identify the basic strategies of the frm. Identify the frms primary
competitors. If possible (try searching MIS magazine Web sites), identify the
technology level used in the company.
19. Research the rivalry and strategies being pursued by Amazon and Barnes
and Noble. Look at market share and service plans and pricing. What are the
short-run and long-run strategies of each company? Is one frm stronger than
the other?
20. Each person should fnd a company that has used innovation (creativity)
successfully. Identify any use of information technology in the process.
Write a short paragraph about the case. Combine the information from all
team members and vote on which company made the best use of information
technology.
21. Identify the competition in the industry. Who are existing rivals? Who are
potential rivals? Be sure to defne the industry carefully. Consider using
North American Industrial Classifcation System (NAICS) codes.
22. Perform a value chain analysis of the company. Could they improve profts
by expanding vertically or horizontally? Are there additional products we
should consider offering?
23. The management has the opportunity to purchase a chain of retail bicycle
stores. Evaluate the strategic aspects of this proposed acquisition. What
will be the effect on the information systems? Can the existing information
system be used to improve the operations of the retail stores? What additions
and improvements would be needed?
24. How can the company use social networks to increase ties to customers?
25. Examine the value chain in the bicycle industry. How many levels are there
and which levels are the most proftable?
26. A key element in the bicycle industry is that sales are driven by the
popularity of the sport, which in turn is infuenced by demographics.
Consequently, sales can decline over time or suddenly expand for particular
bicycle models. How can a manufacturer like Rolling Thunder Bicycles
strategically deal with these issues?
27. Use Porters Five Forces model to examine Rolling Thunder Bicycles
relationships and long-term strategic potential.
Additional Reading
Applegate, Lynda M., Robert D. Austin, and F. Warren McFarlan, Corporate
Information Strategy and Management, 2007, McGraw-Hill/Irwin: New
York. [Business strategy in information technology from a Harvard
perspective.]
Melymuka, Kathleen. State Street Banks change in direction required a new
IT approach, Computerworld, February 15, 1999. [Changing strategy with
information technology.]
Morison, Robert E. and Kirtland C. Mead, A Hard Look at Strategic Systems,
Indications, January 1989. [Myths and issues in IT strategy.]
Porter, Michael E. Competitive Advantage: Creating and Sustaining Superior
Performance, New York: Free Press, 1985. [Early discussion of strategy and
competitive advantage.]
Porter, Michael E. Strategy and the Internet, Harvard Business Review, March
2001, Vol. 79 Issue 3.
Strassmann, Paul A. The Squandered Computer: Evaluating the Business
Alignment of Information Technologies, 1997, New Canaan, CT: Information
Economics Press. [A detailed, but controversial discussion of the value of
information technology, with some excellent cases.]
Cases: Te Airline Industry
Te Industry
To understand the state of the airline industry, you have to look at some of the
history. The most important historical issue is that the airline industry was heavily
regulated until 1978. Until then, the civil aeronautics board (CAB) controlled all
of the business aspects of the airlines, including landing rights and fares. Several
large airlines were established and grew during that time. In 1978, the CAB was
disbanded and the airlines were free to select routes and set prices as they chose,
without government oversight. However, the FAA still controls several safety is-
sues, including work rules for fight crews, the number and spacing of fights into
airports, and national fight lanes, as well as several other details.
So, what happened when fights were deregulated? This is a good time for you
to start thinking like an airline executive. Were you making money? Did your pas-
senger mix suddenly change? Are your revenue and costs any different initially?
The simple answer is that initially, very little changed. As an established company,
with procedures and a system that works and makes money, you would not be in-
clined to change anything.
0
5
10
15
20
25
30
35
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
B
i
l
l
i
o
n

$
Annual Revenue
American Airlines
Delta Airline
Southwest
Sabre
Travelport Ltd
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
R
a
t
i
o
American Airlines
Delta Airline
Southwest
Sabre
Travelport Ltd
Deregulation and Competition
But, all it took was a little time. A few airlines started offering new fights on high-
demand routes. Then, Donald Burr founded People Express, the airline that scared
every other business manager, starting in 1981. As a no-frills airline (it charged
$3 to check luggage and sold snacks on the plane), the airline charged incredibly
low prices to fy tourists. You no longer had to be in the upper class or funded by
a business to fy. Based out of the unused international terminal in Newark, New
Jersey, the company picked up non-union crews, negotiated with the FAA to fy
with only a pilot and copilot (no fight engineer), and charged fares of a paltry $59
to Chicago or $99 to Londona tiny fraction of the fares charged by the tradi-
tional airlines.
With their substantially higher cost structures, the mainstream airlines could
not compete. So, why did People Express disappear and some of the other airlines
survive? Part of the answer is mismanagementPeople Express tried to grow too
rapidly and did not have a solid management structure in place to handle the prob-
lems. But, a big part of the answer is that American Airlines fought back with
information technology. The company quickly realized that People Express ap-
pealed to tourists who were very price sensitive. Business people stayed away
from the discounter and wanted the amenities and reliable service provided by
the mainstream airlines. So, American Airlines turned to its reservation system
(Sabre) and estimated the number of people on every fight who were business
people. It then came up with the differential pricing system (yield management)
that is still in use today. Business people, those who book a couple of days before
the fight and do not want to stay Saturday night, get charged thousands of dollars.
Tourists, who book early, are offered low fares that compete with the discounters.
The basic premise is that the large airlines want to charge higher prices and fy
business people, but there is no point in fying empty seats, so they discount the
seats until they are all sold. Today, it is likely that almost every passenger on a
fight has paid a different price (McCartney 2003).
Growth of the Discounters
People Express did not have the information systems and could not compete.
The big airlines had wonbut only for a while. They did learn to adopt the hub-
and-spoke system pioneered by People Express. Small feeder fights brought pas-
sengers to a few major airports, where passengers made connections to the next
hub. The cost advantage of the hubs is that at the end of the day, all the planes and
fight crews returned to a hub city. Consequently, maintenance facilities are cen-
tralized into a few key airports instead of scattered around the country. Plus, the
key airports handle multiple fights each day, simplifying routing decisions.
But, the story is not over. Southwest Airlines was also an upstartbeginning in
Texas in 1971. By fying within the state, the airline was not subject to CAB regu-
lations. Founded and largely run by Herbert D. Kelleher, the airline grew slowly
to maintain control. Part of the simplicity lay in its fee structurefar simpler than
the major airlines, Southwest prices fights so that they are proftable. By sup-
porting several daily fights to each destination, with reasonable fares, Southwest
promotes itself as the company plane. By holding down costs (fight attendants
often help clean the airplane, and the company fies only Boeing 737s to minimize
maintenance costs), the company grew steadily. More important, Southwest Air-
lines has been proftable every single year for over 30 yearsa statement that no
other airline can match (www.southwest.com).
Recent Changes
To understand the current state of the airline industry, you also have to look at
September 11, 2001. The event shocked the American people. The industry and
government response provided almost as much of a shock to the industry. Sud-
denly, fying became considerably more complicatedpassengers and airlines
wrestled with long security lines and new rules. The U.S. government bailed out
most of the airlines by providing billions of dollars in grants and loan guarantees.
Yet the major airlines were still saddled with strict working rules and high wag-
es through negotiated contracts. In December 2002, United Airlines fled for re-
organization under Chapter 11 of the bankruptcy code. In 2004, the company was
still struggling to obtain fnancing to emerge from bankruptcy (www.pd-ual.com).
On September 15, 2005 both Delta and Northwest fled for bankruptcy protec-
tion. A large reason for fling for bankruptcy appeared to be a negotiation strategy
aimed at reducing labor contract costs. The airlines also declared their pension
plans bankrupt and foisted them onto the federal pension insurance program, re-
moving the liability and reducing the payouts to retired pilots and employees.
Several other airlines were created in the early 2000s to pick up passengers on
highly traveled routes. JetBlue, founded in 2003, is one of the largest. JetBlue
is based out of JFK airport in New York City and initially focused on East Coast
destinations. It has expanded to cross-country fights and is slowly adding some
Western cities to its routes. It has been successful (passengers like the DirectTV
broadcasts available on the seat-back sets) and proftable (www.jetblue.com).
Airtran, a successful start-up from 1998, has been hammering Deltas fares in the
Atlanta airportDeltas home base (www.airtran.com). While still only carrying
a fraction of the passengers of the big airlines (maybe 10 percent), JetBlue and
AirTran have been growing (The Wall Street Journal 2002). The secret to success,
as explained by Joe Leonard, the CEO of AirTran, to a congressional committee
is not to be a low-fare airline, but to be a highly effcient airline, whether you are
a big airline or small, and to adapt to the changing marketplace (Leonard 2004).
In 2011, Southwest purchased AirTran.
The obvious question at this point is that since Southwest has been so success-
ful with its strategy, why dont the airlines simply copy their formula? Yes, 9/11
and the resulting gasoline price increases due to the war with Iraq have put added
pressure on the airlines. But, the big airlines were struggling even before these
events. The answer is that it might not be possible for the big airlines to make the
huge changes needed. Delta and United started some experiments. Delta created
a subsidiary called Song in 2003 designed to compete directly with the discount-
ersfocusing on leisure travel between high-demand cities (www.fysong.com).
United formed Ted (a part of United) in early 2004 for the same purpose (www.
fyted.com). Although the fare rules for Song and Ted are usually simpler than for
the parent company, the interesting part is that it is sometimes cheaper to fy the
parent airline. In 2004, the CEO of Delta, Leo Mullins, resigned, and the interim
CEO suspended any expansion of the Song airline. And the Ted airline was sub-
sumed back into United.
Airline Costs
If airlines are going to cut costs, it helps to know where those costs are. The Air
Transport Association (ATA) tracks these costs and releases summary data. The
easiest way to understand the costs is to examine them as a percentage of the to-
tal. By far, labor costs are the largest share (38.4 percent). All other costs, mostly
administrative overhead, are second (23.7 percent). Fuel costs vary over time, but
most airlines use futures markets to hedge their costs (11.6 percent). The cost of
the planes seems reasonable (10.2 percent), but you also have to add in the interest
costs of the debt for the planes. Since interest rates were at historical lows in 2003,
the cost was relatively small (3.0 percent). Maintenance material (2.5 percent) is
usually steady, unless an airline is running a large percentage of older planes. Pas-
senger food costs have been declining as airlines drop their food programs (2.3
percent). Airlines still pay some commissions to travel agents, but that number is
dropping steadily (2.2 percent). Airlines have to pay landing fees to airports to use
their gates (2.1 percent). Insurance costs (1.5 percent), communication (1.5 per-
cent), and advertising (1.0 percent) are all relatively small components (McCart-
ney 2002). These industry averages conceal the differences between the airlines.
Through cheaper labor contracts, newer planes that need less maintenance and use
less fuel, and a focus on costs, the discount airlines average a cost of 7.3 cents per
mile, compared with 11.7 cents for traditional airlines (Rosato 2004).
Te Possible Future
A large part of the differences with Song and Ted are the union contracts. The ma-
jor airlines and the fight crews need to fnd some way to negotiate contracts that
enable the carriers to compete with the discounters. The airlines also need to fnd
a way to reduce costs and provide a service that people are willing to pay for. The
good news for passengers is that airlines should be more willing to try innovations
and search for ways to provide better service.
For business travelers, some new options are on the near horizon. Several com-
panies are building small, personal jets that will cost less than $1 million. Some
companies will purchase the jets for company use, but the annual upkeep and
crew costs still make them expensive. On the other hand, they might enable a type
of air taxi service to be created. Flying from small regional airports, the jets could
offer quick service directly from one city to another, bypassing the big airline
hubs and their driving, parking, and security hassles. This service will probably
be too expensive for leisure travelers, but would appeal to a business person who
is currently being charged $2,000 for a major airline fight. That would leave the
big airlines running a type of low-cost mass-transit bus service (McCartney 2003).
The micro-jet option never emerged as a major player. The technology still ex-
ists, but high fuel costs probably make it too expensive to fy a small number of
passengers.
Additional Reading
Leonard, Joe, U.S. Government Subsidizes Ineffciency in Airline Industry,
Aviation Week & Space Technology, June 21, 2004, Vol 160(25), p. 86.
McCartney, Scott, Which Costs Airlines More: Fuel, Labor or (Ugh) Meals?
The Wall Street Journal, November 6, 2002.
McCartney, Scott, Airline Visionary Burr Sees Future in Personal Travel, The
Rosato, Donna, The Plane Truth About Flying Cheap, Money, May 2004, Vol
33(5), pp. 83-86.
The Wall Street Journal, The New Normal: Big Carriers Lose Market Share to
Upstarts, May 14, 2002.
Case: Delta Air Lines
Delta (ticker: DAL) is one of the original airlinestracing its history back to its
frst passenger fight in 1929. The company, headquartered in Atlanta, had rev-
enues of $13 billion in 2003a steady decline from 2000. With over 60,000 em-
ployees, the company few about 100 million passengers in 2003. The company,
and its code-share partners, fy to almost 500 cities in 86 nations (www.delta.
com). Deltas books showed a net proft in 2000, but the company has been in the
red since that time. Delta has reported a total loss of $3.7 billion in 2001, 2002,
and 2003. For most quarters, the company is losing money even on operations
(annual report). Fred Reid, president of Delta knows the problems: Theres no
question whatsoever in Deltas mind that we are in the midst of a profound, fun-
damental and truly irreversible series of changes in which the value proposition
to the investors and the customers is changed unalterably. In this new world some
airlines are winning and some airlines are losing, and I would say that Delta is still
positioning for long-term success (Bond March 2004).
In April 2007, Delta emerged after spending 19 months in bankruptcy. The air-
line reduced capacity by about 10 percent and 20 percent fewer employees, as well
as reduced operating expenses by about $3 billion a year. The rest of the industry
also reduced fights so Delta was fying at 84 percent capacity on average fights
compared to 72 percent when it fled bankruptcy. Fares also edged up, improv-
ing revenues (Isidore 2007). The airline also started expanding into international
marketsassuming there was less competition on international routes, enabling it
to charge higher prices. The company still lost $6.2 billion on operations in 2006
Pilots and Costs
As of 2004, Deltas pilots were the highest paid in the industry. After acrimonious
negotiations, then-CEO Leo Mullins granted the union huge pay raises. In 2004,
Leo Mullins was removed as CEO. Given the bankruptcies at the other airlines,
Delta then attempted to renegotiate all salaries. The company executives have dis-
cussed the possibility of fling bankruptcy to force the union to negotiate lower
salaries (Bond May 2004). Many of Deltas older pilots are abandoning ship be-
fore it sinks. In mid-2004, 288 pilots applied for retirement, 266 of them were for
early retirement. An additional 300 had already applied for retirement in Septem-
ber 2003. Despite the retirements, Delta has plenty of pilots available: 8400 on ac-
tive duty, plus 1,060 on furlough (Aviation Week & Space Technology June 2004).
Deltas total costs of fying domestically are 30 percent higher than Southwests
(McCartney October 2002).
Subsidiaries
To fght the discount airlines, Delta established or purchased several subsidiary
airlines. Delta Express and Skywest are two regional subsidiaries that operate out
of Delta hubs. These airlines fy smaller planes or regional jets and deliver pas-
sengers to the hub, providing access to Delta in smaller cities. A key arrangement
with the pilots union allows the regional airline pilots to be paid less and fy under
less limiting rulesas long as the planes carry a limited number of passengers.
In 2003, Delta launched service on Song. Although the goal was to keep fares
relatively low, Song was competing more with JetBlue than with Southwest. In
particular, the airline offered seat-back television for satellite programs, pay-per-
view movies, and video games. However, the company had trouble getting the
systems installed and had to begin operations without the technology for sever-
al months. Another drawback faced by Song was that they used the same Delta
personnel, with the same contracts and lack of focus on cost cutting (Melymu-
ka 2003). When Leo Mullins was replaced by Gerald Grinstein as CEO in 2004,
Grinstein halted expansion of the Song division.
In 2008, Delta bought Northwest Airlines and merged the two companies
(Crawley 2008). It took several years to merge the names and information systems
of the two companies.
Information technology has an impressive role in the history of airlines. At the
pure transaction level, IT is necessary to handle the millions of reservations a day.
Running from huge databases, the large systems fnd fights, check seats, and bill
customers when seats are reserved. When American Airlines used its system to
create yield management and differentiate prices, it represented a milestone in
the competitive use of technology. But today, the reservations systems have been
moved out to subsidiaries and passengers can book online or through third-party
reservation systems. Although the systems are still used by some of the major
airlines to jockey prices, companies are increasingly looking at simpler fare struc-
tures. In 2004, Qantas, the Australian airline, decided it was pointless to run its
own system. CIO Fiona Balfour, said I dont think airlines get a competitive ad-
vantage form IT anymore. They get competitive advantage from how they use it.
Running IT at a low unit cost becomes a competitive advantage (Kontzer 2004).
Nonetheless, Delta (and most of the other major airlines) continues to run its own
systems. Deltas CIO, Curtis Robb, believes that outsourcing will not necessarily
reduce costs or improve service. He observes that we get payback on a yearly
basis for in-house IT work and we get it at a fraction of the cost (McDougall
2003). Note that the in-house reservation systems are separate from the aggrega-
tors including Sabre and Orbitz.
Away from the reservation system, Delta is working harder to use the transac-
tions to increase the operational effciency. For example, the company runs the
Delta Nervous System (DNS) that pushes data from the transaction system out to
the desktop and even handheld computers. For example, as a plane arrives at an
airport, the gate agent and luggage handler are notifed electronically. A frst step
in creating the DNS was to set common defnitions for data stored in over 70 dif-
ferent databases and place every item into one of 15 subject categories. This sys-
tem forms the foundation of all the real-time applications. For example, the gate
information display simply queries the DNS data layer for real-time information
about the specifc fight, formats it, and presents it to the customers. New applica-
tions can be built relatively quicklyprogrammers need only identify the data
they require and can focus on the application and display. At Delta hubs, the DNS
data is supplemented with data from the federal air traffc control (ATC) system.
This data is examined for late fights and enables managers to spot problems and
reschedule fights. DNS is a $1.8 billion project started in 1998 (Air Transport
World 2004.
The company also installed customer relationship management software for
the call center. The CRM system enables clerks to see the entire picture of the
customer who called (McDougall 2003). In the process, the company consoli-
dated 30 customer databases into one, with three data warehouses to support an-
alytical functions (Gareiss 2003). The newer jets offer additional ways to save
money through integrating with information technology. The Boeing 777 collects
information from various subsystems electronically. This telemetry is collected
and transmitted to the maintenance engineers. If a problem arises, such as exces-
sive vibration in an engine, managers can spot the problem and correct it before
it causes additional problems. The company is also working to create electronic
documents for all of the maintenance procedures. Electronic versions can be pro-
vided instantly to maintenance workerssaving search time (McDougall 2003b).
Like other airlines, Delta has installed kiosks to encourage departing passen-
gers to record their own check-in, saving money by reducing the number of tick-
eting agents. Delta has also created a telephone system that customers can use to
check in. More importantly, the IT system automatically monitors late fights and
missed connections. If a passenger is not going to make a connection, the system
automatically rebooks the fight (Schwartz 2004).
In 2006, Delta launched a three-year project to update its primary systems with
a service-oriented architecture (SOA). One goal was to replace BEA Systems
Tuxedo middleware with standards-based technology. Middleware takes data
from one system and converts it into a format that can be used by other systems.
By standardizing the data interchange, it will be easier to replace various systems
in the future (Havenstein 2006).
Delta was able to use the bankruptcy fling as a lever to renegotiate contracts,
including a 27 percent drop in software fees from $3.74 million to 2.73 million
charged by SAP. Delta uses SAP software for ordering supplies and tracking spare
parts (McDougall 2006a)
Bankruptcy is not a panaceait can cause more problems than it solves. In
October 2005, Deltas pricing system crashed for 12 hours, costing the compa-
ny more than $4 million in revenue. IT outages also led to several fight delays.
Delta reported that it experienced unprecedented attrition and chaos within the
IT department as workers left in droves following the bankruptcy flingonly
39 percent offered a job decided to stay. In August 2006, Delta outsourced its
operations to IBM under a seven-year contract (McDougall 2006b). With a short-
age of workers, Delta turned to other providers to help with systems, but bank-
ruptcy made it more diffcult to pay the bills. Unisys asked the bankruptcy court
to pay it $37,000 for developing an application to assign a unique product code
to the special handling required for transporting human remains. According to the
work order, Deltas system did not distinguish between human remains and other
priority shipments, leading to service failures and irate customers (McDougall
January 2007). Cisco also had to fle a claim to get paid for $20 million in leased
equipment (McDougall March 2007). If you were a CIO at a major corporation,
you would probably see similar claims and have to deal with these standard is-
sues. The interesting aspect of bankruptcy flings is that the information becomes
public.
On a more personal level of technology, airlines have been adding Wi-Fi ser-
vice to the fights so that passengers can surf the Web aboard planes. The airlines
purchase the service from a third-party such as GoGo (Lawson 2011).
Questions
1. Why do people fy on the discount airlines? What do they not like about the
discounters? Can Delta combine these answers with IT to regain market share
and profts?
2. How does Delta use technology to reduce costs? Is it enough to make a
difference?
3. Can Delta use IT to become more like Southwest? Is that the best strategy?
Additional Reading
Air Transport World, Technology Leadership, February 2004, Vol 41(2), p. 32.
Aviation Week & Space Technology, Delta Pilots in New Retirement Application
Spike, June 14, 2004, Vol 160(24), p. 17.
Bond, David, Work in Progress, Aviation Week & Space Technology, March 1,
2004, Vol 160(9), p. 45.
Bond, David, Bankruptcys Siren Song, Aviation Week & Space Technology,
May 17, 2004, Vol 160(20), p. 24.
Crawley, John, Delta Buys Northwest to Create Biggest Airline, Reuters,
October 30, 2008.
Gareiss, Robin, Technology Takes To the Air, Information Week, April 21,
2003.
Havenstein, Heather, Delta Set to Launch Three-Year SOA Project,
Isidore, Chris, Delta Exits Bankruptcy with Planes Full, CNNMoney.com, April
30, 2007.
Kontzer, Tony, Airline Taps IBM for Flexible Pricing Deal, Information Week,
May 24, 2004.
Lawson, Stephen, Aircell to Boost In-Flight Wi-Fi Speed, Computerworld,
March 9, 2011.
McCartney, Scott, Southwest Sets Standard on Costs for Airlines, The Wall
Street Journal, October 9, 2002.
McDougall, Paul, Tough Climb, Information Week, April 21, 2003.
McDougall, Paul, Prevention: The Cure for High Maintenance Costs,
McDougall, Paul, SAP Agrees to a 27% Reduction in Software Fees Charged to
Bankrupt Delta Air Lines, Information Week, December 16, 2006.
McDougall, Paul, Techie Exodus at Bankrupt Delta Caused Flight Delays,
Broader Chaos, Information Week, December 19, 2006.
McDougall, Paul, Unisys Demands Cash for Helping Delta Air Lines Ship Dead
Bodies, Information Week, January 5, 2007.
McDougall, Paul, Cisco Claims Its Owed $20 Million from Delta Air Lines,
Information Week, March 8, 2007.
Melymuka, Kathleen, Deltas Test Pilot: IT at Song, Deltas New Low-Cost
Airline Unit, Computerworld, August 18, 2003.
Schwartz, Adele C., On the Wings of IT, Air Transport World, March 2004, Vol
41(3), pp. S7-9.
Case: Southwest Airlines
One of Herb Kellehers favorite stories is that back when they started fying to
West Texas, the competitors tried to fght back by matching his fares. Herbs an-
swer was to start offering a bottle of whiskey to every Southwest passenger. With-
in a few weeks, Southwest (ticker: LUV) became the largest liquor distributor in
West Texas. Knowing the customers and not being afraid to take chances (or play
jokes) became hallmarks at Southwest. At one point in 2002, domestic traffc at
the nations fve largest airlines fell by 10 percent. At the same time, travel on the
fve biggest discount airlines increased by 11 percent. In 2002, the discounters
accounted for 20 percent of U.S. passenger traffc (Trottman 2002). Within Cali-
fornia, Southwest few a whopping 63 percent of the passengers (Leonhardt and
Maynard 2002).
To traditional airline executives, the amazing aspect of this change is that even
businesspeople are fying on the discount airlines. Southwest pushed this concept
hard for many yearsemphasizing that a manager could get on the plane at the
spur of the moment and fy to a business meeting at a reasonable fare. One execu-
tive, Brent Harris, managing director of Pacifc Investment Management, notes
that he now fies JetBlue for business instead of one of the big carriers for trans-
continental fights. He observes that While I could afford to pay more, theres a
certain sense of satisfaction, in getting the discounts (Trottman 2002). Southwest
began adding transcontinental fights (from Baltimore) in 2002 (The Wall Street
Journal 2002).
Southwest is an incredibly effcient airline, turning a plane around in as little as
20 minutes. Other airlines average 50 minutes. As a result, Southwest is able to fy
its Boeing 737s an average of nine hours a day, compared with six hours for Delta
and United (Trottman 2002). In general, the major airlines would have to cut 29
percent of their costs ($18.6 billion) to match the cost structure of Southwest (Mc-
Cartney 2002). In 2002, Southwest had become the nations sixth largest airline.
Not all is rosy in the Southwest family. In 2004, Herb Kelleher had to step in to
negotiate a new contact with the fight attendants union. Normally, negotiations
are friendly at the employee-centered company, but these took two years to settle.
In the end, the workers gained an average 31 percent raise spread over six years
(The Wall Street Journal 2004). In July, James F. Parker resigned as CEO.
In 2005, Southwest CEO Gary Kelly commented on the changing competition
in the airline industry, I think all carriers are low-fare carriers. To survive, all
carriers are going to have to be at least lower-cost carriers. In this new strategic
environment, Southwest still wants to be the low-fare airline, and Mr. Kelly wants
to use technology to reduce costs. One step was to shift more reservations onto the
Web site, where costs run $0.50 to $1 versus $10 or more for other methods. By
focusing on airport operations, including customer service, the airline has been
able to reduce costs by hiring fewer employees. In 2000, the airline employed 95
people per aircraft owned. By 2005, the number was down to 73 or 74 (Steinert-
Threlkeld 2005).
In 2007, Southwest announced that it would slow its expansion, including a
reduction in the number of new planes it will add. The airline also reduced some
roundtrip fights in underperforming markets. The company was also planning to
announce a new boarding/seating method at the end of 2007 (press release June
27, 2007). Ultimately, Southwest continues to expand. In 2011, Southwest took a
big step by purchasing AirTran. The company expects the merger process to take a
year, but AirTran dramatically expands Southwests reach on the East coast (press
release May 2, 2011).
One major change for Southwest was the increasing costs of fuel. According to
the 2010 Annual Report, in 2005, when fuel was $1.13 a gallon, fuel costs were
21.4 percent of operating expenses. In 2010 at $2.51 a gallon, they were 32.6
percent of operating expenses. Industry-wide, the market crash of 2008 and 2009
also resulted in a drop in demand for fights, so all airlines struggled during the
recession.
Even in the early days, Southwest managed to avoid many of the hassles of res-
ervation systems. For years, the company was able to list its fights for free on
the industry reservation systems. It required travel agents to call the company
to book the fightthe systems carried listings for free but charged when fights
were booked. One estimate places the cost of booking through the Sabre system
at $12 to $14 per transaction (Hoffman 2002). Today, Southwest has benefted
considerably from the Internet. Customers can see all fares quickly and make their
own reservations. Southwest gets over 40 percent of its bookings online, com-
pared wit Uniteds 5 percent (Trottman 2002). The Internet also makes air fares
more transparent. Even businesspeople can quickly see the difference in the prices
across airlines. Using the online tools, travelers can arrange fights to avoid the
huge fares that airlines are trying to hoist onto business fiers. David Weiner, cor-
porate travel manager for DaimlerChrysler, explains that although the company
has preferred discounts with Northwest, most of his managers would rather fy
Southwest. He claims that with companies like Southwest, their fares are afford-
able as is. Its not like you would require any additional discounts (Leonhardt
and Maynard 2002).
Southwest also pioneered business reservations on its Web site (www.swabiz.
com). In addition to letting managers and workers book fights, the system tracks
companywide data for travel managers. Kaiser Permanente, a major health care
organization, spends $6 million on Southwest travel. The business site enables
Margy Skinner, the travel manager, to identify all tickets and use them to obtain
quarterly bulk rebates (Kontzer 2004).
The Web site is so critical to Southwest that it rebuilt the system in 2002. The
new UNIX-based system uses servers from Fujitsu in a cluster to ensure that
the system keeps running and can scale up. Steve Taylor, director of interactive
marketing observes that the bedrock of customer service in the online world is
always being there. Thats what were aiming for with this clustering project
(Greenemeier 2002). The online marketing team consists of 70 employees, with
50 of them IT professionals and the rest marketing specialists.
Southwest does use information technology to help analyze its data. Since it
does collect a large share of the transaction data directly, it has purchased business
intelligence tools to help analyze it. In particular, the company is using Hyperions
Essbase OLAP application and budgeting software to make fnancial forecasts.
Mike Van de Ven, vice president of fnancial planning and analysis at Southwest
notes that after the September 11 attacks, the company knew it faced huge uncer-
tainty and needed help. He adds that we were asked to give some sort of fnancial
insight for a variety of decisions the company might make (Songini 2002). Be-
fore installing the $1 million software, managers relied on writing custom data-
base queries and evaluating the data in spreadsheets. Essbase provides immediate
analysis and charts, cutting analysis time to as little as two minutes. Managers use
the system to evaluate best and worst case scenarios to determine how to respond
to problems. The software paid for itself within the frst year by providing more
accurate forecasts and saving time for managers.
Unlike the other big airlines, Southwest is growingthat means it needs to hire
more workers. Growth is good for both the company and the employees, but how
can Southwest deal with the 200,000 resumes a year that it receives? Southwest
is turning to Deploy software to handle the major tasks of hiring. The software
tracks everything from job requests by departments, to application progress, and
candidate matching and ranking. The tool also integrates with the U.S. govern-
ment system for verifying job candidate fngerprints and validating drug-test re-
sults (Hayes 2003).
Radio systems in airlines are ancient. Even if the planes and radios are new,
they use technology developed in World War II. As one of the few airlines making
money, Southwest is one of the frst to install a new digital technology (VDLM2)
to transmit data between its dispatchers and aircraft captains. Because the tech-
nology is restricted to a portion of the VHF bandwidth allocated to airlines, it
provides a data rate of 31.5 K bits/second. But that rate is 15 times the rate of the
companys old system. The system primarily handles short text messagesoften
instructions on the best route around weather problems. Ultimately, the system
will be used to transmit high-resolution weather radar images to the cockpit. The
company is also planning to collect telemetry information on the planes perfor-
mance and route it to maintenance personnel at the next airport if problems are
detected (Brewin 2004).
With high prices for fuel, airlines began focusing on methods to reduce fight
distances and times. In 2007, several airlines, including Southwest, began imple-
menting fight-planning software from Jeppesen. Jeppesen has long been known
as a provider of fight maps and related technologies. In early trials, the system
was able to reduce fight times on hundreds of fights from 4 to 7 minutes. A cou-
ple of minutes might not sound exciting, but do it enough times and an airline
saves millions of dollars. EDS also sells fight-planning software to more than 40
airlines worldwide (Weiss 2007).
Because Southwest focuses on costs and proftability, it also tries to hold down
spending on technology. Rarely is it the frst adopter of a new technologyin-
stead, it fnds ways to get passengers from point A to point B as simply as pos-
sible. For instance, the fights do not assign seats and it was one of the last to print
paper boarding passes. By 2005, the CEO reported that IT costs had fat-lined to
about $170 million per year, for equipment and wages. The company also applies
its knowledge of operations to IT development. Mr. Kelly, the CEO noted that
we have a single data architecture. We have a standard testing approach, where
we will rarely allow shortcuts (Steinert-Threlkeld 2005).
In 2010, Southwest announced plans to implement a new reservations sys-
tem (2010 Annual Report). The main goal was to be able to handle international
fights, either with Southwest planes, or with a code share arrangement through
another airline. Prior to this point, Southwest focused only on domestic fights
and the reservation system was designed only for those fights. Yet, as Southwest
grows, it feels it needs to be able to compete with the larger airlines who continu-
ally emphasize their international reach. In 2011, Southwest altered its frequent
fyer program to be based on a point system instead of the number of trips. The
points are based on dollar value of the tickets, presumably to reward business
travelers who pay more money for fights than tourists. The new program required
a complete rewrite of the software and Web site to support the changes.
In 2010, Southwest fnished integrating the SAP ERP system which replaced
the old general ledger, accounts payable, accounts receivable, payroll, benefts,
cash management, and fxed asset management systems (2010 Annual Report).
The new systems are integrated, providing managers with a comprehensive view
of the accounting data.
Questions
1. How does Southwest use information technology to establish fares?
2. How has Southwest used the Internet to gain at least a short-term competitive
advantage?
3. Can Southwest maintain its competitive strengths as it continues to grow?
What risks will the company face?
Additional Reading
Brewin, Bob,Data Takes Flight, Computerworld, June 21, 2004.
Greenemeier, Larry, IT Pros Plus Marketing Experts Equal Better Service,
Information Week, March 18, 2002.
Hayes, Mary, Southwests People Plan, Information Week, September 8, 2003.
Hoffman, Thomas, IT Investments Grounded At USAir, Other Carriers,
Kontzer, Tony, Low-Cost Airlines Build Portals For Business, Information
Leonhardt, David and Micheline Maynard, Troubled Airlines Face Reality:
Those Cheap Fares Have a Price, The New York Times, August 18, 2002.
McCartney, Scott, Southwest Sets Standard on Costs for Airlines, The Wall
Songini, Marc L., Southwest Expands Business Tools Role, Computerworld,
July 12, 2002.
Steinert-Threlkeld, Tom, Southwest Airlines: High Tech, Low Costs, Baseline,
April 10, 2005.
Trottman, Melanie and Scott McCartney, The Age of Wal-Mart Airlines
Crunches the Biggest Carriers, The Wall Street Journal, June 18, 2002.
Trottman, Melanie and Scott McCartney, Southwests CEO Abruptly Quits a
Draining Job, The Wall Street Journal, July 16, 2004.
The Wall Street Journal, The New Normal: Big Carriers Lose Market Share to
Upstarts, May 14, 2002.
The Wall Street Journal, Business and Finance, July 2, 2004.
Case: Sabre
Sabre (ticker: TSG) is the monster real-time reservation system originally created
by American Airlines. The frst system went live in 1960 and handled 84,000 tele-
phone calls a day. In 1964, the system ran on its own private networkreducing
American Airlines staff by 30 percent in the frst year. In 1976, the system was in-
stalled into travel agent offcesquickly reaching 130 locations. In 1985, preced-
ing the Internet adoption by several years, easySabre gave dial-up access to users
with personal computers. In 2000, the company was spun off as a separate entity,
but tracking-stock data is available back to 1996. In 2004, the system handled $70
billion of travel products and connected 53,000 travel agents. It also forms the
foundation of the Travelocity Web site (www.sabretravelnetwork.com).
Technology
For years, the Sabre system ran on large IBM computers. Much of it was joint-
ly developed with IBM, since the demands and technologies needed continually
pushed the available hardware and software. The system handles 15,000 transac-
tions per second and tracks 79 million air fares. In 2001, the company announced
it would migrate the massive system to a completely new architecturebased on
a UNIX platform (Anthes May 2004), but the transfer would take several years.
Throughout its history, Sabre has been a leading-edge system, handling huge
transaction volumes and providing detailed data for analysis. Almost all of the
code has been custom written. In 2000, Sabre produced Release 8 of its AirFlite
Proft Manager. The modeling and forecasting package is used by airlines to es-
timate demand for seats on every fight. It is the core of the yield management
system. Release 8 had about 500,000 lines of code. The problem is that it was four
months late because fnal system testing turned up 300 bugs. The frst customer
found 26 more bugs in the frst three days, and additional joint testing turned up
another 200 defects. Sabre and its development team were embarrassed. However,
the situation mirrored other development projects. The catch is that Sabre has 62
software products with 13 million lines of code. It cannot afford defect rates that
high. For Release 10, shipped in December 2002, Sabre turned to extreme pro-
gramming (XP). With XP, programmers work in pairs, but more important, they
defne testing procedures for each module before writing the code. The fnal ver-
sion turned up only 100 defects after 16 months of use. At the same time, pro-
grammer productivity increased. The reduced defect rate also cut the number of
support programmers needed (Anthes March 2004).
Partly from the increased competition, partly from the need to support new
standards, Sabre moved in 2005 to replace its 15-year-old EDI system with Web
services applications. Web services uses standards such as HTTP and XML to ex-
change data more easily with remote systems. The standards reduce the cost and
the use of the Internet reduces the transmission costs and make it easier to develop
applications that work together. Transferring data from airline systems onto Sa-
bres system and ensuring timeliness and accuracy represent substantial costs for
airlines (Havenstein 205).
Competition
Despite the technological advances and prowess of Sabre, it faces competition.
Its biggest competition is undoubtedly the Internet. Although the company runs
Travelocity, one of the big Internet travel sites, the site is not proftable yet. Fur-
thermore, when customers book fights through Travelocity, the airline or hotel
pays a fee to Sabre. Consequently, airlines have been encouraging customers to
book their fights directly at the airline Web sitesbypassing Sabre entirely. The
Orbitz Web site was created by the fve largest airlines specifcally for that pur-
pose. It searches the companys individual databases and routes the customers
choice directly to the airlines server.
In 2004, Alaska Airlines, tired of waiting for new services from Sabre, decided
to create its own itinerary-planning and fare-searching system on a Linux-based
system. Alaska simply purchased the system from Cambridge, Massachusetts ITA
software. Steve Javris, vice president of e-commerce at Alaska, noted that we
couldnt wait on Sabre. ITAs algorithms are widely regarded as the best in the
industry (Verton 2004). Alaska will continue to use Sabre to book reservations,
but will use the new system for data analysis. The ITA system is primarily used as
the back-end processor for Web siteshelping customers identify routes. Jeremy
Wertheimer, ITAs founder and CEO, comments that it processes and confrms
availability for [trip] pricing in less than one-tenth of a second (Verton 2004).
But, Sabre is not standing still. The company is creating new tools to encour-
age airlines to continue using the system, and new products that will help travel
agents. In terms of helping the airlines, Sabre launched its interline e-ticketing
(IET) hub in 2004. Since Sabre serves multiple airlines, it is in an ideal position to
provide links between them. Previously, customers had trouble booking e-tickets
for fights that involved multiple airlines. Each system was separate and the air-
lines had to reidentify the passenger at each step. With the interline system, the
passenger data and validation are shared across airlines. Essentially, Sabre serves
as an EDI consolidator and translates data from each system into a common for-
mat that is accessible to all systems. The system is based on Web services, making
it easy to expand and change as airline systems change (Rosencrance 2004).
Sabre also introduced a new feature for travel agents in 2004. What happens if
a customer wants an aisle seat, but the only seat available is a middle seat? A good
travel agent would book the available seat to keep it, and then periodically check
the fight to see if a more desirable seat opens up. But that requires considerable
time and effort by the travel agent. Sabres answer was to create an event model
that will alert the travel agent when a seat opens up. Loren Brown, CIO of Carlson
Wagonlit Travel, a company with 8,000 agents, notes that agencies would likely
pay extra for that feature alone, stating that would be a much more elegant solu-
tion than we have in place now (Kontzer June 2004).
Sabre also faces competition from more traditional rivals: Amadeus Global
Travel Distribution, Cendants Galileo International, and Worldspan. All of these
were originally developed by other airlines and spun off as well. All of them are
facing similar problems and working to cut costs and offer new services. That
is one of the main reasons driving the switch in servers. Sabre estimates that by
running the open-source MySQL database on open-source Linux servers, the new
system will cost 80 percent less to operate (Kontzer June 2004).
Several startups offer new ways to search for fights, hotels, and cars. In 2006,
ITA Software received $100 million in venture capital funding. Its system drops
processing costs to a couple of dollars a ticket, compared to the $12 a ticket that
airlines pay to Sabre for each reservation. ITA CEO Jeremy Wertheimer observed
that almost every ticket you buy is still being handled by assembler code running
on a mainframe. His systems use inexpensive x86 PC hardware running open
source software (Kontzer 2006). Many airlines pass that cost onto customers, so
savvy customers use the search systems (such as Sabres Travelocity) to locate
the lowest-cost fights, and then book the fight directly with the airlineso Sabre
gets paid only for the listing, not for the purchase.
Low-cost airlines have been particularly reluctant to book fights with Sabre.
Southwest relied on its own system and Jet Blue dropped the Sabre listings when
more customers began booking directly with the airline. AirTran bucked the trend
in 2005 when it added all of its fights to the Sabre system (Kontzer 2005). By
2007, facing increased competition for low-fare fights, Southwest added its list-
ings to Sabre and to the Galileo search system (Wall Street Journal 2007).
Experts have questioned how long the big reservation systems will compete
with the startups. In 2005, Sabre took one step by purchasing European company
Lastminute.com for slightly over $1 billion. The site primarily gives Sabre ac-
cess to the European market (Information Week 2005). Competitor Expedia
had already purchased Hotels.com and Hotwire.com. Orbitz owns the Web site
CheapTickets.
In 2008, Sabre added a social network component to its system (Havenstein
2008). The objective was to enable people to share advice and recommendations
about their travel plans. It is not clear if anyone uses the systemparticularly
since anyone could use existing social networks instead.
Te Future
With so many competitive factors, it is diffcult to guess who is going to win the
battles for customer reservations. Sabre and its direct rivals have some amazing
technology as well as the developers to build complex systems. One of the things
that might be changing is that passengers will want to book package deals. But
not quite like the old heres a tourtake it or leave it. Instead, customers will
want to go to a site and select their own custom bundle: pick low-cost air fares,
choose hotels that offer the desired amenities at an acceptable price, then add in
some adventure excursions. Yes, the big travel sites support these steps in a lim-
ited way now. However, they rarely allow anything more than limited customiza-
tion. Ultimately, customers want a more intelligent system that can create desired
bundlesat a discount. So, airlines, hotels, and other providers will need to co-
operate. Then, the reservation sites will have to become sophisticated enough to
balance the various choices and compute all of the package deals (Kontzer May
2004).
Sabre Holdings Corp. was purchased by private equity frms in March 2007 for
$4.5 billion, so it is no longer a public company (Fox News 2006). Operations are
likely to continue as before. Private-equity frms went on buying sprees in 2006
and 2007, searching for companies with relatively low stock prices, that the inves-
tors believed could be used to generate solid cash fows.
In 2010 and 2011, the reservation industry faced several major challenges.
First, the recession affected bookings so revenue was down. More importantly, the
airlines began rebelling against the fees charged by the big sites (e.g., Travelocity
and Orbitz). American Airlines went so far as to remove its listings from all of the
major sites. Which was interesting because Travelocity (Sabre) was once a divi-
sion of American Airlines. The goal at AA was to encourage customers to book
fights directly at the companys Web site, bypassing the fees charged by the tick-
eting sites. Southwest uses this technique to save money and AA managers want-
ed to do the same thing. Eventually, AA renegotiated with most of the systems, but
passengers are well-advised to double check fares on individual airlines to get the
best prices. Some sites, notably Kayak, automatically check fares through the big
travel sites as well as the individual airlines.
Another interesting issue arose in 2010 when Google proposed to buy ITA Soft-
ware (Perez 2010). ITA is a company that provides the airline data to many of the
other travel sites, airlines, and Microsoft Bing. The company was founded in 1996
and has about 500 employees. Google wants to own the company so that it can
integrate fight data and reservations directly into the Google search engine. But,
other providers worry that this step would reduce competition, meaning take away
their customers. Eventually, the Department of Justice approved the purchase but
imposed a few conditions to encourage continued competition.
Despite the advances, travel search engines are still somewhat primitive. For
starters, they rarely fnd the best prices. But, remember it is the airlines that pay
the initial fees, so they have an incentive to arrange the systems to support their
objectives. The fexible searches that check multiple departure dates and different
airports do help users fnd better prices, but in the end, a human has to do a large
amount of work to fnd the best fares and fight times.
Questions
1. Who are Sabres competitors?
2. What risks does the company face? Will travel agents continue to exist and
will they use Sabre?
3. What factors are needed for an intelligent agent reservation system to be
created? Who would create these elements?
Additional Reading
Anthes, Gary H., Sabre Takes Extreme Measures, Computerworld, March 29,
2004.
Anthes, Gary, Sidebar: Sabre Timeline, Computerworld, May 31, 2004.
Havenstein, Heather, Sabre replacing EDI with Web services, Computerworld,
August 22, 2005.
Havenstein, Heather, Sabre Launches Web 2.0 Social Network for Corporate
Travelers, Computerworld, February 20, 2008.
Information Week, Sabre Agrees to Buy European E-Travel Company, May 12,
2005.
Fox News, Private-Equtiy Groups Buying Travelocity Parent Sabre for $4.5
Billion, December 12, 2006.
Kontzer, Tony, Travel Execs Get A Glimpse of Their Future, Information Week,
May 7, 2004.
Kontzer, Tony, Bound For Industry Upheaval--With a Layover in Dallas,
Kontzer, Tony, Sabre Lands Low-Fare Carriers Airfares, Information Week,
October 14, 2005.
Kontzer, Tony, VC Deal Signals Turbulence Ahead for Airlines Legacy IT,
Perez, Juan Carlos, Travel Search Players Unite to Oppose Google-ITA
Merger, Computerworld, October 26, 2010.
Rosencrance, Linda, Sabre Launches Interline E-Ticketing Hub,
Verton, Dan, Alaska Airlines Switches to Linux-based Fare Searching,
The Wall Street Journal, Southwest Airlines Adds Sales Outlet, May 17, 2007.
Case: American Airlines
American Airlines (ticker: AA) is one of few original airlines still in existence.
The company traces its history back to mail fights by Charles Lindbergh in 1926.
By February 1937, American had carried one million passengers. At the start of
1959, American was the frst to offer coast-to-coast jet service with the Boeing
707. In 1959 and 1960, American teamed up with IBM to develop and imple-
ment the SABRE (Semi-Automated Business Research Environment) system for
handling reservations and business data. By 1964, the SABRE network extended
across the U.S., Canada, and Mexico. American was proud to be at the leading
edge of aerospace and information technology (www.aa.com--history).
When the airline industry was deregulated in 1978, low-cost carriers arose
across the nationled by People Express in New Jersey. For four long years, the
major carriers did not know how to compete. In 1985, American turned to the data
held in the SABRE system to devise a winning strategy. Introducing Super Saver
faresfor travelers who booked in advance and stayed over SaturdayAmerican
was able to compete dollar-for-dollar with the low-cost carriers. The secret was to
hold back enough seats on each fight to sell at substantially higher rates for busi-
ness travelers who booked at the last minute. The years of fight and passenger
data held in the SABRE system enabled American to perform yield management
to determine the optimum number of seats to hold back on each fight. In 1981,
American Airlines started AAdvantage as the frst frequent fier program. It was
initially targeted to business travelersto encourage loyalty (CNN Online 2006).
Over time, despite the various attempts by American and the other big carri-
ers, competition changed the airline industry. The expansion of Southwest drove
lower fares into new markets, which dramatically increased the number of pas-
sengers. Increasingly, American had to focus on cutting costs, and emphasizing
its global access. But fights became commodities, where passengers simply want
to get to their destination on time. But airlines cannot just throw in the towel and
give up. In the 1980s, Bob Crandell, the CEO at American, thought about trying
to become like Southwest Airlines, but realized it was not possible. Instead, the
company focuses on using technology and creating systems to reduce costs and
provide better service to customers (Rosencrance 2005).
In 2011, American began offering Samsung Galaxy Tab 1.0 devices to premium
class passengers on its fights. The tablet is confgured to provide in-fight enter-
tainment and will allow passengers to browse the Web or check e-mail on fights
with Wi-Fi installed (Hamblen 2011). Like the other carriers, American lost $2.1
billion in 2008 and $1.5 billion in 2009. But, unlike some of the others, American
continued to suffer losses ($0.5 billion) in 2010 (2010 Annual Report).
Technology
Sabre drove American Airlines for 40 years, handling transactions and provid-
ing information about fights and crews. But, American became dependent on the
system, and it was a system designed in the 1960s. When employees have new
ideas, it can be diffcult or impossible to reprogram Sabre to incorporate innova-
tions. Captain Doug Pinion, scheduling chairman for the Allied Pilots Association
at American, has had several ideas for saving money. For instance, he suggested
back-up pilots could use the Web to bid on fights scheduled for the next day. But
pilots suggestions are overruled as too expensive, and American selects pilots
based on seniority; phoning each person in turn. The problem is that the informa-
tion system is too diffcult to reprogram for new tasks. Americans CIO in 2003,
Monte Ford, said that all of the major airlines face the same issues. In a lot of
ways were better advantaged because of the kinds of upgrades weve completed
or are about to complete (Gage 2003). Some operations have been moved onto
Unix and Windows serverspulling data from the Sabre transaction processing
system. However, Ford believes that the mainframe system is still the best way to
handle millions of transactions.
In many ways, the advent of the Internet accelerated the conversion of airline
seats into commodities. Customers, tourists and businesses, can quickly and eas-
ily see the prices. Some Web sites, such as farecast.com, even track patterns for
fights, and others can provide immediate notifcations of any price changes. Hen-
ry Harteveldt, a vice president at Forrester Research noted that yield-manage-
ment systems made it possible to have more fare types than seats on an airplane.
But the airlines got greedy, and their bad pricing strategies caught up with them
when the Internet exposed the pricing structures. The bottom line is that the es-
tablished carriers have substantially higher operating costs than the discount carri-
ers, as much as 7.2 cents per seat-mile (Rothfeder 2005).
Reducing costs will entail more than wringing additional concessions from la-
bor, retiring old planes, and closing a few hubs. Even after exiting bankruptcy,
Delta and United only reduced costs by small amounts. Ultimately, the big airlines
like American will have to fnd a way to simplify operations and reduce the huge
administrative costs. The micro-jets are waiting in the wings. Robert Crandall,
ex-CEO of American Airlines, and Cameron Burr, son of Donald Burr, ex-CEO
of People Express, have teamed up to start an executive taxi service called Pogo
based on microjets (Meehan 2007). Depending on the progress of FAA certifca-
tions, operations were planned to start in 2008. For about the same price as busi-
ness-class seats executives will be able to charter a fight that fies directly to the
desired airport on any desired time schedule. The jets can fy into smaller, more
convenient airports, and passengers will have no lines for security, check-in, or
baggage. Ultimately, the jets could eat away at the passengers that yield-manage-
ment relies on for profts.
American faced the same problems as every other airline during the recession.
In an attempt to reduce costs, it tried to encourage passengers to book fights di-
rectly on its Web site and to send booking data directly to travel agents (2010
Annual Report). The company ran into contract-violation problems when it tried
to remove its listings from the major ticketing agencies, and eventually relisted
its fights on most of the systems (Cameron 2011). In 2011, American sued Orbitz
and Travelport LP alleging anticompetitive practices. Overall, the company is fo-
cused on reducing costs.
Questions
1. How can American compete with Southwest?
2. Is it possible to rebuild Americans information system?
3. How can American compete with the coming microjets?
Additional Reading
Cameron, Doug, American Ups Ante in Dispute, The Wall Street Journal, April
14, 2011.
CNN Online, Frequent Flier Programs Straying Off Course? June 29, 2006.
Gage, Deborah, American Airlines: Hitting Turbulence, Baseline, July 1, 2003.
Hamblen, Matt, American to Offer Galaxy Tab 10.1 to Premium Class
Passengers, Computerworld, June 13, 2011.
Rosencrance, Linda, IT Gives Airlines a Lift, Computerworld, September 5,
2005.
Rothfeder, Jeffrey, Can Information Technology Save the Airlines? CIO
Insight, February 5, 2005.
Meehan, Rod, Getting a Move On, Conntact.com, May 14, 2007.
adopted by rivals?
sectors?
715
How can small businesses use information technology?
How are small businesses different from larger ones?
What information and technology do small businesses need?
How do you start a business?
How will your business be different from the existing frms?
How do you turn an idea into money?
Is it true that genius is 1 percent inspiration and 99 percent perspiration?
What additional steps are required to start and EC frm?
Why did thousands of dot-com frms fail?
How does cloud computing help startup frms?
Chapter Outline
Entrepreneurship
11
Chapter
Introduction, 716
Small Business Constraints, 717
Money, 718
Workers and Specialists, 720
Information Technology Expertise, 721
Strategic Power, 723
Information Tasks, 724
Operations and Transactions, 724
Tactics and Decisions, 727
Strategies, 728
Selecting Technology Levels, 729
Managing Consultants, 730
Internationalization, 731
Idea, 732
Strategy, 732
Research, 734
Plan, 736
Strategy, Competition, and Market
Analysis, 738
Forecasts, Cash Flow, and Investment
Budget, 738
Marketing, 740
Online Advertising, 741
Organization and Timetable, 742
Implementation, 742
Ownership Structure, 743
Financing, 744
Accounting and Benchmarks, 747
Starting an e-Commerce Firm, 748
Analysis of Dot-Com Failures, 749
Pure Internet Plays, 749
Proft Margins, 750
Advertising Revenue, 751
Summary, 752
Key Words, 754
Exercises, 755
Cases: Entrepreneurship, 758
716 Chapter 11: Entrepreneurship
Petz Enterprises, Inc.
Introduction
How can small businesses use information technology? In many
ways, this question has already been answered in the other chapters. You can ap-
ply the same analyses and tools to any size business. However, small businesses
often face a shortage of investment money and skilled IT workers. Business own-
ers also do not have time to spend hours thinking about IT solutions. Yet, technol-
ogy applied correctly can save business owners a considerable amount of time and
money. With the relatively low cost and tremendous power of information tech-
nology, it is possible to run relatively complex businesses with a small number of
people.
A second major question addressed in this chapter involves starting a new busi-
ness. What technologies do you need to start a business? And how can you create
a company to make money online? Entrepreneurship is a growing feld in the busi-
ness discipline and you can probably take many courses to learn how to analyze
product ideas and create a new business. This chapter reviews some of the basic
concepts and shows how technology is used to make it easier and provide more
information. Starting from scratch forces you to focus on the information you will
need to run the business. It is a good opportunity to evaluate the technologies
and data you will need to operate a business. Selling products online is relatively
straightforward today. Building new Web 2.0 applications or even phone-based
applications requires developing custom programs.
How can small frms use the Internet to improve their business? Petz Enterprises
in Tracy, California, is a family-run company that specializes in income taxes. The
company became a leader in providing software to professional accounting frms and
storefront tax offces. The tax-accounting staff at Petz keeps up with the hundreds of
tax-code changes made by the federal and state governments. It then writes the rules
that compute the taxes and print the forms. As a closely held company, Petz does not
report its fnancial data.
With the growth of the Internet, Leroy Petz, Sr., decided that the frm had to ex-
pand and create a system that could be used directly by individuals. The initial Tax-
Brain site attracted several respondents in the frst two years. After it was given a
substantial facelift, and backed by online advertising, the site became the third most
popular tax site on the Internet. TaxBrain faces some serious competition with the
well-known brands of H&R Block and TurboTax (Intuit).
Entrepreneurship, the development of new frms, is an idea that foats in the minds
of many businesspeople. In 2003, over 500,000 new frms were started. On the other
hand, almost the same number of frms were closed in the same time period. The
Small Business Administration (SBA) reports that about two-thirds of new employer
frms survive at least two years. Fifty percent survive past four years (Bounds 2004).
Information technology plays an important role in many small frmsbecause en-
trepreneurs need higher productivity. The Internet in some ways makes it easier to
start or expand a business. In other ways, it increases the marketing and technical
costs. And the Internet is certainly no guarantee of profts or success.
Even if you do not intend to start or run a small business in the next year or
two, you should read this chapter. First, with the number of small businesses cre-
ated each year and the growth rates, there is a good chance that someday you will
work for one. Second, the chapter serves as an integrated review of many of the
concepts covered in the other chapters. In a small business, the manager/owner is
responsible for putting everything together to solve problems and grow the frm.
All of the aspects of information technology are used in this process.
What is a small business? The simple defnition is any independent business
that is not dominant in its industry. However, the U.S. Small Business Adminis-
tration uses more specifc defnitions for its fnancial lending programs. The base
defnition is frms with fewer than 500 employees. In some industries, the number
of employees drops to 100, in others it goes as high as 1,500. Using the 500-em-
ployee defnition, the SBA concludes that about half of the employees in the U.S.
work for small businesses, and 99.9 percent of businesses are small. The year
2007 saw 688,395 new employer frms created, but 592,410 were closed for a net
gain of about 96,000 new frms. An employer frm is a company that has employ-
ees. It does not include single-person self-employment. Figure 11.1 shows the
basic totals for 2009 (the most recent data available).
Small Business Constraints
How are small businesses different from larger ones? In many
ways, all businesses have the same tasks to perform regardless of the size of the
frm. On the other hand, small businesses face several constraints that can often
determine whether the company succeeds or fails. The biggest issue is money
or investment capital. Partly tied to the issue of money, small businesses have
smaller numbers of workers and generally lack specialistsparticularly in man-
agement. Consequently, the owner, or a designated manager, handles most of the
management tasks; and usually runs out of time each day. Outside the technology
industry, few small businesses have any expertise with information technology.
Finally, small businesses often survive by selling to one or two large companies.
Looking at Porters Five Forces model, it means that small businesses often lack
strategic power. They must rely on goodwill and mandates of their largest custom-
ers. All of these constraints make it diffcult for the small business to thrive and
expand.
Figure 11.1
Small business numbers. In total, small businesses in the U.S. employee about half of
the total workers and constitute 99.9 percent of the number of frms. Source: www.
sba.gov. *The most recent birth and termination data is from 2007.
Estimates 2005 2009
Employer frms (nonfarm) 5,992,400 5,815,800
Employer frm births 671,800 688,395*
Employer frm terminations 544,800 592,410*
Self-employment, nonincorporated 10,500,000 9,800,000
Self-employment, incorporated 5,300,000 5,500,000
Business bankruptcies 39,201 60,837
Money
Small businesses by defnition do not generate as much revenue as large frms and
have fewer assets to borrow against. The numbers show that small frms also gen-
erate less revenue per employee than large frms do. Many have been in existence
only a short time. With these factors, small businesses often have problems acquir-
ing the money needed to expand a business, hire workers, or purchase technology.
Financing options are examined in more detail in the Entrepreneurship section,
but every small business has problems raising money. Eventually, businesses that
are successful will generate profts that can be used to pay for items that fuel ad-
ditional growth. But, except for some rare cases, this money is incremental and
growth is limited to a small annual percentage.
In the meantime, businesses face ongoing expenses. Many of them are fxed
costs such as leasing space, employee salaries, professional services (legal and
accounting), insurance, vehicles, marketing, and information technology. Small
business owners have a tendency to look at the frms expenses as deductions from
their take-home proft. Buying a new server might be nice, but at $5,000 it could
represent 5-10 percent of the annual proft available to the owner. Consequently,
small businesses are often reluctant (or unable) to invest in new technologies un-
less there is an immediate payoff in terms of reduced expenses or increased rev-
Trends
Small businesses make up about half of the U.S. economyparticularly in
terms of employment. There is a high probability that you will work for a
small business at some in your career. Perhaps you will even want to start
your own company or become self-employed. The Census Bureau keeps sta-
tistics on the number of people employed by small businesses. The ratio of
the number of small businesses to large has been relatively constant since the
late 1980s. The percentage of people working for small versus large compa-
nies has changed slightly over time. In the late 1980s, about 9 million more
people worked for small frms than for large ones. In the late 1990s, despite
a growth in the number of people employed in businesses hiring 20-99 em-
ployees, larger frms began hiring a greater percentage of the total employ-
ees. Perhaps because of the collapsing dot-com economy, workers switched
to the security of larger frms so that the share was balanced in 2001. Since
that time, large and small frms have hired about equally.
However, you see more interesting results by looking at the total payrolls
and the revenues of the two frm sizes. From the early 1990s, large frm an-
nual payrolls have grown faster than small frms. In 2007, the average salary
per employee was $47,400 for large frms compared to $37,300 for small
frms (79 percent). It is possible that these numbers are distorted by the huge
salaries paid to CEOs at large frms, but it is clear that large frm payrolls are
higher. Total revenues are not reported every year, but in 2007, large frms
earned an average of $302,000 per employee compared to only $190,000 for
small frms (63 percent). Small frms spend a higher percentage of revenue
in payroll. The lower revenue and higher salary expenses strongly affect the
way small frms are managed.
enue. Figure 11.2 shows that large frms pay more on average, but small frms
spend a higher percentage of revenue on employee payrolls.
Figure 11.3 shows another perspective on money by comparing small frms to
large frms in terms of receipts and payroll. The base numbers consist of total re-
ceipts divided by total employees and total payroll divided by total employees. In
every case, the values for the small frms are lower than those for the large frms.
The table shows the value for the small frms as a percentage of those for the large
frms. For example, in 2007, average receipts per employee in small frms were
62.9 percent of the value for large frms. And payroll per employee for small frms
was 79.3 percent of the value for large frms.
Figure 11.2
Large v. Small Payroll. Large frms have a higher total payroll. Since the number
of employees is about equal, the difference arises because large frms pay higher
salaries.
Source: http://www.census.gov/econ/susb/historical_data.html
0
500
1000
1500
2000
2500
3000
3500
Payroll (Billion$)
500+
Small
Figure 11.3
Small v. Large receipts and payroll. The table shows the percentage of small over
large frms. For example, in 2007, the average receipts per employee in small frms
were 62.9 percent of those in large frms.
Source: http://www.census.gov/econ/susb/historical_data.html
Year Receipts/Emp Payroll/Emp
1997 64.4% 80.8%
2002 63.0% 81.6%
2007 62.9% 79.3%
Larger frms could have higher receipts per employee due to economies of scale
which includes a broader base of customers and more effcient use of technology.
Typically, they have also been around longer, so they have an established base of
customers. Note that although small frms pay less per employee than large frms,
the revenue per employee is even lower. Consequently, proftability is going to be
squeezed at smaller frms.
Fortunately, with the declining costs of technology, it is easier for frms to pur-
chase powerful tools at relatively low costs. More importantly, with the expansion
of service frms, small businesses can pay monthly fees for high-end software and
services. A large frm might pay hundreds of thousands or millions of dollars and
hire experts to customize an ERP system. A small business cannot afford the hard-
ware, much less the software and personnel to build an ERP and keep it running.
Instead, the small business can contract with an online provider to host and run the
ERP system. For a few dollars a month, the small business can obtain almost the
same effectiveness. And, the system can be scaled up as the company growsso
the small business pays a small fee now but can expand the system later when
revenues increase.
Workers and Specialists
In part because of the money constraints, small businesses do not hire as many
workers. They will have to hire workers on the production sidefactory or ser-
vice workers and salespeople. But, it is expensive to hire back-offce administra-
tors or managers. Workers that do not directly generate revenue are seen as an
expense. They are hired only when the owner needs additional support or specifc
expertise. Accounting, marketing, and production managers are the most common
employees.
Small businesses rarely hire information technology specialists (outside of
technology and EC frms). Workers are expected to know how to use the basic
personal productivity tools. When things break, small businesses rely on service
contracts or local consulting frms to solve problems. As a result, it is diffcult for
small businesses to keep up with technology changes or to know what tools are
available to help them. They are also at the mercy of the local repair companies in
terms of pricing for service and support.
On the other hand, small businesses in specifc industries can often purchase
the information technology they need as an off-the-shelf package. For instance,
dentists, physicians, attorneys, restaurants, and so on can choose from many turn-
key systems. The name derives from the concept that the business buys the pack-
age of hardware and software, plugs it in, turns the metaphorical key, and has
everything necessary to run the business. The systems company provides support,
training, and upgrades. Turn-key systems exist for a surprising number of busi-
nesses. You can often fnd them through advertisements in trade magazines for the
specifc industry. Or, you can visit similar businesses to see what they are using.
The prices sometimes seem a little high, but they are often based on monthly pay-
mentsavoiding the need for large amounts of money up front.
To offset these drawbacks, small businesses should consider paying a consul-
tant every year or two to evaluate their operations and security. Some accounting
frms provide this service and some will require EDP audits to ensure basic secu-
rity provisions are being met. IT consulting to evaluate technology levels, costs,
and effectiveness is also available.
Information Technology Expertise
Figure 11.4 shows that the lack of IT expertise affects more than purchasing and
managing systems. As noted, day-to-day operations can be handled through pur-
chased systems and support contracts. Often the bigger constraint is that small
businesses do not know how to use IT to solve business problemsparticularly
in terms of searching for a competitive advantage. For example, a law frm can
purchase a time-and-billing system to track paperwork and bill clients. But, the
system and the attorneys know nothing about using technology for marketing or
teamwork. Perhaps the frm specializes in specifc complex cases. It could beneft
from a teamwork system that saves all of the prior work in a knowledge manage-
ment systemreducing the time it takes to prepare the next case. Such a system
could be created relatively cheaply using a SharePoint server, but if there is no
turn-key system for it and no in-house IT specialist, the frm might never know
such a system is possible or affordable.
Computer security is another critical area of expertise that is unavailable in most
small businesses. Properly installed, good turn-key systems include provisions for
security. Most use passwords to grant permissions, and most have semi-automated
backup systems to make it easy for businesses to maintain regular backups of the
data. But, with a small number of employees, managers often circumvent the in-
ternal security provisions. For example, it is common to assign a single account
with password for everyone to share. It simplifes creating accounts and solves
Figure 11.4
Lack of IT expertise. Small businesses need to hire someone to help with basic IT
tasks. Operations and transactions generally work because they have to work. Tactics
and strategies along with security tasks are often skipped because of the perceived
cost.
Service contracts for hardware.
Consultants for network and troubleshooting.
Assign someone to do backup.
Security is usually forgotten.
Upgrades and decision software are rare.

problems if an employee calls in sick and other workers need access to the data.
But, sharing accounts and passwords pretty much wipes out any security controls.
The example shows that even if the system has good security provisions, users
(or owners) can bypass them, leaving the system open for attack. Many aspects
of computer security today derive from best practices that establish procedure for
everyone to follow to ensure the security rules are followed. But few small busi-
nesses are aware of these best practices, and often fnd them inconvenient.
Choosing the timing and level of information technology is another challenge
for small frms. Figure 11.5 shows the challenges faced by every business, but
large companies have an MIS staff, money, and opportunities to experiment with
new technologies. Small businesses (not directly involved in technology or EC),
tend to purchase standard technology when they begin. Managers then keep the
hardware and software as long as possiblebecause the costs are generally fxed
and have already been paid. The problem is that older technology begins to die
and software rarely gets upgraded. Problems often arise at the least convenient
times causing problems and expenses for emergency repairs or even lost data. On
the other end of the spectrum, buying leading-edge technologies presents different
benefts and costs. Being the frst to implement new technologies can lead to new
business methods and opportunities to provide products and services not offered
by competitors. But, new technologies are expensive and diffcult to predict. New
technologies often have multiple versions or competitors. Typically, one version
will succeed and the others will disappear. It is hard to bet a company on a new
technology that might or might not survive. And, continually switching between
new technologies carries high costs of disruption.
Large frms have an MIS staff that follows technology news. Many have bud-
gets for experimental projects and often obtain demonstration units from vendors.
Figure 11.5
Choosing technology timing. It can seem cheaper and easier to keep old technology
for long periods of time but there are costs to keeping it too long. On the future side,
shiny new technologies can appear to be useful but buying early is expensive and
means understanding and choosing new technologies that might not survive. Without
an MIS staff, it is diffcult to keep up with and evaluate potentially useful new
technologies.
past future
Fixed cost/paid for.
Know how to use.
Mostly works.
Hardware failures.
Limited features.
Hard to repair.
New hardware features.
New business methods.
Improved security.
High prices for new technology.
Uncertain life, adoption rates.
Difficult to follow tech industry.
Harder to find support.
Constant switching and costs.
Existing small frms typically stick with older technologies. But, new small frms
are probably going to buy newer hardwarealthough probably not leading-edge
tools. Still, with half a million frms starting each year, some smaller frms will be
using relatively advanced technologies.
Strategic Power
Recall Porters Five Forces model. One method of achieving strategic gains is
to build alliances with suppliers or customers. To build these linkages, you frst
gauge the relative power of your frm versus these external entities. But, it is pret-
ty clear that small businesses will have minimal power in these strategic relation-
ships. Many small businesses survive by working closely with one or two large
frms. In almost any situation, the small business is going to be at the weak end of
the power relationship. That means that instead of your frm initiating an alliance,
you will respond to the needs or demands of the larger customer. As shown in Fig-
ure 11.6, the good news is that most large businesses already have ERP systems
that provide Web-based support for suppliers and customers. You will be able to
use the clients computer system to track orders, returns, and payments. The small
business simply needs a Web browser and an Internet connection. The Internet
connection can be simple and low cost, similar to a household link.
The drawbacks to relying on a customer (or supplier) ERP system arise when
the small business grows larger and acquires more clients. Each customer will
expect you to use their system for interactions. You will want a way to consolidate
the data from each of the systems so you can see everything in one location. Even
if the small business has its own ERP system, it can be diffcult to get the systems
Big Supplier Big Customer
Use supplier and customer ERP
systems to track purchases and
sales and replace the need for
your own system.
Figure 11.6
Lack of strategic power. As a small business facing 800-pound gorillas for customers
and suppliers, accept their strength and use their systems. They have ERP systems
that will track purchases, sales, payments, receipts and other data you need. Use their
systems for detailed record keeping and consolidate the totals in a basic accounting
package.
to work together to download the data. Lacking the strategic power, the small
business pays the cost to hire someone to copy data from these other systems and
re-enter it into its own system. There is no good answer to this problem. In some
cases, the best answer is to install the same ERP system that is used by the major-
ity of the customers. It is easier to connect systems and share data if they are built
by the same provider (such as SAP).
Buying software is another place where lack of strategic power hurts small
businesses. Companies such as Microsoft give generous discounts to large com-
panies that buy hundreds of copies of software. As a small business, you will not
qualify for most of those discounts and can end up paying quite a it more money
per copy than your large competitors. And, since most salespeople are paid on
commission, do not expect any of them to pay much attention to you if you want
information.
Information Tasks
What information and technology do small businesses need?
Figure 11.7 summarizes that like any other business, a small business needs to
deal with decisions and information at three levels: operations, tactics, and strat-
egies. The information and technology requirements at each level are different,
so you should examine each area separately. In most cases, you can examine the
business as if it is starting from scratch. Even if a small business has been operat-
ing for a few years, it most likely needs to re-examine its information systems.
Operations and Transactions
The basic transaction tasks are sometimes the most diffcult for small frms to han-
dle properly. Large frms have many people and established rules. Small frms of-
ten record data when the managers get around to it, and procedures are perceived
Operations
Accounting
Sales
Purchases
Cash fow
Production
Shipping
Payroll
Basic accounting software
but it must use double-
entry.
Possibly online ERP
service.
Outsourcepayroll.
Tactics
Customer analyses
Product analyses
Employee evaluations
Forecasting
Growth
Teamwork
Often skipped because of
lackofexpertise.
Possible with spreadsheets
andSharePoint.
Strategies
Smallest frm
Minimal power
Limited funds
Rely on larger customers
andpartners.
Useonlineservices.
Figure 11.7
Information tasks. Small businesses face the same decision levels but tend to
concentrate on the day-to-day operations. Technology must be affordable and
demonstrate a direct contribution to the bottom line. Support for tactics and strategies
are often skipped or deferred.
as unnecessary constraints that take too long. But without correct data, it takes
longer to put reports together, to analyze the frms condition, and to make deci-
sions. Many small-business owners complain about having to spend long nights
searching for and collating data. Even an inexpensive accounting systemwith
the proper procedurescan save huge amounts of time and effort.
Most small businesses can handle day-to-day transactions through accounting
systems. The standard fnancial data is often the most important information to
the managers, and the accounting system stores everything in a database that has
basic search and reporting capabilities. The key is to use at least a double-entry
accounting system (as opposed to a check register), and to establish procedures so
that all data is entered promptly, correctly, and verifed.
Most businesses will want to hire accountants to ensure the fnancial data and
procedures are sound and to handle annual tasks such as income taxes and state
flings and permits. In these cases, the business is generally better off adopting
the accounting system recommended by the accountant. The accountant will need
to access the data and will be familiar with the standard operations of the soft-
ware. However, small businesses should also look at the online accounting/ERP
systems. The online Web-based systems are particularly useful when multiple
managers need access to the data, or if there is a need to access the data while
traveling. For defned monthly fees, the online systems handle the software, data
backup, and basic security. The online systems offer more features than low-end
standalone accounting packages and can easily expand as the company grows.
Plus, the business pays a fxed monthly fee instead of paying large upfront and
annual maintenance costs.
Choosing an accounting package does not automatically solve all of the trans-
action issues. Every accounting system has to be confgured to match the spe-
cifc company. The managers, perhaps with the advice of the accountants, need to
Reality Bytes: IPO Fever Spreads Again
Starting in 2010 and continuing through 2011, several well-known technology com-
panies used initial public offerings (IPOs) to make money. Many of them generated
huge interest and stock prices rapidly increased on the day of the IPO. Pandora Me-
dia is one of the early companies that generated strong initial demand. The Pandora
app is one of the fve most-used apps on all of the major smartphone platforms in
the U.S. The company has a free app and provides free access to many streamed
songs. It also has a pay-model where for a small fee users get more control over the
music they stream. It also gains some revenue from advertising. However, since the
company was created in 2000, it has never been proftable. The problem is that as
the company gains users, it streams more songs, which means it has to pay more
royalties for the songs. Rick Summer, a senior analyst at Morningstar Inc., noted
that Pandora faces more problems in the long runthe contract with music publish-
ers that sets the royalty rates expires in 2015. And if listener demand remains high,
the record companies are likely to renegotiate for an even greater cut of the revenue.
But, investors still pumped money into the company in the early days of the IPO.
Adapted from Lynn Cowan, Pandora Plays a Hot Tune, The Wall Street Journal,
June 13, 2011.
identify the information needs. Does the frm, like Rolling Thunder Bicycles, sell
items by categories? Are prices fxed or negotiated on every item? For services,
are they billed to customers in advance (like a cable TV bill) or at the end of the
month (like a lawn service)? Who is responsible for entering data and when will
each item be entered? Do entries have to be approved or monitored? Who deposits
money in the bank and who records and verifes the deposits? Similar questions
apply to handling purchases and receipt of supplies. All of these questions have to
be identifed, answered, and expressed as procedures.
Hiring employees and handling payroll is another challenging transaction task.
Today, even small businesses need to verify applicant information such as taxpay-
er ID numbers. Tiny frms can probably get by with simple payroll systems, but
you still need to handle federal, state, and local tax withholding, plus unemploy-
ment compensation and so on. With dozens of employees, you might want to offer
more advanced payroll options such as health insurance, direct deposit, and retire-
ment plans. There is no way a small business should handle these details itself.
Almost all companies contract payroll services to an outside frm. Several nation-
wide payroll frms exist and many can handle even small frms at reasonable costs.
Automated Data Processing, Inc. (ADP) is one of the oldest and largest. Providers
Reality Bytes: Industry Salaries
Industry Median Pay
at 5 Years
Median Pay
at 15 Years
Finance/Securities $65,200 $114,000
Federal Government $61,500 $81,900
Hospitals $58,900 $70,800
Information Technology $58,300 $104,000
Manufacturing $57,500 $89,900
Professional/Science $57,000 $92,100
Construction $55,600 $84,400
Transportation $53,300 $79,000
Retail Trade $51,300 $91,000
Management $50,300 $76,800
Banking $49,300 $83,900
Administrative $48,400 $73,400
Real Estate $46,300 $82,200
Educational $43,000 $59,400
Arts, Entertainment, Rec. $42,200 $67,700
Hospitality $42,000 $68,800
Food Services $40,200 $61,800
Adapted from Carolyn Bigda and Donna Rosato, 5 Secrets of Successful Career
Changers, Money Magazine, May 18, 2010.
of small-accounting packages such as Intuit also provide these services, but you
should compare services and prices with several vendors.
Communication systems and networks should also be included in an analysis of
operations. In the old days, you would have to set up a business account with the
local telephone company and probably purchase an expensive telephone system.
You can still take this route, but it is probably cheaper today to use cell phones.
They have the added advantage of mobility so your employees can be reached at
any location. But, like every other technology, prices change and you will have
to investigate the various options to ensure you are not missing some low-cost
alternative.
Internet connections are an important factor in any business today. The biggest
decision to be made is whether you need full-time high-speed access lines, such
as frame relay or T1 lines. Chapter 3 shows that these are relatively expensive
at hundreds of dollars a month. But they are necessary if you want to run your
own Internet servers. For most frms, it is going to be substantially cheaper to
pay someone to host Internet servers and go with a simple DSL, cable modem,
or Wi-Max Internet connection. These services provide relatively fast download
speeds for a reasonable monthly fee. They are generally fast enough to be shared
with several employees, but you will have to monitor performance. If your frm
transfers huge fles or you gain dozens of employees, you will have to increase the
Internet bandwidth through additional services.
Tactics and Decisions
Few small businesses have the expertise to analyze data. Many might not even
have the data or know how to get it out of their accounting systems. Yet, busi-
ness survival can depend on knowing which customers are the most proftable,
which products and services make the most money, or identifying regional or an-
nual selling patterns. Fortunately, the basic tools are relatively inexpensive. Most
small-businesses have a limited amount of data that can be easily handled with
basic toolsoften just a spreadsheet. You will fnd that most small businesses
do not have copies of Microsoft Access. Fortunately, the Offce 2007 version of
Excel can handle substantially more rows than in the past. Queries are not quite as
convenient or powerful, but you can use a worksheet as a small database.
Even a basic PivotTable can be used to help managers explore subtotals and
search for patterns in the data. Chapter 9 shows that more complex analyses can
be performed using spreadsheets, but it might be helpful to have an expert look
at the business and set up the initial spreadsheets. At a minimum, an owner with
limited knowledge of the statistical tools should have an expert examine and test
any complex spreadsheets. Several companies have experienced major problems
due to faulty analysis in spreadsheets.
Service frms should seriously investigate the teamwork tools. Some of these
(such as customer relationship management) can be purchased as online services.
Similarly, you can use the online offce tools (such as those provided by Google)
to support project teams. However, it is relatively inexpensive to run your own
SharePoint server and provide complete teamwork solutions. Even without these
tools, frms should fnd ways to store knowledge documents in a location and for-
mat that can be searched and retrieved to save time on the next project.
Decision support tools and expert systems can be added to a frm at almost any
time. The challenge lies in identifying areas that can beneft from these tools, and
then determining whether the potential benefts exceed the costs. Complex ana-
lytical tools can carry high up-front costs, but they can easily pay for themselves if
they save money on operations that are repeated hundreds or thousands of times.
The important point is to understand the business well enough to know what areas
need to be fne-tuned and then to know the current capabilities of the tools.
Strategies
By defnition, small businesses have relatively weak strategic power compared
to rivals, suppliers, and customers. A small manufacturer is not going to go to
Wal-Mart and insist they use a specifc online ordering system. On the other hand,
you can use this disparity to your advantage. If you want to build electronic links
to large customers, simply ask to use their systems. Most large companies have
sophisticated ERP systems that are already Web based and support links to sup-
pliers. Your customers could easily confgure an account for you on their systems
that enables you to track all orders, shipments, and payments online. All you need
is an Internet connection and a Web browser. Similarly, you can set up online ac-
counts at major suppliers. For instance, you can buy all offce supplies online at
large chains such as Offce Depot. Their system tracks purchases, handles elec-
tronic payments, and supports manager approval so you can monitor and con-
trol all of your offce supply purchases. Since the technology is readily available,
many other large suppliers should offer similar services. Think about it for a min-
Reality Bytes: Self-Publishing Pays Off
For decades, even centuries, publishers have controlled access to bookstores. Al-
most anyone can write a book, and anyone can hire an editor and graphics artist to
make the book look nice. But publishers controlled the distribution to the main book-
stores. Without a contract, the number of books any person could sell was tiny. And
publishers denigrated self-published books as books that were not worthy of read-
ing because they had likely already been turned down by a big publisher. Of course,
big publishers had fnancial incentives to sign a small number of authors, so many
good authors fell through the cracks. The shift to electronic booksheavily pushed
by Amazonhas changed the world. Karen McQuestion is a good example. She
took one of her books to Amazon and eleven months later had sold 36,000 copies for
the Kindle. Amazon is going to print a paperback version of one book and she has
a flm option with a Hollywood producer. Several sites help authors create, edit, or
distribute e-books. Some of the large sites are: Amazon, Apple, Lulu, Smashwords,
Scribd, and FastPencil. Compared to the almost trivial 12-15 percent that traditional
publishers pay to authors, Amazon pays 70 percent of the revenue to the author, de-
pending on price. Apple likewise takes a 30 percent cut. Even top-selling authors
have moved some of their back-list titles to e-booksmaking them widely available
at lower prices. Joe Konrath, writing as Jack Kilborn says he earned about $30,000
selling his novel Afraid in all forms through a traditional publishers. He estimates
he could have made that much in 18 months selling 800 e-books a month on Ama-
zon. He notes that he is already making more from self-published Kindle e-books
that had been rejected by traditional publishers. He commented that Im outselling a
bunch of famous, name-brand authors. I couldnt touch their sales in print.
Adapted from Geoffrey A. Fowler and Jeffrey A. Trachtenberg, Vanity Press Goes
Digital, The Wall Street Journal, June 3, 2010.
uteyou have just shifted most of your transaction processing onto your sup-
pliers and customers, giving you access to your data at any time with no major
out-of-pocket costs.
Strategy is diffcult for any frm, and harder for small businesses that lack mon-
ey and power. Probably the two most important concepts are (1) Identify a strat-
egy and purchase technology to support that strategy, and (2) Be fexible and will-
ing to change strategies if something is not working. For instance, if your strategy
is to be the least-cost provider, then select technologies that will reduce costs. But,
keep an eye on new technologies as they appear because something might arrive
that offers new benefts. Least-cost does not mean cheap or that you cannot buy
anything. For example, a service frm might beneft from using SharePoint, even
though it carries a hardware cost, the improved productivity could more than off-
set those costs.
That line between low-cost and useful technology is always diffcult, and often
requires additional training. Many small businesses eventually purchased personal
computers, but gained little because they did not know how to use the tools to
share and analyze data. The PCs replaced typewriters and provided some useful
features, but even today many businesses could gain more with the analytical and
collaboration tools.
Selecting Technology Levels
As summarized in Figure 11.8, much of the discussion of technology in small
business reduces to a decision of choosing the appropriate level of technology.
Does the frm need state-of-the-art technology or a dusty PC sitting in the corner?
It is tempting for small frms the save money and stick with minimal costs and
old technology. On the fip side, suppliers and vendors will try to sell the latest
Figure 11.8
Selecting technology levels. Small businesses often lack money and expertise. It is
important to match the technology level to the resources of the business. Different
applications can choose different technology levels. Operations are critical and can
justify more expensive tools. Systems for tactics and strategies might need to search
for lower-cost options.
Level Description Cost Perspective
Build and manage
technologyyourself.
Often necessary for leading
edge ideas or to customize to
yourmanagement.
Expensive, difcult to
controlcosts.Requires
considerableITexpertise.
Buy commercial of-
the-shelf hardware and
software.
Relatively fexible today
butstillrequiresinitial
customization.
Mostly up-front costs and
probably need consultants
atthebeginning.
Onlineserviceproviders. Moreoptionstoday.Easyto
sharedata.Lessworryabout
securityandbackups.
Minimalup-frontcosts.
Flexible pricing and
growth.
Buy low-end tools and
patchtogether.
Many small businesses
survive with Microsoft Ofce
andanaccountingpackage.
Can be low-cost, even go
withopensourcesoftware.
complex technologies to every business, regardless of size or needs. The deci-
sion is diffcult because the answer depends on the people as well as the indus-
try and technology. If managers are not comfortable sharing fles and working on
documents together, there is no point in building a collaboration system, even if
it could make everyone more productive. Larger companies might be able to train
workers and simply order them to use the new system. Companies with a limited
number of employees are more likely to let the users make decisions. So you have
to talk to all of the users to evaluate the level of technology to choose.
Buying hardware and software is only one factor in creating an information
system. For small businesses, often the most critical problem is the cost of MIS
workers. Remember that support workers are back-offce expenses that do not ap-
pear to contribute directly to increased sales. Sometimes they can be justifed in
terms of reducing operations costs or improving worker productivity. But small
business owners will always be critical of hiring workers.
One of the most important aspects to technology today is the ability to obtain
complex services online. It would rarely make sense to put an ERP system into a
small manufacturing frm with 10 employees. Among other reasons, the company
will not have the MIS personnel to run the system. On the other hand, it is possi-
ble to use a sophisticated ERP system through online services for a fxed monthly
fee, without needing to add expensive MIS workers and hardware.
Turn-key systems are the other popular way to reduce the need for MIS work-
ers. Most turn-key systems provide service contracts and at least telephone sup-
port. Although the contracts might seem expensive, they can be justifed by avoid-
ing the fxed-cost of hiring MIS workers.
Managing Consultants
Because small businesses are reluctant to hire permanent workers, they often end
up hiring consultants. Or, they try to get free advice from consultants, friends,
and academics. The big consulting frms tend to charge relatively high fees. The
work might be worth the amount of these fees, but managers need to think about
what they are getting for the price. Specifcally, in terms of technology, managers
need more than the fnal product or report produced by the consultants. Managers
should look at consulting as an educational opportunity and use the time to extract
as much knowledge as possible. Then if you need support in the future, you have a
head start and might even be able to perform the basic tasks yourself.
Consider a simple example where you need a LAN installed. Assuming you
know nothing about networks, you hire a frm to select the hardware and install
everything for you. You, or one of your employees, should observe the consult-
ing team, ask plenty of questions, and get full copies of documentation. When the
business expands and you want to add a couple of computers to the network, you
should be able to consult your notes and the documentation and add the new com-
puters yourself instead of having to call the consultants back and pay more money.
Similarly, if the cleaning group unplugs the network switch some night and noth-
ing works in the morning, you can perform basic troubleshooting and solve the
problem quickly and inexpensively.
Like any other purchase, you need to evaluate multiple consultants before hir-
ing one. Check references, get recommendations from other clients, and negotiate
contracts. Be careful of low-cost bids. Be sure that contracts completely spell out
all work and costs. Most small businesses prefer to work with fxed-price con-
tracts, but many consulting organizations will balkpreferring to charge by the
hour. If you cannot get a fxed-price contract, you will have to be careful about
what you ask for. Any changes or requests you make will cost additional time and
money. So, begin with a concise statement of what you want done and refer back
to it often.
Internationalization
The declining price of information technology makes powerful tools accessible
to even small businesses. The Internet and VoIP make it possible to communicate
globally at minimal costs. Web and auction sites make it easy to sell products
around the world, and large payment systems like PayPal and credit card pro-
cessors handle currency conversions automatically. The giant shipping companies
(UPS, DHL, and FedEx) handle worldwide delivery and customs regulations al-
most automatically. With these tools, small businesses can compete with almost
any large company.
Small businesses can even hire workers in foreign nations. Without leaving
your desk, you can use online worker-matching Web sites (e.g., rentacoder.com)
to advertise and fnd someone in a low-wage country to work on your project. Yes,
you have to stay up late at night to talk directly to the workers, but the tools are
available and if that is what it takes to save money and acquire the expertise you
need, it is a small price to pay. If you need a more traditional relationship within
a country, you can use the services provided by most state governments. Most
of the states have established exchange and international promotion programs in
various countries. You can use these programs to initiate contacts with foreign
government offcials, potential manufacturers, and distributors in various coun-
tries. Sometimes all you need is a phone number and initial contact to get started
in your search.
Entrepreneurship
How do you start a business? How does technology help? Entre-
preneurship is the act of building and running a business. The term is generally
applied to new businesses, but it is becoming common for large businesses to
encourage entrepreneurship within the main organization. For instance, a manager
who comes up with an idea for a product might seek support to run a new project
within the larger company.
Entrepreneurship is built on three broad fundamentals: (1) an idea, (2) a busi-
ness plan, and (3) implementation. Risk is a fourth important element. You should
not consider becoming an entrepreneur unless you are willing to deal with risk of
lossloss of time and loss of money. However, having a good idea, building a
solid business plan, and managing the implementation carefully can reduce risk.
Flexibility is another important characteristic of successful entrepreneurs. The
problem with having a new idea is that it is diffcult to forecast exactly how it will
be received by customers. Hence, many ideas and plans began in one direction,
and only succeeded when the managers used the information to change directions
and fnd a more proftable solution. Along the same lines, a thousand unforeseen
obstacles can leap into the path of any good plan; fexibility and perseverance are
important to circumventing these problems.
Idea
How will your business be different from the existing frms?
Ideas are the foundation of entrepreneurship. There is little point in starting a busi-
ness or project just to copy someone else. The idea could be a new product or
service, or it could be a better method of production, or a better marketing or
fnancial system.
Ideas are closely tied to strategy. As a start-up frm, yours will be small and
must have a clear focus. Are you trying to be the least-cost producer to attract cus-
tomers from older frms? Or are you planning to offer radically new products and
services that provide greater benefts than the competition? A successful strategy
will depend on a careful analysis of the industry and your role within it.
Strategy
The key as an entrepreneur is to examine the many aspects of strategy to fnd and
clarify an idea. For entrepreneurs, three key strategic issues are (1) an understand-
parts
supplier
parts
supplier
parts
supplier
warehouse warehouse
tool
manufacturer
Manufacturer
workers
wholesaler
wholesaler
distributor
retail store
Consumers
Figure 11.9
Production chain. When starting a business it is critical that you examine the entire
production chain to identify the 800-pound gorillas and fgure out where the profts
are made. If the niche you are looking at has too much rivalry, you might consider a
different spot on the chain.
ing of where you will stand on the production chain, (2) identifcation of the com-
petition and substitute products, and (3) barriers to entry.
As a new frm, yours will most likely be one of the smallest. Even if you start
an entirely new concept or industry, you will be dealing with an entrenched base
of suppliers and customers. As a newcomer, your business clout will be small, so
you will not be able to count on discounts or goodwill from your suppliers.
Figure 11.9 shows a generic production chain. When you are looking for ideas,
you should examine the production chain for various industries. Get information
on the leading frms in each step of the chain. Determine the concentration ratios.
Do four frms control 50 percent of the market at a given level, or are there many
small frms with no dominant player? Look at the fnal price and proft of the
product or service, and then trace backward and identify the various costs and
profts at each level. Then combine your analyses. For example, if there is a rea-
sonable proft at the consumer level, it might appear to be a good opportunity. But
if only a few dominant frms supply the product, then these frms might be think-
ing about expanding into consumer salesor they might make it diffcult for you
to create a new retail frm.
Even if you have an idea that creates an entirely new industry, your frm will
face competition. You must carefully identify your closest competitors and also
specify any potential substitute products. You can also use this analysis to gener-
ate new ideas. As a consumer, look at the products and services you buy and iden-
tify the main competitors and the potential substitutes. E-commerce specifcally
looks at the steps that consumers must go through to purchase an item. Can these
steps be simplifed? Can additional services or products be offered at the same
time using information technology?
Figure 11.10
Expand your focus. When searching for ideas, it might be better to partner with a
large frm instead of competing head-on.
customers
Big competitor
You might try to
compete directly.
It might be better to sell
your innovation as a service
to the dominant rm or to
be an intermediary for
consumers.
When searching for ideas, expand your focus to include different aspects of the
problem. For instance, perhaps you can create an expert system to help customers
select features of a product. You might think about creating a retail store or a Web
site to sell that product. But perhaps a few large frms dominate the retail side, or
it requires expensive advertising to enter the market. In this situation, as shown in
Figure 11.10, it might be more proftable to build your system and sell or lease it
to the existing retail frms. Or you could create a service Web site that other sites
can connect to and pay a fee for each use of your system.
When evaluating ideas, you must always consider the issue of barriers to entry.
If you do have a great idea, and your company makes a proft, how are you going
to keep rivals from entering your industry and taking away your customers? If you
create a new business process or new software, what will stop others from emulat-
ing your system? Chapter 10 examines some of the typical areas frms consider
to create barriers to entry. As a new, small frm, the economies of scale, capital
requirements, and control over distribution are not likely to apply to you, except
negatively. Also, remember that there is a fne line between creating barriers to
entry and violating antitrust laws. So far, most IT barriers have been considered
acceptable as long as you do not coerce people to use them.
In the United States it is still possible to obtain patents on business processes.
These patents were popular in the early days of the dot-com expansion, but the
patent offce began to take a closer look and deny some obvious ideas. If you have
a truly new process, you might be able to patent the conceptpreventing anyone
else from copying it for 20 years. Of course, a single patent can cost $10,000 to
$30,000 or more to obtain.
Research
Research is closely tied to idea generation. As you evaluate alternatives, you need
to obtain current data on several items. Figure 11.11 summarizes some of the ba-
sic data that you will need.
Figure 11.11
Business research. You need to collect data on competitors, the size of the market and
how it is growing, production costs, and the legal environment.
Competition
o Number
o Concentration ratios
o Sales by frm
o Technology plans
Size of the market
o Number of customers
o Growth rate
o Market comparison for substitute products
o Consumer focus group interviews
Production costs
o Startup/fxed costs
o Operating costs
Legal environment
Broad industry information can be obtained from various government Web
sites or publications. More specifc data can be obtained from the companies
themselves, if they are publicly traded. Sales data and more detailed comments on
rivals can often be purchased from marketing companies. A few companies moni-
tor Web site traffc, so you can obtain basic online activity data for some of the
larger frms. Customer focus interviews are important. At some point, you need
real-world feedback on your ideas.
Production costs and other hints can be obtained from suppliers and salespeo-
ple. If you are serious about developing a presence in a particular area, scour the
trade journals and fnd some of the leading suppliers. Call the regional sales repre-
sentatives and they will provide detailed information on items that you will need.
But be sure to compare prices from several frms.
Even for retail frms, several legal hurdles must be cleared. Some industries
have more complications than others, so you need to carefully investigate all laws
and rules that might apply to your business. Find out if there are restrictions on
what you will be allowed to do. In terms of permits, identify the permits you need
to obtain, exactly where to get them, how often they need to be renewed, the cost
of the permits, and the time frames between application, inspection, and approval.
Information technology can be used in several areas to help obtain and evalu-
ate research. Clearly, the tools examined in Chapter 9 can be applied to analyze
data. The problem for new frms lies in fnding the data. Today, much of the data is
online. Government agencies provide summary data by industry (particularly the
Census Bureau). The Securities and Exchange Commission (SEC) provides stan-
dard fnancial reports online. Other sites, such as WSJ.com and fnance.Yahoo.
com consolidate that data and provide tools to quickly compare and chart frm and
industry data. Other specialist frms track detailed sales data. For instance, you
can purchase detailed data from supermarket scanners for any level of products or
geographic location.
Another useful source of data is Googlebut in a slightly different way. Yes,
you can search for competitors. You can also use AdWords to estimate the poten-
tial response to different keywords. But, Google provides even more powerful
tools. Figure 11.12 lists the three main tools: Insights, Trends, and Zeitgeist. They
are related, but provide slightly different interfaces. Insights is probably the most
general. It is essentially a search of the keywords people are entering. It has tools
to look through categories, seasonality, geography, and other properties. You can
build custom searches to examine any set of words or features. For instance,
Figure 11.12
Find what people are searching for. Google has several tools that enable you to
search through the searches. You can see what items are popular and how the
searches change over time or season. Tons of data but they provide a view into
consumers minds.
Insights http://www.google.com/insights/search/#
Trends http://www.google.com/trends
Zeitgeist http://www.google.com/zeitgeist
you could examine how often and when people search for a particular prod-
uctand then see how those searches differ by time of year and location. The
zeitgeist site provides a more general look at things people are searching for
around the world. The main Twitter Web page provides a similar concept by dis-
playing the currently most popular tweet subjects.
Remember that people use Google searches for almost everything today. These
tools provide the ability to look into their minds and see what is important. If you
want to know if people might be interested in your product or idea, go see how
many people are searching for it. Because Google stores these statistics, you can
get a view of ideas and topics over time. (Did consumers respond to an advertising
campaign by running searches?) You cannot obtain searches made by specifc in-
dividuals, but marketing research is all about identifying patterns based on similar
locations and times. Oh, and this data is free of charge.
Plan
How do you turn an idea into money? Millions of people have ideas;
only a few are able to create new businesses and turn them into money. Once you
have identifed a reasonable idea and done the basic research so that you under-
stand the industry, you need to write a business plan. The purpose of the plan is to
create a road-map that will help you set up, manage, and evaluate your progress.
It is also critical to obtaining external fnancing. In 2000, the heyday of Web start-
ups, there were stories of entrepreneurs obtaining fnancing on the basis of a short
PowerPoint slide show. Those days are gone. A detailed business plan will con-
vince prospective investors that you are serious and that you know what you are
doing. It will also help them evaluate the true potential of your ideas.
Executive Summary
Strategy, Competition, and
Market Analysis
Forecasts, Cash Flow, and
Investment Budget
Marketing
Organization and Timetable
tasks
time
Figure 11.13
Business plan. A business plans helps you organize your ideas, provides a roadmap
and schedule, and provides fnancial targets to use as benchmarks as you move
forward. It is also critical for obtaining investment funding.
Technology Toolbox: Sending E-Mail Legally (CAN-SPAM)
Problem: You need to send e-mail to potential customers to advertise.
Tools: The CAN-SPAM Act of 2003 (Controlling the Assault of Non-Solicited Por-
nography and Marketing Act) 15 USC Chapter 103 defnes the rules you need to fol-
low to send commercial (advertising) messages.
E-mails that are transactional or relationship messages (i.e,. existing contacts)
are not required to follow the rules. The Act requires
(1) All header information to be accurate. Header information includes the origi-
nating e-mail address, IP address, and the from line. Spammers routinely al-
ter this information or hijack other e-mail accounts to make it more diffcult to
track them down.
(2) Subject headings must be accurate. Hard to defne but usually obvious. You
cannot deliberately lie in the subject heading. Spammers often use false subject
headings to convince people to open the message.
(3) E-mail must contain an electronic opt-out mechanism. It must be functional
for at least 30 days. You must also include a valid physical (postal) address.
(4) You must stop sending messages if a person opts out. You must stop send-
ing messages within 10 business days. The long delay is a problem for users
because they have to track their request date; but it gives senders a chance to
update databases with e-mail senders.
(5) No e-mail address harvesting. You cannot extract e-mail addresses from Web
sites, domain registration lists, random generation, and so on. Spammers com-
monly do these things, but it can be diffcult to prove, unless someone deliber-
ately sets up a trap.
(6) Sexually explicit messages must be identifed. An FTC interpretive rule re-
quires that sexually oriented messages must include the line SEXUALLY-EX-
PLICIT in the subject line. If people actually follow this rule, it is relatively
easy to defne rules that block those messages.
(7) The Act applies to the sender and the advertiser. Even if you hire someone
who sends the message, as the company or Web site being advertised, you can
be held responsible for violations of the Act.
The Act provides for fnes of up to $11,000 per violationwhich could amount to
millions of dollars. Enforcement is charged to the FTC, and the FTC is supposed to
be the one to take charge. Individual users do not have legal standingwhich means
that as a recipient, the most you can do is complain to the FTC or to your ISP. Cur-
rently, most cases have been pursued by Internet service providersnotably Micro-
soft. Paragraph (g) of section 7706 permits civil lawsuits by ISPson the grounds
that the spammers are stealing bandwidth.
Quick Quiz:
1. Has the Act reduced the level of spam?
2. Why would spammers risk violating the law?
3. What other provisions would you want to include?
You can purchase software that will help you organize the business plan, but
you must still collect the data and write the descriptive sections. You must also be
careful when using some software templates. When potential investors see plans
that are simple fll-in-the-blank templates with little additional content, they do not
believe you spent much time on the plan, and are not serious about the business.
As shown in Figure 11.13, the goal of the plan is to precisely describe the busi-
ness you wish to start (or expand), the market environment, and your strategy. You
must also include fnancial analyses using forecasted sales and costs. You should
include a timetable that indicates how the company will need to grow. Based on
these projections, you will be able to determine the amount of money you need to
raise to run the company over the next three to fve years. In terms of presenta-
tion, you must also include an executive summary that is a one-page review of the
major points.
Strategy, Competition, and Market Analysis
The strategy section is based on your research of the market. It contains several
subsections that describe exactly what products or services you will produce. It
should identify the major competitors and estimate the size of the market and how
it will change over time.
If you are creating or distributing products, you need to identify your suppli-
ers, including backup suppliers if something happens to your primary source. For
products, it is also critical that you describe your distribution network. Will you
distribute through standard retail stores? Ship products by UPS? If there are multi-
ple layers, it is particularly critical that you identify how you will track shipments
and sales through the process.
Forecasts, Cash Flow, and Investment Budget
The fnancial section is a primary component of the business plan. For a start-up
frm, it can also be one of the most diffcult to create. This section includes esti-
mates of sales and costs. You will have a separate section for start-up costs and
ongoing coststhis section is relatively straightforward, but you have to contact
several suppliers and contractors to get good estimates of the costs.
The more challenging aspect of the fnancial section is the need to forecast sales
by month for at least three years, and annual sales for fve years. Figure 11.14
shows that the sales forecast is the foundation for the other fnancial data. The
level of sales directly determines the revenue, the marketing costs, and the cost of
goods sold. Once you know the sales level, you can determine the scale of the in-
frastructure needed to support those sales, for instance, the size of the Web server
and Internet connection speeds in an e-commerce world, or size of distribution
facilities in a traditional retail environment. The size of the frm also determines
the number of employees needed, which identifes the cost of salaries. Salaries
can be a signifcant component of some frms. Note that for e-business frms, you
might require a larger number of contract employees to develop software in the
beginning. Once the system is operational, you may be able to run with a smaller
core group of employees. These costs should be recorded in a separate start-up
cost statement.
With sales, cost of goods sold, infrastructure costs (leases and so on), and
salaries, you have estimated the primary costs and can create projected fnancial
statements. You need to estimate a growth factor for each month or year. This
growth factor is one of the most subjective elements in the projection. However,
you should try to estimate growth rates of similar frms and keep your numbers
in a reasonable range. Also, higher growth rates will mean that you need greater
marketing expenses to obtain that increase in sales.
With the basic fnancial statements estimated, you can concentrate on cash
fows. When will money arrive? Will there be delays in payments? Many of your
costs occur up front or on a monthly basis, so calculate these and estimate the
frms cash position for each month. You will need a source of funds to cover
times when the cash fow is negative. You should do the same for proft, so that
you have an estimate of when the frm will become proftable.
Of course, you still face the problem of estimating the level of sales, which can
be next to impossible for new products or services. If there is no way to generate
a plausible sales forecast, it might be better to start with an estimate of the infra-
structure size. From there it is generally easy to estimate the fxed costs. Now, ex-
amine various levels of sales to pick up revenue, cost of goods sold, and market-
ing costs. As shown in Figure 11.15, you can now compute total cost and revenue
for varying levels of sales. The point where the two are equal (where the lines
cross) is the break-even point. You must reach this level of sales before you can
make a proft. Now, compare that sales number to similar frms. If the sales num-
ber is too high, it is unlikely that your venture will be proftable. Conversely, if it
is substantially lower than for other frms, you are probably overestimating the
price you can charge, or you are missing some costs. If the lines never crossyou
have a major problem: the frm will never be proftable based on your estimates.
Customers
and Sales
estimate
Infrastructure
scale
Employees
Sales revenue
Marketing costs
Operating
and selling
costs
Salary
costs
Balance Sheet
Income Statement
Prot and Loss
Cash Flow
Financial
statement
estimates
Financial
statements and
ratios
Figure 11.14
Estimating fnancial statements. The estimate of sales determines the size of the
infrastructure and the marketing and purchasing costs. The number of employees
needed can be determined from the size of the frm and provides the estimate for
salaries. With these major points, the fnancial statements can be projected to identify
the cash fow and the amount of investment money needed.
Marketing
As part of the business plan, you will have to create a marketing plan. The market-
ing plan will begin with the anticipated sales level. Then it will identify the target
customers with as much demographic detail as you can obtain. Income level is
critical. Regional location could be important for some businesses. It is also de-
sirable to fnd out what magazines, newspapers, television shows, and radio pro-
grams the consumers prefer. If you are selling to other businesses, you should
identify trade groups that are likely to represent the target businesses.
The marketing plan should contain an analysis of pricing. At a minimum, you
should survey consumers, or create a focus group, and fnd out how much the
potential customers are willing to pay for your product or service. You should
also identify points for quantity discounts, particularly if you are selling to other
businesses.
You then need to create an advertising plan. You need to fnd prices and viewer
demographic data for newspapers, magazines, radio, television, and Web sites.
You might also consider alternatives such as direct mail, billboards, and tie-ins
with other products. For local promotions, you can contact advertising managers
directly. For nationwide campaigns, you will want to hire an advertising design
and placement frm. While it increases your costs, the experience and contacts of
the frm will save you time and help focus your campaign.
You will also need to create a sales management plan, where you focus on
the internal structure of the marketing department. How many salespeople will
you need? How will they be paid and what additional incentives will you pro-
vide? How will you handle customer complaints? What tools will you use for
Revenue
Cost
Breakeven point
C
o
s
t
s

a
n
d

R
e
v
e
n
u
e

(
0
0
0
)
Sales (000)
200
100
80
60
40
20
10 20 30 40 50 60
Figure 11.15
Break-even analysis. If it is too hard to forecast sales, you can choose an
infrastructure size and estimate fxed costs. Then estimate variable costs and revenue
per unit sold. Compute total cost and total revenue for varying levels of sales. Look
for the break-even point. That is the minimum level of sales you must be able to
reach to be proftable.
customer relationship management? How will you identify and build cross sales
of products?
Online Advertising
As a startup or small business, particularly an e-business, online advertising can
be critical to your success. Online advertising has changed over timelargely led
by Googles innovations in charging for keywords. This process is explored in
Chapter 7. The main step is to analyze the returns from each ad so that you can
correctly set per-click prices as well as daily budgets. As more people search on-
line, the ads can be effective to introduce your Web site to new customers.
To reduce your costs, you want to fnd mechanisms so that customers return
directly to your site, instead of repeatedly going through Google. Every time a
customer clicks through Google ads, you have to pay again. It is better to get them
to return directly to your site. You can improve this process by selecting a Web
site address that is easy to remember and associate with your product. The draw-
back is that many simple site names are already taken, either by competitors or by
domain squatters who do not use the site but will be happy to sell it to you for an
exorbitant fee. Another useful technique is to add links to your site that make it
easy to add your Web site to the customers list of favorites.
Once you have a list of customers, you can send them e-mail messages an-
nouncing sales or special deals. Actually, the U.S. CAN-SPAM Act of 2004 al-
lows you to send e-mail messages to almost anyone. However, in all cases, you
must be careful to follow the rules of the law or you can be fned up to $11,000
per violation. You must also be careful about laws in other nations. Many U.S.
states also have laws against unsolicited commercial e-mail, but these laws are
generally superseded by the CAN-SPAM Act. Consult an attorney if you have any
uncertainty.
The basic provisions of the CAN-SPAM act are listed in Figure 11.16. The Act
has been completely ineffective at stopping or reducing unwanted e-mail (spam).
However, it is important to recognize that the Act makes it legal for businesses
to send e-mail messages to customers and potential customers. So, e-mail can be
Bans false or misleading header information. From, To, and routing
information must be correct.
Subject lines cannot be deceptive.
The e-mail must provide an opt-out method that works and these requests
must be honored within 10 business days.
The message must be identifed as advertising and must include the send-
ers valid physical address.
Sexually explicit messages must be identifed according to FTC rules.
Only ISPs (not individuals) have standing to sue spammersfor wasting
their bandwidth.
Figure 11.16
CAN-SPAM Act provisions. Although the Act is completely ineffective at reducing
unwanted e-mail, businesses should follow all of its provisions to avoid being fned.
used as a marketing toolas long as you scrupulously abide by all of the provi-
sions of the Act. Any deviation could be punished by fnes, but more importantly,
it will annoy potential customers and they will shop somewhere else. E-mail is
particularly helpful in dealing with existing customers and expanding their pur-
chases. Just be sure to include privacy notices and opt-out choices. There is no
point in sending e-mail to customers who do not want it.
Of course, because the Act was unable to make a dent in spam messages, your
real problem lies in writing subject lines and messages that will make it through
the spam flters on your customers e-mail systems. This question has no fxed
answer because spammers and flters continually change formats and rules. You
can keep separate testing accounts with major e-mail providers and check your
messages to see if they survive.
The other major issue you will face is that you should contact your ISP before
sending out hundreds or thousands of messages. Most ISPs have automated tools
to watch for potential spammers on their networks and they might shut down your
account. In most cases, the easiest solution is to contract with a separate compa-
ny to handle your outgoing e-mails. Several companies specialize in maintaining
your e-mail lists and sending thousands or millions of e-mail messages. How-
ever, be careful when evaluating these companiessome are just spammers in
disguise. Be sure that you work with a company where you provide your own lists
and ask for a list of clients so you can verify that they deal with real companies.
Organization and Timetable
You need to specify the organizational structure and create a project timetable to
provide a road-map and benchmarks so that you can evaluate your progress. For
the management, you need to identify who will fll each of the primary roles. If
you are using the plan to raise money, you need to include a brief biography of
each person. If there will be more than a handful of employees, you need to draw
an organizational chart. You also need to indicate how the structure will change as
the frm grows.
For complex start-ups, a project management timetable can be invaluable. The
thousands of management tasks from government flings, construction, purchases,
writing programs, hiring people, managing advertising, and dealing with suppliers
can quickly bury you in details. You need a timetable to record when tasks should
be started, and which tasks depend on others, and to track which tasks have not
been completed. A project management package is a useful tool to handle all of
these details. It can also track assignments by employees and has tools to help you
identify bottlenecks and times when you might need to hire additional people.
The timetable should also be integrated with the fnancial forecasts so that as
the company moves forward, you can evaluate your progress. If you do not reach
a certain sales level at the forecasted time, you can adjust your future growth rates
and recalculate the amount of investment money you will need in the future.
Implementation
Is it true that genius is 1 percent inspiration and 99 percent
perspiration? Yes, most likely, Edison was correct. Ideas and planning are
only the beginning of creating a business. The real work begins with implemen-
tation. The true job of an entrepreneur is to devise the rules and procedures that
make a business successful. Once the plans are in place and you know how much
it is going to cost to get started, you need to form the legal company and obtain
fnancing. You probably need to hire an initial staff, and you need to create an ac-
counting system to record all transactions and monitor your progress.
Starting a frm requires a considerable amount of paperwork. Some of the basic
steps are shown in Figure 11.17. Some types of frms require dozens of licenses,
and if you need to construct or remodel facilities, you will need additional per-
mits. One of the frst decisions you must make is to choose a state in which to
incorporate. Each state has different rules, procedures, and fees. Delaware is a
popular choice because of the way its laws are written. But small businesses may
fnd it easier to fle with the state in which they are located. Then the company can
be its own registered agent and can avoid paying franchise fees to two states. Even
if you incorporate in Delaware, you will still have to register in each state that you
have a physical presence. A few companies specialize in helping you incorporate
a new company for a fee. You answer a few basic questions, and the frm flls out
boilerplate articles of incorporation and bylaws and fles them with the state of
your choice.
Ownership Structure
One of the more diffcult decisions to make is the legal structure of the frm. Sim-
ple partnerships are relatively easy to create and to register with the state. How-
ever, partnerships generally cannot issue stock and it can be harder to protect the
partners from lawsuits. Corporations stand as separate entities and can issue stock,
but the accounting requirements are a little more time consuming, and you face
a double-taxation issue. Any money the company makes is subject to corporate
income taxes, and dividends that you pay to owners are subject to personal taxes.
Most states enable you to create a subchapter S corporation or a limited liability
company (LLC) to combine features of partnerships with those of corporations.
Most small business start-ups choose one of these two structures. With both
forms, income and losses fow directly to the owners income statements and are
only taxed once. Both protect the owners from lawsuits, as long as you keep a
Figure 11.17
Primary forms to start a business. Companies are incorporated by the state, and states
have different requirements and fling fees. Companies must also obtain an EIN from
the IRS. The S corporation election is optional, but popular with small businesses.
State forms
o Articles of incorporation
o Corporate Bylaws
o Registered Agent (self)
o Business Registration Form
State Employer Number
Withholding ID
Sales Tax ID
o Additional licenses
Federal forms
o SS-4 Application for Employer Identifcation Number
o 2553 Election by a Small Business Corporation
solid line between company business and personal funds. The primary difference
between them is that the LLC is not a corporation and cannot sell stock. The S
corporation is also easier to convert into a standard (chapter C) corporation. States
have different interpretations of the LLC, so your choice of structure depends on
the state in which you incorporate.
Financing
Obtaining fnancing is related to the type of business structure and the size of your
frm. The two choices are debt and equity (stock). However, as a start-up, you will
fnd it diffcult to fnd a bank willing to lend you money. Banks will lend money
for relatively liquid assets, such as inventory. But they will generally not lend over
80 percent of the value. Most start-up businesses fnd investors and grant them
partial ownership through shares of stock.
Debt
Firms can borrow money to fnance certain things, but it is diffcult to borrow
money for a start-up. Banks know that many small businesses fail within the frst
year, so they prefer to lend to an established business. Some banks specialize in
merchant loans to cover some of the costs of buying products that will be sold at
retail. But even in these cases, the company will have to provide cash to cover
some of the costs. Larger frms can sell bonds on the market for long-term debt,
but it is unlikely that anyone would be willing to buy bonds from an unknown
start-up company. Borrowing money also entails interest payments, so the cost
can be relatively high.
Equity
Most entrepreneurs search for investors willing to provide start-up capital in ex-
change for partial ownership in the form of shares of stock. Figure 11.18 shows
Reality Bytes: Sales are Good, Now Collect the Money
Ask almost any entrepreneur which problem they hate the most and they are likely
to say something about collecting money from customers. Not just sellingactu-
ally getting the customer to pay. Service-based frms typically contract the work and
then bill the client. Small frms are slow to pay because they are struggling with
cash fow. Big frms are slow to pay simply because they can. When the economy
turns down, delays get longer. Several online sites now make it easier and cheaper to
check on potential customers before signing contracts. Cortera (www.cortera.com)
provides a business directory and basic data about payment history. Larger credit-
reporting bureaus include Dun & Bradstreet (www.dnb.com) and PayNet (www.
paynetonline.com) which track payment history and provide basic credit ratings. All
of them charge fees for their services, but for it is possible to get basic data for rela-
tively low cost. Companies that sell to individual customers can also do background
checks, but most of the time it is easier to use standard payment methods including
credit cards and PayPal. Several tools make it possible to swipe credit cards using
cell phones to get immediate payment data almost anywhere.
Adapted from Raymund Flandez, Three Best Ways to Make Sure Customers Pay,
The Wall Street Journal, November 13, 2009.
that venture capital (VC) frms exist specifcally to provide funding to start-up
frms. An entrepreneur generally presents the business plan to a VC frm, with a
detailed budget and a request for the money to cover development and operating
costs for the frst year. VC frms evaluate hundreds or thousands of proposals in
terms of the strength of the idea, the ability and track record of the management,
and the potential profts. VC frms expect many of the small companies to fail, but
to cover the losses by having a few frms with enormous returns.
Once the frm is established and potentially proftable, the managers take the
frm public by issuing stock at an initial public offering (IPO). The public stock
raises additional money and eventually gives the entrepreneur and the VC frm an
opportunity to sell some of their shares for a personal gain.
In a hot market, start-up frms can often trade initial private shares of stock for
many items they need. For example, many managers are willing to accept smaller
salaries in exchange for stock options. Stock options are granted by the frm at a
specifc price. If the frm goes public and the stock price increases, the employee
buys the stock from the company at the option (low) price and sells it for the
higher public price. As an employee, keep in mind that this transaction is taxed as
current income and can take one-third or more of the proft. But you need to offer
stock options, because you need experienced employees willing to work for low
pay in exchange for future rewards if the company succeeds.
Although equity has many advantages, remember that you give some control of
the frm to the investors in exchange for their money. How much control depends
on your negotiating skills, on the number of frms interested, and on the amount
of money.
Figure 11.18
Start-up fnancing. Venture capital frms and partners are given ownership positions
and sometimes provide management control in exchange for development funding.
If the frm is successful, it issues an initial public offering of stock, which funds
additional operations and rewards the original investors.
Venture Capital
Angel Investor
Partners
Funding for devel-
opment and
operations.
Become owners with
some control over
management.
Successful rm IPO:
Additional funds.
Reward to original
investors.
Technology Toolbox: Creating a Business Plan
Problem: You want to start a business.
Tools: You can buy templates and software that will help you organize your ideas
and generate business plans. In most cases you are better off using a spreadsheet or
accounting software to design your own plan.
The overall structure of the plan should follow a standard organization: Introduc-
tion (a summary of the company), Marketing (market analysis and pricing), Historic
Analysis (sales and profts in the industry), Organization (structure of the frm and
the management team), Financing (detailed budgets and cash needs), and Projec-
tions (estimates of sales, costs, and accounting statements).
Forecasting is an important, but challenging aspect to creating a plan. You need
to create a complete set of accounting statements (income statement, balance sheet,
and cash fow) that refect the anticipated position of the frm for the frst few years.
The key to estimating these numbers lies in starting with the size of the company
in terms of sales. If Rolling Thunder Bicycles anticipates selling only 100 bicycles,
you could run the company with a couple of employees. If sales grow to 2,000 bi-
cycles, you will need more employees to build and sell the bicycles. Consequently,
you will need bigger facilities and more managers. In the case of RT, you would es-
timate sales of each type of bicycle and multiply by the average sale price to obtain
an estimate of revenue. The sales number also tells you the materials and tools you
need, which gives estimates of the major expenses. Now you can estimate the start-
up costs to build or lease space and buy equipment. Combining these numbers with
the sales revenue gives you assets and cash fow. Initially cash fow will probably be
negative. This number tells you how much you have to fnance, which means you
will have to include interest costs. When you build the spreadsheet, you should set
up a page of constants (such as percentage increase in sales) and build the formulas
to refer to these cells. You can change values to see the effect of your assumptions.
Building the accounting statements yourself (perhaps with the assistance of an
accountant), forces you to identify the primary fnancial information items. It helps
you see fnancial relationships and the internal structure that you will have to build.
Quick Quiz:
1. How can you forecast sales? What information would you want to collect?
2. How would the fnancial statements be different for an EC frm (for example, a
website that sells photographs)?
3. What key element would you place in the marketing section for a service frm
(e.g., dentist)?
Rolling Thunder Estimated Sales
Number of Bicycles
Year Increase Hybrid Mountain Race Road Tour Track Annual Total
1 250 250 350 200 350 50 1450
2 10% 275 275 385 220 385 55 1595
3 10% 302 302 423 242 423 60 1752
4 10% 332 332 465 266 465 66 1926
5 10% 365 365 511 292 511 72 2116
Average Sale Price of a Bicycle
$1,000 $1,500 $2,500 $2,000 $1,000 $2,000
Estimated Sales Value
Year Hybrid Mountain Race Road Tour Track Annual Sales
1 $250,000 $375,000 $875,000 $400,000 $350,000 $100,000 $2,350,000
2 $275,000 $412,500 $962,500 $440,000 $385,000 $110,000 $2,585,000
3 $302,000 $453,000 $1,057,500 $484,000 $423,000 $120,000 $2,839,500
4 $332,000 $498,000 $1,162,500 $532,000 $465,000 $132,000 $3,121,500
5 $365,000 $547,500 $1,277,500 $584,000 $511,000 $144,000 $3,429,000
Year
Balance Sheet--projected at year end 1 2 3
Assets
Current Assets
Cash ($193,550) $3,095 $302,395
Receivables 235,000 258,500 283,950
Inventories 98,700 108,570 119,259
Pepaid expenses 1,000 1,000 1,000
Total Current Assets 141,150 371,165 706,604
Property, Plant and Equipment
Land 0 0 0
Buildings 0 0 0
Fixtures and Equipment 250,000 300,000 350,000
Subtotal 250,000 300,000 350,000
Less accumulated depreciation 50,000 110,000 180,000
Net Property, Plant and Equip 200,000 190,000 170,000
Total Assets $341,150 $561,165 $876,604
Liabilities and Shareholders' Equity
Current Liabilities
Acccounts payable 82,250 90,475 99,383
Accured payroll and benefits 0 0 0
Income taxes payable (123,000) (284,100) (350,307)
Other current liabilities 0 0 0
Total Current Liabilities (40,750) (193,625) (250,925)
Other Liabilities 0 0 0
Long Term Debt 0 0 0
Total Liabilities (40,750) (193,625) (250,925)
Shareholders' Equity 0 (193,550) 3,095
Additional paid-in capital 0 0 0
Retained earnings (193,550) 196,645 299,300
Total Shareholders' Equity (193,550) 3,095 302,395
Total Liabilities and Shareholders' Equity ($234,300) ($190,530) $51,470
Money to be raised (equity or debt) $575,450 $751,695 $825,134
Cash Flow--projected Year
1 2 3
Net earnings $184,500 $426,150 $525,461
Depreciation 50,000 110,000 180,000
Net (gain) loss on asset sales 0 0 0
Other 0 0 0
Subtotal from sales 50,000 110,000 180,000
(Increase) decrease in current assets:
Receivables (235,000) (23,500) (25,450)
Inventories 98,700 9,870 10,689
Prepaid expenses (1,000) 0 0
Subtotal from assets (137,300) (13,630) (14,761)
Increase (decrease) in current liabilities
Accounts payable 82,250 8,225 8,908
Other current liabilities 0 0 0
Accrued payroll 0 0 0
Income taxes payable (123,000) (284,100) (350,307)
Total change in current liabilities (40,750) (275,875) (341,400)
Total adjustments (128,050) (179,505) (176,161)
Net cash provided by operations 56,450 246,645 349,300
Cash flows from Investing:
Expended for property, plant, equip (250,000) (50,000) (50,000)
Proceeds from sale of assets 0 0 0
Net cash used in investing (250,000) (50,000) (50,000)
Cash flows from Financing:
Proceeds (payments) from long-term debt 0 0 0
Stock or Additional paid in capital 0 0 0
Cash dividends 0 0 0
Net cash provided by financing 0 0 0
Net increase (decrease in cash) (193,550) 196,645 299,300
Cash and cash equivalents:
Beginning of year $0 (193,550) 3,095
End of year ($193,550) $3,095 $302,395
Income Statement--projected Year
1 2 3
Sales $2,350,000 $2,585,000 $2,839,500
Material 822,500 904,750 993,825
Labor 550,000 550,000 550,000
Lease 60,000 60,000 60,000
Advertising/Promotion 500,000 250,000 250,000
Tools depreciation 50,000 60,000 70,000
Cost of merchandise sold 1,932,500 1,764,750 1,853,825
Operating and Admin Expenses 100,000 100,000 100,000
Operating Profit 317,500 720,250 885,675
Other income (expense)
Interest Income 0 0 93
Interest Expense 0 0 0
Shareholder related expense (10,000) (10,000) (10,000)
Earnings before income taxes 307,500 710,250 875,768
Federal and state income taxes (123,000) (284,100) (350,307)
Net earnings $184,500 $426,150 $525,461
Accounting and Benchmarks
Careful accounting is an important requirement in a start-up frm. You need to
be particularly careful at tracking expenses. A good accounting system is impor-
tant, but you must also establish procedures and policies. In the hectic day-to-day
world of starting a frm, there is little time to stop and analyze every transaction.
You need procedures in place so that everything gets recorded as soon as possible.
If several managers have purchasing authority, a Web-based accounting system
might be helpful to ensure that all items are recorded immediatelywherever the
managers are located.
Your accounting system needs to run comparisons to benchmarks that you es-
tablished in the business plan. For example, if your cash fow is running below
projected levels, you will need to cut expenses or fnd additional funding. You
need to closely monitor these numbers so that you have more time to react and
make corrections.
If your frm requires developing software, you need to track the development
progress. Estimating design and programming time is notoriously diffcult, but
you should still track the progress because you will need to update your target
completion dates. Project management software shows how each task depends on
others, and it can identify bottlenecks and highlight which tasks need additional
resources.
The entrepreneur also needs to provide feedback to investors. Beyond standard
quarterly accounting reports, you will need to keep investors apprised of develop-
ment progress, marketing campaigns, and sales data.
Reality Bytes: A Part Time Job as an Entrepreneur
Jay Brewer had worked as a Web designer at several startup companies. In 2001, he
formed Blogpire, working on the site in his spare time. The site consists of a col-
lection of blogs where people write opinions about consumer products that fall into
specifc niches. He began by writing reviews about single-serve coffee machines.
The site took off and he started including ads from Google and Amazon. Within a
month, Aloha Island Coffee agreed to advertise on the site. Mr. Brewer notes that
all of a sudden I had a little business. The site was making money from Google
ads and Amazon, and it had its own direct advertiser. He moved on to launch three
more blogs (kitchencontraptions.com, justthechips.com, and shavingstuff.com; all
while continuing his day job. He noted that you have to be disciplined and make a
schedule and really get it going each day. By March 2005, his blogs kept increasing
in traffc and revenue so he quit his regular job. His startup company created new
blogs, hired writers, and brought on more direct advertisers. With 14 blogs, 750,000
visitors a month, and 30 direct advertisers he makes more money than at his previ-
ous jobs. The busiest sites can pull in $10,000 a month in revenue. Mr. Brewer noted
that even with smaller volume, this works because we are geared toward consum-
ers that are looking to research something they want to buy. And for that reason, we
get [a higher quality customer] than other sites, and that makes advertisers happy.
His writers receive 50 percent of the revenue for a blog, and vendors now offer free
products for review.
Adapted from Stephen Grocer, Consumer-Product Blogs Spark a Web Empire,
Starting an e-Commerce Firm
What additional steps are required to start and EC frm? An EC
start-up faces the same paperwork, fnancing, and organizational issues as any
other frm. However, as shown in Figure 11.19, frms face additional tasks in
terms of setting up the online elements. Today, these tasks are straightforward for
selling products online. Chapter 7 explores the options in detail, but the simple
answer is to pay someone else to run the servers and handle payments. The costs
are relatively low and they are typically variable costsperfect for startup com-
panies. Most systems include scalability so it is easy to expand as traffc increases.
Entrepreneurs are free to focus on the harder task: marketing.
Life is more challenging for new online businesses that want to develop ser-
vice software, Web 2.0 social interactions, or mobile phone applications. To at-
tract customers, these services need to be new and unique; which requires writing
custom programs. Programming tools continue to improve, so programmers can
be more effcient. However, the more unique the system, the harder it will be for
others to copy, but it will be more diffcult to create. A key element in creating
new online services is the ability to hire, organize, and evaluate programmers.
These topics are explored in Chapters 12 and 13.
Sites and services that have a social network component face another huge
problem: attracting a large set of initial users. These types of services succeed
only by attracting a large number of participantsknown as the network effect.
So, any newcomer must determine how to attract attention away from existing
frms and fnd a way to encourage people to join its network instead. And it will
take time to reach a critical threshold, so the frm must have funding sources that
will carry it far enough to reach that point.
Products
* Use simple HTML, Amazon,
eBayorWebcommerceserver.
* Process payments through PayPal,
Google,orsimilarthirdparty.
* Hosted server is inexpensive
today.
* Build links to get picked up by
searchengines.
Services/Web 2.0
* Hosting: Use a hosted server or
cloudcomputing.
* Software Development
* Do you have the skills to
hire, organize, and evaluate
programmers?
* Will the code have to be
updated and rewritten
continually?
* Can the software be built using
existing modules?
Figure 11.19
New businesses online. Many tools and companies make it straightforward to sell
products online. In general, the hardest part will be the marketinggetting your site
recognized. Building new online services or phone applications is more challenging.
You will need to hire and pay developers and manage the development process.
Analysis of Dot-Com Failures
Why did thousands of dot-com frms fail? This question frst arose in
2001 and 2002. The question was important at that time because of an initial be-
lief by many frms that it was possible to sell products online and survive without
making a proft. The crash changed that beliefat least for a while. Then the in-
teractive Web 2.0 came along and new frms were created. Once again, companies
and investors believed that money could be made online even with no revenue.
The two situations are different but have some similarities. It remains to be seen
if social networking frms can survive in the long term. And the same question is
likely to be raised in the future as new technologies are created.
From about 1996 to 2000, hundreds of dot-com frms were created, many in the
San Francisco area. The excitement of the Internet led people to believe that these
frms were the start of a new economy. Overhyped statements were made about
the death of the old economy and that traditional bricks-and-mortar frms would
soon fail, to be replaced by an online world of competitive prices and advertise-
ments tailored to individual customers. Entrepreneurs believed that if they could
be the frst frm to break into a category, and if they advertised heavily enough,
they would automatically become the dominant player in the new economy. Many
investors felt it was important to get in on the ground foor of these frms. IPOs
were released daily; stock prices soared. Newly minted paper billionaires graced
the covers of business magazines. And then investors woke up and the market
crashed. The NASDAQ market index that covered many of these technology frms
dropped from over 5,000 points to around 2,000 in less than a year. Pundits whip-
sawed to the other end of the spectrum and proclaimed the end of e-commerce. Of
course, reality lies between the two endpoints, but it is worth examining some of
the concepts of the time to understand the role of e-commerce in the future.
Pure Internet Plays
One of the frst types of frms to fail followed a strategy known as pure Internet
plays, where the e-commerce frm relies entirely on Internet transactions for mon-
eywith no ties to real products. Examples include sites that provide services to
other sites, such as the search engines and Web advertising sites. Closely related
sites include some that advertised and sold products over the Internet, but relied
on other traditional companies to produce and deliver the products to customers.
Reality Bytes: Pay Your DNS Bills
The Washington Post newspaper somehow missed renewing a payment on its do-
main name (WashPost.com). Because of the overdue bill, the domain name was del-
isted and readers, writers, and editors were unable to use the e-mail system that relies
on the domain name. After identifying the cause, the newspaper quickly renewed the
registration, and services were restored. The company was fortunate. Many of the
domain registrars allow others to place a hold on a domain name that is not renewed.
If you somehow do not renew a domain name, it could be immediately acquired by
another company.
Adapted from Weiss, Todd R., Overdue Domain Registration Bill Stops E-mail Ac-
cess to Washington Post, Computerworld, February 6, 2004.
These frms were at risk because they depended almost completely on Internet
traffc and funds. When a few of the frms failed, it set off a domino effect that re-
duced cash fow at many other frms. Many of the frms were interlocked with the
others through service agreements and advertising relationships.
Proft Margins
Proft is an important issue for any frm. Yet many managers claimed that there
was no need for traditional profts. This belief was endorsed by a market where
IPOs were celebrated and stock prices on new frms jumped over $100 per share
in the frst day. These IPOs provided the cash that frms used to fund operations.
In many cases, the goal of the frm was to advertise heavily and undercut prices
in an attempt to build name recognition and capture market share. Even Amazon.
com, one of the early entrants and cheerleader for e-commerce, gained its promi-
nence by offering products at 30, 40, and 50 percent discounts. The CFO at Ama-
zon once said that they never intended to make a proft. More recently, the frm
has been working hard to reduce costs to achieve an operating proft. Much of the
proft is the result of a change in strategy: substantially higher prices.
One of the challenges to measuring profts is that many different defnitions of
proft can be used, largely depending on which costs are included. But it is dif-
fcult to make a proft when products are sold for less than the cost of the item. It
Outside
advertising
$
Outside
advertising
$
Cross
advertising
Much of the cross adver-
tising was not real money,
simply agreements that
were sometimes recorded
as revenue.
When the outside money
declined, and weak rms
crashed; others crashed as
well. Then they ran out of
investors money.
Figure 11.20
Dot-com advertising crash. Many of the early frms were linked by agreements
to advertise on other sites. With few actual sales, some frms ran out of investors
money and the outside advertising money dried up. The weak frms crashed, which
cut off the advertising money to others, and the domino effect accelerated the crash.
can be done, if additional revenue is received from other sources, such as advertis-
ing revenue, service contracts, or additional vendor payments. But, in the end, a
frm must make a proft on operations or there is no point to continuing. In 2000,
a McKinsey study revealed that only 20 percent of the e-commerce frms were
making a proft on sales, and the top frms were actually traditional frms, such as
mail-order companies.
Many of the early e-commerce frms attempted to establish market dominance
through heavy advertising and large discounts. The plan was to use the name rec-
ognition to attract customers away from traditional frms and eventually increase
prices to make a proft. The problem was that these frms still had to compete
with traditional frms that already face considerable competition. In this crucible,
the existing frms learned to squeeze costs. These effciencies enable them to sell
products at a proft and to fend off competitors during short-term attacks.
Advertising Revenue
As shown in Figure 11.20, Internet advertising revenue was a dominant factor
in the failure of many of the early e-commerce frms. For some, the Internet ap-
peared to be a new educational mediummuch like television. Web sites would
provide interesting or useful content that would attract viewers (often referred to
as eyeballs). These sites would then sell screen space to other companies to use as
advertising. The interactive nature of the Internet would also make it possible to
track the effectiveness of the adsboth in terms of the number of times an ad was
seen and in terms of the click-through rate for each ad.
This model does apply to some extentthe Interactive Advertising Board re-
ported that U.S. Internet advertising revenue was $8.2 billion in 2000. However,
only a handful of the top frms receive the lions share of the revenue. Moreover,
as click-through rates fell in 2000, advertisers began to believe that browsers no
longer paid attention to the ads, so they began to question the value of advertising
on the Internet and became particularly leery of smaller sites.
The situation was compounded by the fact that many of the frms followed
some shaky accounting practices. For instance, advertising exchange agreements
were often counted as revenue. Firm A would agree to carry advertisements for
frm B in exchange for similar ads on Bs site. Both sites recorded revenue from
the advertisementsbut never received any cash.
With the declining emphasis on advertising revenue, content providers are in-
creasing their attempts to charge consumers for the content. According to a study
by Jupiter Media Metrix, one of the leading e-commerce monitoring frms, in 1999
the top 100 frms were charging for only 6 percent of their content. In 2000, that
number almost doubled to 11 percent. Almost all of these frms anticipate charg-
ing for some content by 2003. Some of the content is paid with service agree-
ments, some through micro-payments for individual pieces of data.
The online world changed considerablypartly from the early crash, but also
because of Google. Remember that Google controls most of the online advertis-
ingbut it enables any Web site to participate. One of the positive effects of this
control is that online advertising money is real and not just an accounting entry.
Still, more frms exist as purely online today. Between social networks, news sites,
and dozens of other types of service-oriented sites, thousands of Web sites depend
almost completely on advertising revenue. Others, particularly Twitter, have yet
to fnd a mechanism to generate signifcant amounts of revenue. Yet, investors are
putting huge amounts of money into some of these frms; so they must believe that
the companies can eventually be proftable. It is diffcult for outsiders to evaluate
the situation when the frms remain private and do not publish fnancial data.
News sites are an interesting example of the challenges. As the traditional
newspapers lose print circulation, they are searching for new ways to survive.
For several years, most newspapers offered online content free (except the Wall
Street Journal). In 2011, led by experiments by a few smaller papers, the New
York Times began charging for some content. The risk is that with free content
elsewhere, users will simply disappear instead of paying. So the paper could lose
advertising revenue because of a smaller number of readers. Selecting a revenue
method is still a diffcult problem for online sites.
Cloud Computing
How does cloud computing help startup frms? One of the main
strengths of cloud computing is that it is designed to price technology as a vari-
able cost. Firms can establish servers and Web sites with no fxed costs. These
sites can be scaled to almost any sizesimply by paying more fees as the usage
rates increase. Cloud computing makes it easy for a startup frm to obtain any
level of computing. Firms still need to purchase personal computers, but these can
be relatively inexpensive. At the extreme end, today a company can run all soft-
ware online, including standard word processing and accounting systems. Even
frms with dozens or hundreds of employees could follow this approach. The per-
sonal computers would just run browsers and would essentially be disposable. All
software and data can be stored online. Technical support, upgrades, security, and
most other tasks would be handled by the online providers. The amazing part is
that the costs are all variable, and the frm does not need to come up with large
sums of money up front. As the company grows, more capacity is added and the
costs increasebut presumably, the increased business revenue will cover the
new costs.
In less extreme cases, some companies will still have to purchase and run com-
puters. For example, a simple restaurant will likely need a turnkey computer sys-
tem. However, even in these cases, most of the elements can be leased by paying
a monthly fee. And technical support is provided by the company that created and
installed the system. The cloud can be used in these situations to provide backup
and access to the data for external owners and managers.
Many frms might want to run their own transaction systems. In these cases,
cloud computing is useful for higher-level tasks such as teamwork and data anal-
ysis. Tools for meetings, sharing data, and even some data analysis and expert
systems can be run as cloud services. Some enterprise features such as customer-
relationship management are also readily available online. The point is that en-
trepreneurs can fnd many levels of tools in the cloud and pay for them as they
need them. This availability makes it possible for companies to compete at a much
higher level than in the past.
Summary
Small businesses need information systems as much as any company. However,
small businesses face constraints on money, limited number of workers, lack of IT
expertise, and strategic power. Many small businesses look at IT as a back-offce
expense that is a necessary evil but rarely contributes to making money. Con-
sequently, many small businesses under invest in information technology. With
declining technology costs and increasing options through online services, more
opportunities exist to use IT to reduce costs and improve productivity.
Small businesses, particularly startups, need to focus on using IT for transac-
tion processing and day-to-day operations. Most will use basic accounting sys-
tems to record fnancial information. Small businesses should increasingly look
at online ERP services to provide more sophisticated tools with minimal effort
and fxed monthly fees. Beyond choosing an accounting system, managers need
to focus on the type of data and information they want to extract from the transac-
tion system. HRM systems are complex because of the constantly changing local,
state, and federal laws. Consequently, most businesses pay a service to handle
payroll and other HRM transactions.
Entrepreneurship requires a good idea and a detailed plan, followed by imple-
mentation. The plan should include fnancial details and budgets. You then have to
raise the money needed to start the business and keep it operational until it gener-
ates suffcient revenue to be self-sustaining. Developing online e-businesses can
be straightforward if you just need to sell products. Developing new systems with
completely new software is far more complex and requires the ability to manage
development projects.
A Managers View
Small businesses face the same decisions and technology issues as large
companies. But they have fewer workers, less money, and less expertise. In-
formation technology can provide many benefts including reduced costs and
improved productivity. However, all technology has to be evaluated careful-
ly. Hiring MIS workers is an expensive fxed cost, so small businesses will
often want to examine turn-key and online service frms instead of trying to
run in-house technology. Creating a plan and successfully implementing it
are critical steps in entrepreneurship and e-business frms.
Key Words
Web Site References
Angel Investors and Venture Capital
Angel Investor Magazine www.spencertrask.com
Business Finance www.businessfnance.com
inc Advice and Lists www.inc.com/guides/fnance
National Venture Capital Association www.nvca.org
Venture Capital Resource Directory www.vfnance.com
Common Statistics
Bureau of Labor Statistics www.bls.gov
Census Bureau www.census.gov
FedStats www.fedstats.gov
General Reference Sites
Library of Congress thomas.loc.gov
Congress www.house.gov
www.senate.gov
Congressional Quarterly www.cq.com
Copyright Ofce www.copyright.gov
Executive Branch www.loc.gov/rr/news/fedgov.html
IRS www.irs.gov
Judicial www.uscourts.gov
Legislative votes www.vis.org/toolbox/default.aspx
Patent Ofce www.uspto.gov
CAN-SPAM Act
entrepreneurship
initial public offering (IPO)
limited liability company (LLC)
network effect
pure Internet plays
stock options
subchapter S corporation
turn-key systems
venture capital (VC)
Review Questions
1. What are the primary constraints facing small businesses?
2. How can small businesses deal with a lack of IT expertise?
3. How do most small businesses acquire IT to handle operations?
4. How could small businesses use affordable technology better at the tactical
level?
5. What are the benefts to designing an accounting system while you are
planning a new business?
6. How can IT be used to help start a business or fnd opportunities?
7. How do the elements of strategic analysis relate to entrepreneurship?
8. What are the main choices for frm structure and fnancing?
9. What additional steps are needed to start an EC frm?
10. Why did so many dot-com frms fail in 2000/2001 and is it likely to happen
again?
Exercises
1. Consider a small service frm such as a physician, dentist, accountant,
or lawyer. Is it possible for such an offce to use computers to gain a
competitive advantage? To start, identify the customers, suppliers, and rivals.
Do you think the natural switching costs are high or low; that is, how often
do customers switch to competitors? Which of the major techniques do you
think would be the most successful (barriers to entry, switching costs, quality
control, lower prices, ties to customers or suppliers, etc.)?
2. Write a business plan for a new company. Choose an existing small company
if you do not have ideas for a new frm.
3. Research the detailed steps needed to start a Chapter S corporation in your
state. Obtain the necessary forms (most states have them on Web sites).
4. Visit a small business or talk to an owner/manager (perhaps a relative).
Identify the use of information technology and write a short report on how
the frm could beneft from using additional technology, particularly for
decision making and collaboration.
5. If you were starting a new restaurant, would you choose a franchise or start
an independent operation? Explain why.
6. Interview a manager of a small business (< 500 employees), possibly a
friend or relative. Identify the use of information technology in the three
decision levels (operations, tactics, and strategies). Just ask about IT, do not
expect the manager to be able to classify the technology for you. Who does
the manager use for IT advice and support? How many MIS employees are
there? Where does the manager believe the company lies on the levels of
technology (leading edge, turn-key, bare-bones)? Where would you classify
the company?
7. How long would it take you to build a spreadsheet/Access application that
could be used to forecast a daily count of customers at a small restaurant?
If you could sell the application, you much do you think you could charge?
Now search the Internet to see if such an application exists.
8. If you (or a friend) want to create a new smart phone application but do not
know how to program, how will you fnd a programmer or two and how
much will it cost? Why would a programmer work for you instead of creating
the application independently?
9. Search the Web and fnd a site or tool that will help you create a business
plan. What features could this tool provide that would be helpful?
10. Assume you want to run a Web site that survives on advertising revenue.
Estimate how much money Google pays for advertising clicks and the
number of visitors per month you would need to generate $1,000.
Technology Toolbox
11. Collect several commercial e-mail messages and count the number of
violations of the CAN-SPAM Act provisions.
12. What provisions could you add to the Act that might actually reduce the level
of unwanted messages?
13. If spammers violate all of the provisions of the Act, how can you track them
down?
14. Select a small business that you might want to start. Choose the type of
business structure and where it will be located. Explain your choices.
15. Choose a company that you would like to start, write the overall strategy
section, and build the projected accounting statements for the frst three
years.
16. Assume that you have been hired by a physician who wants to start a Web
site to help dieters. Create a business plan that focuses on the competition,
marketing, and Web aspects of the business.
Teamwork
17. Choose a frm that the team might want to start. Create a brief business plan
for the company by assigning one section to each team member.
18. Assign each person to investigate a different accounting system, including
online, that could be used for a small business. Write a report that compares
the features, costs, and limitations of each package. If possible, recommend
one for someone with a small business.
19. Select an industry that has several many small businesses such as restaurants
(or have one assigned by your instructor). Find at least two turn-key systems
that are available for businesses in this industry. Without harassing sales
representatives, compare the systems in terms of features, price, and market
share. If possible, talk to a small business owner who uses the system for
additional comments.
20. Find at least two companies that process payrolls for small companies.
Compare their features, and if possible, market share and prices.
21. Construct an e-mail ad that meets the CAN-SPAM criteria. Send the ad to
each other and see if it makes it through the anti-spam flters. Try using a
non-university e-mail account. Report on the results.
22. Find at least two systems that could be used to run a restaurant of about 60
tables. Determine the number of terminals that would be needed. Compare
the two systems in terms of features, and select the better system.
23. Using the existing data, write a business plan to obtain venture capital to
expand the operations of Rolling Thunderfocusing on the need to develop
a marketing campaign and a Web-based ordering system.
24. Outline a plan to develop and create an e-commerce site for Rolling Thunder
Bicycles. Estimate the costs to develop and host the site.
25. Assuming the salaries and capital costs are fxed to start Rolling Thunder
Bicycles, and assuming the average price of a bicycle is $2,250, compute the
break-even number of bicycles.
26. Describe the level of technology that would be appropriate for Rolling
Thunder Bicycles as a startup. What would be an approximate budget?
27. Can the company justify hiring a full-time MIS employee? If not, how should
the MIS tasks be handled?
Additional Reading
Small Business Administration, The Small Business Economy for Data Year
2005, December 2006. http://www.sba.gov/advo/research/sb_
econ2006.pdf. [Basic small business statistics and trends.]
Hal R. Varian, Technology Levels the Business Playing Field, The New York
Times, August 25, 2005. [Comments on the ability of small businesses to use
technology to use global resources.]
Cases: Entrepreneurship
Te Industry
Entrepreneurship is the practice of starting a new business. The business might
start small with one or two employees, become a large corporation, or anywhere in
between. In many ways, the American economy has been driven by entrepreneurs,
people with vision and independence, willing to take risks and try new ideas and
new technologies. But starting a business requires more than just ideas and enthu-
siasm. It requires detailed organizational skills. The entrepreneur is responsible
for creating all of the many procedures and rules that will keep the business run-
ning. You have to determine simple things like how the daily receipts get depos-
ited each day and how employees are evaluated. You also have to determine all of
the big-picture issues, such as the best way to counter a new plan by a competitor
0
0.5
1
1.5
2
2.5
19941995199619971998199920002001200220032004200520062007200820092010
$

B
i
l
l
i
o
n
Annual Revenue
PacWest
Facebook
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
19941995199619971998199920002001200220032004200520062007200820092010
R
a
t
i
o
PacWest
Facebook
and keeping your customers happy. Several familiar companies began as small
frms and grew to dominate their respective industries and even the economy. The
large computer frms of IBM and Hewlett-Packard were once small companies.
The Edward Jones brokerage frm was started by one man.
In 2000, slightly over 5 million corporations fled income tax returns in the
United States, along with about 2 million partnerships and almost 18 million pro-
prietorships. About 12 million of those proprietorships had less than $25,000 in
receipts, compared to 1.2 million of the corporations and 1.1 million partnerships.
Most of the corporations had receipts between $100,000 and $500,000. The total
number of corporations increased from 3.7 million in 1990 to 5.0 million in 2000.
At the same time, partnerships increased from 1.5 million to 2.1 million and pro-
prietorships from 14.8 million to 17.9 million. The actual number of new frms
created will be higher because the totals do not count the frms that disappeared
in that time. Looking only at 1999 to 2000 (the most recent data available), the
number of all frms increased by 559,000 (a 2.2 percent increase) (U.S. Census
Bureau 2002).
Te Internet
The Internet raises the hopes of many of todays entrepreneurs. You can use it
to reach a large audienceperhaps even worldwide. You can start small with a
limited budget. But as the dot-com crash of the early 2000s shows, you still need
a business plan. You need an idea that will generate profts. Even small expenses
can overwhelm a company that has no sales revenue. People are still drawn by the
examples of companies like Amazon.com, Yahoo, and Google. Of course, many
forget about the hundreds of other Web start-ups that failed. Some of those com-
panies actually had good ideas and plans. Part of their failure was due to timing
entrepreneurs and investors thought the world could be changed overnight.
Financing
Starting a business requires moneystart-up capital. Depending on the industry,
you might need to purchase buildings or equipment. You might have to pay pro-
grammers to develop software. You generally have to rent an offce and rent or
buy furniture. You most likely need to buy computers, along with offce supplies.
You might have to pay for licenses or patent fees. You most likely need to pay
employees until the business generates enough cash to cover the day-to-day ex-
penses. Several sources of funding exist, but most of of them have drawbacks. If
you can get by with small amounts of money, the best answer would be to save
it and invest yourself, or with family members. Larger amounts of capital can be
obtained from angel investors or possibly venture capital frms. Sometimes you
can borrow money from a bank, but those amounts are generally limited to spe-
cifc items such as inventory or capital goods that can be resold if your company
fails to survive. No one gives you money for free. Banks require you to pay inter-
est. Venture capital frms often ask you to give up some control, and they install
a manager. In most cases, you give up some ownership, as the investors gain a
negotiated percentage of your company.
Additional Reading
Bounds, Gwendolyn, Lessons of Successand Failure, The Wall Street
U.S. Census Bureau, Statistical Abstract of the United States 2002, Washington,
D.C., 2004 (http://www.census.gov/prod/www/statistical-abstract-02.html).
Case: Petz Enterprises Inc. (TaxBrain)
Leroy E. Petz, Sr., founded Petz Enterprises in 1964 in Tracy, California. Until
2000, the company specialized in providing tax software for accounting frms.
The CrossLink package was commonly used by accountants to complete and print
income tax returns for clients. The accounting frms, or local tax offces, buy the
software annually and then use it to produce even complex returns for customers.
The accountants provide advice and organize the customers data. Clients often
arrive with a shoebox full of receipts and the accountant has to sort it and cat-
egorize the data to determine what items are deductible and classify the income
properly.
In 2000, the company realized that the Internet was going to become an impor-
tant method for individuals to fle taxes. The company created the TaxBrain Web
site using its knowledge of the tax laws (company Web site).
CrossLink
As its original fagship product, CrossLink is an important source of revenue and
name recognition. Petz has continued to improve the product over time to make
it easier to use and to provide features that are useful to large tax-preparation of-
fces. For example, returns that are fled electronically are initially evaluated by
the IRS. If there are errors, the CrossLink package picks up the messages and
integrates them into the form so that the tax preparer can quickly see the problems
with a complete diagnostic statement. In 2001, the company also added the Tax
Return Logging System. Petz notes that tax offces with branches need to keep
track of all tax returns. The Tax Return Logging System will track all returns that
have been created, deleted, and changed (McCausland and Lombardo 2001). The
potential problem is that in a large company, a tax preparer hired in a remote loca-
tion might use the software to generate a clients return, print it, and pocket the
money. With the logging system, all use of the system is tracked.
In 2003, Petz rolled out a new approachhosting the tax preparation software
on its own site. Charles Petz, product development director, notes that the new
product (VTax) is a subset of our professional [CrossLink] product, but were
going after guys who have a lot of offces (Petz Enterprises 2003). In particu-
lar, the system is geared for the hundreds of neighborhood electronic refund origi-
nators (EROs) that have sprung up in the past couple of years. The companies set
up small storefront operations, and customers show up with their W2 form and
basic fnancial information. Clerks enter the data into the system and give the cli-
ents an immediate refund. Of course, the refund is really a short-term loan and the
actual refund is collected by the ERO, minus the interest and fling costs. The key
is that the ERO practitioner has all of the data up-front to make the loan and can
fle the data immediately or hold it until later if additional information is needed.
By using a hosted service, the ERO offce needs only a simple computer with a
Web browser and a basic Internet connection. Even better, all offces of the ERO
have access to the form. A client could drop off paperwork at one site and return
to a different one if needed. The system also prints detailed reports of activity
by offce. Originally, Petz anticipated 500 to 600 offces using the system. In the
end, it looks like well be closer to 2,000 locations, through 10 different EROs
(Petz Enterprises 2003). Another strength of the hosted service is that the ERO
can quickly set up a new offce if demand is high. V-Tax appeared to appeal to
practitioners because of its ease of use and the ability to open new service offces
with just a Web browser and an Internet connection. (Widmer 2007 and Account-
ing Technology 2006).
In 2010, Petz continued to expand the CrossLink professional suite. The com-
pany opened offces in Bellevue, WA and Rome, GA to help support the product
and extend coverage nationwide.
TaxBrain.com
In 2000 and 2001, the TaxBrain site saw limited commercial success. For a fat
fee of $24.95, individuals could fle federal and state returns. The electronic fling
fee was included in the cost. By fling electronically, flers could get a refund in
as little as 10 days. Since they did not have to buy and install software, customers
needed little computer knowledge. Mr. Petz noted that although the company did
only limited marketing, we saw over 10,000 hits a day and we have no idea how
people found us (McCausland and Lombardo 2001).
In 2002, Petz got more serious about the TaxBrain site. The company com-
pletely reworked the user interface to make it easier to use. It also began a more
targeted marketing campaign to bring in more users and track where they were
coming from. In November 2002, Petz observed that in the 2000 tax year, 130
million individual tax returns were fled, with 17.7 percent of those from off-the-
shelf software. By 2005, he forecasted that percentage to drop to 7.1 percent. In
2000, 56.7 percent of individuals paid a professional to fle their forms; his fore-
casts show that by 2005 that number will drop to 48 percent (press release 2002).
By 2003, the site was the third most popular tax system on the Internet, behind
only the giants H&R Block and Intuit (TurboTax). At its peak on April 15, 2003,
TaxBrain attracted 50,000 unique visitors (press release May 2003).
At the same time, the IRS saw a huge increase in the number of people fling
returns electronically, with 53 million fled electronically in 2003 (IRS Web site).
Leroy Petz Sr. explains the popularity, For millions of Americans, fling over the
Internet is now accepted as the best way to go to beat the deadline. TaxBrain can
help you get your taxes fled in less than an hour, provides assistance through live
instant messaging chat and ensures your return is received by the IRS with an of-
fcial electronic reply. Petz also estimates that individuals are saving more than
$200 in time and professional fees doing their returns online (TaxBrain 2004).
Electronic fling is also better at catching errors, particularly since the IRS system
catches most common problems and identifes them immediately. Overall, about
20 percent of paper returns have errors, while less than 1 percent of electronic
returns have errors.
Although electronic fling is increasing, much of the growth is through third-
party systems, such as the local offces and purchased software. For 2004, Petz
initiated a larger marketing campaign. Leroy Petz, Sr., notes that in order to make
people more aware that they can prepare and fle their taxes online, PEI is embark-
ing on an aggressive media campaign to inform taxpayers of the ease, speed, and
security of doing their taxes with TaxBrain. Our biggest challenge is demonstrat-
ing to the taxpaying public that it is truly easy (press release December 2003).
Because of the companys background in writing software for professionals,
TaxBrain benefts by being up to date. Online reviews, such as http://tax-software-
review.toptenreviews.com/tax-brain-review.html, have pointed out the software
for being accurate and up to date. They have also recognized the telephone and
chat support availability. Although, the company suffered in 2009 when a wrong
bank routing number was used and direct-deposit refunds were delayed for many
customers.
Free Filing
Given the new IRS emphasis on electronic fling, along with the automated report-
ing system by employers and banks, many Americans are beginning to raise an
important question: Why do they have to pay to purchase software or fle electron-
ically? Why not let the IRS develop the online software and make it available for
free? To expert observers, the most immediate answer to the question is that the
IRS has historically been terrible at developing software. The second issue is that
the IRS might be tempted to provide tax advice. Although it might seem strange,
many people could be better off using private advice instead of recommendations
from the IRS. The IRS is likely to be considerably more conservative and has
been known to ignore court cases that it has lost (Becker 2002). But over time,
there is a possibility that the IRS could compete with the commercial providers.
In part to answer these questions, several of the commercial providers cooper-
ated with the IRS in 2003 to offer the IRS Free Filing Program. The goal was to
target the low-income and elderly populations. TaxBrain participated in 2003 and
served 50,000 individuals. Strangely, in 2004, the IRS created a new rule that
required all commercial providers to permanently fag every return that was fled
through the program. The IRS apparently wanted to perform some type of analy-
sis on these returns, but did not disclose the reasoning. Consequently, TaxBrain
decided not to participate in the programbelieving that it as an unnecessary in-
vasion of privacy with a high risk to individuals (press release 2004). TurboTax
announced that it would continue to participate, but would ignore the IRS require-
ments and not fag the returns of those participating in the program (Wiles 2004).
The IRS denied that taxpayer privacy might be compromised, but failed to pro-
vide an explanation. The IRS does note that it has always been able to tell how a
return was prepared, whether it was by software, volunteer, or IRS walk-in loca-
tion (Wiles 2004).
In 2005, about 13 million taxpayers fled tax returns over the Internet. The
number was expected to continue growing at double-digit rates (Keizer 2006). For
2007 (2006 tax year), Petz handled almost 38 million individual tax returns, and
30 million of those used e-fling. Of those that were fled electronically, 21 mil-
lion were handled through tax professionals (CrossLink and V-Tax), and 9 million
were self-prepared through TaxBrain (Petz Newsletter March 2007).
Questions
1. How can TaxBrain gain more customers and take market share from H&R
Block and Intuit?
2. How does the CrossLink system give Petz an advantage with the TaxBrain
site?
3. Should the IRS develop online tax software, possibly eliminating frms like
H&R Block, Intuit, and Petz Enterprises?
4. By 2010, what percentage of people will fle tax returns electronically? When
this maximum is reached, how will Petz continue to grow?
5. Did Petz make the right decision with the Free File program? Should the
company rejoin it?
Additional Reading
Accounting Technology, Petz Enterprises, October 2003, Vol 19(9) p. 43.
Accounting Technology, Petz Enterprises, October 2006, Vol 22(9) p. 32.
Becker, David, IRS E-Filing Plans Worry Tax Industry, News.Com, March 11,
2002.
BizWire, TaxBrain Online Tax Center-TM- Braced for Blitz of Millions of Last-
Minute Filers Using Internet to Beat April 15th Deadline, March 19, 2004,
see www.taxbrain.com/about/news.
Keizer, Gregg, Tax Prep Sites Go Live, Information Week, January 3, 2006.
McCausland, Richard and Carly Lombardo, 2001 Tax Software Odyssey
Company Profles: Petz Enterprises, Accounting Technology, October 2001,
Vol 17(9), p. 41.
Petz Press Release, Petz Enterprises Sees Decreased Sales of Tax Software
and Diminished Use of Professional Tax Preparers as Impact of New IRS
Electronic Filing Initiatives, November 12, 2002, see www.taxbrain.com/
about/news.
Petz Press Release, Petz Enterprises Reports Record Tax Season, May15, 2003,
see www.taxbrain.com/about/news.
Petz Press Release, Is Your Neighborhood Tax Offce Going The Way of The
Dinosaur? December 4, 2003, see www.taxbrain.com/about/news.
Petz Press Release, For 2004, the TaxBrain Online Tax Center is not
participating in the IRS Free Filing Program, 2004, http://www.taxbrain.
com/services/service_irsFreeFilingProgram.asp
Widmer, Lori, Accountants Reveal Electronic Filing Software Benefts, CPA
Magazine, December/March 2007).
Wiles, Russ, IRS Free-Filing Rule Raises Ire, Cincinnati Enquirer, February
11, 2004; http://www.enquirer.com/editions/2004/02/11/biz_freefle11.html.
Case: Pac-West Telecomm, Inc.
Located in Stockton, California, Pac-West Telecomm (ticker: PACW) was found-
ed in 1980 by Wally Griffn. The company is a competitive local exchange carrier
(CLEC), which is a technical description for a telephone company that is not one
of the original Bell companies. With the breakup of the telephone monopoly in
1986, the way was cleared for new frms to step in and provide voice and data
services. Of course, it would be incredibly expensive to provide service to indi-
viduals, so the company focuses on communications for businesses. In particular,
Pac-West dominates in providing data connections for Internet service providers
and call centers. The company can provide one-stop service for all telecommuni-
cation needs, including voice and data (company Web site).
Internet Service Providers
Perhaps the most interesting service provided by Pac-West is its extensive tele-
communications for West Coast ISPs. If you have a dial-up Internet account, you
have to call a point-of-presence (POP) phone number. This number connects your
modem to another modem, which is connected to a router and gets you on the
Internet. But where is that modem and router located? One solution might be to
put banks of modems in many cities and then connect them together. But that ap-
proach would require high-speed Internet lines in hundreds or thousands of loca-
tions even in small cities. Remember that most people want to dial local phone
numbers to connect to the Internet. The solution Pac-West offers is to place tele-
phone switches in these multiple areas. You dial that local phone, the switch trans-
fers the call to one of Pac-Wests SuperPOPs, which then connects to the Internet.
Costs are reduced by consolidating the POP connections to a few locations. Reli-
ability is higher because it is easier to provide multiple modems and backup facili-
ties in the limited number of SuperPOPs.
Pac-West does not sell individual Internet connections. Instead, ISPs contract
to use Pac-Wests network. In 2000, over 90 ISPs were using the system, includ-
ing Earthlink and NetZero (press release May 2000). The company had $95 mil-
lion in revenue in 1999, followed by $139 million in 2000 (annual report). For the
most part, Pac-West leases high-speed lines in bulk, and then resells services to
smaller frms. For example, in 2000, the company signed with Qwest for exclu-
sive rights to an OC-48 optical transport ring from central to southern California.
The line provides speeds of 2.4 gigabits per second (Qwest 2000). Jason Mills,
vice president of network operations for Pac-West, notes that ISPs and corpora-
tions need to focus on their core business, not building, fnancing and administer-
ing a widely distributed access network (Alcatel).
In 2003 and 2004, several frms introduced data compression technology to re-
duce the time it takes to transfer data on dial-up lines. Pac-West installed this
equipment on its servers and made it available to its ISP clients. With virtually no
additional effort or change in technology, the ISPs could now offer faster access
to individual customers. Wayne Bell, Pac-Wests vice president of marketing and
sales, observed: The content acceleration technologies that are available through
our partner program offer dial-up subscribers an Internet experience up to 5x fast-
er than traditional dial-up, including Web surfng, fle downloads, and email, at
a fraction of the cost of broadband alternatives, such as DSL, cable, and wire-
less. By partnering with our ISP customers to enhance their dial-up service, we
help them improve customer satisfaction, reduce churn, and generate new revenue
streams, particularly in areas where DSL, cable, and wireless are not available
(press release 2004).
In the mid-2000s, Pac-West needed to fnd a product that could support growth.
Voice over IP (VoIP) was one answer that made sense. Pac-West has solid Internet
connections as well as direct ties into the publicly-switched telephone network.
It is relatively easy for customers with broadband connections to install a VoIP
telephone. And if you are simply calling another user with a VoIP phone, the en-
tire connection can be handled over the Internet. However, to call standard tele-
phones or cell phones, the Internet call has to be moved onto the telephone lines.
Pac-West became a wholesale provider for handling these connections. Sarita Fer-
nandes, VP of marketing and product management in 2007 said that we are still
seeing a lot of people getting into VoIP space (Wilson 2007). In 2005, Pac-West
sold its traditional business telephone clients, along with some supporting equip-
ment, to U.S. TelePacifc Corp. (2006 Annual Report). With the sale, Pac-West
was able to show a net income of $8.5 million in 2005.
Growth and the Telecommunications Crash
In 2001 and 2002, the telecommunications industry suffered a major collapse. In
part, the crash was driven by the dot-com Internet crash. As frms failed, pay-
ments to Internet providers declined. But a big part of the telecommunications
crash happened in the network side, when several large frms fled for bankruptcy.
The problem was that they had overbuilt capacity. Anticipating huge and contin-
ued Internet expansion, some frms borrowed heavily to lay thousands of miles
of fber optic cables. When demand never materialized, they fled for bankruptcy.
Pac-West avoided this fate. In 2002, Griffn noted that we stayed out of fber.
Were not digging any holes in the ground. Thats what kept us at a lower debt
rating. Even so, with sales of $150 million in 2002, Pac-West lost $34.7 million
(Pac-West Telecomm 2002).
Despite the challenges, Pac-West is still a substantial presence in the West
Coast states. In 2004, Pac-West reported that it carries about 20 percent of the
dial-up Internet traffc in California. It also carries over 120 million minutes of
voice and data traffc per day (company Web site).
Te Role of the FCC
The Federal Communications Commission (FCC) both through policy and through
losing some court cases has caused some consternation among the CLECs. One
FCC order phases out the line sharing policy that was imposed on the local tele-
phone monopolies. These companies were originally required to offer a choice
of ISPs to DSL subscribers. If you wanted a high-speed line, you had to pay the
phone monopoly to handle the communication over the copper wire from your
house to a switching facility. But you had a choice of using the phone company
as the ISP to connect to the Internet or a third company. Dropping this rule most
likely will result in less competition and possibly higher prices for DSL services
(press release May 2003). If prices for DSL do increase, demand for higher-speed
dial-up connections is likely to increase. So, Pac-West could beneft.
The other major shift in policy occurred when the U.S. Solicitor General re-
fused to seek Supreme Court review of an FCC order that was overturned in a
federal court. The original order gave the FCC the power to regulate prices for un-
bundled network elements (UNEs) from the local phone monopolies. The original
monopoly breakup decree required that the phone companies lease lines and phys-
ical space to other companies so that they could connect to individual phone lines
and offer new services. The FCC order enabled the agency to control the pricing
for these leases. Since the order was overturned, the monopoly phone companies
are now free to charge any price they choose. These increased costs have the po-
tential to put rival CLECs out of business. However, John Sumpter, Pac-Wests
vice president of regulatory, notes that Pac-West does not employ UNEs in its
current network architecture in any signifcant way, and therefore is not directly
impacted by these actions. But he also gets a nice jab in at the same time: In-
directly, however, perpetual regulatory confusion and biased decision making in
favor of the former monopolies continue to impede the benefts of competition
intended by the 1996 Telecom Act (PRNewsWire 2004).
Investors
A company the size of Pac-West requires decent capitalization. Pac-West was a
closely held company, which means it had only a few key stockholder inves-
tors. In 1998, an investment group led by Safeguard Scientifcs acquired Pac-West
for $115 million. The infusion of cash was primarily used to pay for expansions
(Technology Brief 1998). Most of the money was provided by venture capital
frms: Bay Alarm Co., SCP Private Equity Partners, William Blair Capital Part-
ners, and TL Ventures.
In 2000, at the height of the Internet bubble, the group took Pac-West public
with an initial public offering (IPO). The offering raised about $106 million, pri-
marily targeted for capital expenditures and working capital (Postelnicu 2000).
The company had profts of $3.6 million on revenues of $139 million in 2000. Al-
though revenues increased in 2001 and 2002, the company suffered $34.8 million
in losses in 2001. It recovered slightly to a $2 million proft in 2002. In 2003, the
company downsized to reduce costs, and revenue plunged by almost 18 percent to
$134.6 million with a net loss of $15 million (Corporate annual report).
Adjusted for splits, Pac-Wests stock price peaked at $40 a share early in 2000.
Suffering from the dot-com crash and then the telecommunication industry melt-
down, the stock plunged in 2001 and 2002. In 2003, the company was almost
delisted from NASDAQ when its price fell below $1 per share. The company had
planned a reverse stock split in 2003 to increase the price of its stock, but in Au-
gust, the stock price had recovered suffciently to meet the listing criteria (press
release August 2003).
On April 30, 2007, Pac-West fled for Chapter 11 bankruptcy and most of the se-
nior managers resigned (8-K SEC fled on May 1, 2007). The company planned to
continue its national expansion into the VoIP market in cooperation with Verisign.
Pac-West became a private company after bankruptcy. In 2008, it acquired Tex-
Link Communications, a frm that provided similar services in Texas. Pac-West
also reported that it carried 20 percent of the California Internet dial-up traffc
in 2010 (Web site). In 2010, Pac-West introduced the Telastic service to provide
cloud-based telecommunications. Essentially, Pac-West runs Internet server-based
tools to handle VoIP services. Pac-West provides the services to ISPs and smaller
phone companies.
Questions
1. Can Pac-West survive? What will the company need to do to grow?
2. Since the industry is heavily regulated, what can the government do to
encourage growth in frms like Pac-West? Should this growth be encouraged?
3. What ongoing information technology events will affect the growth and
survival of Pac-West?
4. Can Pac-West make better use of information technology or the Internet to
improve sales or reduce costs?
Additional Reading
Alcatel, http://www.businessweek.com/adsections/broadband/innovation/
publicnet/alcatel.htm.
BizWatch, Qwest Communications Awarded $24 Million Contract From Pac-
West Telecomm For Optical Network Services, September 12, 2000.
Pac-West Press Release, Pac-West Telecomm Increases Revenues 114 Percent,
May 1, 2000, see www.pacwest.com.
Pac-West Press Release, Pac-West Telecomm Launches New Dial Broadband
Service to ISPs, May 13, 2003.
Pac-West Press Release, Pac-West Telecomm Regains Compliance With Nasdaq
Listing Requirements, August 11, 2003.
Pac-West Press Release, Pac-West Telecomm Launches Content Acceleration
Partner Program, May 18, 2004.
Postelnicu, Andrei, Pac-West Telecomm Inc. Venture Capital Journal, January
1, 2000.
PRNewsWire, Pac-West Telecomm Anticipates No Direct Impact From Recent
FCC Triennial Review Actions, June 10, 2004.
Sacramento Business Journal, Pac-West Telecomm Earnings Down, August 7,
2002.
The Wall Street Journal, Technology Brief -- SAFEGUARD SCIENTIFICS
INC.: Investor Group to Acquire Pac-West Telecomm Inc. July 7, 1998.
Wilson, Carol, VoIP Market Still Growing, Wholesalers Say, Telephony, March
26, 2007.
Case: Facebook
As a company with 500 million active users and over 2,000 employees in 2010,
Facebook no longer seems like a startup company. Yet the company was only
founded in 2004 and represents a classic case of successful entrepreneurship.
Some of the startup aspects of the company were accentuated in the popular movie
The Social Network, which presented a loose and somewhat fctional background
of the companys early days. Also, because of the controversy at the startup and
the lengthy lawsuits, a considerable amount of information regarding startup
details and communications have been made public (Carlson 2010). For those
who missed the movie, book, and extended lawsuits, essentially, two other Har-
vard students (the Winklevoss twins), have stated that they initially hired Mark
Zuckerberg to create a site for them and that Zuckerberg stole their ideas and
turned them into Facebook. At one point, Facebook paid them $65 million. Which
probably would have been enough money for most people, but then they argued
that Facebook withheld critical valuation data and they thought they should have
more money. The case and appeals dragged through the courts for years, but Face-
book continued to win. In the end, the political details are probably unimportant
to the business aspects of the company. However, they do highlight some of the
concerns any startup company must face. Any idea for a company is rarely com-
pletely new or isolated. It is always easy for someone to claim that they had an
idea frst. Even though it might be diffcult to prove in court, the legal fees can
become an issue. In fact, Facebook was not the frst social network on the Internet.
Two other issues are more interesting in the case of Facebook: (1) How Face-
book can make money, and (2) Shareholders and investing. The two topics are
closely related. If a company is not making money, it needs early investors to
pump in money in the hopes that someday it will generate revenue. Most startup
companies need to focus on revenue early in the discussions. Eventually, no frm
can survive without revenue and profts. The challenge for Web-based frms is that
two streams of revenue exist: Sales and Advertising. Both of them present chal-
lenges to frms like Facebook that need to attract a large number of users.
A network site can succeed only if it has a huge numbers of users. Who wants to
belong to a network with a small number of users? But, then how does a network
get started? In Facebooks case, the answer was to start the network as exclu-
siveto Harvard, then other top-tier universities. At any one school, it was easier
to entice a small number of students to join the site. As the company spread into
more schools, it then had a base of users that appealed to more schools and could
spread nationwide. For the frst few years, only students at universities (based
on their e-mail addresses) could join Facebook. By the time Facebook opened
the network to anyone (September 2006), it already had a large base of users.
This strong base of young adults gave a strong reputation to the company. People
who belonged to earlier sites (such as MySpace and Friendster), jumped ship and
signed up on Facebook.
Consider social networks from the perspective as a user for a minute. How
would you choose a network site? Do the features matter? Speed? Or just the
number of friends you have on the network? Now, what happens if a new site is
developed and a bunch of your friends move to that site. Would you change? If so,
when? Would you be an early adopter or wait until many people have switched?
How hard would it be to switch? Now, look at Facebooks policy that makes
it diffcult (or perhaps impossible) to automatically transfer your collection of e-
mail addresses and other data to a different site (Sullivan 2010). OK, you can
transfer the data but you have to do it by hand. Facebook prevents anyone from
building an automated tool to pull your data and transfer it to a competitor. But,
how bad would a site have to be, or how good would a new site have to be for
you to consider changing?
For example, Facebook has had several episodes of problems with privacy.
Over time, the company has been trying to add more privacy features and to give
users more control over who sees their data. But, partly because the features ap-
pear to have been added on, instead of built in at the start, they tend to be diffcult
to use. In mid-2011, Google introduced a new social networking service that in-
cluded easer and better privacy controls (Miller 2011). But, will it be enough for
people to switch? Once people do start switching away from a network, a point
is reached where the frst network quickly collapses (see MySpace or Friendster).
Now, go back and think again about the issue of money. If you plan to be an en-
trepreneur, you should always remember the money issue. If a company charges
fees to use a social network (mostly at dating sites), the fee automatically restricts
the number of users. Some fees might be low enough to be unimportant, but the
network will have slower growth if fees are charged. A secondary possibility is to
charge fees for premium items, such as added storage space or customization.
But any base fee runs the risk of driving users to competitor sites.
The other way of raising money is to sell advertising, which is about the only
way social networks can realistically make money. First, the site has to be large
enough to attract big advertisers. More people on the site mean more advertising
money available. But, the site needs some way to display ads; and the people us-
ing the site have to not only tolerate the ads, but actually click on enough of them
and perhaps buy something from the advertisers, or the advertisers will stop. If the
site relies on the big sites such as Google to provide the ads, then it will collect
only a minor share of the advertising money. So, ultimately, the site has to sell its
own advertising, which means it needs the software to display and track the ads,
and the company needs a sales staff to fnd advertisers. If you look at Facebooks
Web site of press releases it shows offces in at least nine U.S. cities and 14 other
cities around the world. Most of the people in these offces are probably dedicat-
ed to selling ads, or the accounting associated with them. Advertising revenue in
2010 was stated to be $1.86 billion, with some additional revenue for on-site sales
yielding about $2 billion in total revenue. Eventually, the actual revenue and cost
numbers will be released in public flings.
An interesting aspect to Facebook is that Zuckerberg appears to have been fo-
cused initially on building the size of the networkor number of usersbefore he
started worrying about income. Given the importance of attaining size and market
share to keep out competitors, this approach seems to have paid off. On the other
hand, the company needed money just to pay operating expenses such as pro-
grammer salaries, offce space, and server costs. Ultimately, server and Internet
costs can be expensive. Servers and bandwidth to handle millions of customers
a day can become expensive. Facebook was able to obtain private fnancing rela-
tively early$12.7 million in May 2005 and $27.5 million in April 2006 (Face-
book Web site/timeline). In 2007, Microsoft invested $240 million in Facebook.
Ultimately, individuals and venture capital frms provide money to compa-
nies because they get ownership in the company. That ownership pays off either
through dividends or when the owners decide to issue public stock in an initial
public offering (IPO). At that time, the investors can sell their holdings to the pub-
lichopefully at a much higher value than the amount they put into the frm. In
the meantime, it is diffcult for investors and entrepreneurs to make moneyother
than through salary. Some early owners of Facebook stock took an intermediate
step and began selling their shares privately. In fact, in 2011, Facebook teamed
with the investment bank Goldman-Sachs to offer private shares of the company.
The company passed the 500-shareholder mark sometime in 2011, which means
that it will be required to publicly fle its fnancial reports with the securities and
exchange commission (SEC) by the frst quarter of 2012 (Press release January
21, 2011).
The stock/ownership issues are only one part of the Facebook storyor for
any startup frm. Ultimately, growth means providing a product that benefts users
or customers. And it requires continual improvementsadding new features and
identifying what customers will want tomorrow. Web software is a particularly
challenging product because people want it free, and it is diffcult and expensive
to fnd good programmers to keep adding features.
Questions
1. What features does Facebook have that might encourage people to keep
using the site instead of moving to a different site?
2. Why is growth so important to a Web-based frm and what costs does it
impose?
3. Can Google or some other frm take customers away from Facebook?
Additional Reading
Carlson, Nicholas, At LastThe Full Story of How Facebook Was Founded,
Business Insider, March 5, 2010.
Miller, Claire Cain, Google Introduces Facebook Competitor, Emphasizing
Privacy, The New York Times, June 28, 2011.
Womack, Brian, Facebook 2010 Sales Said Likely to Reach $2 Billion, More
than Estimated, Bloomberg, December 16, 2010.
Sullivan, Danny, Facebook: Youve No Right to Export Email Addresses
(Unless Its To Yahoo & Microsoft), Search Engine Land, November 9,
2010. http://searchengineland.com/facebook-you-have-no-right-to-export-
email-addresses-55247
adopted by rivals?
sectors?
4
Part
Organizing Businesses
and Systems
How are information systems created? How should information
system resources be organized? How do information systems af-
fect society?
Information technology provides new ways to organize busi-
nesses. In particular, the Internet makes it possible for frms of
any size to conduct business around the world. But creating these
new structures requires thought, effort, and time. Experience
shows that it is easy to make mistakes and waste huge amounts
of money. Organizing and planning are critical to avoiding these
problems.
Because of their importance in a modern frm, information
systems must be carefully planned, designed, and maintained.
Business managers are increasingly involved in designing and
organizing MIS resources. Managers need to understand the diff-
culties faced in systems development to understand the rules and
processes. As technology changes, the organization of business
operations and the MIS resources is changing. By identifying
these changes, business managers can improve their operations
and make better use of new information technologies.
Changes in technology and business cause fundamental chang-
es in society. These changes affect everything from education to
government to our daily lives as employees and citizens. Chang-
ing technology brings new responsibilities and problems. As
managers and citizens we will face many new decisions. We must
always remember our ethical responsibilities to other members of
society.
Chapter 12: Systems Development
Chapter 13: Organizing MIS Resources
Chapter 14: Information Management and Society
How do managers organize and control businesses and informa-
tion systems?
772
How do you create the software tools needed for your organization?
What main options exist for building information systems?
What are computer programs?
How do you control a major development project?
Is SDLC always the best approach? What other methodologies could be used?
How do you analyze and annotate a process-based system?
How is object-oriented design different from process design?
Can software be located in multiple places?
How does cloud computing change software development?
Chapter Outline
Systems Development
12
Chapter
Introduction, 773
Building Information Systems, 776
Custom Programming, 777
Outsourcing and Contract Programmers,
778
Assemble Applications from
Components, 779
Purchase an External Solution, 780
Computer Programming, 781
Programming Logic, 781
Events, 783
Object-Oriented Programming, 785
From Programming to Development,
787
Systems Development Life Cycle, 788
Introduction to SDLC, 789
Feasibility and Planning, 789
Systems Analysis, 791
Systems Design, 792
Systems Implementation, 793
Maintenance, 796
Evaluation, 797
Strengths and Weaknesses of SDLC,
798
Alternatives to SDLC, 799
Prototyping or Iterative Development,
800
Extreme Programming and Agile
Development, 800
Developing Systems Requires
Teamwork: JAD and RAD, 802
Communication, 806
Open Source Development, 807
End-User Development, 808
Development Summary, 809
Process Analysis, 810
Input, Process, Output, 810
Divide and Conquer, 812
Goals and Objectives, 813
Diagramming Systems, 814
Summary: How Do You Create a DFD?,
818
Distributed Services, 819
Summary, 821
Key Words, 823
Exercises, 824
Cases: Government Agencies, 829
773 Chapter 12: Systems Development
Federal Aviation Administration
Introduction
How do you create the software tools needed for your organiza-
tion? Today, most companies have chosen not to develop software. In most cas-
es, it is simpler to buy the tools from existing companies. Still, tools often need to
be customized and sometimes small applications can make a big difference in how
your company operates. As a business manager, you will probably not become
a programmer or systems developer. On the other hand, at some point, you will
have to interact with developers to ensure that you get the systems you need. You
need to be able to communicate with the developers, and you need to understand
their constraints. Furthermore, as a manager, you need to understand the costs and
management issues facing MIS departments. You might have to choose among
development methodologies, so you need to know the options and their strengths
and weaknesses.
As shown in Figure 12.1, creating software systems is a diffcult
task, similar to creating highly-complex buildings. The tools have got-
ten better over time, but the tasks are much harder. Several cases and
stories exist of failed projectssome costing millions of dollars ended
up being discarded. Some studies have reported that half or more of
software development projects are over budget and behind schedule.
However, these numbers are merely estimates because most schedules
are best guesses. Building anything entails guesswork on schedules. If
you are creating something you have built a hundred times before, you
will have a pretty good idea of how long it will take. But, software gen-
erally involves creating totally new systemsoften with new hardware
and new tools. This process is more like the challenges faced by Boeing
Why is it so diffcult to develop software? Developing software for complex sys-
tems is a diffcult task. Even bringing in outside contractors has not solved all of the
problems for the FAA. Cost overruns and missed schedules have happened so many
times in FAA (and other government) projects that completing a project on time is
the rare exception.
The FAA is charged with controlling civilian and military uses of U.S. airspace.
The FAA is also responsible for modernizing the airways, installing radar, and train-
ing air traffc controllers. Probably their best-known function is control over com-
mercial fights and routes to maintain safety and effciency. With 50,000 fights a day
among 300 major airports, the FAA has a huge task.
The FAA has a computer system to help it control the thousands of daily fights.
However, the system was created in the early 1960s. It has been patched and up-
graded, but most of the hardware and software are based on decades-old technology.
On several occasions, the FAA attempted to upgrade the facilities, but complications
have forced the agency back to the old technology.
The FAA systems are particularly diffcult since they entail high-risk operations
that have to be performed accurately. Also, the technology is relatively unique and
development often requires state-of-the-art skills in areas that are rarely taught or
researched in schools. Also, large-scale projects are always diffcult to forecast, be-
cause any number of things can go wrong.
in designing and building the new 787 modelwhich fell several years
behind schedule. Boeing is certainly one of the best companies at de-
signing and building airplanes, yet the challenges faced in using new
materials and new construction techniques caused many problems and
delays. With large complex projects, it simply is impossible to predict
all possible delays.
Small projects that can be built by one person are relatively easy.
The challenge lies with large projects that involve many users and
many programmers. These systems need to be split into smaller piec-
eswhere the pieces are assigned to teams and eventually to individ-
ual programmers. But, splitting a project means that everyone has to
work together to share their designs and progress. The pieces have to
ft back together and still meet the overall objectives. The system needs
a solid plan and control mechansims to ensure everything works to-
gether and stays on schedule.
The essence of managing software development comes down to coordinating
dozens or hundreds of programmers and several methodologies have been created
to handle the common tasks. The most formal approach is known as the systems
development life cycle (SDLC). Its structure and concept were borrowed from
engineering and construction of large physical items such as bridges and dams.
Large organizations that develop several systems use this method to coordinate
the teams, evaluate progress, and ensure quality development. Most organizations
Figure 12.1
It is not easy to create information systems to support business needs (strategy,
tactics, and operations). Three basic development techniques are: systems
development life cycle, prototyping, and end-user development. As a manager, you
will participate in each of these methods. You will sometimes have to choose which
method to use.
have created their own versions of SDLC. Any major company that uses SDLC
also has a manual that is several inches thick (or comparable online documenta-
tion) that lays out the rules that MIS designers have to follow. Although these
details vary from frm to frm, all of the methods have a common foundation. The
goal is to build a system by analyzing the business processes and breaking the
problem into smaller, more manageable pieces.
Improvements in technology also improve the development process. The pow-
erful features of commercial software make it easier to build new applications.
Programmers and designers can work with larger, more powerful objects. For ex-
ample, instead of programming each line in COBOL, a report can be created in a
few minutes using a database management system or a spreadsheet. Prototyping
is a design technique that takes advantage of these new tools. The main objective
of prototyping is to create a working version of the system as quickly as possible,
even if some components are not included in the early versions. A third method of
creating systems, end-user development, relies on users to create their own sys-
tems. This method typically uses advanced software (such as spreadsheets and da-
tabase management systems) and requires users who have some computer skills.
It is important to be careful when you implement any new system. Case studies
show that major problems have arisen during implementation of systems. In fact,
some organizations have experienced so many problems that they will deliber-
ately stick with older, less useful systems just to avoid the problems that occur
during implementation. Although changes can cause problems, there are ways to
deal with them during implementation.
Trends
Internally, computer processors have limited capabilities. The processor has
a set of a few hundred internal instructions that it knows how to perform,
such as moving a number to a new place in memory or adding two numbers
together. Initially, all computer programs were written at this low level; but
writing programs at this level is diffcult and time consuming. Over time,
two major innovations were created to reduce the diffculty of programming
at these low levels: (1) higher-level languages were created that handle many
details automatically for the programmer, and (2) common algorithms used
by many applications were created and sold as operating systems. These ad-
vances enable programmers to focus on the applications instead of machine-
specifc details.
Despite many advances, writing programs is still complex and time con-
suming. Large-scale applications require the teamwork of millions of hours
of programmer time and cost hundreds of millions of dollars to build and
maintain. These high, fxed development costs underlie the growth of the
commercial software industry. By making the software relatively generic
and selling it to hundreds or thousands of frms, the development costs are
spread over a wider group.
Increasingly, frms are moving away from custom-written software. They
are purchasing packages and hiring outside programmers to develop many
components. In these situations, design issues generally consist of choosing
and customizing the software to meet the individual needs of the organiza-
tion.
There have been some spectacular failures in the development of computer sys-
tems. Projects always seem to be over budget and late. Worse, systems are some-
times developed and never used because they did not solve the right problems
or they are impossible to use. Several design methods have been created to help
prevent these problems. All methods have advantages and drawbacks. As a result,
they tend to be suitable for different types of problems. As a manager, you need
to be aware of these different tools and their strengths so you can ensure the right
methodology is used for your projects. Sure, you could call in EDS, Accenture, or
IBM to work on all of your projectsbut it would be absurdly expensive to use
these formal methods for tiny projects. You know not to call Boeing to help you
build a paper airplane, but how do you know when you need help turning your
hundreds of spreadsheets into a Web-based application?
Figure 12.2 shows one of the initial issues you will face when looking for new
systems. Buying a system is the easiest, fastest method of getting a new applica-
tion that works. It might have to be customized to meet the needs of the company,
but customizing existing software is easier than starting from scratch. If the exact
tool you want is not available, you might be able to build something close using
other components. For instance, a Web site could be created by adding a shopping
cart, chat room, and payment processing components. If you need a more spe-
cifc component, you could hire a contract programmer to write it. Creating small,
clearly-defned components is usually straightforward and can be done without
huge management costs. For truly unique systems, it is still necessary to hire pro-
grammers and developers and build most of the elements from scratch. This type
of development requires the most oversight and control.
Building Information Systems
What main options exist for building information systems? Fig-
ure 12.3 illustrates the three primary methods of creating or customizing a busi-
ness application that you are likely to encounter as a manager. When you need to
create or customize an application, you ultimately have to choose some variation
of these three. Many alternatives exist, but these three represent anchors. SDLC
tools are used for large, critical projects involving many developers because they
foster control and repeatability. On the other side of the spectrum, small projects
such as reports, queries, charts, and data analysis can be handled completely by
end userswith the appropriate training and experience. As long as you, the user,
has the tools and business knowledge, it will often be faster for you to develop
Figure 12.2
Build or Buy. Buying software is much faster and easier even if it requires some
customization. Other options require more time, and time means higher costs plus
more risks. But sometimes you have no choice but to build a customized application.
Purchase
complete
solution
Assemble
from
components
Outsource or
contract
programmers
Custom
programming
Development time and costs
and customize applications to meet your own needs. If you lack the skills to build
the application yourself but the project is not too complex, a developer with the
skills and tools can work with you to design and build the application.
Another way to look at the problem of software development is to recognize
that ultimately developers have to write detailed program code. If your needs are
similar to those of other businesses, companies might have already created a com-
mercial application that you can use or customize. If you can buy existing soft-
ware or even components, the cost is substantially lower, and you reduce risk by
using code that has already been tested.
Custom Programming
Ultimately, all applications are created by teams of programmers writing detailed
code. Writing your own custom program gives you complete control over the ap-
plication. You can include any features, build in special routines unique to your
company, and integrate the data with your existing systems. The problem with
creating your own code is that programming is diffcult, time consuming, hard
to control, and expensive. Even when the application is completed, you will still
need groups of programmers to fx problems, add new features, and develop fu-
ture versions of the software. Modern development tools make it easier to write
programs today, but every application still requires intense development efforts,
and the tools never seem to provide all of the features you need.
Figure 12.3
Primary methodology choices. Business applications need to be developed or
customized and you have to choose which methodology to use. These three anchor
the primary choices. Large, critical projects need the control of SDLC. On the other
side, individual users can build their own applications if they are small enough.
Prototyping
End user development
Systems development life cycle
Large projects,
formal control
Single user,
reports and one -
time computations
Decision support,
analyses, and
reports
Quality and security are two critical issues with any type of development. If
you build a custom application, you must leave suffcient time to build in quality
tests and to correctly build security controls. Employees also have to be trained in
these two areas, and you generally need specialized mangers to oversee each of
these two areas.
Outsourcing and Contract Programmers
One of the other problems with creating your own software is that you generally
have to hire many programmers at some point in the development. But when the
development is fnished, you will rarely need all of these programmers. So, either
you fnd new tasks for them or you have to release them to reduce your costs.
But hiring and fring workers for a single project is frowned upon. Consequent-
ly, many frms use contract programmers, or even outsourcing, to handle system
development.
With contract programming, you negotiate with a company to provide special-
ists for a given period of time. When their work on the project is complete, they
move on to another job. The process saves you the problems of hiring and fring
Reality Bytes: System Development Methodologies
Accenture is the largest worldwide consulting frm in management information
systems. It conducts major installations using a proprietary methodology called
Method/1. Method/1 uses four phases in the development process: plan, design, im-
plement, and maintain.
McKinsey and Co., a strategic consulting frm, examines organizations with a
copyrighted Seven S model. The Seven Ss are structure, systems, style, staff,
skills, strategy, and shared values.
Electronic Data Systems (EDS), started by Ross Perot, purchased by GM, and
now independent, is the largest outsourcing company. It also develops systems using
a traditional SDLC methodology. The detailed methodology has thousands of indi-
vidual steps spelled out on separate pages that must be signed off at each step.
Rational Rose, a tool to support object-oriented development, was designed by the
leading OO gurus. It is a graphical tool designed to show the object details and their
relationships. It supports reverse engineering, by reading existing code and convert-
ing it into the corresponding diagrams. The Rose tool can also generate the fnal code
from the diagrams.
Microsoft, as a commercial software vendor developing highly complex systems
used by millions of people, has developed its own methodologies for quickly creat-
ing code. The company has been a proponent of Rapid Application Development to
reduce the time it takes to complete large projects. The methodology is designed to
segment the code so that hundreds of programmers can work on it simultaneously.
The methodology also relies on creating code that can be modifed later, for im-
provements and patches.
With Open Source, led by Richard Stallman and the GNU project, thousands of
programmers around the globe are building complex projects in a loosely-knit orga-
nization. Generally, one person is a project leader responsible for setting strategies
and resolving disputes. All of the source code is publicly available to anyone (hence
the name). Interested programmers suggest improvements and submit the modifca-
tions. Some complex systems have been built this way with minimal central control.
large numbers of employees. On the other hand, contractor salaries are usually
higher than traditional employees. More important, several lawsuits have made it
critical that you clarify the exact role of contractorsotherwise they can sue to be
classifed as regular employees to gain additional benefts such as stock options.
Outsourcing goes a step further than contract programming. When you out-
source project development, you transfer most responsibility to the outside frm.
Typically, you negotiate a development price, provide detailed specifcations,
and the outsourcer hires workers and develops the system. A huge variety of out-
sourcing arrangements are available, including situations where the outsourcers
run your entire MIS department or just your servers or networks, or handle PC
maintenance.
The primary advantage of outsourcing is that the external company takes re-
sponsibility for managing the process and producing the application. You still
have the responsibility to clearly defne exactly how the application should work,
but the outsourcer bears more of the riskparticularly with fxed-fee contracts.
The one thing you want to avoid with contractors and outsourcers is uncontrolled
hourly fees.
Assemble Applications from Components
A good way to reduce development time and costs is to buy portions of the system
from other companies. Even if you need a custom solution, you can purchase a
variety of software components that handle many of the diffcult tasks for you.
Components are a powerful feature of modern operating systems. They are blocks
of code that are integrated into custom applications. For instance, you could pur-
chase a security control to handle encryption on a Web site. Whenever your ap-
plication needs to encrypt or decrypt some data, it simply calls the components
Reality Bytes: It Doesnt Get Easier
Developing software is hard. The bigger the project, the harder it is to manage the
team. Engrave these rules in stone, developers have faced them for decades, and still
the process does not get easier. Fred Brooks (The Mythical Man Month) described
the problems in detail regarding a 1960s IBM project. Scott Rosenberg (Dreaming
in Code, 2007) explains how the problems still exist. He was project manager on an
open source project to create a new personal information manager called Chandler.
Two dozen programmers struggled on the project for over three years. The project is
still unfnished (0.7) and searching for more developers. In an interview, Rosenberg
points out two major insights into the development process. If we try to design a
system that is fully featured and complex and does everything everyone wants it to,
we'll never have any system. Break things up into small bites. Any opportunity
you have to do that should be seized. He also emphasizes that it is diffcult, expen-
sive, and time-consuming to develop new software. For example, even at Microsoft
prices its cheaper to use Word than to write a new word processor. New software is
developed when people fnd new tools and need better ways to work with data. But
developing software is hard. Start with a reasonable goal and get something out the
door. If it works, improve it.
Adapted from Edward Cone, Scott Rosenberg: What Makes Software So Hard,
CIO Insight, January 5, 2007.
methods. Similarly, if you need to process a credit card application, you can install
a component (or link to a Web service) that handles everything for you. Thou-
sands of useful components are available for a few hundred dollars each or less.
You simply install the component on your server and your programmers can begin
using the functions within their code. This approach relies on the capabilities of
commercial off-the-shelf (COTS) software. As the number and quality of soft-
ware packages have increased, it has become easier to build a system based on
COTS.
Increasingly, tasks currently performed by components are being offered as
services over the Internet. The same principles will apply, but the external com-
pany will maintain the application, install upgrades, and add new features. In ad-
dition, new services will become available. For example, services provide instant
exchange rate conversions so that you can list prices in any currency. The conver-
sion rates will always be current with no effort on your part.
Components have many substantial advantages and only minor drawbacks (pri-
marily the price hassles with upgrades). They can signifcantly reduce develop-
ment time and provide powerful features that are beyond the capabilities of many
staff programmers. In fact, many outsource specialists develop their own collec-
tion of components to use in developing custom solutions. By integrating com-
monly used features, they can build new applications faster with fewer errors.
Purchase an External Solution
Taking the concept of components and outsourcing a step further, many commer-
cial software companies sell prepackaged applications. Some are turnkey systems
where you simply load your data, select a few preferences, and the system runs
(much like buying an automobile, you turn the key and no assembly is needed).
Other applications require detailed customization. The ERP packages (such as
SAP) are classic examples. The system handles all of the basic operational data of
the frm, including generation of fnancial reports. You purchase the software from
the vendor and install it on top of a database management system. You still have
to set up your accounts and some custom details for reports. The application can
then be used by your company to track all fnancial and manufacturing data and
produce standardized reports.
On the other hand, you can also customize most of the features. If you need
unique manufacturing reports, you can write code to generate them. The degree of
customization often depends on the attitude of management. The drawback to ex-
tensive customization is that it requires specially trained programmers and delays
the entire project. Moreover, when the DBMS vendor or the ERP vendor upgrades
the underlying software, you may have to rewrite all of your custom programs.
Prewritten packages can have high price tags (SAP costs can easily run into
millions of dollars). But it could take millions of hours of programmer time to cre-
ate a custom system with the same functionality.
In general, it is almost always preferable to buy solutions, but keep a close
eye on prices. The commercial software essentially spreads the development costs
across thousands of frms. Unless you have a truly unique application and are will-
ing to pay a staff of top-notch programmers, it is better to share the development
costs. And if you do have a radically different application, you should consider
packaging it and selling it to other frms to reduce your costs.
Computer Programming
What are computer programs? To begin to understand the development
process, it helps if you know a little bit about how programs are written. The goal
of this section is to help you understand the level that programmers deal with and
to help you read the initial products. Programmers often write pseudocode or out-
lines that you need to be able to read to ensure the program will work correctly.
Of course, it takes several books, considerable practice, and good math skills,
to become a programmer. But, the foundations of programming are relatively easy
to understand. You should check out the Toolbox on programming to see how
managers can write a few lines of code to help solve some relatively complex
tasks. Figure 12.4 shows the fundamental concepts of programming. Essentially,
processors execute one line of code at a time and follow your instructions. Parallel
processing is more diffcult because you can have several sections of code execut-
ing at the same time, which means you have to worry about which one might
fnish frst; but it leads to faster programs. Variables, computations, conditions,
loops, subroutines, and input/output are common features of almost every lan-
guage. Once you learn the program logic of these commands, you can write code
in almost any language. Computer languages also possess a syntax that dictates
exactly how you have to type in the commands, including spelling (case-sensi-
tive), commas, dots, and other annotations. Development tools (such as Microsoft
Visual Studio and the open source Eclipse) make it easier to enter code correctly
by prompting with syntax options and verifying the items as you type them.
Programming Logic
Figure 12.5 shows a tiny sample program. Instead of following the syntax of a
particular language, it uses pseudocode to illustrate the logic. Several additional
statements would be needed to turn this code into an application that runs. On the
other hand, almost no one would ever write this code because you can use SQL to
accomplish the task easier and faster. But, its simplicity makes it a good starting
point. Pay attention to the role of the Total variable as an accumulator. It repre-
sents one location in memory that acts as a container. Initially, zero is placed in
that container. Each time a row of data is read, the number currently stored in the
Sequential execution: Statements are executed in order.
Parallel execution: Groups of code are executed at the same time.
Variables: Containers to hold data
Computations
Conditions: If Then Else
Loops: While End
Subroutines and Functions: Break code into manageable pieces.
Input/Output: Transferring data
Objects: Code pieces purchased or created to do specifc tasks.
Figure 12.4
Programming structures. Computer programs are developed by using these basic
concepts. You can purchase prewritten objects and functions to handle many common
tasks. But, you have to understand the objects and learn how to use them.
Total location is retrieved and added to the value from the query. This computa-
tion is performed within the computer processor itself. The result is transferred
back to memory. The loop causes the processor to retrieve each row from the
query and execute the accumulation statement until the end of fle is reached. The
program code itself is also stored in memory while it is being executed. Loops and
conditions are common elements of any program. A conditional statement consists
of an IF statement, such as if (Total < 0), and a set of statements to execute only
if the condition is true. Many programs also contain else statements that are ex-
ecuted only if the condition is false.
Writing code is partly science and partly art. It takes practice to learn how to
translate what you want into detailed code statements. But, writing code can be
fun and creative. You get to tell the computer exactly what you want it to do, and
you get to create something that people can see and use. Programmers also have to
be detail-oriented. Small errors in logic or typing can be diffcult to fnd and cor-
rect. This combination of creativity, logic, and detail-orientation is diffcult to fnd;
which partly explains why good programmers receive relatively high salaries.
Figure 12.6 lists a few of the programming languages in use today. Many of
the languages have a similar syntaxloosely based on the earlier C language. But
subtle syntax and critical structural differences exist among the languages. A key
difference exists with C and C++ because they are both compiled down to assem-
bly code which provides detailed control over the computer and has high perfor-
mance. Many of the other languages run on an intermediate level which handles
data and interfaces differently. It is easier to write code in these languages (such as
Java and C#), but performance can be an issue for some types of problems. Other
languages (notably Javascript) are dynamic or script languages which are often
Figure 12.5
Sample code. Note the role of the variable Total as an accumulator. Also notice how
the loop walks through the query fle one line at a time until all elements have been
read. Each time a new row is read the accumulation statement is executed to add
the current value of Total to the query value and store the results back into the Total
variable.
Declare and initialize variable
Memory Space
Total: 0
Sales Query/File
Value
13
22
18
17
Open/start
Total = Total + Sales.Value
13 = 0 + 13
13
Loop through le/query one row
at a time until end of le.
Retrieve value and accumulate
in total variable.
Variable Total = 0
Open Sales
While NOT Sales.EOF
Total = Total + Sales.Value
Sales.MoveNext
End
Print Total
Close Sales
parsed from text only when the code is executed (run time) and can run even more
slowly.
Programming requires learning more than just the syntax and structure of a giv-
en language. Each program runs within a framework or environment that can limit
the actions available and also provide detailed tools and options. For instance,
the Windows operating system provides detailed controls for interacting with the
screen, the user, and other devices. Writing code for a Web server uses a different
set of tools, libraries, and events.
Events
Today, understanding programming logic and syntax is only the frst step in learn-
ing to program. Most programming environments are event-driven systems where
programmers create code function that are executed when some event is triggered.
You are familiar with these systems as a user, but you might not have realized
what was happening. Most of the user interfaces today are based on graphical
screens and forms, including Web pages. Each of these environments defnes doz-
ens or hundreds of events. Programmers can write code that attaches to an event
and then action is taken when that event occurs. As a simple example, Figure 12.7
shows a partial Web form and lists the events that can be activated by the Submit
button. A programmer can write Javascript code for any of those events that will
Figure 12.6
Some programming languages. Religious wars have arisen among geeks over which
language is better. Many of the languages have a similar syntax, but the framework
and structures can be radically different.
Language Main Purpose or Context
Java General purpose.
Designed to run on servers and clients.
C#, VB, ASP .NET Microsoft used for Web servers and
applications. Managed code.
Javascript, HTML5 Script language for Web page interactivity.
C++, C Compiled, powerful, lower-level languages
often used for systems and tools or where
speed is critical.
PHP, PERL, Python, Ruby Web server scripting/dynamic languages often
used on Linux/UNIX or Apache Web servers.
Objective-C Apples language for iPhone/iPad applications.
Flash (Adobe)
Silverlight (Microsoft)
Special purpose Web add-ins with code to
control interactivity. Silverlight uses C# or VB.
COBOL Older business applications, and SAP.
FORTRAN Older scientifc programming applications, but
check out F#.
be executed when the event is fred. If you think 56 events is too many, go online
and check to see how many exist now. Adding events creates more fexibility and
granularity in responding to subtle changes, so new events are periodically added
to the Web standards.
In most situations, as the programmer, you have to identify which event most
closely matches an action that you want the code to take. Rarely will you need
more than one or two events for a given object, and the onclick event is popular.
In this example, the onclick event is fred whenever the user clicks the Submit
button. However, it is often better to use the onsubmit event associated with the
form object instead. The onsubmit event is triggered regardless of how the form
is submitted (via a click or via some other lines of code). The distinction is subtle,
but one of the tricky aspects of programming is that you need to completely un-
derstand all of these subtle differencesfor hundreds or thousands of events. To
make it easier to learn to program and to reduce mistakes, most programmers and
companies develop a set of best practices that spell out the best way to handle
typical problems. Good programming books list and follow these best practices.
onactivate
onafterupdate
onbeforeupdate
onbeforecopy
onbeforecut
onbeforedeactivate
onbeforeeditfocus
onbeforepaste
onbeforeupdate
onblur
onclick
oncontextmenu
oncontrolselect
oncopy
oncut
ondatabinding
ondblclick
ondeactivate
ondisposed
ondrag
ondragend
ondragenter
ondragleave
ondragover
ondrop
onerrorupdate
onflterchange
onfocus
onfocusin
onfocusout
onhelp
oninit
onkeydown
onkeyup
onload
onlosecapture
onmousedown
onmouseenter
onmouseleave
onmousemove
onmouseout
onmouseover
onmousewheel
onmove
onmoveend
onmovestart
onpaste
onprerender
onpropertychange
onreadystatechange
onresize
onresizeend
onresizestart
onselectstart
onserverclick
onunload
Figure 12.7
HTML Javascript events for the Submit button. Any event can cause code to execute.
Having more events gives programs more fexibility, but rarely are more than three or
four used for a specifc problem.
Figure 12.8 shows a couple of simple lines of code that are executed when a
user rolls the mouse cursor over the Submit button. You would never write this
code on a real Web page because alerts are exceedingly annoying. However, it is
an easy way to verify that the code is executing properly. With a little thought, you
could write some conditions for the function that are usefulsuch as checking to
see if the user entered data in all of the boxes. The art of programming revolves
around building applications that solve problems and are easy to use.
Object-Oriented Programming
Object-oriented programming (OOP) evolved in the 1990s as a way to make
it easier and faster to create new applications. With this approach, the focus is
on defning objects that have specifc properties (data) and methods (code). All
of the common applications you use; including graphical operating systems, per-
sonal productivity tools, and Web browsers; were built using OO techniques. The
point of this approach is that the internal objects created to build these tools are
available for you to use in other programs. For example, you can write a new busi-
ness program that uses a spreadsheet object and its internal tools. Reusability of
objects is an important feature of the OOP approach.
Business programmers can use similar techniques to create their own objects.
Objects are defned by a set of properties (or attributes). The properties defne the
object. They also represent data that will be collected for each object. Consider
<input type="submit" value="Submit" name="submit"
onmouseover="ShowMessage('Areyoureallydone?');"/>
<script language="javascript">
function ShowMessage(msg)
{
alert(msg);
}
</script>
Figure 12.8
Simple Javascript example. Whenever the user rolls the mouse over the Submit
button, a message box pops up. It is exceedingly annoying to the user, but it shows
the effect of the mouseover event.
the small example of a banking system. One primary object will be Accounts.
A generic account object would have basic properties such as Account Number,
Account Name, Client, Manager, Date Opened, Beginning Balance, Current Bal-
ance, and Interest Rate. Each object also has functions, which describe actions
that can be performed by the objects and defne how to alter the object. In the
bank example, there would be functions to Open Account, Close Account, Accept
Deposits, Pay Withdrawals, and Pay Interest. Note that each type of account could
have a different method for computing interest payments. One account might
compound them daily, another weekly, and so on. With the object-oriented ap-
proach, the properties and functions are combined into the defnition of the object.
The goal is to describe a system so that if you change a function, you only have to
change one object. All of the other objects and processes remain the same.
Objects are related to each other. Typically there is a base class of objects, and
other objects are derived from the base defnitions by adding properties and alter-
ing functions. This process results in an object hierarchy, illustrated in Figure
12.9, that shows how the classes are derived from each other. The bank exam-
ple has several types of accounts with each of these categories containing further
subdivisions.
Figure 12.9 also shows detail in the classes by including some of the proper-
ties and member functions. The accounts have elements in common that are an
Figure 12.9
Objects: encapsulation, hierarchy, inheritance, polymorphism. Object-oriented
design focuses on individual objects and the data within the organization. Processes
are secondary and they are usually embedded in the object. By encapsulating these
defnitions, the objects can be used to develop related systems with less effort. It is
also easier to modify a system by making small changes to an objects behavior.
Object name:
Object attributes/properties:
Object functions/methods:
Interest Rate
Monthly Fees
Pay Interest
Compute Charges
Print Quarterly Statement
Number Beginning Balance
Name Ending Balance
Client Current Balance
Manager Interest Rate
Date Opened
Open Acct Accept Deposits
Close Acct Withdrawal
Pay Interest
Lowest Balance in Month
Bad Check Charges
Authorized Signature
Print Monthly Statement
Send Bad Check Notice
Pay Interest
Budget Saver
Savings
Account
Checking
Money Market
Senior Citizen CD
Fixed Fee
Student
Volume
inheritance from the base class (account), such as the balance attributes. Each
level adds additional detail. Each account class also contains member functions
to perform operations, such as paying interest. Because the interest computations
can be different for each of the accounts, the method is stored with the original
defnition of each account.
To see the usefulness of the object approach, consider what happens if the bank
decides to collect additional data for the checking accounts. The only change
needed is to add the new data items (and the associated functions) to the checking
account class. All checking accounts will then inherit those properties and func-
tions. None of the other operations are affected. Changes to the information sys-
tem will only affect the specifc accounts; the rest of the system will remain the
same.
From Programming to Development
Now that you have an idea of what it takes to create even a simple program, mul-
tiply that task by thousands to get a feel for the challenges of building a large,
complex application. Plus, the application needs to do what the users want, not
necessarily what the programmers want it to do. So someone has to identify busi-
ness needs and translate those into concepts that the programmers understand.
This role is handled by a systems analyst. In small projects, the programmers also
perform the systems analyst tasks, but in large projects the discussion and design
issues are handled by specialists.
In theory, a single programmer could develop almost any applicationbut you
will probably have to wait several years for the project to fnish. Keep in mind that
the development tools are continually improving, so actually a single programmer
today can build a relatively complex system in a short period of time. Projects
Figure 12.10
Runaway projects. Managers fear runaway projects, but they still occur. Some
projects end up two to fve times over budget and behind schedule. Some projects
are canceled because they never meet their objectives. Some fail because of design
problems and conficts among users, management, and developers. An important step
in managing projects is to identify when the project becomes a runaway project.
$
2013
2014
2012
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31
2015
* Technical measures
* 2-5 times over budget
* 2-5 times behind schedule
* Missing technical objectives
* Design problems
* Dupilcation of eorts
* Incompatibilities
* User/designer conicts
that once required teams of programmers weeks or months to build can be built
in a short time by even a single programmer. However, there are dozens of other
details that need to be handled, including testing, installation, documentation, and
training. The overall time can be reduced if additional people are hired to handle
these tasks, and if the programming can be divided into pieces so additional pro-
grammers can work on the pieces at the same time. Whenever you add people to a
project, you need a way to assign tasks, communicate ideas, monitor progress, and
control the entire process. At this point, you need a development methodology.
Systems Development Life Cycle
How do you control a major development project? Runaway proj-
ects illustrated in Figure 12.10 are a substantial problem in any development ef-
fort, but they are particularly important for new designs. Building a project from
scratch means it is hard to estimate the amount of time and effort needed to build
the system. As projects become larger, they become more diffcult to monitor and
control. Several major projects, including e-commerce frms, failed because they
were unable to produce a working system.
A factor in many runaway projects is the concept of scope creep or expanding
features. Once development starts, users and programmers start thinking of new
ideas that they would like to see in the project. So a simple, two-month project for
one division suddenly expands into a two-year companywide project costing mil-
lions of dollars. A key role of any IT project manager is to politely avoid adding
features that are not immediately necessary.
SDLC was designed to overcome the problems that arose with large projects
that involve many users and require thousands of hours of development by mul-
tiple analysts and programmers.
Before the use of the SDLC method, several related problems were common.
It was hard to coordinate and control the various programmers and analysts, so
efforts were duplicated. Individual programmers created portions of a system that
would not work together. Users did not always have much input into the process.
When they did have input, there were conficts between users, and analysts did
not know which approach to use. With long-term projects, programmers were pro-
moted to other areas or left for different companies. New employees had to learn
the system and determine what others had done before they could proceed. Simi-
larly, new users would appear (through promotions and transfers), and existing
users would change the specifcations of the system. These problems often lead
to runaway projectsprojects that are signifcantly late and over budget. Even
today, there are many instances of runaway projects.
Figure 12.11 shows some of the main risks in software development as outlined
by Boehm in 1991. The one at the bottom of the list is an interesting question
because it keeps changing. Developers and academics know how to solve certain
types of problems. Computer scientists also know some problems that do not yet
have good solutions. Any project that pushes these boundaries is going to be dif-
fcult to create and hard to control. Organizations need to be careful when entering
into projects that are experimental. Dealing with large projects that use standard
tools is hard enough to control. Fortunately, most business projects use relatively
common technologies. Still, it is easy to get carried away and start believing that
computers can perform impossible tasks.
These problems are related through the issue of control. It is impossible to pre-
vent users from changing the specifcations and to prevent employees from tak-
ing other jobs. Likewise, large projects involving many analysts and program-
mers will always have problems with coordination and compatibility. The goal of
SDLC was to design a system that can handle all of these problems.
A key value in SDLC is project management. An important aspect of project
management consists of identifying the dependencies among the various tasks.
Project management tools exist to help evaluate these dependencies and show
how the overall schedule is affected by delays in individual tasks.
Introduction to SDLC
An important feature of the SDLC approach is that it is a comprehensive method.
Some organizations (such as EDS) that specialize in systems development have
hundreds of pages in manuals to detail all the steps and rules for using SDLC.
Fortunately, it is possible to understand SDLC by looking at a smaller number of
steps. As illustrated in Figure 12.12, the SDLC approach encompasses fve basic
stages: (1) feasibility and planning, (2) systems analysis, (3) systems design, (4)
implementation, and (5) maintenance and review.
Actually, just about any systems-development methodology uses these fve
steps. They differ largely in how much time is spent in each section, who does the
work, and in the degree of formality involved. The SDLC approach is by far the
most formal method, so it offers a good starting point in describing the various
methodologies.
Feasibility and Planning
The primary goal of systems analysis is to identify problems and determine how
they can be solved with a computer system. In formal SDLC methodologies, the
frst step in systems analysis is a feasibility study. A feasibility study is a quick
examination of the problems, goals, and expected costs of the system. The objec-
tive is to determine whether the problem can reasonably be solved with a comput-
er system. In some cases, maybe there is a better (or cheaper) alternative, or per-
haps the problem is simply a short-term annoyance and will gradually disappear.
In other cases, the problem may turn out to be more complex than was thought
and to involve users across the company. Also, some problems may not be solv-
Figure 12.11
Project Risks (Boehm 1991). Some of the main risks that can delay or prevent
the successful completion of software development projects. Good development
practices can reduce some of these risks, but project managers must constantly watch
for signs of problems.
Personnel shortfalls
Unrealistic schedules and budgets
Developing the wrong functions and properties
Developing the wrong user interface
Gold plating (adding more functionality/features than necessary)
Continuing stream of requirements changes (scope creep)
Shortfalls in externally furnished components
Shortfalls in externally performed tasks
Real-time performance shortfalls
Straining computer-science capabilities
able with todays technology. It might be better to wait for improved technology
or lower prices. In any case, you need to determine the scope of the project to gain
a better idea of the costs, benefts, and objectives.
The feasibility study is typically written so that it can be easily understood by
nonprogrammers. It is used to sell the project to upper management and as a
starting point for the next step. Furthermore, it is used as a reference to keep the
project on track, and to evaluate the progress of the MIS team. Projects are typi-
cally evaluated in three areas of feasibility: economical, operational, and techni-
cal. Is the project cost effective or is there a cheaper solution? Will the proposed
system improve the operations of the frm, or will complicating factors prevent it
from achieving its goals? Does the technology exist, and does the frm have the
staff to make the technology work?
When the proposal is determined to be feasible, the MIS team leaders are ap-
pointed, and a plan and schedule are created. The schedule contains a detailed list-
ing of what parts of the project will be completed at each time. Of course, it is ex-
tremely diffcult to estimate the true costs and completion dates. Nonetheless, the
schedule is an important tool to evaluate the status of the project and the progress
of the MIS teams. As shown in Figure 12.13, planning and scheduling provides
the blueprint or structure for the rest of the project. It is a crucial step that provides
control for the remaining project.
Existing System
Systems Maintenance: Incremental changes
Feasibility &
Planning
Systems
Analysis
Systems
Design
Systems
Implementation
Problems &
Improvements
New System
Proposal
Goals & plans
Business requirements
Technical Design
problems
revisions
problems
revisions
Figure 12.12
Systems development life cycle. Sometimes SDLC is known as the waterfall
methodology because each step produces outputs that are used in the next step. The
existing system is studied for problems and improvements. A new design is analyzed
for feasibility. In-depth analysis generates the business requirements. Systems design
turns them into a technical design that is implemented, creating a new system. This
new system is analyzed and the process continues.
Systems Analysis
Once a project has been shown to be feasible and is approved, work can begin
on a full-fedged analysis. The frst step is to determine how the existing system
works and where the problems are located. The technique is to break the system
into pieces. Smaller pieces are easier to understand and to explain to others. Also,
each piece can be assigned to a different MIS team. As long as they work from the
same initial description and follow all of the standards, the resulting pieces should
ft back together. Of course, it still takes time and effort to integrate all of the piec-
es. The key objective in this stage is to understand the business organization and
determine the specifc requirements for the new project. It is also useful to collect
test cases. Modern development systems can be programmed to automatically run
test cases as the code is being developed and changed. The test data (with the cor-
rect results) ensure that the system always produces accurate results.
Diagrams are often created to illustrate the system. The diagrams are used to
communicate among analysts and users, other analysts, and eventually the pro-
grammers. Data fow diagrams are a common method to display the relationships
that were determined during systems analysis. The diagrams represent a way to
divide the system into smaller pieces.
Graphics tools provide a useful way to communicate with the user and to doc-
ument the user requirements. However, they do not speed up the development
process. Producing, changing, and storing documentation can be a signifcant
problem. Yet these tools are necessary because they make it easier for the user
Figure 12.13
Development controls. A complex system requires careful management. Without
planning and control, any project will become a runaway. Control begins with a
detailed plan and performance targets that enable managers to evaluate progress
and identify problems. System control is provided by standardized practices and
procedures to ensure that teams are producing compatible output. User input and
control ensure that the fnal project will actually be useful.
Detailed work plan
Performance targets
Practices & procedures
User input & control Blue Print/Planning
to control the fnal result. One increasingly common solution is to keep all of the
documentation on the computer. This method reduces the costs, makes it easier
for everyone to share the documentation, and ensures that all users have up-to-
date information for the system.
At the end of the analysis phase, the MIS team will have a complete description
of the business requirements. The problems and needs are documented with text,
data fow diagrams, and other fgures depending on the methodology followed.
Systems Design
The third major step of the SDLC approach is to design the new system. During
this step, the new system is typically designed. The objective of systems design
is to describe the new system as a collection of modules or subsystems. By sub-
dividing the total project, each portion can be given to a single programmer to
develop. As the pieces are completed, the overall design ensures that they will
work together. Typically, the diagrams created during the analysis phase can be
modifed to indicate how the new system will work. The design will list all of the
details, including data inputs, system outputs, processing steps, database designs,
manual procedures, and feedback and control mechanisms. Backup and recovery
plans along with security controls will be spelled out to ensure that the database is
protected.
In traditional SDLC methods, managers and users will be shown various com-
ponents of the system as they are completed. The managers will have to sign off
on these sections to indicate that they meet the user needs. This signature is de-
signed to ensure that users provide input to the system. If there are many diverse
users, there can be major disagreements about how the system should function.
Sign-offs require users to negotiate and formally agree to the design. It is rela-
tively easy to make design changes at this stage. If everyone attempts to make
changes at later stages, the cost increases dramatically.
In terms of physical design, some of the hardware and software will be pur-
chased. Programmers will write and test the program code. In most large projects,
the actual coding takes only 15 to 30 percent of the total development time. Initial
data will be collected or transferred from existing systems. Manuals and proce-
dures will be written to instruct users and system operators on how to use the
system.
Reality Bytes: What Tech Skills do Employers Want
Because technology continually changes, students are always asking what skills are
needed. Actually, even experienced IT employees ask the same question. In 2011,
employers said they wanted the several basic skills: 77 percent want programming
skills, 82 percent want database skills, 76 percent want analytical and architectural
skills, and 80 percent want general problem solving and technical skills. Many em-
ployers also want experience, but when the job market tightens, they have few op-
tions. In the same study, 50 percent said they were willing to hire new IT graduates;
but two-thirds wanted at least some college internship experience.
Adapted from Michael Cooney, IT Graduates Not Well-Trained, Ready-to-go,
Network World, February 25, 2011.
Design tools can be used to create prototypes of major system elements. For
example, a designer can quickly piece together displays that illustrate how each
screen might look and how the user will see the system. These prototypes can be
used to help users walk through aspects of the proposed system and make changes
while it is easy and inexpensive. The walkthroughs also provide management with
feedback regarding the time schedule and anticipated costs of the project, because
they are often scheduled in the original feasibility study.
The output of the design stage consists of a complete technical specifcation
of the new system. It includes as many details as possible, sometimes leading to
thousands of pages (or computer fles) of description.
One of the diffculties in the design stage is sometimes called creeping el-
egance. As the system is being built, analysts, programmers, and users all want
to include additional features. Although many of the features are good ideas, the
continual evolution of the system causes additional delays. It also complicates
testing, because changes in one section can affect the rest of the system.
Systems Implementation
Systems implementation involves installation and changeover from the previous
system to the new one, including training users and making adjustments to the
system. Many nasty problems can arise at this stage. You have to be extremely
careful in implementing new systems. First, users are probably nervous about the
change already. If something goes wrong, they may never trust the new system.
Second, if major errors occur, you could lose important business data.
A crucial stage in implementation is fnal testing. Testing and quality control
must be performed at every stage of development, but a fnal systems test is need-
ed before staff entrust the companys data to the new system. Occasionally, small
problems will be noted, but their resolution will be left for later. In any large sys-
tem, errors and changes will occur. The key is to identify them and determine
which ones must be fxed immediately. Smaller problems are often left to the soft-
ware maintenance staff.
Reality Bytes: You Won. Just Kidding
In 1994, the U.S. Congress created an international green-card lottery. The annual
competition randomly selects 50,000 people to live and work legally in the U.S. with
minimal other criteria. Almost 15 million people applied for the lottery in 2011. In
May, the State Department posted a list of 20,000 winning numbers on its Web site.
A few lucky people, such as Max, a 28-year-old German were ecstatic. Until a few
days later. The State Department identifed a glitch in the computer program that se-
lected the names. Apparently, 90 percent of the numbers selected had applied in the
frst two days. Deputy assistant secretary of state David Donahue noted that These
results are not valid because they did not represent a fair, random selection of the en-
trants as required by U.S. law. The lottery has been conducted electronically for 15
years, but the State department blamed a coding error in a new computer program.
The department did not provide an explanation as to why no one tested the program
ahead of time. The lottery was rescheduled.
Adapted from The Wall Street Journal, Computer Glitch Voids Green-Card Lot-
tery, May 14, 2011.
Technology Toolbox: Creating Forms with InfoPath
Problem: You want employees to use digital forms to collect basic data such as ex-
pense reports.
Tools: You can use InfoPath in Microsoft Offce to create, exchange, and store data
from digital forms.
InfoPath is included with several business versions of Offce 2007. It has a forms-
builder tool to create forms and defne the data that will be collected. Once the form
is created, it can be published to a SharePoint server, e-mailed to people, or built into
a Visual Studio project. Simple forms can be built that can be opened with just a Web
browser, so recipients do not need InfoPath installed on their computers. However,
InfoPath forms are more powerful than browser forms.
Creating forms is relatively easyespecially if you can use one of the sample
forms included with InfoPath. The basic steps are to defne the data elements, then
add text and input boxes using the layout tools. After checking the design for errors,
you publish the form so employees can fll it out as needed.
The easiest method is to centralize the forms and store them on a SharePoint serv-
er. You can use additional tools to build in workfow procedures. The expense report
is one of several sample forms included with InfoPath. Employees go to the Share-
Point server and create a new form flling out the relevant data. The data is saved
on the SharePoint server, and the form is e-mailed automatically to the employees
manager for review. At any time, the manager can see summaries of specifed data
columns on the SharePoint server site. All data is stored in XML fles, so it can be
transferred to other systems fairly easily.
Quick Quiz:
1. What standard business forms would you want to create electronically?
2. What security conditions would you impose when installing expense report
forms on a SharePoint server?
3. What are the benefts of using digital forms instead of paper forms?
Change is an important part of MIS. Designing and implementing new sys-
tems often cause changes in the business operations. Yet many people do not like
changes. Changes require learning new methods, forging new relationships with
people and managers, or perhaps even loss of jobs. Changes exist on many levels:
in society, in business, and in information systems. Changes can occur because of
shifts in the environment, or they can be introduced by internal change agents.
Left to themselves, most organizations will resist even small changes. Change
agents are objects or people who cause or facilitate changes. Sometimes it might
be a new employee who brings fresh ideas; other times changes can be mandated
by top-level management. Sometimes an outside event such as arrival of a new
competitor or a natural disaster forces an organization to change. Whatever the
cause, people tend to resist change. However, if organizations do not change, they
cannot survive. The goal is to implement systems in a manner that recognizes
resistance to change but encourages people to accept the new system. Effective
new
new
new
Direct cutover
Parallel
Pilot
new
Phased
old
dept or component 1
dept or component 2
dept or component 3
dept or component 4
store 1
store 2
store 3
store 4
new
new
new
old
old
Figure 12.14
Conversion options. When you implement a new system, there are several possible
conversion methods. In most cases, direct cutover should be avoided because
of the disruptions and potential for lost data. Parallel conversion entails running
both systems simultaneously, which is safe but can become expensive and time
consuming. With multiple stores or business units, pilot introductions of phased
implementations are common. For pilot testing, designers can bring extra workers,
managers, and systems designers to one location and work out the problems with the
system. Once the system is running well, it can be implemented at other locations.
With a phased implementation, a system can be introduced slowly throughout the
company (e.g., by department). Projects can also be phased in by modules.
implementation involves fnding ways to reduce this resistance. Sometimes, im-
plementation involves the cooperation of outsiders such as suppliers.
Because implementation is so important, several techniques have been devel-
oped to help implement new systems. Direct cutover is an obvious technique,
where the old system is simply dropped and the new one started. If at all possible,
it is best to avoid this technique, because it is the most dangerous to data. If any-
thing goes wrong with the new system, you run the risk of losing valuable infor-
mation because the old system is not available. The various methods are displayed
in Figure 12.14.
In many ways, the safest choice is to use parallel implementation. In this case,
the new system is introduced alongside the old one. Both systems are operated at
the same time until you determine that the new system is acceptable. The main
drawback to this method is that it can be expensive because data has to be entered
twice. In addition, if users are nervous about the new system, they might avoid the
change and stick with the old method. In this case, the new system may never get
a fair trial.
Several intermediate possibilities are called phased implementation. For ex-
ample, if you design a system for a chain of retail stores, you could pilot-test the
frst implementation in one store. By working with one store at a time, there are
likely to be fewer problems. But if problems do arise, you will have more staff
members around to overcome the obstacles. When the system is working well in
one store, you can move to the next location. Similarly, even if there is only one
store, you might be able to split the implementation into sections based on the area
of business. You might install a set of computer cash registers frst. When they
work correctly, you can connect them to a central computer and produce daily
reports. Next, you can move on to annual summaries and payroll. Eventually the
entire system will be installed.
Maintenance
Once the system is installed, the MIS job has just begun. Computer systems are
constantly changing. Hardware upgrades occur continually, and commercial soft-
ware tools may change every year. Users change jobs. Errors may exist in the
system. The business changes, and management and users demand new infor-
mation and expansions. All of these actions mean the system needs to be modi-
fed. The job of overseeing and making these modifcations is called software
maintenance.
The pressures for change are so great that in most organizations today as much
as 80 percent of the MIS staff is devoted to modifying existing programs. These
changes can be time consuming and diffcult. Most major systems were created by
teams of programmers and analysts over a long period. In order to make a change
to a program, the programmer has to understand how the current program works.
Because the program was written by many different people with varying styles, it
can be hard to understand. Finally, when a programmer makes a minor change in
one location, it can affect another area of the program, which can cause additional
errors or necessitate more changes.
One diffculty with software maintenance is that every time part of an applica-
tion is modifed, there is a risk of adding defects (bugs). Also, over time the appli-
cation becomes less structured and more complex, making it harder to understand.
At some point, a company may decide to replace or improve the heavily modifed
system. Several techniques can be used to improve an existing system, ranging
from rewriting individual sections to restructuring the entire application. The dif-
ference lies in scopehow much of the application needs to be modifed. Older
applications that were subject to modifcations over several years tend to contain
code that is no longer used, poorly documented changes, and inconsistent nam-
ing conventions. These applications are prime candidates for restructuring, during
which the entire code is analyzed and reorganized to make it more effcient. More
important, the code is organized, standardized, and documented to make it easier
to make changes in the future.
Evaluation
An important phase in any project is evaluating the resulting system. As part of
this evaluation, it is also important to assess the effectiveness of the particular
development process. There are several questions to ask: Were the initial cost es-
timates accurate? Was the project completed on time? Did users have suffcient
input? Are maintenance costs higher than expected? The assessment items are
summarized in Figure 12.15.
Evaluation is a diffcult issue. As a manager, how can you tell the difference
between a good system and a poor one? In some way, the system should decrease
costs, increase revenue, or provide a competitive advantage. Although these ef-
fects are important, they are often subtle and diffcult to measure. The system
should also be easy to use and fexible enough to adapt to changes in the busi-
ness. If employees or customers continue to complain about a system, it should be
reexamined.
A system also needs to be reliable. It should be available when needed and
should produce accurate output. Error detection can be provided in the system
to recognize and avoid common problems. Similarly, some systems can be built
to tolerate errors, so that when errors arise, the system recognizes the problem
Feasibility Comparison
Cost & Budget Compareactualcoststobudgetestimates.
Time Estimates Was project completed on time?
Revenue Efects Does system produce additional revenue?
Maintenance Costs How much money and time are spent on changes?
Project Goals Does system meet the initial goals of the project?
User Satisfaction How do users (and management) evaluate the system?
System Performance
System Reliability Are the results accurate and on time?
System Availability Is the system available on a continuous basis?
System Security Does the system provide access only to authorized users?
Figure 12.15
Evaluation of completed projects. When projects are completed, the design team
should evaluate the project and assess the development procedures. Cost and time
estimates can be used to improve estimates for future projects. System performance
issues can be addressed with future upgrades. It is important that the system achieve
project goals and provide users with necessary tools and support.
and works around it. For example, some computers exist today that automatical-
ly switch to backup components when one section fails, thereby exhibiting fault
tolerance.
An important concept for managers to remember when dealing with new sys-
tems is that the evaluation mechanism should be determined at the start of the
project. Far too often, the question of evaluation is ignored until someone ques-
tions the value of the fnished product. It is a good design practice to ask what
would make this system a good system when it is fnished, or how we can tell a
good system from a bad one in this application. Even though these questions may
be diffcult to answer, they need to be asked. The answers, however incomplete,
will provide valuable guidance during the design stage.
Recall that every system needs a goal, a way of measuring progress toward that
goal, and a feedback mechanism. Traditionally, control of systems has been the
task of the computer programming staff. Their primary goal was to create error-
free code, and they used various testing techniques to fnd and correct errors in the
code. Today, creating error-free code is not a suffcient goal.
Everyone has heard the phrase The customer is always right. The meaning
behind this phrase is that sometimes people have different opinions on whether
a system is behaving correctly. When there is a confict, the opinion that is most
important is that of the customer. In the fnal analysis, customers are in control
because they can always take their business elsewhere. With information systems,
the users are the customers and the users should be the ones in control. Users de-
termine whether a system is good. If the users are not convinced that the system
performs useful tasks, it is not a good system.
Strengths and Weaknesses of SDLC
The primary purpose of the SDLC method of designing systems is to provide
guidance and control over the development process. As summarized in Figure
12.16, there are strengths and weaknesses to this methodology. SDLC manage-
ment control is vital for large projects to ensure that the individual teams work
Figure 12.16
Strengths and weaknesses of SDLC. The SDLC methodologies were created to
control large, complex development projects. They work fairly well for those types of
processes. They do not work as well for small projects that require rapid development
or heavy user involvement with many changes.
Strengths Weaknesses
Control. Increaseddevelopmenttime.
Monitorlargeprojects. Increaseddevelopmentcosts.
Detailedsteps. Systemsmustbedefnedupfront.
Evaluatecostsandcompletiontargets. Rigidity.
Documentation. Hardtoestimatecosts,projectoverruns.
Well-defneduserinput. Userinputissometimeslimited.
Easeofmaintenance.
Developmentanddesignstandards.
TolerateschangesinMISstafng.
together. There are also fnancial controls to keep track of the project expenses.
The SDLC steps are often spelled out in great detail. The formality makes it easier
to train employees and to evaluate the progress of the development. It also ensures
that steps are not skipped, such as user approval, documentation, and testing. For
large, complex projects, this degree of control is necessary to ensure the project
can be completed. Another advantage of SDLC is that by adhering to standards
while building the system, programmers will fnd the system easier to modify and
maintain later. The internal consistency and documentation make it easier to mod-
ify. With 80 percent of MIS resources spent on maintenance, this advantage can
be critical.
In some cases the formality of the SDLC approach causes problems. Most im-
portant, it increases the cost of development and lengthens the development time.
In many cases less than 25 percent of the time is spent on actually writing pro-
grams. A great deal of the rest of the time is spent flling out forms and drawing
diagrams.
The formality of the SDLC method also causes problems with projects that are
hard to defne. SDLC works best if the entire system can be accurately specifed in
the beginning. That is, users and managers need to know exactly what the system
should do long before the system is created. That is not a serious problem with
transaction-processing systems. However, consider the development of a complex
decision support system. Initially, the users may not know how the system can
help. Only through working with the system on actual problems will they spot er-
rors and identify enhancements.
Although some large projects could never have been completed without SDLC,
its rigidity tends to make it diffcult to develop many modern applications. More-
over, experience has shown that it has not really solved the problems of projects
being over budget and late. As a result of this criticism, many people are searching
for alternatives. One possibility is to keep the basic SDLC in place and use tech-
nology to make it more effcient. Other suggestions have been to replace the entire
process with a more effcient development process, such as prototyping. Consider
the assistance of technology frst.
Several researchers at Carnegie Mellon University have created (and trade-
marked) the capability maturity model integration (CMMI) to help develop-
ment organizations evaluate their abilities. Figure 12.17 shows the various levels
of maturity. The goal is to improve the development process within an organiza-
tion so that everyone follows a process that is measurable and sustainable. In stan-
dard management terms, quantifying the development process makes it possible
to fne-tune and improve. Possibly the greatest strength of the CMMI approach
is also one of its weaknesses. The overall approach is designed to support and
encourage mediocrity in development. Programmers are considered interchange-
ablean organization that succeeds by relying on star programmers is consid-
ered to be inferior. For some large organizations (particularly governments), this
characterization makes sensethe system should function even with staff turn-
over. The weakness is that some software development requires creativity and
fexibility to create new approaches.
Alternatives to SDLC
Is SDLC always the best approach? What other methodologies
could be used? The two primary drawbacks to SDLC are that (1) it takes a
considerable amount of time, and (2) all the system details have to be specifed
up front. The project management and control features add paperwork and delays,
making SDLC unsuitable for small projects. SDLC works reasonably well for
transaction-processing systems that are well defned where the design elements
can be specifed up front. It does not work well for decision support systems par-
ticularly when users do not really know exactly what they want the system to do.
Prototyping or Iterative Development
Prototyping has been proposed as a method to use for systems that are not overly
complex and do not involve too many users or analysts. Just as automobile engi-
neers design prototypes before attempting to build the fnal car, MIS programmers
can build early versions of systems. These systems are then continually modifed
until the user is satisfed.
The frst step in designing a system via prototyping is to talk with the user.
The analyst then uses a fourth-generation language and a DBMS to create ap-
proximately what the user wants. This frst step generally requires only a couple
of weeks. The business user then works with the prototype and suggests changes.
The analyst makes the changes and this cycle repeats until the user is satisfed or
decides that the system is not worth pursuing. The emphasis is on getting a work-
ing version of the system to the user as fast as possible, even if it does not have all
the details. Figure 12.18 illustrates the cycle involved in prototyping.
The major advantage of prototyping is that users receive a working system
much sooner than they would with the SDLC method. Furthermore, the users
have more input so they are more likely to get what they wanted. Finally, remem-
ber that a large portion of MIS time is spent making changes. A system designed
with the prototyping method is much easier to change because it was designed to
be modifed from the start.
Extreme Programming and Agile Development
In some ways, extreme programming (XP) is a new concept; in other ways it is
an extension of the prototyping ideas. The main premise of XP is that SDLC and
its variants are too large and cumbersome. While they might provide control, they
end up adding complexity, taking more time, and slowing down top programmers.
Figure 12.17
Capability Maturity Model. Based on standard management techniques. A
development organization should strive to install processes, measure progress, and
improve the development methodology.
1. Initial. Ad hoc development with undefned processes. Often driven by
individual programmers.
2. Managed. Standard project management tools to track costs and sched-
ules. Basic processes to ensure development is repeatable.
3. Defned. Management and development is defned and standardized.
Processes are documented and followed.
4. Quantitatively Managed. Detailed measures are collected and evalu-
ated.
5. Optimizing. Continuous improvement methods are applied to fne-tune
and improve the development process.
XP and agile development simplify the development process by focusing on small
releases (similar to prototyping) that provide value to the customer.
As shown in Figure 12.19, XP and agile development were pushed heavily in
the development of new Web-based systems. In this highly-competitive environ-
ment, getting applications out the door and on the Web quickly was more impor-
tant than loading on tons of features. Yet, everything had to work correctly. To
improve quality XP adopted a relatively new principle from computer science:
building test cases frst. The basic system is designed in terms of what each mod-
ule should accomplish. These features are defned with a set of test cases. Pro-
grammers then write code and feed the test cases through to ensure the modules
work correctly. Whenever the system is changed, the programmers rerun the test
cases to ensure nothing else was broken. Tools were developed to store the test
cases and the results, making it easy to develop new cases and modules quickly.
One new aspect to XP is paired programming, where two programmers work
together constantly. Generally, one is the lead programmer and the other is re-
sponsible for testing, but the jobs can overlap and be defned by the team. Making
testing a key element of programming is an important part of XP. However, paired
programming is seen by many as an ineffcient use of resources. The second pro-
grammer is often a less experienced developer and can slow down an experienced
developer. Besides, it can be more effcient to have one person test large sections
of code at a time, instead of multiple people testing separate pieces.
Figure 12.18
Prototyping. Prototyping typically involves one user and one developer. The
developer interviews the user and designs an initial system using a DBMS. The user
works with the prototype and suggests changes. This process repeats until the user or
developer is satisfed or gives up.
Build Initial Prototype
Use Prototype
Modify Prototype
Request changes
New Version
MIS Designer
Initial Interview
Process repeats until:
User
1) User is satised.
2) User and designer give up.
3) Formal system is built from prototype.
4) Need for application is removed.
5) Process never ends.
One of the most challenging aspects to development is that there is a tremen-
dous difference between individual programmersin subject area knowledge,
speed of programming, number of defects, and code maintainability. Some meth-
odologies work well when an organization has top-notch developers, but fall apart
in other companies. In choosing a methodology, managers must be aware of the
capabilities of the individual programmersand beware of turnover.
Developing Systems Requires Teamwork: JAD and RAD
Designing and developing systems is much easier if the entire system can be built
by one person. In fact, that is one of the strengths of recent toolsthey enable a
single person to build more complex systems. However, many information sys-
tems, especially those that affect the entire organization, require teams of IS work-
ers. As soon as multiple designers, analysts, and programmers are involved, ev-
eryone encounters management and communication problems. MIS researchers
have measured the effects of these problems. One study by DeMarco and Lister
showed that on large projects, 70 percent of a developers time is spent working
with others. Jones noted that team activities accounted for 85 percent of the de-
velopment costs. There seem to be substantial areas for improvement in systems
development by focusing on teamwork.
One of the most diffcult steps in creating any new system is determining the
user requirements. What does the system need to do and how will it work? This
step is crucial. If the designers make a mistake here, the system will either be use-
less or will need expensive modifcations later. Prototyping and SDLC take dif-
ferent approaches to this problem. With SDLC, analysts talk with users and write
reports that describe how the system will operate. Users examine the reports and
Figure 12.19
Extreme programming. Set target release dates and build test cases so you know
when modules work. Use paired programming or other fast techniques to build
modules and test them. Add features for next release.
time
Inputs: 16, 7, 19
Output: 91
Test cases
Paired programming
Release 1.0 Release 1.1
Target release dates.
Build test cases.
Write code and test it.
Release product.
Add features for next release.
Technology Toolbox: Programming a New Function in Excel
Problem: You need to add a tricky function to a spreadsheet.
Tools: Microsoft Offce contains a language to create your own tools and functions.
Many programming languages exist for different jobs, but they all have similar fea-
tures. The common tasks are (1) defne variables and perform calculations, (2) create
functions and subroutines, (3) use conditional (if) statements to make choices, (4)
write loops to repeat steps, and (5) input and output data (to the spreadsheet in this
case). You use these building blocks to build functions and applications.
A fnancial option is a contract that enables you to purchase (or sell) shares of
stock in the future for a specifed exercise price. The stock currently trades at some
other price, so you are gambling the price will increase or decrease. The Black-
Scholes equation from fnance is often used to calculate a value for option prices.
You need a place to write your new function. Open a new worksheet. Choose
Developer/Record Macro. Start recording, click a cell in the worksheet and stop re-
cording. Now choose View Code and delete the Macro1 subroutine you created in
Module1. Add the new BlackScholes function.
Return to the spreadsheet and enter some sample data for stock price (60), ex-
ercise price (65), time left (0.25), rate (0.08), and volatility (0.3). To determine the
value of a call option, enter the formula: =BlackSholes(call, B2, B3, B4, B5, B6),
where the cell values match the locations you put the sample data. The spreadsheet
will call your new function and return the results.
stock price 60 call 2.133368
exercise price 65 put 5.846282
time left 0.25
rate 0.08
volatility 0.3
Quick Quiz:
1. What does the statement If (Left(CallPut, 1) = c) do in the code?
2. What security setting do you need for this function to work?
3. How can a function directly alter several cells in a spreadsheet?
Function BlackScholes(CallPut As String, StockPrice As Double,_
ExercisePrice As Double,TimeLeft As Double, rate As Double, _
volatility As Double) As Double
Dim d1 As Double, d2 As Double
d1 = (Math.Log(StockPrice / ExercisePrice) + _
(rate + volatility ^ 2 / 2) * TimeLeft) / _
(volatility * Math.Sqr(TimeLeft))
d2 = d1 - volatility * Math.Sqr(TimeLeft)
If (Left(CallPut, 1) = "c") Then
BlackScholes = StockPrice _
* Application.WorksheetFunction.NormSDist(d1) _
- ExercisePrice * Exp(-rate * TimeLeft) _
* Application.WorksheetFunction.NormSDist(d2)
Else
BlackScholes = ExercisePrice*Exp(-rate*TimeLeft) * _
Application.WorksheetFunction.NormSDist(-d2)-StockPrice _
* Application.WorksheetFunction.NormSDist(-d1)
End If
End Function
make changes. This approach is time consuming and diffcult for users because
they only see paper notes of the proposed system. Prototyping overcomes some
of the problems by letting users work with actual screens and reports. But use of
prototyping is hard to expand beyond one or two users.
Some companies overcome the problems of SDLC by prototyping each input
screen and report with one or two primary users. Once the main concepts have
been designed, the analysts formalize the system and get approval from other
users. The designs are then given to programmers to create with the traditional
SDLC development methods.
Recall that an important reason for using SDLC is to obtain the views and
agreement of many users. Using traditional interview methods and paper docu-
mentation, this process often takes several months. Each change has to be reexam-
ined by other users, and disagreements have to be resolved.
A technique known as joint application design (JAD) was created to speed
up the design stage. With JAD the main system is designed in an intense three-
to fve-day workshop. As shown in Figure 12.20, users, managers, and systems
analysts participate in a series of intense meetings to design the inputs (data and
screens) and outputs (reports) needed by the new system.
By putting all of the decision makers in one room at the same time, conficts
are identifed and resolved faster. Users and managers gain a better understanding
of the problems and limitations of technology. The resulting system has greater
Goals
Primary Needs
Computations
Forms
Database
Get everyone together to
identify the primary
elements of the design
with no distractions.
Figure 12.20
Joint application design. Application design can be accelerated and simplifed by
putting key users and developers together for a few days. By focusing on the single
project, everyone gets input and can reach a consensus in a shorter time.
value for users and managers because it more closely matches their needs. There
is less need for changes later, when they become more expensive, so the system is
cheaper to create.
The biggest drawback to JAD is that it requires getting everyone together at the
same time for an extended period of time. Even for moderately complex systems,
the meetings can run eight hours a day for three to fve days. Most managers (and
users) fnd it diffcult to be away from their jobs for that length of time. Higher-
level managers are also needed at these meetings to ensure the system provides
the appropriate reports and information. Finally, the meetings can succeed only if
they are led by a trained facilitator. The facilitator keeps the discussions moving
in the right direction, minimizes conficts, and encourages everyone to participate.
At the end of the sessions, the systems development team should have a complete
description of the proposed system.
Rapid application development (RAD) applies the value of teamwork to the
developers. By providing advanced development tools, prebuilt objects, and col-
laboration tools, some companies have found it is possible to reduce the overall
development time. The key is to target steps that can overlap and be performed
by multiple teams. By improving the collaboration tools, more steps can be com-
pressed. Many e-commerce projects were developed with RAD techniques. Firms
were concerned about being the frst in the market and felt they needed to develop
software rapidly. The goal of being frst was later shown to be pointless, but the
techniques of using small groups of programmers with advanced tools, collabora-
tion, and intense programming sessions were relatively successful at quickly pro-
ducing thousands of new applications.
Figure 12.21
Communication. A project with two people has simple communication needs. Adding
more users or more developers multiplies the number of potential communication
problems. Different methodologies are used to handle and improve communication
among the participants.
Communication
Communication is a challenge in any projectbut it is particularly critical in pro-
gramming or systems development projects. Figure 12.21 shows the basic issue
that as the number of participants increases, the need for communication multi-
plies. One of the key goals of a methodology is to defne and improve the way that
everyone communicates. Some formal systems, such as SDLC, defne specifc
communication pathsoften with a top-down approach where project managers
produce daily or weekly statements. The fgure also explains why adding people
to a project slows it down. More people means more communication needs to take
place.
Other methodologies work to reduce the communication needs by reducing the
number of peopleor by focusing the immediate communication needs among
a smaller number of participants. For example, prototyping works because one
developer works with one user, emphasizing the input and communication of
that person. Extreme programming and modern tools work because they enable
a single developer or small team to produce systems instead of relying on a large
number of programmers. Even SDLC relies on reducing communication needs by
emphasizing the importance of splitting a problem into smaller, independent piec-
es. Each piece can be developed by one person or a small teamwithout needing
to see all of the details from every other programmer. The goal is to isolate the
details and communicate only the formal interface connections.
Object-Oriented Design
Based on the value of OOP, it was natural for organizations to try and extend the
concepts of objects to the entire design process. The goal is to defne business
objects that will apply to the entire organization. The objects would include the
ability to retrieve data and perform standard functions. For instance, a Customer
Reality Bytes: Thousands of Apps not Much Money
In 2011, Google claimed that 450,000 developers had produced 200,000 applica-
tions for Android-based smart phones. In March, 2011, third-party counts estimated
250,000 Android apps versus 350,000 for Apple. But, the count quickly becomes
meaninglessno one even has time to look at even a review of that many applica-
tions. Distimo, an analytics frm, noted that only 72,000 of the Android apps were
sold for money; and of that list, only two had been downloaded more than 500,000
times worldwide. The company reported that of the iPhones 211,000 paid apps,
six were downloaded more than 500,000 times in April and May 2011. In terms of
games, the most popular category, the Android market has fve paid games down-
loaded more than 250,000 times and the iPhone has 10 paid games downloaded in
the U.S. over two months. Of course, the iPhone has been sold for several more
years than the Android phones. For developers and entrepreneurs, the point is that
despite the potentially large market, very few smart phone applications make any
money. In the Android market, 20 percent of all free apps and 80 percent of all paid
apps were downloaded less than 100 times.
Adapted from Clint Boulton, Android No Threat to Apple in Paid Apps, eWeek,
May 29, 2011; and http://www.businessinsider.com/charts-of-the-week-ipad-compe-
tition-is-toast-2011-3#google-is-closing-the-gap-on-apples-app-store-3
object could be defned one time and stored in a central location. Whenever a pro-
grammer builds an application, the appropriate objects could be retrieved from the
central store and pieced together to create the fnal application. With most of the
work defned ahead of time, it would be easier and faster to build the application
compared to starting from scratch each time.
Although the idea was sound, object-oriented design did not work very well in
practice. Part of the problem is the necessity of defning all of the objects ahead
of time. Most organizations are so busy trying to get basic tasks completed they
do not have time to create objects that might be used someday. Another problem
is that organizations change relatively quickly, often resulting in the need for new
and modifed objects. Finally, most big organizations simply decided to purchase
ERP systems to integrate all of the organizations applications. The ERP systems
use a DBMS to store data, and have internally-defned objects that can be used
for custom programming. In a sense, the ERP systems solved the problem by pre-
defning standard business objects that can be customized for any organization.
But the initial development costs are borne by the ERP vendors and spread across
thousands of customers. For any other application, OO design has generally been
limited to single applications.
Open Source Development
Open source development is an interesting new method of developing complex
software. With this approach, developers from virtually any company or loca-
tion work on portions of the code. Usually, one person coordinates the efforts and
identifes major changes and structure. The individual programmers write, debug,
or test sections of code. If a programmer fnds a better way to implement a func-
tion, the newer version is incorporated into the code. Hundreds or even thousands
of programmers can contribute to the development of a project.
So far, this methodology has been used only to develop free software that is
available for everyones use. Many of the techniques were pioneered by Richard
Stallman who developed emacs, a programmers text editor. He later founded the
GNU project (www.gnu.org) that uses the same methods to create and distribute
Reality Bytes: Random Hacks of Kindness
Computer programmers can be expensive. Developing software requires training
and creativity and demand for programmers is high, so it can be diffcult for not-for-
proft frms to hire or pay programmers. In 2009, Google, Yahoo!, the World Bank,
NASA, and Microsoft created the Random Hacks of Kindness organization to en-
courage programmers and hackers to devote a weekend to helping fnd solutions to
old problems and help not-for-proft organizations. In 2011 in Nairobi, developers
stayed on site for 36 hours to develop programs to assist the Kenyan Red Cross man-
age volunteers. Another application helped people infected with HIV/AIDS stay on
their medication during a disaster. An application created in Atlanta was similar to
one built in Toronto. MessageCarrier can be used in a disaster to collect messages
from people in remote areas without Internet or phone access. When the message
phone is reconnected to the network, it will transfer all of the messages.
Adapted from Alden Mahler Levine, Brains Collide During Hackathon for Cli-
mate Change, Disaster Relief, CNN Online, June 8, 2011.
tools and systems software. Linus Torvalds uses a similar approach to create and
distribute the Linux operating system.
Open source development is interesting in terms of both the sophisticated soft-
ware that has been created and the development methodology. Using Internet
communications, and only a small team to coordinate and review the work, thou-
sands of individuals have been able to work together to create complex software
that rivals commercial products costing millions of dollars to create. In theory,
similar techniques could be used to improve development within business. On
the other hand, the technique requires the cooperation of hundreds of developers,
often some of the best programmers in the world. It might be possible to hire these
programmers on a freelance basis. A few companies offer Web sites that enable
you to auction contracts for various portions of a programming job. But it is not
entirely clear that this approach is cheaper than just hiring the best programmers.
Another issue with open source development also affects your decision about
whether to use open source products such as Linux. How is the software going
to be maintained and updated? Creating the initial software is only the frst step.
Bugs have to be fxed and new features added on a regular basis. As long as there
is a core group of people willing to continue working on the project, these issues
can be handled. Or if you have a staff with the skills to modify the software, you
can make any changes you wantbecause you have the source code. But what
happens 20 years later? Or even in the short run, can open source projects devote
the time and money to usability testing and radical improvements as hardware
changes? A commercial company has a fnancial incentive and the cash fow to
keep products moving forward. Open source development has only the personal
motivations of the prime organizer and the world developer community. Some-
times these motivations are enough to ensure the longevity of a product; some-
times they are not.
End-User Development
The term end user development simply means that users do all of the develop-
ment work themselves. In many ways, it resembles prototyping, except that users
(instead of analysts from the MIS department) create and modify the prototypes.
Clearly the main advantage is that users get what they want without waiting for an
MIS team to fnish its other work and without the diffculty of trying to describe
the business problems to someone else.
Two basic reasons explain why end-user development is increasingly popular.
First, most MIS organizations are facing a two- or three-year backlog of projects.
This means that if you bring a new project to MIS, the designers will not even
start on it for at least two years (unless you give up some other project). The sec-
ond reason is that software tools are getting more powerful and easier to use at
the same time. Today it is possible for users to create systems with a spreadsheet
in a few hours that 10 years ago would have taken MIS programmers a month to
build with third-generation languages. As tools become more powerful and more
integrated, it becomes possible to create even more complex systems. Think about
some of the database and Web toolssuch as PivotTables. Ten years ago, most
users would not dream of being able to create these reports. Today, you can get
reports and analyze data any way you want with a few clicks of the mouse. The
advantages of end-user development are similar to those in prototyping. In par-
ticular, users get what they want, and they get working systems sooner.
The potential problems of end-user development are not always easy to see.
Most of them arise from the fact that users generally lack the training and experi-
ence of MIS analysts and programmers. For instance, systems produced by end
users tend to be written for only one person to use. They are oriented to working
on stand-alone personal computers. The systems are often customized to ft the
needs of the original users. Moreover, most users do not write documentation, so
others will have diffculty using the products. Because of lack of training, users
rarely perform as much testing as they should. The systems lack security controls
and are hard to modify. Think about the problems you encounter when you are
given a spreadsheet that was created by the person who held the job before you.
Other problems stem from the bottom-up approach inherent in end-user de-
velopment. People in different areas of the company will wind up working on
the same problem, when it could have been solved once by MIS. Data tends to
be scattered throughout the company, making it hard to share and wasting space.
Not following standards generates incompatibilities among systems, making it
diffcult to combine systems created by different departments or even by people
within the same department.
End users are limited by the capabilities of commercial software. The initial
systems may work fne, but as the company grows and changes, the commercial
software might be unable to support the necessary changes. As a result, some us-
ers have created systems that produce incorrect answers, take too long to run, or
lose data.
The last, and possibly most important, complication is that end-user develop-
ment takes time away from the users job. Some users spend months creating and
modifying systems that might have been created by MIS programmers in a frac-
tion of the time. One of the reasons for creating an MIS department is to gain ef-
fciency from using specialists. If users are spending too much time creating and
revising their own applications, the company needs to consider hiring more MIS
personnel.
Development Summary
As a manager, one of the more diffcult IT decisions you make is the choice of
development methodology. As a business manager in a large organization, you
might not have a vote. But within a smaller company, you will certainly have to
look at the alternatives to help identify the most effcient means of creating proj-
ects. Even within a larger company, you might be in a position to suggest alterna-
tives when price tags or time frames get too high.
Figure 12.22 summarizes the characteristics of the primary development meth-
odologies. The chart is basically organized on a scale of formality. Large, formal
projects are built using SDLC to control the development and record progress.
Small scale reports and analyses can be created with end-user development. Pro-
totyping is similar to end-user development, but is a step toward more control and
formality because it uses trained MIS developers who follow established proce-
dures and internal standards.
Remember that the various methodologies can be combined. For example, a
JAD session might be used to defne the initial goals and attributes of a large
project. The forms might be refned through prototyping. But the overall project
could be controlled through an SDLC project management system. Remember
that each technology has different costs. SDLC provides the most control, but
adds overhead costs that you have to recognize. On the other hand, prototyping
might appear to be inexpensive, but the costs could skyrocket if the project is nev-
er completed or requires huge amounts of developer and management time. One
key issue in modern development is to identify these possible risks and threats up
front. Then, each day, managers should evaluate the risks and see if the project has
headed in the wrong direction. It might be impossible to prevent all risks, but at
least if you are alert to the symptoms and recognize the problem earlier, you can
correct it before the costs escalate and kill the project.
Process Analysis
How do you analyze and annotate a process-based system? If
you are examining a transaction-processing system or dealing with a system that
is largely noncomputerized, you should consider creating a process diagram. The
purpose of a process diagram is to describe how the individual processes interact
with each other. It concentrates on the business activities instead of the objects.
A data fow diagram is a process-oriented technique used for investigating in-
formation systems. The method can be used to look at the big picture and see
how a system works in total. It also can be used to examine the details that occur
within each process. Examining organizations at both levels provides a relatively
complete picture of the problems and potential solutions. The use of systems anal-
ysis is illustrated by evaluating a small system for a zoo.
Input, Process, Output
One useful approach to systems is to look at them as a collection of processes
or activities. The most important step in solving problems is to fnd the cause of
the problems. Identifying the major processes in a system will help you under-
SDLC RAD XP JAD Prototyping End User
Control Formal MIS Time Joint User User
Time frame Long Short Short Medium Short Short
Users Many Few Few Few One or two One
MIS Staf Many Few Many Few One or two None
Trans/DSS Trans. Both Both/DSS DSS DSS DSS
Interface Minimal Minimal Good Crucial Crucial Crucial
Documentation,
Training
Good Limited Variable Limited Weak None
Integrity,
Security
Vital Vital Unknown Limited Weak Weak
Reusability Limited Some Maybe Limited Weak None
Figure 12.22
Comparison of methodologies. Each methodology has different strengths and
weaknesses. You need to understand these differences so that you can choose the
right tool for each project. Note that you can combine methodologies on large
projects. For example, you could use prototyping to develop initial forms and reports
that are incorporated into a larger SDLC project.
stand how the system works. Examining input and output objects helps you spot
problems and trace them back to their source. As illustrated in Figure 12.23, sys-
tems receive input, which is processed to produce output. The process could be
mechanical, such as manufacturing using raw materials, workers, and power. Al-
ternatively, it might be a process involving symbolic processing instead of physi-
cal activity. For example, accounting systems receive sales data and process it
into cash-fow statements. In many cases, there are two types of input and output:
physical and data. Physical fows are often accompanied by data. For instance, raw
materials are shipped with an invoice that describes the products and the shipping
information. Systems theory can be used to examine both types of fow. However,
this is an MIS text, so most of the problems presented here deal with fows of data.
Systems are described by collections of these processes. Each system operates
in an environment that is somewhat arbitrarily defned by the boundaries of the
system. For most problems, anything directly controlled by the frm is consid-
ered part of the relevant system. Everything else exists in the environment outside
of the frm. The environment typically includes at least the physical space, laws,
customs, industry, society, and country in which the frm operates. The frm can
infuence the physical environment, laws, and customs, but it does not have direct
control over them.
Consider the example of a zoo: input and output are less concrete because a
zoo primarily produces services instead of products. Figure 12.24 shows the basic
inputs of food, money, and health data for potential new animals. Output objects
include education, educational materials, and baby animals for other zoos. For
most purposes, the system boundary is relatively clear. Visitors, suppliers, and
other zoos are outside the direct control of the zoo, so they are in the environment.
If the zoo was operated by a governmental agency, it would be harder to identify
the boundary. Government systems tend to reach into many different areas, and it
can be hard to identify their exact limits, especially since they can be extended or
contracted by political decisions.
If a system is entirely self-contained and does not respond to changes in the
environment, it is called a closed system. An open system learns by altering itself
as the environment changes. Systems are almost never completely closed because
closed systems cannot survive for long. However, some systems (or companies)
are more responsive to changes in the environment than others.
Most large frms face a certain amount of inertia. It is easier for these frms to
keep operating the way they always have than to continually introduce changes.
But if a frm becomes too static, it can no longer respond to changes in the envi-
ronment. Much like the U.S. railroad companies in the 1960s, closed frms will
lose ground to frms that are more open and responsive to the environment. Re-
Input
Process
Output
Figure 12.23
Each system can be decomposed into three major components: input, process, and
output.
member that a key component of strategy is to search the environment for poten-
tial advantages.
Divide and Conquer
Most problems are too complex and too large to deal with all at once. Even if you
could remember all the details, it would be hard to see how everything was sup-
posed to ft together. A crucial step in analyzing a system is to carefully break it
into smaller pieces or a collection of subsystems. Each subsystem is separate from
the others, but they are connected and interdependent.
Figure 12.25 shows the fve primary subsystems within the zoo. Of course,
there could be many possible subsystems for the zoo. The actual division depends
on how the organization operates. Each subsystem is defned by identifying the
input and output fows. How do you know how to divide a system into smaller
parts? Fortunately, most complex systems have already been subdivided into dif-
ferent departments and tasks. Many companies are organized by business func-
tions: accounting, fnance, human resources, marketing, MIS, and production.
Others are split into divisions according to the type of product.
Once you have determined the major components of the system, each subsys-
tem can be divided into even smaller pieces. An accounting department, for ex-
ample, might be split into management reporting, tax management, quarterly re-
porting, and internal auditing groups. Each of these areas might be split into even
more levels of detail. At each step, the subsystems are defned by what they do
(process), what inputs are used, and what outputs are produced.
There are some drawbacks to the divide-and-conquer approach. It is crucial
that you understand how the components work together. If a project is split into
small parts and given to independent teams, the teams might lose sight of the
Members
Donors
Visitors
Other Zoos
Education
Visitor Counts
Educational
Materials
Baby
Animals
Registration
Papers
Health Data
The
Zoo
Money
Suppliers
Requests
& Comments
Animal Feed
Invoices
Figure 12.24
System boundary at the zoo. As we build systems, we must identify the components
that make up the primary system. There will be many other entities that interact with
the system. However, these entities are beyond our control, so they are outside of the
system.
overall goals. Important components might not be completed, or the individual
pieces might not meet the overall objectives of the system.
Goals and Objectives
Subsystems have goals or purposes. A goal of a manufacturing frm might be to
sell more products than any rival (increasing sales). Or it might be to make as
much money as possible for its owners (increasing revenues). Another goal might
be to fnd an entirely new area in which to sell products (new market segments).
The owners of the system should defne its goals. If the system does not have a
goal, it has no purpose and there is no way to evaluate it. In fact, by defnition,
it would not be a system. When you observe a system, you will need to evaluate
performance, which means you have to identify the goals.
Typical spreadsheets give us the ability to ask what-if? questions. For exam-
ple, you might want to know what happens if you increase sales commissions by
10 percent. Goals help focus the answer by providing the ability to ask Why? and
So what? The answer to the What-if? question involving commissions might
be that revenue increases by 5 percent. But what does that result mean? If we
also know that a goal of the company is to increase profts, we could look more
carefully and fnd that increasing commissions by 10 percent leads to a 3 percent
increase in profts. That result is important because it brings the system closer to
one of its goals. Hence, it would make sense to increase the commissions.
It is clear that to solve business problems, you must frst identify the organi-
zations goals. The catch is that there are often conficting ways to measure the
goals. For instance, improved customer satisfaction or product quality might be
useful goals. But how do we measure them? Managers who measure customer sat-
isfaction by the number of complaints they receive will make different decisions
than those who actively survey customers. In other words, the measurement of our
performance with respect to the goals will depend on the data we collect.
Animal
Care
Donor &
Public
Relations
Manage
Facilities
Produce
Management
Reports
Manage
Human
Resources
Figure 12.25
Primary subsystems of the zoo. The frst step in analyzing a system is to identify
the major subsystems. In most organizations, this step is relatively easy because the
organization will consist of several departments or functions.
Diagramming Systems
We often represent systems graphically to gain insights and spot problems. We
also use diagrams to communicate. Communication is of critical importance in
MIS and all areas of business. Users describe their problems to systems analysts,
who design improvements and describe them to programmers. Ideas and com-
ments can be misinterpreted at any step. We can minimize problems by using a
standard diagramming technique. The data fow diagram approach presented in
this section is commonly used because it focuses on the logical components of the
system and there are few rules to remember, so almost anyone can understand the
diagrams.
Although you could invent your own diagramming technique, a method called
a data fow diagram (DFD) has been developed to represent information sys-
tems. It is designed to show how a system is divided into smaller portions and
to highlight the fow of data between those parts. Because there are only three
graphical elements (fve if you count the dashed control fows separately), it is an
easy technique to learn. The DFD illustrates the systems topics in this chapter.
The basic elements of a DFD are external entities (objects), processes, data
stores (fles), and data fows that connect the other items. Each element is drawn
differently, as shown in Figure 12.26. For example, data fows are shown as ar-
rows. Feedback and control data are usually drawn as dashed lines to show that
they have a special purpose.
Figure 12.27 presents the main level of subsystems for the zoo. Notice that
it contains external entities, processes, and data fows. This level generally does
not show data fles or control fows. They can be incorporated in more detailed
presentations.
External Entity
Data Flow
Feedback and Control Data
External Entity
Process
Data Store (le)
Data Flow
Feedback and Control Data
Figure 12.26
Only four or fve objects are used to create a data fow diagram. External entities
are objects that are independent and outside the system. Processes are functions and
actions applied to data. A data store or fle is a place to hold data. Data fows are
shown as solid lines with arrows to indicate the data movement. Control fows are
marked with dashed lines.
External Entity
When you identify the boundary of a system, you will fnd some components in
the environment that communicate with your system. They are called external en-
tities. Although each situation is different, common examples include customers,
suppliers, and management. External entities are objects, so they are labeled with
nouns.
In the zoo example, the primary entities are management, certifcation agen-
cies, other zoos, and members of the public (visitors, donors, and members). All
relevant external entities need to be displayed on the frst-level diagram.
Process
In a DFD, a process is an activity that involves data. Technically, DFDs are used
to show systems that involve data, not products or other materials. However, in
business today, virtually all physical processes have data-processing counterparts.
When customers buy something, they get a receipt. In a manufacturing process,
the amount of raw materials being put into a machine, measures of the volume of
output, and quality control values are recorded. The DFD process is used to repre-
sent what happens to the data, not what occurs with the raw material.
Because processes represent actions, they are typically labeled with verbs, such
as sell products or create tax reports for management. There are two im-
portant rules involving processes. First, a process cannot invent data. That means
every process must have at least one fow of data entering it. Second, a process
cannot be a black hole; every process must transfer data somewhere else. If you
public/
zoo visitors
donors
zoo booster
members
1
donor and
public relations
2
produce
mgt. reports
3
manage
human
resources
4
animal
care
5
manage
facilities
other zoos
& breeders
certication
agencies
management
money &
requests
PR data
receipts
etc.
public requests
needs & budgets animal requests
health research
maintenance
schedule
maint. &
building
request
specialist
request
employee
schedule
emp. reports
employees
hours,
benets,
etc.
pay data, requests
receipts
Usage
reports
Management
reports
e
m
p
. s
c
h
e
d
u
l
e s
p
e
c
i
a
l
i
s
t
r
e
q
u
e
s
t
animal status
Figure 12.27
The zoo. The primary processes and data fows of the zoo.
look at your DFD and fnd one of these two problems, it usually means that you
missed a connection between the processes. On the other hand, processes that do
not export data might be data stores or external entities.
Data Store
A data store or fle is simply a place to hold data for a length of time. It might be a
fling cabinet, reference book, or computer fle. In a computerized system, data is
likely to be stored in a database management system (DBMS). Chapter 5 provides
more detail on the capabilities and uses of a DBMS. For now, it is important to
note that data is a valuable resource to any company. In drawing a DFD, try to list
exactly what needs to be stored, how long it should be held, and who should be
able to read or change the data.
Data Flow
The data fows represent the inputs and outputs of each process or subsystem. The
data fows are easy to draw. They are simply arrows that connect processes, enti-
ties, and data stores. Be sure to label every data fow. The diagram might seem
obvious now; however, if someone else reads it or you put it away for several
months, it can be hard to fgure out what each fow represents.
public/
zoo visitors
donors
zoo booster
members
1.2
handle
donor
requests
1.1
produce PR
& outreach
programs
1.3
org. booster
services &
meetings
1.4
track needs
and donor
programs
1.5
produce
accounting
& reports
adopt an animal
les
receipts
money &
comments
PR data
receipts
etc.
money &
requests
newsletter
notices, etc.
money &
suggestions
booster
requests
needs &
plans
public requests
needs & budgets
needs &
plans
donor
requests
donor
lists
expenses & budget
money, data,
visitor statistics
accounting
reports
public requests
animal
needs
Figure 12.28
Each process can be expanded into more detail. This diagram shows the interactions
with various members of the public. Note that data fows from the higher level must
appear on this level.
Division of the System
A DFD provides an excellent way to represent a system divided into smaller com-
ponents. First, each task is shown as a separate process. The data fows between
the processes represent the inputs and outputs of each subsystem. Second, the
DFD for a complex system would be too large to ft on one page. Hence, the DFD
is displayed on different pages or levels. The top level, or context diagram, acts
as a title page and displays the boundaries of the system and the external enti-
ties that interact with the system. The next level (level zero) shows the primary
subsystems. Figure 12.26 is an example of a level zero diagram. Each of these
processes is then exploded into another level that shows more detail. Figure 12.28
is the exploded detail for the frst process (donor and public relations). These ex-
plosions can continue to any depth until you have displayed all the detailed opera-
tions needed to explain the system.
Processes Description . . .
Animal care Feed, clean, and vet care
Donor & public relations Handlepublicrequestsandprovideeducationalinformation
Employee relations Schedule employees, process benefts, handle government reports
Facility management Handle maintenance, new construction, planning
Producemgt.reports Collect data and produce summary reports for management
Entities
Certifcation agencies Government and private agencies that create rules and regulate
zoos
Donors People and companies who donate money to the zoo
Employees Primary (paid) workers, full-time and part-time
Other zoos and breeders Zoos we trade with and share data with
Public/zoo visitors Daily visits, we rarely keep data on individuals
Zoo booster members Members who donate money and time for minor benefts
Data
Accounting reports Standard (GAAS) accounting reports for management
Certifcation reports Reportsforcertifcationagencies;producedannually
Facility reports Summaries of work done and plans, mostly weekly
Needs and budgets Budgetsandspecialrequestsfromanimalcare
Publicrequests Suggestions and comments from the public
Figure 12.29
A few sample entries from the zoos data dictionary. A data dictionary records details
on all of the organizations objects. It is typically organized by type of object. It is
easiest to maintain if it is stored in a computer database.
Data Dictionary
In any project, you need to remember additional pieces of information about each
object. You might want to keep a sample report for a management tax report data
fow, along with any deadlines that must be met. For data stores, you need to re-
cord information such as who controls it, who needs access to the data, how often
it should be backed up, and what elements it contains.
A data dictionary, or repository, contains all of the information that explains
the terms you used to describe your system. A good computer-aided software en-
gineering (CASE) tool will maintain the dictionary automatically and help you
enter longer descriptions for each item. Without these tools, you will have to keep
a notebook that contains the full descriptions. For convenience, the entries should
be sorted alphabetically. A word processor can be used to hold and print the dic-
tionary. Figure 12.29 shows sample entries for the zoo system.
Summary: How Do You Create a DFD?
The frst step in creating a DFD is to identify the environment and boundaries
of the system by asking the following questions: What problems do you need to
solve? What areas do you want to avoid? What are the goals? What are the main
external entities? The second step consists of identifying the primary processes
that defne the system. Keep the list short (fewer than 10). Then answer these
questions: What are the main activities in the system? What are the inputs and out-
puts of each process? How are these processes interconnected by the data fows?
The third step is to look at each process in more detail and draw an expanded sub-
system on a new page. What activities take place within a given process? What de-
tail is needed in the reports and data inputs? The fourth step is to build the control
fows. What processes are used to monitor progress toward the goals? What addi-
tional data is collected to monitor the environment and the systems performance?
The key to analyzing systems is to start small. You can begin with one detailed
subsystem and build your way up, or you can describe the general system pro-
cesses and work down by adding increasing levels of detail.
Object-Oriented Design
How is object-oriented design different from process design?
One way to begin your analysis of a business is to focus on the business objects:
what they are and what they do. Objects could be anything from people to raw
materials to data fles or schedules. The key to object-oriented design is to focus
on defning what an object is and what it can do. A class is a generic description
of a set of objects. This distinction is not crucial in this book, but you might want
to know there is a difference. For example, the Bicycle class describes any bicycle
that could be built by the company. A specifc bicycle (e.g., serial number 15) is
an object.
The concept of object-oriented design has received considerable attention dur-
ing the past few years. In some ways, the base design techniques are not much
different from traditional SDLC techniques. In other ways, object orientation re-
quires a completely new way of thinking about systems development. The ulti-
mate goal of the object-oriented approach is to build a set of reusable objects and
procedures. The idea is that eventually, it should be possible to create new systems
or modify old ones simply by plugging in a new module or modifying an existing
object.
One key difference between object orientation and other development meth-
ods is the way processes or functions are handled. With objects, all functions are
embedded in the defnition of the objectthe object comes frst. The object ap-
proach reverses the treatment of processes and data. With SDLC, illustrated by
a data fow diagram, the emphasis is on processes, and data (attributes) is passed
between processes.
One goal of an object-oriented approach is to create a set of information system
building blocks. These objects and procedures could be purchased from commer-
cial software companies (such as a spreadsheet from Microsoft or a database sys-
tem from Oracle). MIS programmers or consultants can create additional objects
tailored for your specifc company or department. Once the basic blocks are in
place, end users or MIS analysts can select the individual pieces to create a com-
plete system. Hence, as Figure 12.30 indicates, less time is needed for implemen-
tation, as long as the analysis and design are performed carefully. On the other
hand, the up-front costs of designing and building these objects can be quite high.
Furthermore, the tools and techniques tend to require substantial retraining of the
existing MIS staff. Both of these types of costs have caused some companies to
avoid object-oriented methods.
Distributed Services
Can software be located in multiple places? A major question in
decentralization is where the software needs to be located. Substantial benefts
arise from centralizing dataproviding access across an intranet or the Internet.
But users need some type of distributed hardware to access the data. Does that
mean all of the software has to be installed on each machine? Some new technolo-
SDLC
Object Oriented
60
50
40
30
20
10
Requirements Analysis Implementation Design
Figure 12.30
SDLC versus object oriented. Initial design of an object-oriented approach takes
more effort than with an SDLC approach. However, once the objects are properly
defned, it is much easier to create and implement a new system.
gies are being developed that provide support for distributing software functions
across the network. While basic tools such as a word processor are still needed on
individual machines, the complex business and analytical tools can be installed on
central servers.
One of the primary technologies is the simple object access protocol (SOAP).
It is a standard being pushed by several vendors to defne how objects can be used
across the Internet. It relies heavily on the extensible markup language (XML)
to transfer data between diverse computers. As a general manager, you do not
need to know the details of how these two technologies work, but you should re-
member their purpose. Ultimately, you will want to select applications that fully
support these standards so that you can build and use systems that work transpar-
ently across the Internet.
The purpose of SOAP is to enable frms to build application services that can
be used by other organizations across the Internet. For instance, as shown in Fig-
ure 12.31, a bank (e.g., www.oanda.com) might offer a currency conversion ap-
plication. Your companys accounting application could call the banks program
whenever it needed to convert money to a different currency.
Applications that use the SOAP and XML protocols can interact with other
services across the Internet. However, a big question that remains to be resolved is
how frms will price their services. Firms that create service objects will ultimate-
ly be able to bill clients a usage fee or a monthly charge with unlimited access.
But a standardized billing mechanism has not been implemented yet.
Cloud Computing
How does cloud computing change software development? In
one respect, cloud computing has made computing easier for companies. Essen-
Bank Server
Your Company Server
Manager or Client
The Internet
Your Financial Application
Year Revenue Cost
2000 34,231 27,333
2001 39,983 32,132
2002 36,421 31,456
Sales and Costs
-
10,000
20,000
30,000
40,000
50,000
2000 2001 2002
Revenue
Cost
Bank Service
Currency
converter:
Date,
InCurrency,
Outcurrency
SOAP
Figure 12.31
Simple object access protocol. SOAP enables frms to offer application object
services that other frms can use across the Internet. In this example, your application
can call the banks currency converter object to get the correct exchange rates.
tially, cloud computing offers software as a service (SaaS), where a few frms
hire developers and create and host software and other frms simply lease the
service. That means that only a few frms develop software. This trend has been
evolving for several yearspartly because new technologies make it feasible, but
partly because most companies do not want to spend money and time trying to
create custom software and pay large groups of programmers. The economics of
software development also explain cloud computing because the high develop-
ment costs are spread across many other companies. Still, not all companies are
moving to publicly-run cloud computing. Issues of privacy and customization
have held back some of the acceptance. Instead, smaller frms opt to buy software
and install it on their central computersand larger companies might run their
own private clouds.
The second issue of cloud computing consists of developing software that
runs on the cloud. First, most of the software relies on Internet technologies and
databases. Initial technologies (Web browsers) provided limited support for in-
teraction and usability. But, the Internet technologies continue to evolve, which
provides more capabilities but requires learning and testing new tools and new
designs. For instance, HTML 5 was introduced by Web browsers in 2011, with
advanced interaction features, but the standard and full support are not expected
for at least three years. Consequently, developing software for cloud computing
remains somewhat experimental. Any experimental technology is more diffcult to
predict, and it is harder to estimate development time and cost. Also, it is harder
to develop software that is designed to be used by multiple companies. Additional
security elements need to be designed and tested, and the system has to be built to
be scaled up as more companies use the products. Scalability and expansion are
key elements in designing any cloud-based system. Not just the hardware, but the
software and the administration tools have to be designed so they can be managed
without signifcant increases in the number of administrators. Effectively, most
cloud-based software carries additional development costs because administration
tools need to be built into the system.
Summary
Systems development can be a diffcult task. Many projects have failed because
they cost much more than anticipated or they did not produce useful systems.
Large projects are especially diffcult to control because there can be conficting
goals, it is hard to ensure that subsystems work together, business needs change
during the development process, and there is turnover among the MIS employees.
The systems development life cycle evolved as a means to deal with the complex-
ity of large systems and provide the necessary controls to keep projects on track.
Systems analysis techniques are used to break projects into manageable pieces.
Various graphing tools, such as data fow diagrams, are used to display the rela-
tionships between the components. Systems design techniques use the results of
the analysis to create the new system. The new system consists of interconnected
modules. Each module has inputs, outputs, processing steps, database require-
ments, manual procedures, and controls. At various stages in the design process,
managers and users are asked to sign off on the proposed system, indicating that
they will accept it with no further changes.
In contrast to the rigid control embodied in the SDLC method, the prototyp-
ing approach is iterative and creates an early working model of the system. Users
and managers can see the proposed input screens and reports and make changes
to them. As the project develops, the prototype goes from a simple mockup to a
working system. Prototyping is sometimes used in conjunction with SDLC during
the design phase to lay out input screens and reports.
A third way to build systems is for end users to develop their own projects us-
ing fourth-generation tools such as database management systems, spreadsheets,
and other commercial software. As the capabilities of commercial software tools
increase, users can develop more complex systems. The backlog facing MIS also
encourages users to develop their own systems. The potential dangers of user de-
velopment, such as lack of testing, incompatibilities, and unnecessary duplication,
can be controlled by having MIS teams available to assist end users.
All methods of developing systems involve fve basic steps: feasibility and
planning, systems analysis, design, implementation, and maintenance. Prototyp-
ing and end-user development typically focus on the design stage. However, man-
agers need to remember that implementation problems can arise with any new
system, regardless of how it was created. Similarly, there will always be a need
to maintain and modify existing applications. It is easy to forget these steps when
users develop their own software.
A Managers View
As a manager in a large company, you will work closely with the MIS de-
partment to modify and build systems that support your operations. You need
to be aware of the problems facing MIS staff to understand the reasons for
their rules and methods. Managers are increasingly being asked to develop
their own systems and to participate more heavily in the design of new re-
ports and forms. The details of analysis, design, testing, and implementation
will be useful regardless of the method used. As a manager, you also need
to know the advantages and drawbacks of various development methods;
you will often have to choose the method that is best suited to solving your
problems.
Key Words
Web Site References
Computer Industry News
ACM Digital Library portal.acm.org/dl.cfm
CNet News news.com.com
Computer Economics www.computereconomics.com
Computerworld www.computerworld.com
eWeek www.eweek.com
Federal Computer Weekly www.fcw.com
Gartner Group www.gartner.com
Government Computer News www.gcn.com
IDG www.idg.com
IEEE www.ieee.org
Infoworld www.infoworld.com
Internet.com www.internet.com
PC World www.pcworld.com
Te Industry Standard www.thestandard.com
Review Questions
1. What fundamental methods are available to build information systems?
2. What are the main elements of programming logic?
3. How does object inheritance simplify programming?
4. Explain why the frst step in most business projects should never be
programming.
5. What is the primary purpose of the systems development life cycle
development methodology?
best practices
capability maturity model in-
tegration (CMMI)
change agents
commercial off-the-shelf
software (COTS)
data dictionary
data fow diagram (DFD)
end-user development
event-driven
extensible markup language (XML)
extreme programming (XP)
fault tolerance
feasibility study
inheritance
joint application design (JAD)
object hierarchy
object-oriented programming (OOP)
open source development
outsourcing
phased implementation
program logic
prototyping
rapid application development (RAD)
reusability
scope creep
simple object access protocol (SOAP)
software as a service (SaaS)
software maintenance
syntax
systems analysis
systems analyst
systems development
life cycle (SDLC
6. What are the main steps in the systems development life cycle methodology?
7. What drawbacks are created with the systems development life cycle
methodology?
8. What alternative methods are being used to develop information systems?
9. How does agile or rapid application development speed up the development
process?
10. What is the role of a data fow diagram in analyzing systems?
11. What are the main components of a data fow diagram?
Exercises
1. Interview a local manager to determine the requirements for a new system.
Explain which method would be the best approach to develop the system.
Estimate how long it would take to complete the project and how much it
would cost. Advanced option: Illustrate the new system with a data fow or
object-oriented diagram. More advanced: Create the system.
2. Create a Web page form with 5 text boxes and add Javascript code to check
that each box is not empty when the form is submitted.
3. Find a small example of a business or Web program, choose a function or
subroutine and explain its purpose.
4. A regional bank offce generates loans for builders. The offce has several
bankers who form alliances with regional builders and negotiate loans and
other services. The manager wants a system to track the leads, including the
potential amount of the loan and the probability of the loan going through.
Every month, the main offce sends a spreadsheet fle with current loan
information. The manager and the main offce want the regional bank offcers
to project the amount of money that will be loaned in the coming months.
The manager wants a system to help collect and track this data. Identify the
best development methodology. Assuming no one in the regional offce has
the skills to create the application, do some research to fnd at least two frms
that could handle the job for a reasonable price.
5. A large lawn-maintenance company wants a new system that uses smart
phone applications to collect data for its operations. The company has ten
trucks and 20-30 people who mow lawns and other yard work for businesses
and homes in a large city. To help reduce travel time and gas costs, the
company wants to use GPS to track the daily routes of the 30 trucks and then
later optimize the routes. It also wants to use this information to contact the
nearest truck when a customer has a special request, or to direct the nearest
truck to a new client who is requesting a quote. The phone application also
needs a screen for employees to enter the time they spend on each project,
and other screens to handle purchases of items such as gas, fertilizer, and
tools. Maintenance for each of the mowers and other power tools should also
be handled on the smart phone. Ideally, bar codes will be placed on each
piece of equipment and the phones camera can be used to scan the code to
identify the equipment. The system also needs to support communication
between the workers and the clients. For instance, clients should be able
to enter special requests, and workers can enter problems, such as broken
sprinklers. Create a design for this application.
6. For each of the following information system projects, identify the
development method that would be the best approach for most companies.
a. An electronic system to handle travel requests and data entry for
reimbursements.
b. A smart phone app that is a cross between a map and a social network,
where people enter locations of products they fnd in stores so other people
can follow the map. (Where are the pickles in a Safeway store?)
c. An application to use traffc cameras to track the location of specifc cars
using vision-recognition tools to follow license plates.
d. A program to track health and feeding data at a zoo (for animals not
visitors).
e. An application for a couple of people in marketing to track viewings of
radio and TV ads.
f. Modifying an ERP system by adding forms, rules, and reports to track
expenses related to a new product design team created from an acquisition.
7. Assume that you are on a project to build a new Web site for a midsize
company. The frm sells materials to home buildersusually contractors,
but some individual sales as well. The company wants to take orders
over the Web and enable customers to track the status of current orders.
Contractors also want the ability to look at old orders when placing new
ones. For example, if they build the same style of house twice, they will
need approximately the same materials. The manufacturer is not completely
certain on many of the details yet, and you will have to interview customers
to get additional details and feedback. Set up a schedule for developing this
system using the SDLC approach. Then, identify ways that RAD might be
used to reduce the overall development time.
8. You work for a company that is increasingly asking employees to develop
their own applications using Microsoft Offce tools connected to the
corporate database. This process has not been working very well, and
employees are grumbling. But the company has decided it cannot afford to
hire all of the MIS people that would be needed to develop all applications
and reports within the MIS department. How can the company improve the
process? What tools and capabilities should the company add?
9. Check out www.sourceforge.net. Briefy explain its purpose. Find two
programs that might be useful to businesses.
10. After several decades of challenges building systems, improving tools, and
new development methods, why are many IT projects still over budget and
late?
Technology Toolbox
11. Write a short macro program in Excel that adds all of the numbers between
the values in cell A1 and cell A2 and puts the result in cell A5. For example,
if A1 = 1 and A2 = 5, then add 1 + 2 + 3 + 4 + 5 to get 15. Hint: You can read
or write to a cell with the command Range(A1).
12. Write an Excel macro that looks at each item selected to see if any cells
are blank. If any are blank, display a message notifying the user how many
blank cells there are. (Hint: Use the IsEmpty function to test and the MsgBox
command to display a message.)
13. Create a form in Microsoft Access (or Visual Studio). Place text boxes on
the form for amount to borrow, interest rate, and number of months. Add a
fourth text box to hold the resulting payment amount. Set the properties to
format and name each of the boxes. Add a button to calculate and display the
payment amount based on the entered data. Use the Pmt function to do the
calculation.
14. Use InfoPath to create and publish the expense report form. If possible, save
it to a SharePoint server. Enter sample data and save and submit the form.
E-mail it to your instructor as your supervisor.
15. Talk with a manager or employee to identify internal forms and data that are
collected. Make a list of at least 5 forms for a business that could beneft by
using InfoPath.
16. Find a tool on the Web that could be used to create forms similar to InfoPath
(but probably simpler), but runs on a Web browser and stores all the data on a
Web site.
Teamwork
17. Interview computer users to determine how they feel about their current
system. Do they like it? What are the major advantages and drawbacks? How
long have they used it? When was it changed last? Are there changes users
want to see? Are they willing to accept changes? How are relations with
the MIS workers? Who initiates changes, users or MIS? If users proposed
a new project, how long would it take for MIS to get to it (how long is the
backlog)? Each team member should interview a different person (some
users, some in MIS). Combine your results to get a picture of the entire
company. Do users agree with each other? Does the MIS department agree
with the users? Do they see the same problems? (Hint: If you do not have
access to another company, you can always fnd computer users in the
university.)
18. Choose one person in the team who has an interesting job. Create a data fow
diagram for the job and organization. Be sure to label everything and provide
a data dictionary.
19. Find a manager who needs a computer project completed or updated.
Interview the person and record the comments and notes. Create a basic
design for the system. Identify the best methodology to create the system. If
possible, fnd someone to build the system.
20. Rolling Thunder Bicycles wants to create a new Web site to enable customers
to build and order bicycles online. It should also enable customers to
view the progress of the construction and make payments online. Assign a
different development methodology to each person who will perform a basic
design and argue why that method should be used for the project. Select one
of the methodologies in the end.
21. Examine the sample pseudocode used to compute the total of a set of
numbers. Choose at least two programming languages and assign team
members to a specifc language and fnd or write the code needed to compute
the total in that language. Submit and comment on the differences in the
languages. Hint: The code does not need to run and you can skimp on
the fle and print statements.
22. Create a simple form using InfoPath that collects at least three pieces of data
such as name and e-mail address. Test the form by sending it to each person
on the team. Collect the data in a spreadsheet or simple database table.
23. As an exercise in creativity, each person should write down a business or
personal task that they would like to see computerized or available on smart
phones. Consolidate the list and search the Web to see what tools already
exist to handle the task.
24. Rolling Thunder bicycles needs a new Web site to sell its custom bicycles.
How should it be developed? What methodology could be used?
25. Using the help system and Web site description of Rolling Thunder, create
a data fow diagram to show the main processes directly involved with the
customers (taking orders, sending notices and bills, and receiving payments).
26. Rolling Thunder Bicycles needs a new system to generate and track
electronic orders (EDI) to its suppliers. What methodology should be used to
develop the system?
27. Assume that the managers of Rolling Thunder bicycles have decided to
purchase and implement an enterprise resource planning system. You have
been selected to help determine which system the company should purchase.
Outline the steps you will have to perform to select a vendor.
28. Identify at least three information processing tasks at Rolling Thunder
Bicycles that could be handled with InfoPath. Compare the benefts and
drawbacks of using InfoPath versus adding new forms to the existing
application.
Additional Reading
Boehm, B.W., Software Risk Management Principles and Practices, IEEE
Software, 8(1), 32-41. [Top 10 risk items in development.]
Brooks, Frederick P. The Mythical Man-Month, Anniversary Edition, 1995,
Boston: Addison-Wesley. [Classic book on software development problems
and why adding people to a project often slows it down.]
DeMarco, T. and T. Lister, Peopleware, New York: Dorset House, 1987. [Hints
and problems developing useful systems.]
Jones, T.C., Programming Productivity, New York: McGraw Hill, 1986.
[Evaluating and measuring productivity, costs, and teamwork.]
Jeffries, Ronald, Ann Anderson, Chet Hendrickson, Extreme Programming
Installed, 2001, Boston: Addison-Wesley. [A detailed explanation of how to
use extreme programming. Also look at other books in the series for theory.]
Naumann, Justus and Milton Jenkins, Prototyping: The New Paradigm for
Systems Development, MIS Quarterly, Spring 1982. [Description, uses and
advantages of prototyping.]
McConnell, Steve. Rapid Development: Taming Wild Software Schedules,
Redmond, Microsoft Press, 1996.
Paulk, Mark C., Bill Curtis, Mary Beth Chrissis, and Charles V. Weber,
Capability Maturity Model for Software, 1993, Carnegie-Mellon University,
Software Engineering Institute. http://www.sei.cmu.edu/pub/documents/93.
reports/pdf/tr24.93.pdf [Description of how to evaluate software development
programs in terms of their capabilities on a relatively standardized maturity
scaleessentially an evaluation of how well an organization follows the
SDLC.]
Wallnau, Kurt, Scott Hissam, and Robert C. Seacord, Building Systems from
Commercial Components, 2002, Boston: Addison-Wesley. [Description
and cases of developing systems using commercial off-the-shelf software
components, from the Carnegie Mellon Software Engineering Institute.]
Cases: Government Agencies
Most U.S. citizens know the overall structure of the federal government: the presi-
dent, Congress, and the Supreme Court. These groups are responsible for creating
and interpreting the laws to govern the nation. What many people do not realize
is that both the president and Congress are supported by a huge set of government
agencies. These organizations form a bureaucracy that is ultimately responsible
for carrying out the laws. Governmental agencies have several unique problems.
The most important one is that funding is subject to changes in the political cli-
mate. With each election, an agency runs the risk of having to change direction,
cancel projects, or provide support for new tasks.
On the other hand, most government agencies are not subject to economic
pressures. Consequently, they have not been faced with the same incentives to
economize and minimize costs that have faced businesses. Another critical feature
of most government agencies is that they tend to serve large numbers of people,
especially at the federal level. These large organizations collect huge amounts of
data. Increasingly, these agencies are converting to electronic storage and access.
Publicly available data can often be found on Web sitesat least at the federal
level.
Most governmental agencies have dealt with the size issue by maintaining large
staffs, and combining decentralized management with centralized controls. Tra-
ditionally, government organizations have paid lower salaries than commercial
businesses. Although the salaries are supplemented with benefts and job security,
governmental agencies often face high turnover rates and changes in personnel.
To compensate for these problems, the agencies rely heavily on procedures. These
rules seek to predict and then direct what to do in circumstances that may arise.
As new situations and decisions present themselves, new rules are created. Given
these challenges, there is no surprise that most people perceive government agen-
cies as large bureaucracies, flled with endless forms and strange rules.
0
2
4
6
8
10
12
14
16
18
20
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
B
i
l
l
i
o
n

$
Annual Budget
NASA
FAA
IRS
There are many obvious uses for computers in government agencies. During a
few minutes of observation, anyone can generate ideas that could improve agency
performance, making life easier for the workers and citizens. However, the real
challenges have always come in creating and implementing these ideas. In the
2000s, the number of students graduating with degrees in computer science and
information systems plummeted. As demand for technical skills began to increase
in 2010, many companies found it diffcult to fnd qualifed new employees. This
problem was even worse for government agenciesmany of the top-notch, in-
novative programmers do not want to work for government agencies. Plus, many
of the graduates of tech programs are not U.S. citizens, so they cannot get federal
jobs (Piemonte 2011).
Although many success stories exist regarding computer implementation with-
in government agencies, there are also some costly failures. The Federal Aviation
Administration and the Internal Revenue Service cases present some of the dif-
fculties that have arisen.
Be careful when you read these cases. Do not simply blame the problems on
typical government mismanagement. Many of these problems also exist within
businesses. Always remember that the challenge is to search for and implement
answers and methods that will overcome the obstacles and complications.
Size and Growth
The federal government employs 2 percent of the U.S. workforce, with about 2
million civilian employees in 2010, not counting the Postal Service. In 2003, it
spent $757 billion, rising to 808 billion in 2006 and to $1.26 trillion in discretion-
ary spending in 2010. With the tax cut of the early 2000s, federal receipts declined
in 2002 and 2003, while expenditures continued to increase about 6 percent a year.
The federal debt rose to 2.7 percent of GDP in 2003. Part of the economic balance
is due to the war in Iraq; part is due to the attempts to prevent a major recession.
In 2010, total federal government expenditures were $3.456 trillion, with receipts
of $2.163 trillion, leaving an annual defcit of $1.293 trillion. To put it in context,
expenditures amount to almost 24 percent of gross domestic product (GDP) for
2010. And budget projections for future years showed increasing expenses and
defcits (2012 Federal Budget).
In some ways, the size of the federal government is shrinking. The number of
elected offcials remains constant but the rest of the federal government is down-
sizing. Federal government employment peaked in 1990 with a total number of
3.233 million employees. This number decreased through 2007 and increased in
2008, 2009, and probably 2010. In 2009, the total was 2,804 million. On the other
hand, several widely circulated reports from most federal agencies indicate that by
2008, over 50 percent of the federal workforce will be eligible for retirement. Not
all of them will actually retire immediately, but over the course of a few years, a
substantial percentage of the federal workforce will need to be replaced. A consid-
erable amount of internal knowledge could potentially be lost in the process. The
Census Bureau reports that in 2007 excluding Postal Service employees, the Fed-
eral Government employed 1.845 million full-time civilians. In 2008, that number
was 1.885 million and it dropped to 1.789 million in 2009.
In comparison, for the United Kingdom in 2003, public spending amounted
to 41.1 percent of its GDP. Defcits in European nations routinely run about 5
percent of GDP (The Economist 2004). However, remember that the UK pays for
health care. In the United States, if you include state and local government spend-
ing, total government spending was over $2 trillion in 2003 or 18.7 percent of
GDP. If you include current healthcare spending, the total would be about 35 per-
cent of U.S. GDP.
Like any business, government agencies increasingly rely on technology to im-
prove productivity. Most agencies are under constant pressure to reduce costs
often to the point of having their funding cut. In large agencies, such as the IRS
and the FAA, monster technology projects are funded separately. Consequently,
the agencies have often been able to obtain funds specifcally set aside to acquire
or build new systems.
Most tasks performed by government agencies are unique. As a result, they re-
quire custom-developed software. Since the projects are huge and involve a large
number of users, they are diffcult to develop. Throw in a few bureaucratic turf
battles, and it is amazing that anything gets done. Unfortunately, the result has
been that many agencies are operating with technology that is 10, 20, or more
years out of date. These archaic systems create their own ongoing problems. The
government cannot just stop what it is doing, throw the old systems away, and
build new ones. Instead, most of the government IT workers keep the old systems
running. So, who is going to build the new systems? And how do you obtain the
detailed knowledge from the users and workers of the old systems? Then, how do
you transfer all of the data and keep both systems updated and running while test-
ing the new system? However, we know that information technology constantly
changes. Somehow, agencies have to balance the costs against the capabilities of
the new systems.
In December 2010, Vivek Kundra, the U.S. CIO established a policy called
cloud frst which told federal agencies to move at least three services to cloud
computing within 18 months. The goal was to decrease costs and improve fex-
ibility. Some agencies found it relatively easy to comply; others were focused
on evaluating operating costs, security, and bandwidth needs. For example, the
agency responsible for the American Recovery and Reinvestment Act of 2009
used Amazons EC2 to run the Recovery.gov Web site for sharing data related
to the Act and saved about $750,000 in the frst year [Pratt 2011]. Likewise, the
U.S. Treasury Department moved the Treasury.gov site to Amazon cloud services.
For more secure operations, the Defense Information Systems Agency (DISA)
and NASA have created their own cloud servers (RACE and Nebula) that can be
used to host operations from other agencies. The General Services Administration
(GSA) is defning standard contracts and schedules to make it easier for agencies
to purchase cloud services from commercial vendors. In the meantime, any classi-
fed data is unlikely to be moved to commercial clouds.
Oversight
Congress is charged with appropriating all federal money and controlling spend-
ing. Yet members face an enormous bureaucracy, plus constant reelection wor-
ries to distract them. Consequently, the General Accounting Offce (GAO) was
created to help monitor the spending and procedures at the various agencies. As
a nonpartisan offce, the GAO is free to collect whatever data it wants from the
agencies and yell at them as necessary. You can obtain GAO reports directly from
www.gao.gov. The reports often contain detailed information on specifc projects
and audits. The agencies generally provide responses to GAO concerns within the
report. These responses are always interesting because the agency director tiptoes
a line. No one wants to accept all of the GAO criticisms (and look like there is no
control), yet no one wants to totally disprove the GAO, because that would mean
no more funding is needed.
Additional Reading
Bureau of Economic Analysis, http://www.bea.gov/
The Economist, Has Tony Wasted Your Money? July 10, 2004, p. 12.
Piemonte, Phil, Many Grads, But Not the Right Stuff, Government Computer
News, June 21, 2011.
Pratt, Mary K., Feds Race to the Cloud, Computerworld, July 13, 2011.
Statistical Abstract of the United States, http://www.census.gov/compendia/
statab/
Case: Federal Aviation Administration (FAA)
The FAA is charged with overseeing all public (nonmilitary) fight operations in
the United States related to safety and access to the air. They establish safety crite-
ria, issue licenses for pilots, and create air worthiness certifcates for planes. They
also operate the air traffc control system throughout the United States. Funding
for the agency is generated through user fees and taxes on aircraft fuel, tires, and
airline tickets. The FAA is an executive agency that operates under the budget-
ary control of the president. Appropriations and organizational structure for the
agency are approved by the Congress.
The increase in air traffc in the United States has made air traffc control a
complex issue. In 1990, 466 million passengers a year were fying on U.S. air-
lines. In 2002, the airlines carried 714 million paying passengers. With more air-
lines and more daily fights, the air traffc control system is dealing with more dif-
fcult problems every year. The busiest airports (Atlanta tops the list) cause even
more complicationstrying to schedule hundreds of fights per hour.
Traffc control is organized into three levels: nationwide U.S. airspace, 20 re-
gional air traffc centers, and individual airports. Air traffc control operators at
each airport have immediate control over takeoffs and landings. Regional opera-
tors watch traffc within their defned airspace. They hand off planes as they fy
across the country into the next airspace. Systemwide control is provided by the
Central Flow facilities located in Washington, D.C. The Central Flow managers
examine traffc across the entire United States and resolve conficts and problems
that arise among regions. The 40 traffc management specialists plan each day in
advance, devising alternative routings for aircraft that may be needed because of
problems arising from snowstorms, accidents, and closed runways.
Early Systems and Ongoing Problems
The early traffc control system was built with hardware and software from Sper-
ry-Rand/Univac, a computer company that was purchased in the mid-1980s by
Burroughs, and now named Unisys. The airport-based traffc control computers
were based on 256K bytes of main memory and performed 500,000 instructions
per second. The original systems were installed in the early 1960s. The 20 region-
al centers had their own computersIBM 9020 machines that were custom made
for the FAA in the 1960s.
Air traffc controllers have been reporting problems with existing systems for
years:
In 1992, West Coast air traffc was delayed for several hours. An IBM 3083
at the regional station crashed. In the process, it removed the identifcation
labels from the radar screens of controllers from Oregon to Los Angeles. The
controllers switched to an older backup system but had to increase plane sepa-
ration from the typical 3 miles to up to 20 miles. Pilots and controllers used
radio communication and manually fled fight plans to compensate for this
loss. Ron Wilson, a spokesman for the San Francisco airport, noted that al-
though there were frequent disruptions, the FAA computer failures generally
dont last long, just long enough to screw things up.
In Oakland, California, the controller screens fail an average of three times a
month. When this happens, the controllers have only a few seconds to memo-
rize the position, speed, course, altitude, and destination of the 12 planes they
are typically directing. Then their screens go blank for at least 10 seconds.
Sometimes when the screens come back online, they are missing critical data.
Joel Willemssen, assistant director of the U.S. GAOs Information Manage-
ment and Technology Division, reported that 70 percent of the 63 largest air-
ports in the United States have experienced problems with blank or fickering
computer screens. John Mazor, a spokesman for the Airline Pilots Associa-
tion, notes the problems cause delays, diversions, andin the worst possible
casesaccidents. Its not as dangerous as you might think, but its not some-
thing you want to have happen to you.
The Los Angeles basin region consists of 21 airports handling 6.5 million
fights a year. The GAO notes that the FAA computers in the region have
repeatedly suffered from the loss of critical data and slow responses because
of the overload.
In 2007, the FAAs National Airspace Data Interchange Network (NADIN)
fight-planning system in Atlanta shut down for several hours, causing fight
delays and cancellations across the East Coast. Investigators were unable to
fnd the cause of the problem, but NADIN was scheduled for replacement in
2008 (Weiss 2007).
Improvements
In 1981, the FAA was given approval to upgrade to a comprehensively new com-
puter system. New airports, such as Dallas-Fort Worth, and the deregulation of the
airline industry in 1978 led to huge increases in air traffc. The $12 billion plan
called for replacement of 12 major systems over the course of 12 years. An addi-
tional 80 smaller projects were included in the plan.
By 1990, only 1 of the 12 systems had been replaced and the project was $15
billion over the original budget. The one project that was completed was known
as Host, because it called for replacement of the mainframe computers at the 20
regional control centers. IBM installed its 3083 mainframes on schedule but was
$16 million over budget. The 3083s were technologically obsolete at the time they
were installed because the newer IBM 3090-class machines had already replaced
them over a year before.
The FAA has been criticized for a lack of oversight and control in developing
new systems. In 1980, the Senate Appropriations committee noted that the FAA
has no ongoing, welldefned, and systematic management approach to evaluating
software and operational cost, capacity, and performance of the current system to
meet projected shortrange workloads. The General Accounting Offce (GAO),
the watchdog of Congress, echoed that sentiment several times later.
Advanced Automation System
One of the more visible components of the plan to refurbish the system is the Ad-
vanced Automation System (AAS). It was designed to provide updated tracking
displays for the controllers. It was supposed to be completed by 1990, but at that
time was delayed until 1993. The system was designed to use IBM RS-6000 com-
puters to display fight information, schedules, and current location along with
weather fronts. The color systems were to have higher resolution, be easier to
read, and carry more information.
In 1994, an internal study of the AAS showed that the project was still two
years behind schedule and probably would fall back another two years before
completion. Up until that time, the project had cost $2.3 billion. It was estimated
to eventually cost about $7 billion. David Hinson, FAA administrator, announced
that he was replacing top managers on the project, dropping portions of uncom-
pleted work, and demanding performance guarantees from the contractors. The
Area Control Computer Complex was canceled at this time. It was designed to
interconnect the host computers at the airport with those at the regional levels.
Global Positioning System (GPS)
GPS is a satellite-based navigation system that was developed by the Pentagon
and previously available for use only in connection with military air travel. GPS
allows pilots to navigate based on satellite signals instead of radar signals. It al-
lows real-time fight planning for pilots. As more satellite technology becomes
available, the integration of air traffc as well as weather information and other
data communication will become a necessary technological step. Four-dimen-
sional GPS readingslongitude, latitude, altitude, and timeenable an aircraft to
come within a few feet of any given target. Encryption technology is currently in
place to protect security in the transmission of the satellite messages.
In 2004, the FAA began testing GPS tracking for air traffc control in Alaska.
Because of the vast rugged terrain, it would be impossible to put radar stations
across Alaska. Moreover, the onboard GPS units can report position data every
second, while radar hits a plane only once every six seconds. The ADS-B technol-
ogy GPS systems connect through the Iridium satellite system (Jackson April 26,
2004).
Standard Terminal Automation Replacement System (STARS)
STARS is the next big step in the FAAs comprehensive effort to upgrade air traf-
fc control facilities across the nation. The new system will provide the platform
for improvements to handle the ever-growing volume of air traffc safely and eff-
ciently well into the 21st century, said FAA administrator David R. Hinson (Dorr
1996). STARS will standardize all air traffc control equipment at the 172 FAA
facilities as well as the 199 Department of Defense facilities. STARS will supply
new hardware and software to these facilities. The program will be a complete
replacement for the aging systems currently in use.
The most important feature of the STARS system will be the ability to display
transmissions. The Automated Radar Terminal System (ARTS) that is currently
in place was developed in the 1970s and 1980s. The FAA believes that interim
programs are limited in their ability to extend the ARTS life in the short term. It is
generally accepted that this system does not have the capabilities to take air traffc
into the next century. ARTS software contains various versions and languages that
are very labor intensive as well as expensive to support.
The STARS program includes a commercial standard system that the FAA be-
lieves will be much cheaper and easier to maintain. A key feature is the ability
to extend and advance the capacity of the system without reengineering the ba-
sic architecture. By building on commercially available hardware and software,
the development time for the software will be reduced signifcantly. The resulting
maintenance costs will also be lower than those associated with the current ARTS
system.
By 2003, the STARS project was behind schedule by at least six years and
millions of dollars. The system was supposed to be completed in 1998 for $12
million. But after more than six years of development the system was still not
implemented (McCartney 2003). However, an initial version of the system was
installed in Philadelphia in late 2002. The system gathers data from several radar
systems on color displays. However, not everyone was happy with it. Control-
lers in El Paso noted that the system could not distinguish between planes sitting
on the runway and trucks on a nearby highway (CNN November 17, 2002). In
mid-2004, the FAA announced that it was ready to begin implementing the new
system. The Phase I rollout would take place at airports with the oldest equipment
and cost $1.4 billion. Nineteen of the ffty sites were online as of 2004. But, the
last of the 50 airports were not scheduled to receive the new equipment until the
end of 2007. There was no budget or schedule for the remaining 100 plus airports.
In 2004, the GAO and inspector general urged the FAA to gain control over costs.
The project was already seven years behind schedule and estimated costs had ris-
en to $1.9 billion (Mosquera April 26, 2004).
Wide Area Augmentation System (WAAS)
The Wide Area Augmentation System is used in conjunction with GPS. Using a
network of 36 ground stations to distill satellite GPS signals, WAAS will allow
commercial aircraft to pinpoint a location within seven meters. With the use of
WAAS/GPS, the FAA hopes it can close many of its ground control centers and
allow pilots to fy more direct routes. Consolidated, these tools are projected to
lead to the concept of free fight.
The WAAS system fell even further behind than the STARS project. The sat-
ellite-based system was pushed back by fve years and the estimated costs were
tripled (McCartney 2003).
Free Flight
Free fight is a consolidated goal toward which the FAA is working. Free fight
would enable pilots to control their own navigation procedures. The pilot would
use the WAAS and GPS systems for navigational purposes and choose their own
routes, speed, and altitude. Ground support will be held to a minimum and would
be most important when fights are in congested airport areas, when airplanes ap-
proach restricted airspace, or when safety is at stake.
Two principles that drive the free fight plan are the protected and the alert
airspace zones. The sizes of these zones are determined based on aircraft speed,
performance characteristics, communications, navigation, and surveillance equip-
ment. The protected zone is the zone closest to the aircraft. No aircraft should
overlap the protected zone of another aircraft. The alert zone is one that extends
far beyond an aircrafts protected zone. The distance between planes will be moni-
tored closely. If a plane touches another planes protected zone, the pilots and the
air traffc controllers will determine the course corrections that are needed. Under
the free fight system, interference will be minimized until the alert zones collide.
Of course, after September 11, the issue of free fight is probably obsolete. The
FAA and security agencies are even more interested in controlling and restricting
fights. Nonetheless, the FAA and the GAO continue to investigate free fight op-
tions. A main step in the process is the Traffc Management Advisor. This software
helps controllers effciently regulate the space between airplanes as they arrive at
airports. Under Phase I of the free fight program, fve software tools are being
tested at various sites. Phase II represents the expansion of the systemsif they
work. One system, the User Request Evaluation Tool (URET), was deployed late,
so it will require additional testing. It is designed to identify conficts and respond
to pilot requests for route changes. Another tool, the Final Approach Spacing Tool
(FAST), has been abandoned because of risks found in testing. It was designed to
assign runways and schedule landings (Langlois October 2001)
Some researchers note that reducing fight times will not be suffcient to speed
up the system. Delays are also created by slow operations at the terminals, includ-
ing refueling, baggage handling, and unscheduled gate changes. These researchers
suggest that signifcant changes are needed to improve communications among
airport terminals. One possibility is wireless PDAs carried by all personnel and
updated by the airlines.
The September 11 attacks caused the FAA to delay implementing some aspects
of the free fight (CPDLC) deployment. A major reason for the delay was due to
the costs that would be imposed on the airlines. The FAA was also not ready to
implement the new technologies (Vasishtha 2002)
Technology Innovations
The FAA has suffered through several failed projects over the years, including
the Advanced Automation System (AAS) that was designed in the mid-1990s and
thrown away in favor of the STARS project. The FAA also designed and imple-
mented new radio communication technology. The goal was to transfer data by
text, to reduce the use of voice communications. The Aircraft Monitory System
(ACMS) was designed to collect data on the plane and send it to controllers. The
Aircraft Communication Addressing and Reporting System (ACARS) was intro-
duced to cut down on the use of spoken radio messages to transmit information to
the ground. It was thought that if the fight crew could save time by transmitting
data to the ground rather than conveying it by voice to the air traffc controllers,
they would be better able to concentrate on fying the plane. ACARS directly in-
terfaces with ACMS and sends and receives messages directly to and from the
pilot. The pilot punches the message, such as fight plans, in an alphanumeric
keypad or touch screen. Both systems operate on the Aeronautical Radio system
(ARINC) that runs on VHF radio waves and handles the data transmission be-
tween the plane and the ground controllers. The system is owned and operated
by the major airlines. The main drawback to ARINC is that because of limited
bandwidth, the system transmits data at 2.4 kbps. In 2004, some airlines (notably
Southwest) began installing a newer data service called VHF Digital Link Mode 2
(VDLM2), which can transmit data at rates up to 31.5 kbps (Brewin 2004).
Launched in 2003, the FAA budget for 2008 contained $175 million earmarked
for the Next Generation Air Transportation System (NGATS). The system relies
more on satellite (GPS) navigation. It also encourages airports to upgrade their
facilities
Network
In 1998, the FAA replaced its mainframe-based system for acquisition manage-
ment with a distributed architecture. The old system ran on 1980s-era minicom-
puters at 12 centers nationwide and processed more than 200,000 purchases per
year. It was not updated for more than three years and was not Year 2000 ready.
Mounting problems in the old system led many FAA offcials to revert to paper to
track agency purchases.
The new system is called Acquire. It uses Oracle Corporations Alert software
and the Discoverer/2000 querying tool. The FAA must also use Oracle Federal
Purchasing software to get Acquire to run on a network that links headquarters to
regional offces and feld centers.
The FAA also began preparing a communications system overhaul aimed at
readying the agencys infrastructure to meet the needs of the 21st century. The
FAA Integrated Communications Systems for the 21st century (FICS-21) program
is projected to cost an estimated $2.75 billion.
FICS-21 will provide ground-to-ground transmission switching and network
management control for voice, data, and video communications. The new initia-
tive will replace at least 11 major programs, including FAA-owned and leased
networks. FAA FICS-21 program manager Jeff Yarnell says it is a good time to
rebuild the FAAs telecommunications infrastructure because many telecommuni-
cations contracts expired at the turn of the century.
In 2004, the FAA fnally began rolling out its new communication backbone.
The new FTI system was installed at 27 facilities. Steve Dash, FAA telecom
manager, said that the system is replacing fve disparate networks. He noted that
its the frst phase. The backbone will tie together the major operation facilities
(Jackson January 26, 2004). Ultimately, the system will be connected to the other
5,000 FAA facilities and save $700 million in telecommunication costs over 15
years. Installation of the system was contracted to Harris at an estimated total cost
of $3.5 billion. As much as possible, the system will use off-the-shelf networking
and telecommunication products. The new FTI system fell more than a year be-
hind schedule and in 2006 and 2007 the agency ended up paying for both systems
simultaneously because they had to maintain the old one while transitioning to the
new, incomplete system (Sternstein 2006).
The FAA also provides services to pilots (and the public) through its Web site.
Pilots account for 30 percent of the site traffc. To provide faster service, the FAA
installed an expert system from RightNow Technologies that examines questions
posed by visitors. The software compares the question to answers that have been
provided to other users. Matches that are close are immediately displayed to visi-
tors. Other questions are forwarded to the appropriate FAA authorities. Greg Gian-
forte, CEO with RightNow comments that we use a series of both implicit and
explicit learning capabilities, which include artifcial intelligence and machine
learning, to observe the historical usefulness of each knowledge item and provide
greater visibility to knowledge. Typically, the system can automatically handle
90 percent of the inquiries (Chourey April 26, 2004).
In conjunction with NASA, the FAA is using a simulation system called Fu-
tureFlight to test changes to airport control systems. Researchers testing confg-
urations of the LAX airport found that safety could be improved by moving a
taxiway to one end of the airport. John Bluck, speaking for the Ames Research
Center, notes that the idea is to try it [changes] in a safe way thats as close to
reality as we can make it. You dont have to try something new on a real airport,
where you have thousands of fights coming and going (Langlois October 2001).
Te Future
The FAA has faced considerable criticism over the delays and cost overruns asso-
ciated with replacing its primary systems. The agency makes heavy use of outside
contractors, which is probably a necessity. However, the agency needs to write
better contracts so that it can maintain control over costs and schedules.
The successful implementation of STARS is becoming critical. Like other fed-
eral agencies, by 2014, as many as half of the air traffc controllers can retire
(about 7,000 people) (Chourey July 5, 2004). These workers require intensive
training, and their salaries represent a signifcant expense. In 2002, more than
1,000 controllers earned over $150,000 (McCartney 2003). The FAA is going to
need better automated systems that are easier and safer. With increased traffc de-
mands, the FAA will have to fnd a way to improve productivity.
In the 2008 budget year, the FAA is pushing for a new funding mechanism. In
part led by the commercial airlines who are afraid of the microjet market, the FAA
is trying to push for a cost-based user fee system. Direct fees to commercial carri-
ers and fuel excise taxes collected from general aviation would be determined by
the FAA using some unspecifed process to match the fee to costs of the services
provided to the two user groups. The FAA also wants the authority to levy addi-
tional fees for the most congested airports (2008 U.S. Budget). The agency claims
that the NextGen project cannot be built without more funding. Planned for com-
pletion in 2025, the project is estimated to cost $15 to $22 billion (Bain 2007). In
2007, the FAA awarded an initial design contract to ConceptSolutions, LLC for
a fve-year $32 million project to design the NetGen system (Hardy 2007). The
FAA claims the system is vital to increase fexibility and handle the anticipated 30
percent increase in fights. The 2012 proposed Federal budget called for $1.24 bil-
lion funding for NextGen, an increase of $370 million compared to 2010.
In 2000, Congress approved creating an internal manager to oversee the fight-
control operations, but the position went empty for three years until the FAA hired
Russell Chew in 2003. Coming from business, Mr. Chew has removed layers of
bureaucracy, instituted cost measurement programs, and attempted to instill busi-
ness management into the system. He ordered the frst inventory of equipment
ever conducted at the FAA. With measurements in place, the FAA determined that
it cost $457 to handle one jet on one fight in 2003. With cost-cutting measures,
he reduced the number to $440 in 2004. He has also tried to reduce costs, by urg-
ing for consolidation of facilities. But, Congress ultimately controls spending and
representatives tend to fght plans that call for a reduction of jobs in their districts.
Mr. Chew also faces resistance from other FAA managers. Marion Blakey, FAA
Administrator observed that many FAA employees see any kind of change as
very threatening (Meckler 2006).
The 2010 discretionary spending for the FAA was $12.478 billion. The fscal
year 2012 proposed budget calls for a slight increase to $12.883 billion. These
numbers are on top of the mandatory spending of $3.515 billion in 2010 and pro-
posed $2.424 billion in fscal year 2012. The big challenge is that the FAA does
not plan to fully replace the existing radar tracking system until at least 2025. In
the meantime, problems still arise, such as the minor crash in 2011 when an Air
France Airbus A380 (a huge commercial airliner) collided with a regional CRJ-
700 jet while taxiing at the JFK airport in New York. Bill Voss, former FAA air
traffc development director and current president of the Flight Safety Foundation
noted that Youd be surprisedalmost all of this is done with pieces of paper,
an eyeball and a pencil. It is a very visual and manual activity. There were also
a string of incidents (but not crashes) in 2011 when late-night traffc controllers
were reported to be sleeping on the job (Patterson 2011). In terms of good news,
most ocean fights controlled by the U.S., Canada, New Zealand, Australia, and
most of Europe, are already monitored and controlled using a GPS system. The
Advanced Technology and Oceanic Procedures (ATOP) system was adapted by
Lockheed Martin from a system developed in New Zealand. After four years of
use, with pilots and airlines defning their own routes, the system has saved nearly
10 million gallons of fuel. (McCartney 2009).
Questions
1. After 20 years, why is the FAA still having so many problems building new
systems?
2. What tools or methodologies might be useful to help the FAA complete its
remaining tasks?
3. Is it possible to speed up the NextGen schedule?
4. Will outside contractors (Computer Science Corporation) help the projects?
How can you monitor and control the work by the contractors?
Additional Reading
Air Traffc Control - Good Progress on Interim Replacement for Outage-
Plagued System, But Risks Can Be Further Reduced, GAO Report, October
17, 1996.
Bain, Ben, FAA Funding Battle Threatens to Ground NextGen, FCW, June 14,
2007.
Brewin, Bob,Data Takes Flight, Computerworld, June 21, 2004.
Chourey, Sarita, Getting an Instant Response, FCW, April 26, 2004.
Chourey, Sarita, Air Traffc Controllers Head To the Beach, FCW, July 5, 2004.
CNN Online, New Air Traffc Control System Tested, November 17, 2002.
Dorr, Les. FAA Selects Raytheon for Next-Generation Air Traffc Control
System Upgrade, FAA News, September 16, 1996.
FAA Radar Glitches Found. Computerworld, November 2, 1998, p. 12.
FAA Ready for Free Flight, Advanced Transportation Technology News, April
1996.
FAAs $500 Million Navigation Contract Takes Flight, Federal Computer
Hardy, Michael, FAA Taps ConceptSolutions for Next Generation Air System,
Government Computer News, April 6, 2007.
Jackson, Joab, FAA Considers Commercial Satellites for Traffc Control,
Government Computer News, April 26, 2004.
Jackson, William, FAA and GSA renew their dogfght over air traffc control
modernization, Government Computer News, September 18, 1995, p.73.
Jackson, William, FAA Clears Rollout of Net Backbone, Government
Computer News, January 26, 2004.
Langlois, Greg, Researchers: Dont Forget Airport Ops, Federal Computer
Langlois, Greg, Data Determining Free Flight Future, Federal Computer Week,
October 2, 2001.
Langlois, Greg, NASA: Simulation Proves Its Worth, Federal Computer Week,
November 5, 2001.
Leopold, George, Study: GPS can Fly as Commercial Air Navigator, Electronic
Engineering Times, February 8, 1999, p. 18.
McCartney, Scott, Labor Costs, Mismanagement: Airlines Again? No, the
FAA, The Wall Street Journal, February 19, 2003.
McCartney, Scott, Building a Better Air-Traffc-Control System, The Wall
Meckler, Laura, FAAs Mr. Chew Tackles High Costs, Chronic Lateness, The
Wall Street Journal, January 30, 2006.
Mosquera, Mary, FAA Slows Down Display System Deployment, Government
Computer News, April 26, 2004.
Patterson, Thom, Can Technology Fix the FAAs Air Traffc Troubles? CNN
Online, April 26, 2011.
Sternstein, Aliya, FTI Delay Leaves FAA Paying for Two Systems at Once,
FCW, December 8, 2006.
Tiboni, Frank, FAA Begins Upgrade Project On Its Controller-Pilot Comm
System, Government Computer News, February 8, 1999, p. 8.
Vasishtha, Preeti, FAA Defers CPDLC Deployment, Government Computer
News, January 4, 2002.
Weiss, Todd R. FAA Still Seeking Cause of Flight Plan System Shutdown,
Case: Te Internal Revenue Service (IRS)
For the 2009 tax year, the IRS processed more than 144 million individual and
2.4 million corporate tax returns (www.IRS.gov/taxstats). Many of the returns
are simple one-page forms; others run to thousands of pages of supporting docu-
ments. Overall, the service handles more than 1 billion information documents a
year. The IRS processes more than $1 trillion in tax revenue a year. The IRS has
10 regional service centers that are responsible for processing and storing indi-
vidual forms.
Until 1990, all documents at the IRS were stored as paper records in a cen-
tral warehouse. In 1989, it cost the IRS $34 million just to store the paper docu-
ments. Documents were organized according to the year of fling. As a result, if
a taxpayer had a problem or question that covered multiple years, the citizen had
to schedule multiple meetings with IRS offcials to correct problems for each of
the years. In some cases, it could take weeks or months just to get the fles. Oc-
casionally, the IRS found it was faster to ask the taxpayer for a copy of the return.
By the early 1990s, this problem was resolved by having each of the 10 service
centers store digital images of the tax returns, making them available to agents on
their terminals. While a step in the right direction, this approach did not give the
IRS the fexibility it would receive from the ability to scan the returns directly into
a computerized information system.
Automation sometimes causes problems in addition to solving them. Such
was the case of Dickie Ann Conn. The IRS determined that she owed $67,714 in
back taxes. She was sent a bill for more than $1 billion in interest and penalties.
After being challenged, the IRS admitted that there was an error in the interest
computation.
The IRS operating budget in 2010 was $12.146 billion, with a proposed fscal
year 2012 budget of $13.284 billion.
A History of Automation Problems
The IRS seems like a logical candidate for improved automation. The benefts of
faster processing, fewer mistakes, and easier access to data ought to save a consid-
erable amount of money. The computers ability to search the data, automatically
match transactions, and analyze each return presents several additional opportu-
nities that can either cut costs or raise additional revenue. Managers at the IRS
are fully aware of the potential, and they have proposed several systems over the
years. The problem has been in implementation of the plans and in getting Con-
gress to fnancially support the changes.
In the late 1960s, the IRS knew it needed to redesign its basic systems. In re-
sponse, it began to plan for a system to be installed in the 1970s. The IRS did not
get the needed support in Congress because of fears that it would be too expensive
and too invasive into individual security and taxpayer privacy. As a result of this
lack of support, the IRS turned its attention toward keeping its existing computers
running.
In 1982, the existing system was nearing capacity and the IRS established the
Tax System Redesign program. It promised a complete redesign of the system.
According to the GAO, changes in management resulted in the system never get-
ting past the design stage. A new assistant commissioner in 1982 embarked on the
design of a new system that promised to carry the IRS through the 1990s. Initial
costs were estimated at $3 billion to $5 billion over the entire project. The primary
objective was to replace the old central tape-based system with an online data-
base. Eventually, optical technology would be used to scan the original documents
and store the data in the database. A new communication system would carry the
data to any agents workstation. By 1989, initial planning had already cost the
IRS more than $70 million, with no concrete proposal or results.
The main computer systems were replaced at the IRS service centers in 1985.
The change in the systems was almost disastrous for the IRS. It delayed returns
processing and led to delays in refunds that cost the IRS millions of dollars in
interest payments. IRS employees worked overtime but still could not keep up.
Rumors were fying that some employees were dumping returns to cut down their
backlog. Because of the delays and backlogs, the IRS managed to audit only
about half the usual number of returns on which it conducted audits.
In 1986, the IRS initiated a plan to provide 18,000 laptop computers to enable
its feld auditors to be more productive with its Automated Examination System
(AES). Unfortunately, the service bought the Zenith laptops a full year before the
software was ready. The system was written in Pascal and was delivered to agents
in July 1986. It was designed to examine Form 1040 returns. The biggest draw-
back was that it used 18 different diskettes. This required agents to be constantly
swapping disks. Based on the privatization directives from the Reagan adminis-
tration, the system was subcontracted to outside developers. As IRS funding was
cut, programmers with experience in Pascal were cut. This led the system to be
rewritten in C.
A survey in 1988 revealed that 77 percent of the agents were dissatisfed with
the software. Only 33 percent said that they used it. By 1989, the IRS revised
the software and managed to reduce it to eight disks. By this time, the AES proj-
ect was more than six years behind schedule and, according to the GAO, was
$800 million over the original budget. The IRS originally anticipated that the
AES would produce $16.2 billion in additional revenue over nine years by making
agents more productive. The GAO disputed those numbers, noting that the IRS
has been unable to verify that the use of laptops has actually resulted in the exami-
nation of additional returns or increased tax revenues. In 1990, the White House
cut the funding for the program from $110 million to $20 million.
In 1999, the IRS implemented a new network to connect computers throughout
the organization. Twenty staffers were dedicated to the project and took four years
to complete it. IBMs Tivoli software is a key tool to manage the 132,000 net-
worked devices in 87 locations. The software enables network managers to con-
tinually monitor all aspects of the network. They can also push changes down to
the desktop computers if problems arise or they need upgrades. Before the system
was available, it took an IRS staff member 20 minutes to update each device. With
Tivoli live in 2003, a single network administrator sent one update to 400 desk-
tops in one minute. Jim Kennedy, program manager for enterprise systems man-
agement at the Austin, Texas, support center estimates that the system has saved
$2.6 million in the frst quarter alone (Dubie 2003).
Technology Innovations
By 1989, the IRS knew that it desperately needed to redesign its entire system for
collecting taxes and processing information. In hearings before Congress, Senator
David Pryor (D-Ark.) noted that the 1960s-era IRS computers were headed for a
train wreck in the mid-1990s. The GAO estimated the total project would cost
between $3 billion and $4 billion. The projected date for implementation slipped
from 1995 to 1998.
The overall design for the Tax System Modernization program (TSM) called
for a centralized online database, smaller departmental systems containing local
information, and linkage through a nationwide network. Tax return data would be
entered through a combination of electronic fling and optical scanners.
By 1991, the estimated cost of the plan had expanded to $8 billion. Although
the IRS projected that the system would cut $6 billion in costs, the plan was rapid-
ly attacked by members of Congress. Three studies of the TSM plan by the GAO
were released in early 1991:
The GAO was concerned that optical technology was not suffciently ad-
vanced to perform the tasks demanded by the IRS. The GAO urged greater
emphasis on electronic fling.
The GAO was concerned that management issues such as transition planning,
progress measurement, and accountability were not suffciently addressed by
the plan.
The GAO and Senator John Glenn (D-Ohio) voiced concerns about data secu-
rity and integrity.
GAO offcial Howard Rhile noted, This is a serious omission in view of the
fact that the IRS intends to allow public access to some of its systems and
because concerns over the security of taxpayer information helped doom the frst
[IRS] modernization effort in the late 1970s.
Despite these misgivings, the IRS was committed to the TSM plan. Fred Gold-
berg, IRS commissioner, agreed with the GAO fndings but observed that
We have been running our business essentially the same way,
using essentially the same computer and telecommunications
systems design for 25 years. [Existing systems] will perform
well and achieve incremental improvements for the next few
years Our best judgment is that [OCR] technology will be
there when we need it, by the end of the decade.
By 1992, the situation grew worse. Shirley Peterson, the new commissioner of
Internal Revenue, stated at a congressional hearing that
Our systems are so antiquated that we cannot adequately
serve the public. The potential for breakdown during the
fling season greatly exceeds acceptable business risk. Some
components of these computers are so old and brittle that
they literally crumble when removed for maintenance.
In December 1991, the IRS awarded a 12-year, $300 million contract to TRW
to help manage the process and provide planning and system integration servic-
es. The recommended system was ambitious. It called for 60 major projects, two
dozen major purchases, 20 million lines of new software, and 308 people just to
manage the purchasing process. Despite the best efforts of the administrators, ele-
ments of the IRS modernization plan were stalled because of purchasing diffcul-
ties. In July 1991, the IRS awarded a billion-dollar Treasury Multiuser Acquisi-
tion Contract (TMAC) to AT&T. The goal was to standardize purchasing for the
IRS and the Treasury Department by routing all purchases through one vendor.
The contract was challenged by other vendors and overturned. The contract was
rebid and AT&T won the second time. IBM (one of the original protesters) again
objected to the process, noting that the IBM bid of $708 million was less than the
$1.4 billion bid by AT&T.
In 1993, the IRS acknowledged that the TSM Design Master Plan needed to
be rewritten. In particular, it had to focus on business aspects instead of technol-
ogy elements. To better coordinate technical planning with IRS needs, the agency
established a research and development center funded by $78.5 million of federal
money but run by the private sector. The center was responsible for providing
technical assistance and strategic planning for the TSM. The IRS also established
a high-level architect offce to evaluate technologies and direct their proposed
uses.
Throughout calendar year 1992, the IRS spent $800 million on TSM. In 1993,
new IRS estimates indicated that TSM would cost $7.8 billion above the $15.5
billion needed to keep existing systems running. The new system was projected to
generate $12.6 billion in total benefts by 2008 through reduced costs, increased
collections, and interest savings. Moreover, the improved process was supposed
to save taxpayers $5.4 billion and cut 1 billion hours from the collective time they
needed to spend with the IRS.
In 1996, the IRS asked Congress for a $1.03 billion appropriation. This was a
substantial increase over the $622 million it spent on automation in 1995. Hazel
Edwards from the General Accounting Offce noted, After eight years and an
investment of almost $2 billion, the IRSs progress toward its vision has been
minimal.
IRS Commissioner Margaret Milner Richardson denied the GAO claims. She
noted, I think we have made signifcant progress, not minimal progress but we
do know we can and must do more (Birnbaum, 1998).
The IRS situation represented a dilemma for Congress. The IRS claims that
the only way to make a system that works is to spend more money. The GAO has
set forth that it is impossible to complete the entire project envisioned by the IRS.
The GAO believes the IRS should, instead, concentrate on smaller, more focused
projects that can be completed in a one- to two-year timeframe.
In 2001, Congress passed tax-cut legislation to stimulate the economy, and or-
dered the IRS to send refund checks to all taxpayers. It took several months to
create and mail the tens of millions of checks, but most of them were correct. On
the other hand, about 523,000 taxpayers received notices that they would be get-
ting a check for the full refund amount, when they were actually eligible for only
part of the refund. The mistake was attributed to a programmer error, and the fnal
checks were correct; but some taxpayers were confused by the misleading letter.
Electronic Filing
The IRS introduced electronic fling in 1986, when 25,000 forms were fled elec-
tronically. By 1990, 4.2 million people fled for tax refunds electronically. In 1992,
the number increased to 10 million flers. In 2003, 49 percent of the personal tax
returns were fled electronically (www.irs.gov).
The primary target for electronic fling is the millions of individual taxpayers
who are slated to receive refunds. To control the process and ensure that docu-
ments are properly fled, electronic fling is available only through authorized tax
preparers. The IRS is deliberately avoiding providing access to individual taxpay-
ers. As a result, taxpayers who use the system pay an additional charge to the pre-
parer. However, the electronic fling system provides refunds within a few days.
Forms that have been electronically fled cost the IRS one-tenth the processing
cost of paper forms. This approach also eliminates the cost of paper storage. The
IRS notes that it is able to store 800,000 returns on one side of a 12-inch optical
disk.
For taxpayers with easy returns, the IRS is simplifying the process even fur-
ther. Short forms can now be fled over the telephone. In a 1992 pilot, 117,000
Ohio taxpayers fled for refunds using push-button phone calls. The system was
expanded nationwide in 1994. The push-button system can be used only by tax-
payers who are able to use the 1040EZ form. A replacement form (1040-TEL)
must still be signed and fled with the IRS, along with the W-2 (withholding)
statements.
In the 1998 IRS Restructuring and Reform Act, Congress required the IRS to
encourage the use of electronic fling. The IRS has made it easier for people to
fle electronicallyparticularly for those who use computer software to compute
their taxes. In 1998, about 20 percent of individuals fled electronically; in 2000
the number was 28 percent; in 2001 about 32 percent (45 million). The IRS goal
is to increase this number to 80 percent by 2007 (Dorobek 2001). For the 2001 tax
year (fling in early 2002), the IRS used the Digital Signature law to send PINs to
several million taxpayers, enabling them to legally sign their tax forms electroni-
cally. However, the one important catch is that taxpayers who fle electronically
must pay an additional fee to do so. Hence, only those who receive refunds (about
70 percent of the flers) are interested in paying the fee, because it enables them
to get their money faster. Most experts believe it is unlikely that the IRS will meet
the congressional goal of 80 percent by 2007.
In 2009, the IRS noted that it cost $3.29 to process a paper return and only 19
cents to process one fled electronically (Saunders March 2011). In 2011, the IRS
stopped mailing paper forms to people to encourage them to fle electronically.
For 2009, about 70 percent of the forms were fled electronically. But e-fling is
least used by wealthy taxpayers, and they are the ones with the most complex
returns. Some people with complex forms cannot fle electronically. Other people
are wary of turning over massive amounts of data to a third-party company to
handle their taxes online and submit them to the IRS.
Te Internet
In late 2001, the IRS announced plans to offer electronic payments by businesses
over the Internet. A major portion of the money received by the IRS comes from
withholdings collected by businesses. This money has to be forwarded to the IRS
at regular intervals, so the IRS is trying to reduce handling costs by moving these
transactions online. The Electronic Federal Tax Payment System (EFTPS) is a
Web-based system that can also be used by small businesses and by taxpayers
who make estimated quarterly payments. Using modern strong encryption tech-
nologies, the IRS is confdent the system will be secure.
Relatively early in the dot-com and dot-gov restructuring, the IRS realized the
importance of putting information on its Web site. In fact, a huge amount of infor-
mation is available online. And that is a problem. In 2001, IRS executives were
asked to search the site for common tax information. It generally took 20 or 30
clicks to fnd any piece of information. To improve its Web site, the IRS hired
Gregory Carson in 2001, a designer from private industry who helped to launch
the Priceline.com Web site.
The IRS also signed a contract with the consulting group Accenture to rede-
sign the IRS Web sites. In 2001, the site received 80 million hits a day. Gregory
Carson, director of electronic tax administration modernization at the IRS, notes
that the development of an intuitive, intentions-based design will make it con-
siderably easier for taxpayers and tax preparers, who pull forms from the site, to
obtain the information and documents they need to fle tax returns. (Rosencrance
August 2001) Accentures goal is to make the site easier to use so that users can
reach the desired information within three clicks. Furthermore, Accenture will be
hosting the site on its servers.
In 2003, more people turned to the Internet to fle their tax returns. Several
companies provide online systems that automatically e-fle the data with the IRS.
A few offer free fling. In 2003, 3.4 million taxpayers used the Free File service.
In total over 14 million people used their personal computers to e-fle their taxes
(Mosquera May 10, 2004).
Automated Under-Reporter (AUR)
The Automated Under-Reporter (AUR) is another component of the TSM. The
AUR is a system designed to monitor returns and identify people who are most
likely to underpay their taxes. The system was frst installed in 1992 at the Og-
den, Utah, regional center. The system pulls data from the service centers Unisys
1180 mainframe. The data is downloaded across a local area network to a Sequent
Computer System S-81 minicomputer. From there the information is sent to one
of 240 networked UNIX workstations on the employees desks.
The system automatically matches distribution documents (such as 1099s and
W-2s) with the flings of individual taxpayers. Mark Cox, assistant IRS commis-
sioner for information systems development, noted that in trials with the AUR,
weve been able to cut down the rework of cases from 25 percent to less than 5
percent. We see this type of work enabling us to share in more of a connectivity
mode (Quindlen, 1991).
The system uses an Oracle database running SQL to match data from various
sources. It also performs basic tax computation and helps agents send notices to
taxpayers. Managers have noted that even though the new system has not im-
proved the speed of the agents, it has cut down on the error rates. As agents be-
come familiar with the system, productivity is expected to improve.
In 1991, the Ogden center processed 26 million tax returns and collected $100
billion in tax payments. It processed $9 billion in refunds. In 1992, it won the
Presidential Award for Quality for improved tax processing by saving the govern-
ment $11 million over fve years.
In 2007, the IRS estimated that the compliance rate for individuals paying taxes
on time was 86 percentleading to an estimated $290 billion per year tax gap
(based on 2001 data). President Bush argued (2008 U.S. Budget) that if this tax
gap could be reduced, many programs could be funded without additional taxes.
His budget for the IRS called for an additional $410 million for research, enforce-
ment, technology, and taxpayer services to reduce this gap.
Te Currency and Banking Retrieval System
In 1988, Congress passed a new law in an attempt to cut down on crime (notably
drug dealing) and to provide leads to people who signifcantly underreport their
income. Every cash transaction over $10,000 is required by federal law to be re-
ported to the IRS on a Form 8300. The IRS created the Currency and Banking
Retrieval System to match these forms against the flers tax return. The system
automatically identifes people who had large cash purchases but claimed little
income. Because of a programming error, the system missed forms covering $15
million in cash transactions between 1989 and 1990.
The problem stemmed from the fact that the IRS used the same code number
on the 8300 forms that it used on other cash transaction forms. The IRS later as-
signed separate codes for each form. When programmers wrote the new match-
ing programs, they did not realize there were two codes for each transaction. The
system was corrected in 1991. By 1992 it was used to process more than 1 million
queries a year.
Jennie Stathis of the GAO noted there were additional problems with Form
8300. In particular, the flings were incomplete or contained incorrect taxpayer
identifcation numbers. The IRS is developing software to enable businesses to
verify the taxpayer ID numbers automatically before the customer completes the
purchase.
Document Processing System (DPS) and Service Center Recognition/Image
Processing System (SCRIPS)
In 1994, the IRS awarded a $1.3 billion contract to the IBM Federal Systems
division to design a document processing system. The goal was that by the late
1990s, the system would convert virtually every tax return to a digital format. A
day after the contract was awarded, IBM sold the Federal Systems division to Lo-
ral Corporation for $1.52 billion.
The 15-year systems integration contract was to have the system running on-
line in 1996. The plan called for scanning incoming tax forms. Special software
digitally removed the form layout and instructions, leaving just the taxpayer data.
OCR software was to then convert the characters (including handwritten num-
bers) into computer data.
The system was scheduled for initial installation at the Austin, Texas, regional
center in August 1995. Plans called for installing it at Ogden, Utah; Cincinnati,
Ohio; Memphis, Tennessee; and Kansas City, Missouri, by 1998. Despite the pop-
ularity of electronic fling, the IRS still sees a need for the OCR system. The IRS
received 222 million returns in 2003. Of those, 53 million were electronic.
SCRIPS was the frst scanning project. Presented at a cost of $17 million, it
was approved to cost $88 million when it was awarded in 1993 to Grumman Cor-
porations Data Systems unit. SCRIPS was designed to capture data from four
simple IRS forms that are single-sided. SCRIPS was supposed to be an interim
solution that would support the IRS until the Document Processing System (DPS)
could be fully deployed. However, delays pushed back the delivery of the SCRIPS
project. By the time it was declared fnished, the project cost $200 million (Birn-
baum 1998)
DPS was the second scanning project. It has a projected cost of $1.3 billion. In-
terestingly, Grumman Data Systems was the loser in the contest for the DPS con-
tract. The IRS noted that Grumman failed a key technical test. When completed,
DPS was quite complicated to use. In this program, the IRS developed nine sepa-
rate databases, most of which could not communicate with each other.
In 1996, Art Gross, a veteran of the New York State revenue department, be-
came the new IRS chief information offcer. He stated that the IRSs computers
didnt work in the real world and that its employees lacked the intellectual
capital to transform them. When he arrived in 1996, the IRSs Year 2000 conver-
sion project had a budget of $20 million and a staff of three; by 1998, it had grown
to a $900 million project with 600 workers, many of them consultants (Birnbaum
1998).
Gross tried to get control of the system. He ended the DPS or Bubble Ma-
chine project as being over budget and behind schedule. With help from TRW,
he devised a new top-to-bottom computer architecture. The architecture was built
around a centralized database to coordinate information at the IRS.
When Charles Rossotti arrived as the new commissioner, he proposed an even
more ambitious plan. In addition to Year 2000 changes, computer updates from
the 1997 tax law, and the overall modernization, Rossotti proposed to restructure
the entire organization. This proved to be too much for Gross, who resigned.
In 1998, Congress passed the Government Paperwork Elimination Act, part of
which forces the IRS to move to more electronic transactions. Since then, the IRS
has created electronic versions of its forms that can be downloaded from its serv-
ers. In 2001, the IRS signed a contract with ScanSoft Inc. for OmniPage Pro 11
for use in its federal tax offces around the nation. The goal is to convert the mass-
es of paper fles into electronic documents. Instead of taxpayer fles, the system is
designed more to convert internal forms and documents so that all employees will
have immediate access to up-to-date forms and policies on the IRS intranet.
In late 2001, the IRS began installing customer relationship management (CRM)
software that it purchased from PeopleSoft. A key element of the kinder, gentler
approach is the ability to track customer issues. CRM software can collect all of
the customer interactions into one locationmaking it easier for multiple agents
to see the entire history of a particular problem. The system will also enable the
agency to create Web portals for professional tax preparers, IRS employees, and
taxpayers. The portals will securely provide individual information to these groups
over the Web. In addition to faster service, the IRS hopes to reduce the costs of its
call centers by moving more access online.
The IRS also developed the e-help system in 2002 to provide a central point
for customer service. The system was designed to ensure service representatives
provide consistent and accurate responses to customer questions. The Inspec-
tor General in 2007 reported that the system had made progress, but still needed
improvement. Notably, the system lacked quality measures and procedures, in-
cluding a failure to survey customer opinions and train employees. In a random
sample of 19 employees, the Inspector found that none of them had completed
required training. Michael Phillips, IRS deputy inspector general for audit noted
that ensuring assistors complete required training will be of greater important as
the IRS moves forward with implementation of the next available technology
(Cranmer 2007).
Security Breaches
In 1983, Senator John Glenn (D-Ohio) released an IRS report indicating that 386
employees took advantage of ineffective security controls and looked through
tax records of friends, neighbors, relatives, and celebrities at the Atlanta regional
IRS offce. Furthermore, fve employees used the system to create fraudulent re-
turns, triggering more than 200 false tax refunds. Additional investigations turned
up more than 100 other IRS employees nationwide with unauthorized access to
records. Glenn observed that the IRS investigation examined only one region and
looked at only one of 56 methods that have been identifed to compromise secu-
rity. Glenn expressed the concern that this is just the tip of a very large iceberg.
The IRS noted that the TSM program greatly increases the risk of employee
browsing, disclosure, and fraud, because of the online access to the centralized
databases.
Margaret Richardson, commissioner of the Internal Revenue Service, noted
that the system used by the perpetrators was 20 years old and was used by 56,000
employees. It met all federal security standards, including the use of passwords
and limited access based on job descriptions. The IRS found the problems in At-
lanta by examining records of database access from 1990 to 1993. Because the
system generates 100 million transactions a month, the data is stored on magnetic
tape, making it diffcult to search.
In 1989, the IRS arrested Alan N. Scott, of West Roxbury, Massachusetts, for
allegedly submitting 45 fraudulent returns via the new electronic fling system.
The IRS claims Scott received more than $325,000 in refunds.
The IRS requires tax return preparers to fll out an application before it issues
an access code. Scott apparently used a fake taxpayer ID number and lied on the
application form to gain the access number. The IRS claims he then submitted
false returns using bogus names and taxpayer ID numbers to get refund checks
ranging from $3,000 to $23,000.
IRS offcials noted that the electronic flings actually made it easier to identify
the problem, because the computer could scan the data earlier than the data would
have been scanned if it had been submitted by hand. Once the situation was iden-
tifed, the IRS was able to immediately lock out further transactions from Mr.
Scotts access number.
In May 2007, the Treasury Departments Inspector General reported that the
IRS lost 490 computers between 2003 and 2006. Of these, 111 occurred within
IRS offces. Most of the machines lacked encryption and strong passwords. The
IRS has 100,000 employees and has issued 47,000 laptops. It was unable to iden-
tify the data that was lost because the agency has no records of what data was
stored on the machines, but the audit report claims data was compromised for at
least 2,300 taxpayers. Deputy inspector general Michael R. Phillips, stated that
we believe it is very likely a large number of the lost or stolen IRS computers
contained similar unencrypted data. Employees did not follow encryption proce-
dures because they were either unaware of security requirements, did so for their
own convenience, or did not know their own personal data were considered sensi-
tive. We also found other computer devices, such as fash drives, CDs, and DVDs,
on which sensitive data were not always encrypted. An audit in 2003 reported
similar problems, and the IRS has taken no action to change procedures (Gaudin
2007).
Modern Disasters
In 1998, the message in congressional hearings was to Do something. Anything.
The hearings into IRS dealings with the public revealed several problems within
the IRS. They emphasized the negative perceptions the public has toward this im-
portant agency. After listening to these criticisms, the IRS eventually agreed to
change some of its policies to improve its treatment of citizens. The 1998 IRS
Restructuring and Reform Act was aimed at changing IRS attitudes and provid-
ing citizens with more control in the tax-collection process. Charles Rossotti, the
new IRS commissioner, described the process of upgrading the vacuum tube-era
technology as being similar to rebuilding Manhattan while were still living in
it. The $7 billion agency has attempted the same gargantuan task of modernizing
its computers for 25 years and continues to fail. The total cost in the 1990s alone
has been projected to be nearly $4 billion (Birnbaum 1998).
In 2002, the system included 80 mainframes, 1,335 minicomputers, and
130,000 desktop boxes that were largely unable to communicate with each other.
Before his appointment as commissioner of the IRS in November 1997, Rossotti
served as chairman of the computer consulting frm American Management Sys-
tems. In early 1998, Arthur Gross, the chief technology offcer, who drafted the
latest modernization blueprint, resigned in frustration. Shortly thereafter, Tony
Musick, the chief fnancial offcer, resigned to become deputy CFO at the Com-
merce Department (Birnbaum 1998).
Unfortunately, the IRS has been even less successful at implementing new
technologies. By 1998, nearly all of the earlier systems development efforts were
canceled. In late 1998, the IRS signed a 15-year development contract with Com-
puter Science Corporation (CSC) that was worth $5 billion. By contract, CSC is
responsible for helping design new systems, indicating that the ultimate goal is
still to be determined. Outside experts note that the contract does not necessarily
solve all the IRS problems. The IRS must still deal with the contract management
issues, which have proved diffcult to the IRS in the past.
In 1999, the IRS launched yet another attempt to modernize its systems. The
$8 billion Business Systems Modernization (Bizmo) program was supposed to re-
place the infrastructure and over 100 applications. A key element is to replace the
Master File systemwhich is an ancient tape-based system that holds customer
data that the IRS has been using for over 40 years. The system runs an archaic
programming language with code written in 1962. The heart of the new system
is the Customer Account Data Engine (CADE) designed to run on IBMs DB2
database system. As of 2004, the project is way over budget and years behind
schedule. Even the system to process the simple 1040EZ form is three years late
and $36.8 million over budget (Varon 2004).
The system design actually started out well. The IRS hired CSC as the prime
contractor. But the IRS did not maintain control of the contract, and there are
serious doubts that even CSC was capable of handling the complex project. Paul
Cofoni, president of CSCs Federal Sector business, testifed to the U.S. House
Ways and Means Oversight Subcommittee that I have never encountered a pro-
gram of the size and complexity as the Business Systems Modernization program
at the IRS (Varon 2004). Several times, the IRS considered fring CSC, but kept
deciding that it would not be cost effective. One of the problems is that the IRS is
not providing suffcient oversight of CSC or the project. They originally planned
a relatively hands-off approach to let the company use best practices in its devel-
opment. The problem is that CSC needed the expertise of the IRS agents and IT
workers. The other problem is that the IRS went through fve CIOs in seven years.
In the meantime, CSC has gone through four managers to lead the project.
The CADE system is an impressive piece of technologyif it ever works.
Once the database is active, the IRS will use a customized version of the Sapi-
ens eMerge rule-processing engine. Congressional tax laws are coded as business
rules that the system applies to evaluate each tax return. The system includes an
XML-based RulesScribe layer that handles changes and additions to the rules.
The simplest 1040EZ tax form requires about 1,000 rules. Red Forman, associate
IRS commissioner for business systems modernization, notes that we are certain
we will have tens of thousands of business rules once CADE rolls out, and thats
just for individual flers (Mosquera May 17, 2004).
In May 2003, Mark Everson was appointed IRS commissioner, and three weeks
later, he appointed W. Todd Grams as the CIO. As of 2004, the project is nowhere
near completion. The IRS and CSC have been repeatedly blasted by congressional
reports. Relationships between CSC and the IRS are tense. CSC has been banned
from participating in additional IRS projects (Perez 2004). On the other hand, in
mid-2004, the FAA did hire CSC for up to $589 million to help build an enhanced
Traffc Flow Management system (McDougall 2004).
In 2006, the IRS issued more than $318 million in refunds on phony returns
because of a software failure. The IRS planned to replace the old screening system
with a Web-based application by January 2006, but the organization spent $20.5
million with no progress being developed by CSC. The IRS tried to restore the
old application, but could not get it running in time. In 2005, the software caught
133,000 fake returns, stopping $412 million in refund checks from being sent.
Without the software, the IRS halted a mere $94 million in fraudulent refunds
(Keizer 2006).
In 2007, the CADE system was operational, but not perfected. The IRS intends
to use system in conjunction with its 45-year old Master File system until 2012.
In 2007, the GAO accused CADEs slow processing times or delaying refunds for
millions of taxpayers by several days. The CADE system also signifcantly exceed
costs again in 2006 (Mosquera 2007).
The IRS is trying to balance a fne line when auditing people, particularly small
businesses. In 2011, the IRS started demanding that small businesses turn over not
just the data but also the accounting software and complete electronic fles. Most
small businesses use simple off-the-shelf accounting systems that do not segregate
data. Providing the information requested by the IRS means turning over all of
their records, including customer data. Professionals such as physicians are con-
cerned that the request would violate HIPAA rules about patient confdentiality.
But, even non-healthcare businesspeople are concerned. Benson Goldstein, senior
technical manager of taxation at the AICPA noted that Believe me, small busi-
nesses dont want the IRS calling their customers (Saunders May 2011).
In a letter to accountants, the IRS claimed that the request was part of its mod-
ernization program. Chris Wagner, commissioner of the Small Business/Self-
Employed Division, wrote that the audits require unaltered metadata so examin-
ers can properly consider the integrity and veracity of the electronic fles. Put
another way, the IRS believes that small businesses are responsible for 20 percent
of the estimated $345 billion tax gap, and it wants to completely examine business
records to fnd any errors and inconsistencies.
Te Future
In the meantime, the IRS still has to process taxes. So far, electronic fling is prob-
ably the only thing keeping the agency alive. Yet if anything goes wrong with the
ancient Master File system, the IRS is dead in the water. The IRS, CSC, and IBM
have no choice but to get CADE running correctly as soon as possible.
Talk to citizens about paying taxes and you get lots of interesting responses.
Yet a critical feature of the system is that everyone has to believe that they are
being treated fairlymeaning the same as everyone else. If people somehow per-
ceived that millions of others are getting by without paying taxes, everyone will
revolt. For years, since the 1980s, the IRS has relied on a relatively simple sys-
tem to automatically scan returns and identify possible tax cheats. The problem is
that the rules are based on data and an economy from 20 years ago. The system
is no longer catching the real tax cheats. In 2002, the IRS began collecting new
data and designing new rules to identify which returns should be scrutinized more
carefully. With a 13 percent increase in tax returns and a 29 percent decline in the
auditing staff since 1995, the IRS has to rely on automated systems to analyze the
returns. Charles O. Rossotti, the IRS commissioner at the time, could not give
details of the new rules but did note that the fact is, people who make more than
$100,000 pay more than 60 percent of the taxes, and we need to focus there
(Johnston 2002).
The electronic fling system is critical to improving productivity at the IRS.
Without it, thousands of people have to enter data from the paper forms into the
computer system. Yet, the existing system has several problems. Notably, it often
rejects 1040 forms because of errors. The errors are anticipated, because people
often make simple mistakes while entering data into their systems. The problem
is that the IRS system tends to reply with cryptic messages that users have trouble
decoding. The IRS is aware of the problems, but cannot decide how to fx them.
Terry Lutes, associate chief information offcer for information technology ser-
vices, notes that they cannot decide whether to fx the existing system or build a
new one. He asks, Its a question of how much money do you spend on a system
thats going to be a throwaway? But the drawback to a new system is that opti-
mistically, it would not be in place until 2010 (Olsen 2004).
To top off all of the operational problems, the IRS is being criticized by the
GAO because of problems with its internal accounting procedures. The GAO has
been nagging the IRS about problems with their fnancial management system for
a decade. Of 100 recommendations from the 2003 audit, the IRS has implemented
only 24. The IRS is planning to address many of the other problems with another
part of its modernization system, the Integrated Financial System (Mosquera April
30, 2004).
The IRS reported (2008 U.S. Budget) that electronic fling has increased from
31 percent in 2001 to 54 percent in 2006. This trend simplifes data collection and
processing at the IRS, but it probably has limits.
The IRS held preliminary negotiations with vendors to outsource management
and maintenance of its 100,000 desktops, but canceled the plan in late 2006. The
IRS also scaled back its separate plan to outsource the handling of paper returns
to a contractor. IAP Worldwide Services ran two processing centers, the other
fve were rescheduled for transfer after the 2006 tax year (McDougall November
2006). However, formal dates for transfers were not scheduled.
In 2007, the IRS awarded a fve-year, $9.6 million contract to General Dynam-
ics to help manage the business systems modernization project. Although details
were not provided, the role appears to be implementation-oriented instead of de-
velopment (Hardy 2007).
The fscal year 2012 budget calls for modernizing the IRS to improve custom-
er service and boost tax collections (2012 Federal Budget). Most of the added
money in the budget was allocated for revenue-generating tax enforcement initia-
tives. Much of the budget discussion in Washington has revolved around the tax
gap, the amount of money currently being paid and the amount that Congress and
the President believe people should be paying. The difference is the amount that
some people and frms might be underpaying on their taxes. Conservatively, the
president believes that better enforcement can generate $1.3 billion a year. Part of
the increased enforcement has been geared towards U.S. citizens with off-shore
accounts. The IRS has been pushing people to report all of their income from
other countries. In fscal year 2010, the IRS reported that it collected $57.60 bil-
lion through the enforcement division (www.irs.gov).
The IRS defned an IT Modernization plan in 2006 and 2007 (IRS 2007) that
presents a fve-year plan for improving service and enforcement. A key highlight
of the problems is that many systems are based on the Master Files which were
designed in the 1960s for slow computers. Most of the processing is still handled
through weekly updates to the master fles with programs and patches written over
the course of three decades. The lack of fexibility makes it diffcult to adapt to
the ever-changing tax codes. On the negative side, the 2007 document notes that
the BSM program proposed in 1999 could not be completed in the proposed 10-
15 years because of resource limitations. The new plan calls for incremental
releases and leveraging existing systems. Most of the plan simply defned busi-
ness domains and IT domainsessentially a restructuring of the IRS operations.
But, it crucially lacks any details of how the systems will be revised to support
those operations. The plan does list dozens of potential future projects within
each business domain. That is, projects that might be useful for compliance, but
have not yet been designed. Interestingly, within the technical domains, convert-
ing to IPv6 seems to be an important priority. Most of the document is a laundry
list of features and tasks that commercial organizations already perform.
Interestingly, the IRS Web site has links to its YouTube, Twitter, and Facebook
pages. Most of the information is geared towards professional tax professionals,
but the videos do contain basic information on tax related topics (IRS Web site/
New Media).
Questions
1. What problems have been experienced by the IRS in developing its
information systems?
2. How are these problems related to the services systems development
methodologies?
3. How is the IRS going to get more people to fle electronically? Is there an
upper limit?
4. Are there any ways to speed up the development of systems for the IRS?
What would be the costs and risks?
5. Are the IRSs problems the result of technology or management diffculties?
6. Why was the IRS unable to manage and control the CSC contract? What can
the managers do differently to get the projects fnished?
Additional Reading
Birnbaum, Jeffrey H., Unbelievable! The Mess At The IRS Is Worse Than You
Think, Fortune, April 13, 1998.
Cranmer, Jana, IRS Troubleshoots E-Help Program Government Computer
News, February 13, 2007.
Dorobek, Christopher J., IRS: E-File Goal Is A Stretch, Federal Computer
Dubie, Denise, IRS Touts Savings Project, Network World, April 15, 2003.
GAO, Tax Systems ModernizationManagement and Technical Weaknesses
must be Corrected if Modernization is to Succeed, GAO/AIMD-95-156,
July 1995.
Gaudin, Sharon, IRS Fails Security Audit, 490 Computers Missing in 3 Years,
Government Computer News, All 208 Million Tax Returns Get Electronic
Scrutiny On Arrival, March 6, 1995.
Government Computer News, Automation Failed To Meet Requirements, July
17, 1995, p. 6.
Hardy, Michael, IRS taps General Dynamics for Modernization Work,
Government Computer News, March 12, 2007.
Hasson, Judi, IRS Takes On Old-Style Paperwork, Federal Computer Week,
December 10, 2001.
IRS, Internal Revenue Service IT Modernization Vision & Strategy, October
2007.
Johnston, David Cay, Hunting Tax Cheats, I.R.S. Vows to Focus More Effort on
the Rich, The New York Times, September 13, 2002.
Keizer, Gregg, IRS Gives Away $318 Million Because Of Bungled Software
Upgrade, Information Week, September 5, 2006.
Masud, Sam, New Bells Are Ringing At IRS, Government Computer News,
October 16, 1995, pp. 44-45.
Mayer, Merry, Interim Systems Will Tide IRS Over As It Modernizes,
Government Computer News, January 25, 1999, p. 21.
McDougall, Paul, FAA Taps CSC to Ease Air Traffc Congestion, Information
Week, July 2, 2004.
McDougall, Paul, IRS Deep-Sixes Outsourcing Plans as Tax Season
Approaches, Information Week, November 30, 2006.
McNamee, Mike, A Kinder Gentler IRS? Business Week, February 1, 1999, p.
128.
Modernization, Government Computer News, January 25, 1999, p. 20.
Mosquera, Mary, IRS Financial Management Weaknesses Linger, Government
Computer News, April 30, 2004.
Mosquera, Mary, E-Filers Just Shy Of Half Of All Tax Filers, Government
Computer News, May 10, 2004.
Mosquera, Mary, IRS Will Put Business Rules Engine To The Test,
Government Computer News, May 17, 2004.
Mosquera, Mary, IRS Database Delay Affects Refunds, FCW, April 5, 2007.
Olsen, Florence, IRS Weighs E-Filing Upgrades, FCW, May 11, 2004.
Perez, Juan Carlos, IRS Commissioner Bars CSC from Upcoming Projects,
Quindlen, Terrey Hatcher. Recent IRS Award Puts TSM in Motion,
Government Computer News, December 23, 1991, p. 3-4.
Rosencrance, Linda, IRS Sends Out 523,000 Incorrect Refund Check Notices,
Rosencrance, Linda, Accenture To Redesign, Host IRS Web Site,
Rothfeder, Jeffrey, Invasion of Privacy, PC World, November 1995, pp. 152-
161.
Saunders, Laura, E-File or Else: Whats New for Tax Season, The Wall
Street Journal, March 19, 2011.
Saunders, Laura, Small Businesses Fight IRS Over Data, The Wall Street
Journal, May 26, 2011.
Smith, James M., IRS Spends $1b for Next Five Years Of Systems Support,
Government Computer News, July 17, 1995, p. 82.
Thibodeau, Patrick, Private Sector To Tackle IRS Mess, Computerworld,
January 4, 1999, p. 4.
Varon, Elana, For the IRS Theres No EZ Fix, CIO, April 1, 2004.
Weiss, Todd R., Peoplesoft Sells CRM Software to A Friendlier IRS,
Case: National Aeronautics and Space Administration (NASA)
NASA has experienced some serious problems over the past decade and a half.
Some have been due to software development errors, such as the 1999 Mars Po-
lar Lander that failed because a programmer made a calculation in miles instead
of meters. Systems as complex as the ones that NASA builds are bound to have
problems and most people are willing to tolerate the problems. But it turns out
that rocket science must be easy compared to accounting. NASA has an ongoing
problem with budgets and monitoring expenditures.
For years, NASA has routinely run over budget on projects. For years, the GAO
has criticized NASA for poor accounting practices. One of the problems has been
the decentralized nature of NASAwith offces scattered across the country and
around the world. The fnancial system relied on 10 separate systemssome still
written in decades-old COBOL. NASA knew it had to fx the fnancial reporting
system, so it contracted with SAP to install their R/3 system to create a new Inte-
grated Financial Management Program (IFMP). However, NASA has a target date
of 2007 for completing the conversion, and even that number seems optimistic
(GAO November 2003).
In 2004, the real nature of the mess was highly publicized by CFO Magazine.
The article was largely driven by a disclaimed opinion by PriceWaterhouseCoo-
pers (PwC), NASAs fnancial auditor. In that statement, PwC revealed that NA-
SAs bookkeeping was so bad, that they were unable to account for $565 billion
in year-end adjustments. That is billions, not millions, of dollars. Patrick Ciganer,
program executive offcer for integrated fnancial management at NASA, explains
that we had people in each of the [10] NASA centers who knew they had to make
the year-end adjustments. The problem was, they had never done them before.
They had been trained, but in some cases, that was six or eight months before,
and they did it wrong (Frieswick May 2004). The data was so bad that PwC had
no choice but to discredit the data and note that it could not be fxed. PwC later
declined to bid for the right to audit NASA in the following year.
In echoes reminiscent of Parmalats fnancial scandal, PwC also discovered that
NASA was $2 billion short in its cash account with the Treasury Department. NA-
SAs books claimed $2 billion that did not exist. After seven months, NASA CFO
Gwendolyn Brown was still unable to fnd the missing money. In testimony before
the House Subcommittee on Government Effciency and Financial Management,
she denied that the loss was the result of fraud, waste, or abuse, but she had no
clue about where the money went. Subcommittee chair Rep. Todd Platts (R-Pa.)
retorted, If my checkbook is off by 10 cents, Ill stay up all night until I fnd
that 10 cents. Your checkbook is off by $2 billion (Frieswick July 2004). Brown
noted that the agency was going back and reconciling balances from 2000 on-
ward. But it is unlikely that they will fnd the $2 billion. PwC found several other
problems involving hundreds of millions of dollars and failure by NASA to follow
its own procedures and those required by the government rules.
Decentralization is a huge source of NASAs problems. Each of the 10 centers
(including Kennedy and Marshall space centers and the Glenn research center) is
run independently, with separate fnancial systems and personnel. When Brown
arrived, she attempted to integrate the fnancial procedures, noting, Ive told them
that from now on, the agency will set policy, process, and procedure, and you, the
centers, will do implementation. If were going to be accountable and credible,
thats what we have to do here (Frieswick May 2004). One of the goals of the
SAP implementation is to embed all of the central rules and consolidate the data
so that all of the centers have to follow the same procedures. In particular, each
center will have to reconcile its spending to the Treasury balances every month.
NASA has been extremely optimistic about implementation of the new IFMP
system, or deliberately misleadingit can be hard to tell the difference. The GAO
commented that when NASA announced, in June 2003, that [the core fnancial]
module was fully operational at each of its 10 centers, about two-thirds of the
fnancial events or transaction types needed to carry out day-to-day operations
and produce external fnancial reports had not been implemented in the module
(GAO November 2003). Many of the features that NASA claimed to implement
had never been tested. The GAO was also concerned about NASAs lack of over-
sight of its contractors and its equipment. As of 2003, NASA planned to continue
using manual journal entries to handle all transactions with respect to its $37 bil-
lion in property (GAO November 2003). About $11 billion of that equipment is
located at contractor facilities, and NASA has almost no documentation or reports
on the property.
Along the same lines, NASA is unable to apply costs to specifc projects. In
particular, the agency is supposed to report all costs of the International Space
Station (ISS). In their 2003 report, NASA did not even list the ISSallegedly an
editorial oversight. The GAO report indicates that the NASA system is inca-
pable of correctly assigning costs to projects.
NASA has made several attempts to improve its accounting. The GAO notes
that NASA has made two efforts in the recent past to improve its fnancial man-
agement processes and systems but both of these efforts were eventually aban-
doned after spending a total of 12 years and a reported $180 million (GAO No-
vember 2003). NASA is already over budget and behind schedule with the new
IFMP system. The ERP system was originally budgeted at $982.7 million and is
already $121.8 million over budget (Dizard 2004).
By 2006, NASA was still trying to implement the SAP software. The $1.1 bil-
lion Integrated Enterprise Management Program included a $116 million fnan-
cials upgrade. But NASAs Offce of the Inspector General found several faws in
the management and implementation of the program. It also found end-user resis-
tance and data integration problems. End users have pointed out that the system is
unnecessarily complex, does not integrate well with some existing software, and
does not appear to be increasing productivity (Songini 2006).
The NASA discretionary expenditures for 2010 were $18.912 billion or about
$187 million over budget. The agency anticipated exceeding the 2011 budget
by about $500 million. But the fscal year 2012 proposed budget was reduced to
match the budget for 2010 or about $18.724 billion. Most of the budget was allo-
cated to replacing or modernizing existing buildings. Most of the space programs
had already been cut, and the president stressed cooperating with private agencies
and foreign nations (Russia) to provide lift capacity to space.
NASA is working to build a next generation space-based telescope. Although
complex projects are diffcult to forecast and to manage, an independent review
panel noted that mistakes by NASA management are likely to increase the costs
by $1.5 billion. The fnal costs are likely to be near $6.5 billion for a project that
can be launched in 2015 at the earliest (Harwood 2010). The main problem is that
the project cost estimates were too low because the initial budget did not fully un-
derstand the project, and the budgets never contained enough money to cover the
costs. Chris Scolese, an associate NASA administrator commented that Our main
goal right now is to strengthen the management, which were doing, to strengthen
the oversight, which were doing, and develop a good, strong estimate that we can
defend. We arent in the business of cost overruns. Were taking this very, very
seriously.
NASA IT has had problems with security. Of course, every hacker on the planet
dreams of breaking into NASA computers. But, fnancial systems are equally im-
portant, both to monitor for external attacks and to ensure security for internal
operations. In 2010, the Inspector General found that only 24 percent (7 or 29)
systems had met Financial Information Security Management Act (FISMA) stan-
dards for annual testing. Only 52 percent (15 of 29) met the requirements for an-
nual contingency plan testing. And only 2 of 5 of the external systems had been
certifed and accredited. The problems arose largely because NASA did not have
an independent verifcation function for IT securitya common requirement in
commercial sites. The CIO had purchased an information system in 2005 for $3
million to help with the security plan, but the implementation failed and the Agen-
cy is spending money for a replacement system.
Purchases
NASA purchases billions of dollars of supplies a year. Desktop computers and ac-
cessories are important items for most of the offces. In 1998, to simplify purchas-
es, NASA signed a nine-year $1.3 billion deal with seven companies to provide
server, desktop, communication equipment, and support services. The Outsourc-
ing Desktop Initiative for NASA (ODIN) required the vendors to maintain inven-
tory at each major site. The vendors created ODIN catalogs that listed all of the
parts available. Some NASA sites required workers to order only from the cata-
logs; others allowed workers to fnd different deals. Normally, you would expect a
centralized system to provide negotiated discounts and better prices. In this case,
NASA agencies were able to save thousands of dollars by not using the catalogs.
The Goddard Space Flight Center purchased 5,000 copies of antivirus software
without using the catalogs and saved $200,000. In 2003, the agency decided to tell
all divisions that they were not required to purchase from the catalogs (Cowley
2003). In 2006, NASAs basic budget was $14.5 billion (2008 U.S. Budget). In
2007, NASA awarded a new computer-purchase contract to HP. The contract cov-
ers desktops, Blade PCs, printers, and similar products and allows up to $5.6 bil-
lion in purchases. The contract stipulates that prices must be below the schedule
prices in the U.S. General Services Administration (Singer 2007).
Te Future
In terms of operations and projects, NASA is moving forward with new design
ideas. In particular, NASA is working with Carnegie Mellon University on a ma-
jor software dependability project. The goal is to create software that can toler-
ate hardware faults and security problems. These more intelligent systems would
help prevent problems like those that affected the Mars Polar Lander (Thibodeau
2003).
NASA is also investing in tools to improve collaboration among researchers.
Communication and information sharing have presented operational problems in
the past. If everyone has the same data, it is easier to spot problems, as well as
conduct new research. For the Mars rover mission, the Jet Propulsion Lab (JPL)
installed the DocuShare content-management system from Xerox. Thousands of
experts around the world have access to over 100,000 fles. The system can handle
up to 50 simultaneous projects and has search systems to help researchers fnd the
data they need (Chourey June 2004).
NASA is also installing a new high-speed network to support operations. The
system from Force10 Networks will support 10 gbps, making it possible to trans-
fer huge fles quickly. The data is carried on optical fbers. The high capacity is
needed to handle even the local satellite data. Every day, NASAs 14 satellites in
earth orbit transmit three terabytes of data (Chourey April 2004).
On June 20, 2004, NASA, along with the rest of the world, saw a new piece of
the future: SpaceShipOne, designed and built by Burt Rutan, carried a man into
space and back (Weil 2004). Partly to prove a point, partly to win the $10 mil-
lion Ansari X Prize in October, the ship did one thing extremely well. It pointed
out that it is possible to reach space on a budget of a paltry $30 million. It is not
likely that Rutans company, Scaled Composites, is going to compete with NASA
anytime soon. However, it does indicate that something must be seriously wrong
at NASAwhen a civilian company can get to space on a budget that would not
even be a round-off value in the money that NASA mysteriously loses each year.
The Space Shuttle program ended in the summer of 2011. NASA is part way
through designing a new launch vehicle, but funding and direction are uncertain.
Until a new system is developed, NASA will have to rely on Russian facilities
to deliver people to the international space station or for other purposes. Older
launch vehicles can still be used for satellite launches. Several private companies
have begun developing launch capabilities, but most are currently limited to low-
earth orbit. Given the federal budget crunch, and no agreement on long-term goals
for NASA, the future looks somewhat bleak. On the other hand, maybe it will give
NASA time to fx the computer systems.
Questions
1. How can an organization lose $2 billion?
2. Were the auditor problems ($565 billion) due to technology, management,
individuals, or some other reason?
3. Will the new ERP system be completed? If it is, will it solve NASAs
problems, particularly in terms of centralization?
4. How is NASA going to solve its problems with the development of mission
software? Why is this such a challenging problem?
5. How can Congress get NASA to provide accurate progress information, and
how can Congress determine if it is being told the truth?
Additional Reading
Chourey, Sarita, NASA to Move Data Faster With Bigger Pipes, FCW, April
12, 2004.
Chourey,Sarita, NASA Uses Xerox For Sharing, FCW, June 23, 2004.
Cowley, Stacy, NASA Wastes Money With Desktop Outsourcing Deal, Audit
Finds, Computerworld, August 1, 2003.
Dizard III, Wilson P., NASAs ERP Project Goes Awry, GAO Tells Panel,
Government Computer News, May 24, 2004.
Frieswick, Kris, NASA, We Have A Problem, CFO, May 1, 2004.
Frieswick, Kris, Canary Chorus, CFO, July 1, 2004.
GAO, Business Modernization: NASAs Integrated Financial Management
Program Does Not Fully Address Agencys External Reporting Issues,
November 2003, GAO-04-151, http://www.gao.gov/new.items/d04151.pdf.
Harwood, William, NASA Management Blamed for Space Telescope Cost
Overrun, CBS News TechTalk, November 11, 2010.
Offce of the Inspector General, Review of NASAs Management and Oversight of
Its Information Technology Security Program, September 2010, GAO.
Singer, Michael, HP Lands $5.6 Billion NASA Contract, Information Week,
May 23, 2007.
Songini, Marc, NASAs SAP Launch Drags, Computerworld, March 27, 2006.
Space Flight News, NASAs Finances In Disarray; Auditor Quits, May 16,
2004.
Thibodeau, Patrick, NASA Reinvents Troubled IT With Help of Private Sector,
Weil, Elizabeth, Rocketing Into History, Time, June 23, 2004.
adopted by rivals?
sectors?
860
How is an MIS department managed?
Is the MIS department doing a good job?
What roles and tasks does the MIS department perform?
What MIS jobs are available, and how much will it cost to hire IT employees?
Do you really need to run all of the MIS operations yourself?
Who should control IT resources?
How can Internet technologies be used internally to centralize data but still support
decentralized user access?
Why is the MIS department involved in so many conficts? How do you solve them?
Chapter Outline
Organizing MIS Resources
13
Chapter
Introduction, 861
Managing the Information Systems
Function, 862
Managing Workers, 863
Planning and Integration, 864
Evaluation, Oversight, and Control, 865
MIS Roles, 867
Hardware Administration, 867
Software Support, 868
Network Support, 869
Software Development, 869
Support for End-User Development, 869
Corporate Computing Standards, 870
Data and Database Administration, 871
Security, 872
Advocacy Role, 873
MIS Jobs, 873
Outsourcing, 875
MIS Organization: Centralization and
Decentralization, 880
Hardware, 884
Software and Data, 886
Personnel, 888
Recentralization with Cloud Computing,
889
Networks, 890
Hardware, 893
Data, 893
Confict Management, 894
Summary, 896
Key Words, 897
Exercises, 898
Cases: The Energy Industry, 903
861 Chapter 13: Organizing MIS Resources
ExxonMobile
Introduction
How is an MIS department managed? What tasks are performed by the
MIS department and how do these roles ft into the rest of the organization? What
problems and issues will you have to handle as a manager? As implied in Figure
13.1, the question of how MIS fts within an organization is diffcult, and the an-
swers have been changing along with the technologies. As a non-IS manager, you
will encounter many issues and decisions that are affected by the MIS structure.
You should learn to recognize common problems and possible solutions so that
you can minimize the effect of some of these problems. When you are evaluating
the CIO, MIS, and the overall structure, remember one rule: The job of MIS is to
help the organization and the other managers.
The capabilities of application software are impressive. Because of these tools,
business people using personal computers are solving problems in a few hours
that never would have been attempted ten years ago. With these powerful tools
available to the average business person, it is easy to wonder why a company
needs an MIS department. That is a good question, and the answers keep changing
through the years.
MIS departments provide many important services. At the most basic level,
the department is charged with maintaining the networks and computer servers.
With continual upgrades and evolving security threats, these tasks alone require
considerable time and money. The MIS group is also responsible for maintaining
the transaction processing systems including any ERP, accounting, and HRM sys-
tems. MIS workers also provide support for evaluating and buying new hardware
and software and assisting users.
According to statistics collected by several consulting frms and IT magazines,
large companies spend about 3-5 percent of their sales revenue on the MIS area.
(See: http://www.metrics2.com/blog/2006/06/26/average_company_
spends_34_of_revenue_on_it.html or http://uclue.com/?xq=1133) The
level of technology spending varies by industrywith banking and telecommuni-
How do you create and manage an information system when the organization has
offces and people around the nation or around the world? ExxonMobil and several
other companies faced this question with an additional twist. They had to combine
systems from two huge companies when they merged. Mobil also had to extend its
SpeedPass RFID payment technology to all of the new stations. That meant install-
ing new networks and extending the information system. Ultimately, the answer by
ExxonMobil was to combine the SAP systems used by both companies and run on a
single, centralized system.
ChevronTexaco took a similar approach when those two companies merged. The
new frm has relied on centralization to control technology and reduce costs.
Royal Dutch Petroleum (Shell) faces similar global issues. It has not been as ef-
fective at managing its information and fnancial systems. The company has experi-
enced diffculties dealing with multiple contract employment agencies. The compa-
nys stock also took a beating when the company had to restate its accounts because
it had misstated its oil reserves.
cations frms at the top. For a company with a billion dollars in sales, that amounts
to $50 million a year spent on MIS. This money pays for personal computers,
central computers, communications, software, and MIS personnel to manage it
all. The primary tasks undertaken by the MIS department are software develop-
ment, setting corporate computing standards, hardware administration, database
administration, advocacy and planning, and end-user support.
Small businesses rarely have a separate MIS department. That does not mean
these duties are ignored. Even in small businesses, someone has to be responsible
for these MIS functions. However, small businesses generally do not attempt to
develop their own software. Even relying on commercial software requires that
time be spent on determining data needs and evaluating software packages.
Probably the most important MIS decision facing business today is the issue
of centralization. Because personal computers have a huge price/performance ad-
vantage over larger computers, there is a major incentive to decentralize the hard-
ware. Yet there are some serious complications with complete decentralization.
Several strategies for organizing information resources provide the advantages of
both centralization and decentralization. The management goal is to fnd the com-
bination that works best for each situation. Before examining the alternatives, you
need to understand the basic MIS roles.
Managing the Information Systems Function
Is the MIS department doing a good job? Should the company be
spending more money on MIS? Is it getting a good value for its current spending?
Are there other methods that would be more effcient or save money? Many times
in your career you will fnd yourself heavily involved with members of the MIS
department. In the case of a small business, you might be in charge of the one or
two MIS personnel. At some time, you might be the company liaison to an out-
Figure 13.1
Organizing information system resources. Making effective use of information
systems requires organizing the MIS resources: hardware, software, data, and
personnel. A key decision involves positioning the resources in the organization
which revolves around decentralization versus centralization. The goal is to balance
the need for central control with the value of decentralized decisions.
sourcing vendor, MIS contractor, or consultant. In all of these situations, you will
be responsible for planning, monitoring, and evaluating the MIS organization.
Management of information systems begins by understanding the roles of MIS.
The MIS function is responsible for hardware and software acquisition and sup-
port. The MIS staff provide access to corporate data and build applications. They
support end-user development with training and help desks. MIS workers set cor-
porate data standards and maintain the integrity of the company databases. Some
MIS organizations spend considerable resources developing and maintaining sys-
tems. All of these functions have to be organized, performed, and evaluated on a
regular basis.
Managing Workers
Hiring, evaluating, and promoting workers are key issues in managing an IT de-
partment. MIS employees are relatively scarce and their salaries are often higher
than other disciplines. Consequently, it is important to hire workers who match
your needs. Because technology continually changes, MIS workers need educa-
tion and training programs. Some companies are good at recognizing talented
workers and keeping them happy. Several business publications annually list the
best companies, and best IT companies to work for. Reading these lists and the
description of benefts will give you some interesting ideas on how to attract and
keep the best workers. Sometimes benefts are expensive, other times, simple rec-
ognition of employee contributions is enough to improve a workplace. These is-
sues are covered in more detail in HRM courses.
Trends
In the early days, computers created few management issues. The large,
expensive machines were placed in a central location and serviced by a cen-
tralized group of specialized employees. But as costs dropped over time,
hardware spread throughout the organization. Soon, employees were collect-
ing and creating data on hundreds or thousands of machines across the com-
pany. The local area networks of the late 1980s were installed in an effort to
make it easier to share data and provide more centralized services. The ini-
tial spread of PCs and networks supported decentralization, where workers
became responsible for handling their own data and computers.
Integrated packages (ERP), Web sites, and groupware tools make it easier
to share data and support team work. These tools required corporatewide
standards and began to encourage a recentralization of MIS resources. As
companies periodically focus on cost control, the trade-offs between decen-
tralization and centralization become a source of contention.
The increased attacks and renewed interest in security in the early 2000s
generated additional interest in the management of information resources.
Attempts to increase security led to even more demands for centralized con-
trol over resources. Centralized control and decentralized management are
not necessarily good or bad. They both have strengths and weaknesses. The
key is to fnd the appropriate balance for each organization. The challenge
is that changes in technology and external events (such as security threats)
sometimes cause organizations to leap to one solution without thinking
about the consequences. Interesting arguments often follow.
Hiring workers with specifc skills is always challenging because the desired
skills change. Read a few job ads and you will see a jumble of acronyms. Some
companies insist on trying to hire workers with specifc experience and skills that
ft into a narrow defnitionor they list every possible current acronym they can
think of. Other companies search for bright, motivated workers who can solve
problems and learn new technologies. Regardless of your approach, it is diffcult
to determine the exact skills of any potential employee. IT certifcations exist
for a variety of skill sets. Potential employees obtain these certifcates by taking
examsusually created and monitored by vendors. For example, Microsoft has
several levels of exams ranging from Offce tools to database administration and
software development using Microsoft technologies. The networking giant, Cisco,
has several levels of exams for certifying that people meet basic skills in design-
ing networks and confguring Cisco routers and switches. Companies can also ask
potential employees to take independent examssimilar to fnal exams in college
classes such as programming in C++. The community of workers and employers
continually debates the value of these certifcations. If you only need to ensure
that someone has a basic skill in a particular technology, the certifcations and
exams can help verify that the person meets minimum standards. But certifca-
tions are not replacements for experience. And keep in mind that all technologies
change. You need workers who are willing and able to learn new technologies,
and you need to provide an environment that provides the tools and training to
keep everyone up to date.
IT workers can also be hired on a contract basisto solve specifc short-term
problems. In some cases, these workers might be located in other countries and
found via outsourcing Web sites. Managing contract workers is critical. You need
to be able to communicate effectively and defne precisely what tasks need to be
accomplished.
Planning and Integration
One key issue in managing information technology is organizing the MIS func-
tion so that it matches the structure of the frm. Centralization versus decentral-
ization has been a key issue in the organization of MIS resources. Networks and
powerful personal computers have led to more options supporting decentraliza-
tion of information. The increased options are useful, but they create more issues
that managers must examine. Many arguments arise in organizations over how to
Reality Bytes: The Changing Role of MIS
The role of the MIS department has changed over time. In many respects, it is in the
middle of a fundamental change. In the past, MIS departments focused on creating
information systems and controlling dataparticularly transaction data. Today, as
explained by the Gartner Group (an IS consulting frm), the objectives of MIS are:
Provide transparent access to corporate data.
Optimize access to data stored on multiple platforms for many groups of users.
Maximize the end-users ability to be self-suffcient in meeting individual in-
formation needs.
These changes represent a shift in attitude. It moves toward the goal of increasing
support for workers, not their replacement, so employees can do their jobs better on
their own.
integrate technology. Most organizations have an existing structure, and the IT
support needs to match that structure. Forcing central IT decisions on a decentral-
ized organization might appear to save money, but it can seriously impede worker
productivity. When business workers waste time fghting with central decisions
or not getting the data they need, the business can quickly burn off any potential
savings.
The issue of new technology points out the importance of planning. The only
way to control costs and evaluate MIS benefts is to establish a plan. Plans need
to be detailed so actual results can be compared to the plan. Yet plans need to be
fexible enough to adapt to unexpected events and new technology. You also need
to formulate contingency plans for events that might occur.
Evaluation, Oversight, and Control
MIS is a service organization, which means that technology contributes through
helping other workers do their jobs better. One way to evaluate an MIS organiza-
tion is to ask business users to evaluate the level of support. Several survey instru-
ments have been created to help assess user satisfaction with information systems.
Presumably, if the IT department is doing a good job supporting the users, they
Reality Bytes: KLM Royal Dutch Airlines
Many large companies typically buy PC hardware and software for employees. More
specifcally, the IT department chooses a brand of computer, loads the software, and
gives the PC to the employee to use. Companies that provide cell phones to employ-
ees use the same processchoose one common phone model and give it to every-
one. Rarely does the employee have any options or choice. At KLM Royal Dutch
Airlines, a subsidiary of Air France-KLM SA, employees were unhappy with this
process. In November 2006, Martien van Deth, a senior technology offcer in Am-
sterdam, tried a new approach with 50 IT staff employees. He gave each person a
budget allowance of $203 to purchase a company cell phone for two years. Anyone
wishing to buy a more expensive phone paid the difference in cost, anyone picking
a cheaper phone could pocket the difference. The phones had to run Microsoft Win-
dows Mobile version 5 or 6, and users had to deal with technical support themselves.
If the phones broke, they had to replace it. Employees loved the new plan, and it was
cheaper than the $231 the company typically paid for phones and support. Mr. van
Deth decided to roll out the plan to other employees and expand it to cover laptops.
Eli Lilly is considering going even further. Adrian Seccombe, chief security offcer,
suggested that his company might stop providing PCs to employees. Similar to the
way the company reimburses employees for using their personal cars for travel, the
company will reimburse employees for Internet costs. Other companies are reluctant
to let employees make decisions about technology because of potential security is-
sues and having a huge mixture of products is harder to service. On the other hand,
many corporate applications are based on Web browsers, and employees can use
them with devices ranging from desktops to cell phones. With browser technology,
BP PLC was able to start an employee allowance program, giving them $1,000 a
year to choose and maintain their own computers.
Adapted from Ben Worthen, Offce Techs Next Step: Do It Yourself, The Wall
will be happy with the support. Of course, it is also possible that users will never
be happy.
Part of the problem comes down to economics. If workers look on IT as a mag-
ic genie, they will never be happy until they have the latest toys. But, someone has
to pay for all of the technology. Is Joe in accounting complaining about MIS be-
cause he needs better data or because he wants a new Web-based cell phone? Eco-
nomics explains that when users do not pay for items, they are unlikely to choose
a cost-effcient solution. This problem was more acute when technology was ex-
pensive. Many companies instituted chargeback schemes to solve the problem.
Chargeback or transfer pricing means that the MIS department bills for its ser-
vicesin terms of hardware, software, and personnel. Many different chargeback
methods were developed; including fxed monthly fees, project-based fees, and
fees for individual services, such as printing or consulting. Several large organi-
zations still use chargeback methodsparticularly for large infrastructure items
such as network support. The basic goal of transfer pricing is to obtain economi-
cally effcient solutions by billing departments for their usage of technology. In a
perfect scheme, departments will use the prices to buy exactly as much IT services
as they need to improve their proftability. A side beneft to chargeback methods is
that user departments potentially gain some control over IT. Because they are pay-
ing for service, they can choose to stop buying IT services or go elsewhere if the
costs are too high or the service quality is low.
Figure 13.2
MIS roles. The MIS department is responsible for hardware administration, software
development, and training and support. MIS staff manage networks, provide
computer security, develop Internet and e-business applications, and establish
corporate computing standards The MIS department also plays an advocacy role,
presenting the IT benefts and strategies to the executive offce.
Hardware
administration
and operations
Software
development
Support end user
development
Database
administration
Advocacy
Network
management
Software training
and support
Internet and
e-business
Security
The catch to any transfer price is that it is extremely diffcult to fnd the true
price for the services. When the same service can be purchased externally, a mar-
ket economy can drive the price to effcient levels and it can be used as a compari-
son. Historically, internal chargeback prices tended to be relatively highwith
no incentive for the MIS department to behave effciently. With the increasing
availability of outsourcing, it is now possible to obtain more accurate prices for
many types of services. But many organizations fail to search for and use these
numbers, so reports of excessive overcharging still exist.
Ultimately, someone needs to help the IT organization understand the business
needs of the users. In smaller companies, these discussions can be handled at the
executive level. In larger companies, it is more effcient to establish an oversight
group that is run by the business departments. The oversight group can coordinate
plans, share information, and discuss conficts and costs.
MIS Roles
What roles and tasks does the MIS department perform? Good
information systems do not simply materialize from thin air. Providing timely, ac-
curate, and relevant information requires careful planning and support. Creating
effective information involves maintaining hardware, providing software training
and support, supporting end-user development, defning and controlling access to
databases, establishing corporate standards, and researching the competitive ad-
vantages of new technologies. These basic roles of MIS are outlined in Figure
13.2.
Hardware Administration
In some respects, hardware administration has become a little easier in the last
few years. Computers are more reliable, more standardized, and cheaper. It used
to be diffcult to accommodate users who needed slightly different hardware.
Many companies still prefer to standardize PC hardware to simplify purchasing
and asset tracking. But with common three-year support contracts, standardized
hardware, and low prices, it is less and less important to require that all employ-
ees have the exact same computer. Instead, most companies need to choose a
time frame for updating their computers. For most purposes, three years has been
shown to be a productive length of time to hold a PC. On the other hand, some
companies require groups of employees to have exactly the same hardware. MIS
can then make a standard copy of the confguration and software and reload this
image if anything goes wrong with a users computer.
Mobility has become an increasingly important issue for many business com-
puter users. More managers are requesting the use of laptops instead of (or along
with) desktops. Mobile systems including Internet-accessible cell phones change
the picture even more. Fortunately, the cost of most of these devices has dropped
as well. Consequently, purchasing is less of an issue. Instead, mobile systems pres-
ent more problems in terms of securitybeginning with loss of the equipment.
Purchase costs do not represent the entire cost of hardware. In the mid-1990s,
a few companies began pushing a concept they called total cost of ownership
(TCO). The basic issue is that someone has to confgure the PC, install software,
and troubleshoot problems. At the time, most of these tasks required an on-site
MIS support person, which was expensive. Various attempts were made to esti-
mate these additional support costs and derive the TCO. The slightly hidden ob-
jective was to show that centralized computers were not really more expensive
than personal computers. While most of the numbers from this process were un-
reliable, the process did highlight the diffculties of maintaining personal comput-
ers. Consequently, several frms created tools to install software and troubleshoot
PCs from a central location over the LAN.
The move toward Internet-based applications removes many of the issues of
managing personal computer hardware. As long as the systems support the Inter-
net standards, and as long as they can be purchased inexpensively, there is little
need for detailed MIS oversight. On the other hand, the central servers that run
the Web sites, databases, and applications become more important. MIS personnel
need to monitor the performance of the servers, provide backup plans, maintain
security, and plan for capacity increases.
Capacity planning is a major factor in MIS organizations. Building scalable
systems is an important goal for most organizations. The objective is to purchase
only the level of hardware that is needed. Then, as demand increases, add more
servers or larger servers to handle the increased load. This process holds down
central hardware costs, but it means that MIS has to carefully design the systems
and carefully monitor the usage and predict future demands.
Software Support
Software generally requires more support than hardware does. MIS staff can
help users decide which software to purchase and then install it. Users need to be
trained to use various software features. Whenever workers change jobs or a com-
pany hires new workers, they need to be trained. Similarly, commercial software
versions change almost every year, requiring more training for users. In addition,
someone has to install the new copies on the machines, distribute manuals, and
convert data fles.
When users have diffculty getting the computer to do what they want, it saves
time to have someone to call for help. Most MIS departments offer training and
some level of user supportparticularly for complex systems such as the ERP or
accounting systems. However, as funding for IT get cut to reduce costs, compa-
nies increasingly expect users to answer their own questions. Technical support
often gets reduced to solving basic hardware problems. In most cases, it is easiest
to fnd a co-worker with the answer.
Reality Bytes: The Changing Role of MIS
The role of the MIS department has changed over time. In many respects, it is in the
middle of a fundamental change. In the past, MIS departments focused on creating
information systems and controlling dataparticularly transaction data. Today, as
explained by the Gartner Group (an IS consulting frm), the objectives of MIS are:
Provide transparent access to corporate data.
Optimize access to data stored on multiple platforms for many groups of users.
Maximize the end-users ability to be self-suffcient in meeting individual in-
formation needs.
These changes represent a shift in attitude. It moves toward the goal of increasing
support for workers, not their replacement, so employees can do their jobs better on
their own.
Network Support
Both wired and wireless networks are critical to running a company. Managers
rely on the networks being available 24 hours a day, 7 days a week. Fortunately,
most modern network equipment is reliable and can run continuously for months.
Moreover, network managers can build networks that can correct for failures and
bottlenecks by routing traffc around a switch that has failed.
Network support becomes more complicated when managing Internet con-
nections. Connecting to the Internet requires some specialized skills to confgure
the router. Once the connection is established, confguring the router for secu-
rity becomes critical. Monitoring the connection and keeping up with current se-
curity advisories are even more challenging. Larger companies train specialized
personnel to handle these tasks. Smaller companies often rely on consultants or
contractors.
Wireless networks are relatively easy to installalthough corporate hardware
is often more complex and more powerful than the wireless router you install in
your home. And they are much more diffcult to confgure if you are concerned
about security. Several true stories exist of people sitting in parking lots or even
passing commuter trains and establishing a wireless connection into a companys
network. Network specialists are trained to confgure networks to minimize these
problems.
Software Development
Developing software and business applications is diffcult. Projects can require
teams of hundreds of developers. Even smaller projects and purchases of larger
software applications require devoted attention to detail. Managing software de-
velopment and purchases is a critical role in MIS. Beginning with project evalua-
tion and feasibility studies, through project management and tracking progress, to
evaluating the team efforts, someone has to be in charge of the details.
These tasks are some of the more traditional roles in MIS. They are also the
most diffcult. Unfortunately, most companies have poor track records in manag-
ing software projects. Many development projects fail, and even more are behind
schedule and over budget. Consequently, many frms choose to buy software or
use outside companies to help in development projects. Even in these cases, some-
one in MIS has to be responsible for evaluating choices and monitoring progress.
Support for End-User Development
Many application packages include programming capabilities. For example, a
manager may create a spreadsheet to calculate sales commissions. Each week,
new sales data is entered and the spreadsheet automatically produces summary
reports. It would be better to have a clerk rather than a manager enter this new
data. To make the clerks job easier, the manager uses the macro capabilities in the
spreadsheet to create a set of menus and help messages. Similarly, using a word
processors macro facilities, a legal department can create standard paragraphs for
various contracts. With them, an assistant can type one word to display a prewrit-
ten warranty paragraph. In theory, even complex applications traditionally pro-
vided by the MIS department, such as accounting systems, could be programmed
by end users with prepackaged software.
Several problems can arise from these end-user applications. Techniques that
are acceptable for small projects may lead to errors and delays with large systems.
Programming major applications requires obtaining information from users and
managers. Applications designed for corporate use require extensive checking of
data and security provisions to ensure accuracy. The software often needs to run
on different operating systems and local networks.
The MIS department can provide assistance to minimize these problems. MIS
personnel can assist end users in collecting ideas from other users. They can also
help in testing the applications to verify the accuracy and make sure the software
works with other applications. MIS can provide tools and help end users document
their applications and move them to new operating systems or new hardware. Pro-
grammers can write special routines to overcome any limitations of commercial
software that might arise. MIS staffs also maintain help desks or information cen-
ters to answer user questions and help users debug applications.
Corporate Computing Standards
Over time, MIS has learned that the frm gets into trouble if all of its people work
independently. In the 1960s, applications such as payroll, accounting, and custom-
er order processing were developed independently. During the 1970s, companies
had to spend large amounts of money getting all of the pieces to work together. In
the 1980s, personal computers arrived, and the problems got worse.
Reacting to the problems created by these incompatibilities, MIS professionals
at different companies developed standards. If all vendors used standard formats
for fles, hardware connections, and commands, products from different vendors
could be used together. Today, there are standards for everything: data, hardware,
software, report layouts, and coffee pots.
It is unlikely that the computing world will ever see complete cooperation
among vendors. Three factors prevent products from working together. First, stan-
dards are often ambiguous or incomplete. Human languages always have some
ambiguity, and there is no way to determine whether the description actually cov-
ers every possible situation. A second problem is that standards incorporate what
is known about a topic at the time the standard is developed. Computing technolo-
gies change rapidly. Often, vendors can produce better products by not following
the standards. Then new standards have to be developed. A third problem occurs
because vendors want to distinguish their products from the offerings of com-
Reality Bytes: Google as a Software Developer
Most people in the world use Google software. But few people think about how that
software gets developed, or the number of people it takes to create and improve the
tools you use on a daily basis. In 2011 alone, Google planned to hire more people
than the 6,000 it hired in 2007, its previous record year. By the end of 2011, the com-
pany expects to have over 24,000 employees. Not all of the employees are develop-
ersincreasingly, the frm hires salespeople to work with online and local compa-
nies. But Google continually works to develop new tools and new conceptsinclud-
ing software for self-driving cars. One aspect to software development at Google is
that the company emphasizes small teams of software engineerswith an average
of 3.5 developers on a team. The objective is to encourage communication and get
developers to treat projects as start-ups.
Adapted from Amir Efrati, Google Steps Up Hiring, The Wall Street Journal, Jan-
uary 25, 2011.
petitors. If there were standards that perfectly described a product, there would
be only one way to compete: price. Many vendors fnd that it is more proftable to
offer features beyond what is specifed in the standards, enabling the developers to
charge a higher price.
Even though it is not possible to create perfect industry standards, there are
advantages to creating companywide standards. They enable frms to buy prod-
ucts at lower prices. Most large businesses have special purchase agreements
with hardware and software vendors. Buying in bulk allows them to obtain better
prices. Similarly, it is easier to fx hardware if all the machines are the same. Like-
wise, it is much more convenient to upgrade just one word processing package for
200 computers, instead of 20 different brands. Similarly, training is less expensive
and time consuming if everyone uses the same software and hardware. Finally,
standards make it easier for employees to share information across the company.
The Internet and e-mail create additional demand for standards. To share a fle on
the Internet, you must store it in a standard format (e.g., HTML or PDF). People
sometimes forget that a similar problem arises when attaching fles to e-mail mes-
sages. Particularly when you send a fle to someone in a different company, you
need to remember that the recipient may not have the same version of software
that you are using. Because Microsoft has over 90 percent of the market for basic
offce software, fle sharing had not been an issue for years. However, Microsoft
changed all of the fle formats with the release of Offce 2007. File converters en-
able users to share fles from the newer and older versions, but the conversions are
not perfect and some formatting and layout is usually lost. Since some companies
upgrade before others, it is generally safer to save attached fles in either a stan-
dard format (HTML or PDF) or a previous version.
Some organizations forget that standards cannot be permanent. Hardware and
software change almost continuously; new products arrive that are substantially
better than existing standard items. Similarly, as the business changes, the stan-
dards often have to be revised. Also, there are exceptions to just about any policy.
If one department cannot do its job with the standard equipment, MIS must make
an exception and then fgure out how to support this new equipment and software.
Data and Database Administration
Databases are crucial to the operation of any company. Keeping the databases up-
to-date and accurate does not happen by chance. Larger organizations employ a
database administrator (DBA) to be responsible for maintaining the corporate
databases, monitoring performance of the database management system, and solv-
ing day-to-day problems that arise with the databases.
Companies also need someone to coordinate the defnition of the data. Large
organizations might hire a separate data administrator (DA); smaller companies
will pass this role to the DBA. The DA is responsible for defning the structure of
the databases. The DA has to make certain the data is stored in a form that can be
used by all the applications and users. He or she is responsible for avoiding du-
plicate terms (e.g., customer instead of client). Furthermore, the DA provides for
data integrity, which means that the data must contain as few errors as possible.
The DA is also required to maintain security controls on the database. The DA
has to determine who has access to each part of the data and specify what changes
users are allowed to make. Along the same lines, companies and employees are
required by law to meet certain privacy requirements. For instance, banks are not
allowed to release data about customers except in certain cases. European nations
have much stricter privacy rules. If a frm operates a computer facility in many
European countries, the company must carefully control access to all customer
data. Some nations prohibit the transfer of customer data out of the country. The
DA is responsible for understanding these laws and making sure the company
complies with them.
Finally, because todays databases are so crucial to the company, the business
needs a carefully defned disaster and recovery policy. Typically that means the
databases have to be backed up every day. Sometimes, a company might keep
continuous backup copies of critical data on separate disk drives at all times. MIS
has to plan for things that might go wrong (fres, viruses, foods, hackers, etc.).
If something does affect the data or the computer system, MIS is responsible for
restoring operations. For instance, an alternate computing site might be needed
while the original facilities are being repaired. All of this planning requires con-
siderable time.
Security
Since most of todays business data is stored in computers, computer security has
become a critical role for the MIS department. Often this role is shared with the
accounting department to establish standards and procedures to ensure the integ-
rity of fnancial data. Medium and large organizations have full-time computer se-
curity offcers to set policies, establish controls, and monitor systems for attacks.
Because of the constant evolution of new threats and the large number of systems
and employees, the task can be immense. Attackers often search for one little hole
in one system or one mistake by an employee. Security managers have to keep up
with hundreds of different systems and applications to make sure that all of the
holes are plugged. But, because security systems are never perfect, security ad-
ministrators must also run monitoring tools to watch for and stop ongoing attacks.
Security administration also includes training users, testing system confgu-
rations, and monitoring networks for ongoing attacks. Establishing incident re-
sponse plans and teams is also a major task. When things go wrong, you need
a team and a plan to identify the problem, stop the attack, and restore business
functions as quickly as possible. Along the way, you have to collect and maintain
Reality Bytes: Business Trends: Specialization
Just as in other areas of business, MIS jobs have become highly specialized. For in-
stance, many advertisements for MIS jobs look like someone spilled a bowl of alpha-
bet soup on the page. Companies often search for technical skills involving specifc
hardware and software.
Unfortunately, this approach to jobs causes problems for MIS personnel. In order to
fnd other jobs or to advance in their current position, they have to acquire increas-
ingly detailed knowledge of specifc hardware and software packages. Yet with rapid
changes in the industry, this knowledge can become obsolete in a year or two. These
changes mean employees have to continually expand their knowledge and identify
software and hardware approaches that are likely to succeed.
On the other hand, businesses need to keep their current applications running. With
thousands of hours invested in current systems, companies cannot afford to discard
their current practices and adopt every new hardware and software system that shows
some promise.
evidence that can be used in court cases, so that if you catch the attackers, you can
fle charges against them.
Advocacy Role
The MIS department is headed by a single manager, who often is called the chief
information offcer (CIO). The CIO position might be a vice president or one level
below that. A major portion of this job involves searching for ways in which the
information system can help the company. In particular, the CIO searches for and
advocates strategic uses of MIS. The goal is to use the computer in some way that
attracts customers to provide an advantage over the companys competitors.
The MIS goal is to help the organization and the other managers. But most
business managers are not experts in technology. Whenever a new technology is
introduced, someone has to be responsible for deciding whether it will be worth
the expense to make a change. If there is no one in this advocacy role who evalu-
ates the existing systems and compares them to new products, an organization is
probably not often going to get new equipment. Even when many users are dis-
satisfed with an existing system, they will have a better chance of acquiring new
technology if they can voice their complaints through one highly placed person.
Along these lines, the CIO is responsible for long-run planning in terms of infor-
mation technology.
MIS Jobs
What MIS jobs are available, and how much will it cost to hire
IT employees? A wide variety of jobs are available in MIS. Some of the jobs
require a technical education, such as that for programmers. Specialized positions
are available in data communications and database management. On the other
hand, systems analysts require an extensive knowledge of business problems and
solutions. Some entry-level operator jobs require only minimal training. On the
other end of the scale, analysts may eventually become team leaders or managers.
The entire MIS function is coordinated by chief information offcers.
As you might expect, salaries depend on experience, individual qualifcations,
industry, location, and current economic conditions. Seven basic MIS job tracks
are shown in Figure 13.3: systems development, networks, database, security, user
support, operations, and other specialists. Systems development includes several
levels of analysts and programmers. Network management involves installing net-
work hardware and software, diagnosing problems, and designing new networks.
Database management focuses on database design and administration. End-user
support consists of training users, answering questions, and installing software.
Security tasks include confguring systems, working with programmers, and mon-
itoring ongoing operations. Operations consist of day-to-day tasks such as loading
paper, mounting tapes, and starting long computer tasks. Many of these tasks are
being automated. Entry-level operator jobs do not require a college degree, but
there is little room for advancement without a degree. Specialist positions exist
in larger companies and generally evolve from new technologies. For example,
Web masters who would create and manage Web sites were in high demand for
two or three years; then as the Internet became more important to companies, Web
workers became even more specialized including design, e-commerce, and EDI
specialists.
Every year, Computerworld surveys workers in the industry and publishes
average salaries. Job placement frms such as Robert Half also collect data on
salaries. The Wall Street Journal and several other companies have Web sites
that provide salary information for various jobs and locations. This data can be
useful if you are searching for a job or thinking about a career in MIS. As a busi-
ness manager, the numbers will give you an indication of the costs entailed in
building and maintaining information systems. Basic averages are listed in Figure
13.3.
One way to see the changes occurring in MIS is to look at the types of skills
that businesses are looking for in MIS applicants. Figure 13.4 shows some of the
top skills demanded in 1994, 1998, 2001, and 2004. Notice the demand for CO-
BOL to fx date problems with old software. Then demand shifted to new tech-
nologies (ERP and groupware). In 2004, demand for applications development
and management increased, along with a renewed interest in security.
Figure 13.5 shows another perspective on the costs of programmers by compar-
ing costs across nations. With the worldwide expansion of the Web, it is possible
to hire programmers from almost any area of the world. Economically, the U.S.
has long had a relative shortage of labor. Programming requires considerable edu-
cation and experience, but nations such as India and China have huge populations
and the ability to educate and train programmers. Russia and other Eastern Euro-
Figure 13.3
IS salaries. As in any feld, salaries depend on experience. However, in IS they also
depend heavily on technical skills. Programmer/analysts with current skills and
experience in new technologies fnd it easier to get jobs and obtain higher salaries.
Note that there is a wide variety of jobs in IS, each requiring different types of skills.
Systems Development IS Management
Director
Project manager
System analyst
Senior Developer
Programmer/analyst
Junior programmer
$175,000
102,300
79,200
92,300
77,000
56,800
CIO/VP IS/CTO $191,000
Internet Networks
Director/strategy
Manager
Applicationdevelop.
EC specialist
EDI specialist
$156,400
88,500
66,000
72,700
68,000
Manager
Administrator
Network engineer
Junior analyst
$90,500
64,100
77,300
44,000
Database Security
Manager
Architect
DBA
Analyst
$10,800
117,000
89,300
75,800
Chief security
Manager
Specialist
ISauditmanager.
IS audit staf
$162,000
99,000
89,000
109,000
70,400
Operations User Support
Director
Manager
Systemsadmin.
Lead operator
Computer operator
$104,600
80,000
70,300
53,800
37,400
Manager
Technical trainer
Help desk operator
PC technical support
$75,300
64,600
56,200
51,600
pean nations have had historically high unemployment rates, and educational sys-
tems that at least in the past focused on mathematics and science. Consequently,
at various points in time, salaries for workers in these countries have been lower
those in Western countries.
Of course, employees in countries such as India also have lower costs of living,
so the salary is relatively high. In terms of hiring programmers from one nation
instead of another, it is also important to look at productivity rates. Although it is
diffcult to measure, programmers in the U.S. have some of the highest productiv-
ity rates in the world. Economic theory shows that ultimately wages relative to
productivity will equalize across nations. Some of this effect can be seen in the
increasing wages for programmers in India.
Outsourcing
Do you really need to run all of the MIS operations yourself?
In the past 20 years, many businesses have noted that it has become diffcult to
terminate or lay off employees. In MIS, it has also been expensive for frms to
hire the best people. Consequently, many frms have chosen to outsource various
aspects of their MIS functions. The basic premise is that specialized frms can
offer more effcient service at better prices. For example, EDS runs huge data
centers, and it is relatively easy to add more clients with only minor increases in
costs. As a huge MIS organization, EDS also hires and trains thousands of work-
Figure 13.4
MIS skills in demand. At any given time, some skills are in demandrefecting
demand for applications and a shortage of workers for new technologies. Other skills
are also in demand, but workers with the listed skills generally received premium
wages and bonuses.
Year/
Rank
1 2 3 4 5
2010 Java/J2EE Security Software
Developer
SAP Database
Management
2007 Business
Analyst
Windows
Administrator
Program
Manager
Ofshore
Project
Manager
Vendor
Manager
2004 Application
Development
Project
Management
Database
Management
Networking Security
2001 ERP Object
Engineering
Data
Warehouse and
Visualization
Groupware Wireless
1998 ERP Groupware Database Networking COBOL
Cooney, February 25, 2011, IT Graduates not Well-Trained, Ready-to-Go, Net-
workworld.
Ware, January 2004 IT Staffng Update, CIO Magazine, February 3, 2004; Comput-
erworld, In Demand: IT starts require premium pay, December 10, 2001; Comput-
erworld, November 16, 1998; and Arnett and Litecky, Journal of Systems Manage-
ment, February 1994, www.cio.com/article/101314, 2007.
ers. Outsourcing also is attractive to frms as a temporary measure. For example,
frms might outsource their old accounting systems while designing and installing
a new ERP system. The old system will continue to function and be ably sup-
ported by an expert company. The internal employees can focus on designing and
installing the new system.
Outsourcing can take many forms. Firms might sell their entire computer
center to an outsource specialistand all of the data, software, and employees
would move to the new company. Other frms might contract out other MIS func-
tions such as network management, PC repair, training, security, or development.
Some functions, particularly programming, can even be outsourced to companies
based in other countries. India has several companies, led by Tata, that specialize
in writing programs for American and European frms. Using these frms in other
countries to handle MIS (and other tasks) is known as offshoring.
Two of the leading service providers are Electronic Data Systems (now
owned by Hewlett-Packard) and Global Services, the IBM subsidiary. Some
other leading outsourcing companies are listed in Figure 13.6. Note the continuing
growth in outsourcing. Initially, this trend was partly due to the desire to cut costs,
the inability to hire IT workers, the increasing standardization of IT services, and
the need to focus on core business management. In 1998, Computerworld re-
ported that an average of 20 percent of IS budgets was spent on outsourcing. Gen-
erally, a company signs an agreement to use the services of the outsourcing frm
for a fxed number of years. Depending on the agreement, the outsourcing frm
can be responsible for anything from machine operation and maintenance, to de-
velopment of new systems, to telecommunication services.
Figure 13.7 shows the total value of the revenue for the main outsourcing and
consulting frms. Even without the revenue from the smaller frms, the trend is
clear. Business organizations are increasingly relying on these outside specialists
to develop software and handle other MIS tasks. Throw in software-as-a-service
where companies can lease Web-based software, and it is clear that business orga-
Figure 13.5
Internationalization. In the past few years, U.S. and European frms have turned to
using programmers in other nations. For example, U.S. programmers are paid about
5-10 times as much as Indian and Russian programmers. Both India and Russia have
extensive educational programs. India and China have a large number of people.
Nation Programmer/Analyst Salary
United States
Britain
Russia
China
India/Bangalore
52,100-900,000
45,000-80,000
19,000-34,000
14,000-25,000
9,000-18,000
6,000-12,000-30,000
Data in U.S. dollars. Differences can be affected by other factors, including
benefts, cost of living, productivity, access to equipment, and transporta-
tion and communication costs. See www.payscale.com and www.salarymap.
com.
nizations are interested in having someone else deal with the hassles of running an
MIS department. It is no wonder that two of the top-fve MIS skills in 2007 were
the ability to deal with vendors and project management of offshore providers.
Outsourcing has primarily been used to decrease operating costs or to get the
initial money from the sale of the machines. In particular, the company gains an
infusion of cash through the sale of the machines. Some frms have stated that they
chose outsourcing because it enabled them to focus on their core business and not
worry about maintaining computers and designing systems. Today, outsourcing
Web site hosting and development is relatively common. Few frms have the ex-
pertise to securely confgure the networks and servers required for e-businesses;
so they pay outside frms to handle the technical details.
Figure 13.8 lists conditions under which it is useful to consider outsourcing.
Remember that you can choose which technologies you wish to outsource and
which to keep in house. Each company has a different collection of applications,
hardware, and workers. A key economic point with outsourcing is that you con-
vert upfront fxed costs into somewhat variable monthly fees. Instead of purchas-
ing expensive hardware and software, you pay a monthly fee to have someone
else handle the details. It is also a convenient method for companies to obtain IT
management. Many companies do not know how to run an IT department, Web
site, or ERP system. A major beneft to outsourcing is the opportunity to hire ex-
perts in IT management.
Company 1991 1995 1997 2000 2003 2006
2010
IBM Global Services 0.4 17.7 24.6 37.0 42.6 48.2 56.4
EDS 1.2 12.4 15.2 19.2 19.8 21.3 40.8
Accenture 0.5 4.2 6.3 9.8 11.8 18.2 23.1
CSC 0.4 4.2 6.6 10.5 11.1 14.6 16.1
ADP 0.3 3.0 4.9 7.0 7.1 6.9 8.9
Afliated Computer 0.2 0.4 1.2 2.1 3.8 5.3 9.5
Fiserv 0.2 0.7 1.0 1.7 2.7 4.5 4.1
Perot Systems 0.2 0.3 0.8 1.1 1.5 2.3 3.5
Tata/TCS 1.4 4.2 6.8
Wipro 0.3 1.3 2.4 6.1
InfoSys 1.0 3.2 4.8
Cognizant 0.1 0.4 1.4 4.6
Satyam 0.2 0.5 1.1 1.2
HCL Tech 0.1 0.2 0.4 1.0 2.7
total ($ billion) 3.4 43.2 60.7 88.9 105.4 134.6 188.5
Figure 13.6
Outsourcing revenue. In the latter half of the 1990s, outsourcing with the major
providers accelerated as many companies chose to hire outside frms to run various
MIS functions. For 2000 and beyond, much of the outsourcing is for ERP systems
and Web hosting. Data is taken from annual reports and company Web sites.
On the other hand, situations that are unique or require advanced uses of in-
formation technology are best handled internally. For example, complex markets
that beneft from strategic applications require the knowledge and experience of
employees who work for the company. Likewise, situations that require tight se-
curity are easier to control if they remain in-house. Also avoid outsourcing when
the outsourcing frm will have to pay the same costs that you facebecause they
will charge for an additional proft margin, the fnal cost can be higher. Examples
include applications with high fxed costs or those requiring high levels of expen-
sive state-of-the-art equipment or specialized MIS talent.
Competitive pressures are also leading many managers to consider outsourcing
their information systems. As technology continues to change, it becomes increas-
ingly diffcult for general business managers to keep up with the technology. Each
change brings requests for new hardware and software, and the need to reevaluate
the use of technology within the frm. Changing technology also requires continu-
al retraining of the information systems staff. At the same time, middle-level man-
agement positions are being cut, and managers are asked to take on more tasks.
In these circumstances, companies decide to transfer IS management to an expert
through outsourcing.
Outsourcing has many drawbacks, and several studies have reported that the
majority of companies that sign long-term contracts cancel those contracts within
a few years. As reported in CIO Magazine, a study by DiamondCluster in late
2002 revealed that 78 percent of IT executives had to terminate outsourcing con-
tracts early.
Figure 13.7
Outsourcing trend. The growth of the consulting/outsourcing frms has been
constantly high for over 15 years. Organizations are increasingly turning to these
specialists to develop software and run MIS operations.
0
20
40
60
80
100
120
140
160
180
200
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
2
0
0
7
2
0
0
8
2
0
0
9
2
0
1
0
$

b
i
l
l
i
o
n
Revenue
First, there might be a slight increase in security risk because the MIS em-
ployees have weaker ties to the original company. On the other hand, outsourcing
providers are likely to have stricter security provisions than an average frm does.
A bigger question is the issue of who is responsible for identifying solutions and
new uses of technology for the frm. If MIS workers are employed by an external
frm, will it be their job to search for new applications? If not, who will?
In the past few years, some frms have begun to reconsider the costs of out-
sourcing. The hosting frm has little incentive to strive to reduce its prices or im-
prove its services. Moreover, it can be diffcult to control the decisions of the out-
sourcing frm. Consequently, some frms have become more selective over which
items are outsourced. Before you consider outsourcing, make sure you understand
the answer to three critical questions: (1) How will you ensure adequate service?
(2) How will you control costs? (3) Will it provide the fexibility you need if your
strategies change? Most contracts establish base costs, but additional requests are
charged at higher rates. The industry essentially created the concept of a service
level agreement (SLA). An SLA is a defned performance measure that is speci-
fed in the contract. For example, a contract might specify that an outsource ven-
dor must provide a new network port within 24 hours of the initial request. These
agreements generally contain penalty clauses in the form of reduced payments.
Some internal MIS organizations have mimicked the approach by writing SLAs
for basic services. The main beneft to an SLA is that it provides a defned mea-
sure of the organizations capabilities. The drawbacks are that (1) it is diffcult
to specify all of the detailed SLAs that you might need, and (2) contracts rarely
encourage continuous process improvement.
Condition Outsource In-House
Specialized talent
High fxed costs
Level of technology
Standard workers, hardware, and
software are readily available at
fxedfees.
If you need expensive workers
or technology, you can save
the mark-up profts and keep
controloverselection.
Security and control Providers can aford specialists
andprovidesolidbasicsecurity.
Datathatrequiresabsolute
secrecy needs to be kept in-
house.
Strategic use of IT Providers can handle standard
technology.
Uniqueapplicationsandnew
ideascomefromin-house.
Company size/resources Small companies get access to
specialistsandsharedresources.
You can purchase the level of
technology you need and expand
asyougrow.
Large companies can aford
IT staf and specialists, but
might choose to convert fxed
coststomonthlycosts.
Figure 13.8
Outsourcing evaluation. Outsourcing entails many trade-offs. It means transferring
control of a crucial resource to an outside company. If you are really interested in
development of strategic applications and leading-edge applications, it is usually
better to use an internal development team. If you are dealing with older technology
used mostly for transaction processing, it can be cheaper to hire an outside frm to
maintain your applications.
Sometimes outsourcing contracts can backfre on the vendor. EDS won a con-
tract with the U.S. Navy to supply computers and networks. According to The
Wall Street Journal (April 6, 2004), EDS lost $1.6 billion on the contract and
expected to lose about another half billion before it was completed.
Consultants and contract programmers are a simpler version of outsourcing. If
you have a one-time task that requires workers you do not have on staff, you can
simply hire them for the one job. The cost per hour might seem high, but you only
pay for the specifc job and you do not have to worry about fring anyone when
the job is fnished. One of the biggest issues with contractors is to ensure that your
organization will be able to use and maintain the tools after the contractor leaves.
You might need to arrange for access to the source code or for training sessions.
Managing contractors and outsourced operations are diffcult tasks. It is not like
buying a piece of software or hardware. You cannot simply assume that paying a
fee will get you exactly what you want. Communication is critical. First, you must
precisely defne exactly what you want. Particularly when dealing with workers
from a different country, it is likely that they have different interpretations of basic
concepts. Second, you must build in progress monitoring and feedback to verify
that everyone has the same interpretation of the concepts. Third, do not expect
creativity, or even insight into your business operations.
MIS Organization: Centralization and Decentralization
Who should control IT resources? Two broad trends are slowly creating
signifcant changes in the way people perceive and organize information: (1) de-
Reality Bytes: New Clorox CIO Fixes Problems with Outsourcing
Clorox is a $5.5 billion company headquartered in Oakland, CA with 8,300 employ-
ees. The company might be a familiar name to people because of its bleach, but the
company owns more than 30 other brands including Kingsford charcoal, glad bags,
and Berts Bees. For several years, the company had outsourced much of its IT func-
tion, largely to HP. IT was largely run as a cost center and the company was not
seeing the productivity and effciency gains they had expected. In April 2010, the
company hired Ralph Loura as a new CIO, who noted that the company was under-
invested in people and technology. When he arrived, it was still running Windows
2000 on desktops , and out of date ERP and CRM system, and a version of Lotus
Notes for e-mail that was several versions out of date. Internationally, the company
had 15 different ERP systems in 18 countries. In an interview, Loura commented that
There had been this kind of, If it isnt broke, dont fx it view, which is contract
behavior right? You do the least amount necessary to maintain the SLA with no idea
that certain upgrades create value by unlocking productivity through the features in
the solution. Solving the problems required going through the company to talk to
users about problems and things that worked, then prioritizing tasks and assigning
them to categories. To change attitudes, Loura redefned the role of the business sys-
tems manager to make it a client manager, responsible for identifying and meeting
the needs of the client, instead of focusing on the technology system.
Adapted from John Dix, New CIO Cleans Up Outsourced IT at Clorox, Network
World, January 24, 2011.
clining sizes and prices of technology, and (2) expanding access to the Internet
particularly wireless connections. Internet sites are slowly becoming repositories
of information that is accessible any time from almost any location. External data
on the economy, news, fnancial markets, consumers, competitors, and more is
readily available on the Internet. Increasingly, you will have to pay for this data;
but it is accessible. Advanced companies today are providing similar services in
terms of internal organizational data. Data warehouses collect, clean, and present
data to internal Web browsers. Ultimately, these databases will be integrated, so
managers in a meeting can easily call up current sales data and combine it with
economic forecasts on their handheld computers. The goal is to make integrated
data available anywhere to authorized users. The question is, How should the MIS
system be organized to provide these features?
As shown in Figure 13.9, it is helpful to examine IT management by looking
at major trends. When hardware was expensive, all data, software, and employees
were centralized. As the use of midrange and personal computers expanded, hard-
ware became decentralized and software and data began to follow it. For example,
decision makers stored spreadsheets, analyses, and reports on their personal com-
puters. But this decentralized data is more diffcult to share, even with networks.
At this stage the Internet technologies become importantthe search engines and
browsers were designed to make it easier to fnd and view data in many forms.
Applying these technologies within the company (called an intranet) provided
easier access to corporate data. Network technologies expanded to provide faster
Figure 13.9
MIS organization timeline. Management issues are easier to see in the context of
trends. Computing resources were initially large and expensive, so they were run
from a completely central service. As hardware costs declined, mid-range and micro-
computers spread through the organization causing some decentralization. Software
development shifted to contractors and purchased software. Other tasks were
outsourced. Around 2010, it became possible to re-centralize software and data on
external cloud-based servers.
1950 1960 1970 1980 1990 2000 2010
Central Server
Servers Mid-range PCs Networks Web/Clients
Software,
Development,
Consulting
Servers +
Software
Network
management
Outsource/purchase
TechnologyToolbox:DefningE-MailRules
Problem: You get dozens or hundreds of messages a day.
Tools: Microsoft Outlook and other e-mail client packages have rules and features to
help you deal with the food of e-mail.
Even if you remove the junk e-mail, many business managers receive dozens of
messages a day. Groupware tools and automated project management systems add
to the list. Some are important; others are minor but still worth saving. But even if
you save the messages, you need a way to organize and search them so that you can
respond to the most important ones frst and then search the others later.
The frst step in organizing e-mail is to create additional folders. Right-click the
inbox folder to add new folders. The challenge is to make it easy to fnd the mes-
sages you receive. You can manually move messages into the relevant folders. How-
ever, Outlook enables you to create powerful rules to automatically evaluate and
handle your messages. It effectively allows you to build an expert system agent that
deals with many of your messages automatically. The simplest type of rule is one
that automatically moves messages from a specifed sender into a designated folder.
For example, messages from the CEO could be moved into a critical folder. As you
receive new messages, you can deal with the most important folders frst.
You can create sophisticated rules. The Tools/Rules and Alerts option has a wizard
to help you. Rules have three components: (1) conditions that are evaluated for each
message, (2) an action to take, and (3) exception conditions to exclude certain mes-
sages. You have a couple dozen types of conditions or exceptions, such as choosing
people or matching words in the subject or body. You can select from a couple dozen
actions, including moving the message to a folder, deleting it, forwarding it, or even
running a custom script or opening an application on your desktop. This last option
might be used to open a sales application and generate a new order when a specifc
message is received from a customer. Since you can control the order in which rules
are applied to each message, you can create a decision tree of multiple rules.
Quick Quiz:
1. How is the e-mail system similar to an expert system? How is it different?
2. What is likely to be the most diffcult part of creating a system for handling your
messages?
access to data and services on distant computers. Central computer services be-
came standardized and companies emphasized purchasing software (particular-
ly ERP systems around 2000). Servers became cheaper and simpler to operate.
Some companies outsourced many tasks, including server and network manage-
ment. Decentralization created some security issues, so confict often arose over
control of data. By 2010, the pieces were in place for companies to re-centralize
their operations by using Web servers and software as a service. Software and data
could be stored on cloud-based servers with users accessing everything through
browser-based platforms including tablets and cell phones. It remains to be seen
whether this transfer to cloud-based computing will be a trend or simply a tool to
support certain tasks.
Almost none of the issues of centralization and decentralization are newpoli-
ticians, economists, and organizational theorists have debated them for hundreds
of years. The basic argument for centralization revolves around the need to co-
ordinate activities and effciencies that can be gained from large-scale operations.
Proponents of decentralization argue that moving control to smaller units pro-
duces a more fexible system that can respond faster to market changes, encourage
individual differences, and innovate. Figure 13.10 summarizes the arguments for
centralization and decentralization.
As with many arguments, there are different answers for different circumstanc-
es, and it is rare that the extreme choices are best. Wise managers will attempt to
gain the advantages of both approaches. With information systems, four basic ar-
eas are subject to centralization or decentralization: hardware, software, data, and
staffng. Determining the best way to organize information resources requires that
managers understand the advantages and disadvantages for each of these areas.
Figure 13.10
Summary of benefts of centralization and decentralization. There are advantages to
both centralization and decentralization of the MIS resources. The ultimate objective
is to design an MIS organization to beneft from as many of the advantages as
possible by combining both centralization and decentralization.
Category Centralization Decentralization
Hardware Sharedata.
Controlpurchases.
Controlusage.
Lessduplication.
Efcientuseofresources.
Lesschanceoftotalbreakdown.
Usersgetpersonalizedmachines.
Software Compatibility.
Bulkbuyingdiscounts.
Easiertraining.
Easeofmaintenance.
Diferentuserpreferences.
Easieraccess.
Customization.
Data Easybackup.
Easiertoshare.
Lessduplication.
Securitycontrolandmonitoring.
Only transmit data that needs to be
shared.
Keepsusercontrolandpolitics.
Personnel Similarworkerbackgrounds.
Easiertraining.
Straightforwardcareerpath.
Specializedstaf.
Easiertoseeandcontrolcosts.
Fasterresponsetousers.
Moretimewithusers.
Better understanding and
communication.
Diferentcareerpath.
Hardware
Today, hardware is relatively inexpensive. Even centralized servers have come
down in cost, often using systems based on server farms consisting of hundreds or
thousands of inexpensive computers.
Similarly, on the user side, prices of personal computers have dropped substan-
tially. Even portable devices are relatively inexpensive.
Centralization
The biggest advantage of centralized IS hardware is that it is easier to share hard-
ware, software, and data with multiple users. Complete centralization is shown
in Figure 13.11. Consider a simple example. If a company installs an expensive
printer in one users offce, it will be diffcult for other users to get access to the
printer. On the other hand, with only one central computer, all of the hardware,
software, and data will be located in one place. All users can be given equal access
to these facilities. By keeping all hardware, software, and personnel in one loca-
tion, it is easier to avoid duplication and keep costs down.
Along the same lines, centralized hardware also makes it easier to control user
access to the information system. By storing all data on one machine, it is easy to
monitor usage of the data. In a sense, all user access to data must frst be approved
by the MIS department. Any data alteration or transfer is much easier to control if
it takes place on one machine.
Figure 13.11
Complete centralization. For many years, computers were expensive and there were
few communication networks. Consequently, hardware, data, software, and MIS
personnel were centrally located. Data was sent to the computer for processing and
printed reports were distributed throughout the company. Users only dealt indirectly
with MIS.
Data and
software
MIS personnel
Hardware
User departments
Centralized purchasing can also be used to save money. It is easier to standard-
ize equipment if it is all located in one place. It is generally possible to obtain dis-
counts from vendors by buying larger quantities. Centralized purchases also make
it easier to keep track of the amount of money spent on information technology.
When user departments are responsible for all IT purchases, the lack of central-
ized control can lead to duplication of hardware.
Decentralization
Decentralization of hardware carries its own advantages. First, there is less chance
of a total breakdown. If your computer breaks, everyone else can continue work-
ing. You might even be able to borrow someone elses machine. Figure 13.12
shows the huge adoption rates of personal computers. The lure of total control
over powerful, low-cost computers led users to adopt them and use them for many
computing tasksparticularly decision support analyses. During that time, spend-
ing on all computers, including servers, increased. But the growth of PC-spending
vastly exceeded all others.
The PC technology eventually spread into servers, driving down their costs as
well. Today, the differences between servers and personal computers are subtle;
but some vendors such as IBM, Hewlett-Packard, and Sun sell specialty servers.
Many of these servers are used for Web-based applications, so they do represent a
return to centralization of hardware in some areas.
Figure 13.13 shows another way to look at hardware decentralization. It shows
the growing shift to laptops, where even by count, laptop sales exceeded 50 per-
cent of the shipments by 2009. The chart does not include Web-based cell phones
or tablets. Some people are forecasting that sales of tablets will cut into the sales
Figure 13.12
Growth of personal computers. The huge price and performance gains of PCs
drove their adoption for 30 years. Although servers eventually adopted the same
technologies, hardware became distributed throughout the organization.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1970 1980 1990 1995 2000 2005 2010
S
h
a
r
e

o
f

M
a
r
k
e
t

V
a
l
u
e
Worldwide Computer Sales
Server
Midrange
Personal
of laptops as consumers opt for the smaller devices. This trend to mobile comput-
ing is important because it shows that people want to have their computers, or at
least Web access, with them at all times.
With decentralization, users can obtain personalized equipment. Perhaps a f-
nancial analyst needs an extremely fast machine to process complex equations. Or
maybe a marketing representative needs a portable computer to collect data from
clients. An advertising specialist could use high-resolution graphics to help design
promotions. In each case, the company saves money by buying each user exactly
what he or she needs and not forcing everyone to use one standardized product.
Software and Data
Wherever there is hardware, it is also necessary to provide software. Nonetheless,
it is possible to centralize some aspects of software, even though there are decen-
tralized computers. The goal is to capture the advantages of both methods at the
same time. Data fles are similar to software fles, but there are some additional
features to consider when choosing where to store the data.
Software Centralization
If software applications are standardized and purchased centrally, it is possible
to negotiate lower prices from software vendors. Besides, if everyone uses the
same basic software, fewer compatibility problems arise and it is easy for users
to exchange data with coworkers. Similarly, upgrades, training, and assistance are
much simpler if there are a limited number of packages to support. Imagine the
time and effort involved if the company needs to upgrade different spreadsheets
on 5,000 separate machines. Some companies have reported that by the time they
managed to upgrade the entire company, an even newer version was released.
Software Decentralization
Forcing users to choose identical packages can lead to major arguments between
users and the MIS department. Many times users have different requirements or
perhaps they simply have different preferences. If one group of users prefers the
software that is different from the corporate standard, why should everyone in the
company be forced to use the same tools? Cost becomes less of an issue as soft-
ware prices drop. Support might be a problem, but major software packages today
are similar. Data incompatibilities often can be resolved with conversion software.
To some extent, users should have the ability to customize their software to
match their preferences. Today, most software packages enable users to choose
colors, mouse or keyboard responsiveness, and locations to store personal fles.
If this software is moved to a centralized environment, you have to be careful to
preserve this ability.
One complication with enabling users to choose different software is that it
can be diffcult to determine the confgurations of each machine. If a user has a
Figure 13.13
Mobile computing. Based on unit sales, laptops have become increasingly popular.
Throw in tablets and computer-based cell phones and the extent of decentralization
becomes greater.
0
10
20
30
40
50
60
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Percent Laptop Shipments
problem, the MIS support person needs to know what software is installed on the
machine. When installing new hardware and software, the support team needs to
know what software exists on each target machine. Managers also need to track
software usage when they purchase upgrades and to verify compliance with soft-
ware licenses. Several software tools exist to help the MIS department track soft-
ware usage and report on the confguration of each computer. A small fle is in-
stalled on each computer that reports on the software, hardware, and confguration
of each machine.
Data Centralization
The most important feature of centralized data is the ability to share it with other
users. Large central servers were designed from the ground up to share data. They
were designed to solve the problems of allowing concurrent access and to protect
the integrity of the data. Similarly, they have security facilities that enable owners
of the data to specify which users can and cannot have access to the data. Central-
ized systems also monitor access and usage of the data to prevent problems.
Another important feature of centralized data is the ease of making backups.
When all databases are stored on one machine, a single operator can be hired to
maintain daily backups. If data fles are lost, it is easy to restore them from the
backups. With the data on one machine, it is easy to ensure that all fles are regu-
larly backed up. Contrast this situation with distributed personal computers, where
users are generally responsible for making their own backup copies. How often do
you personally make backups? Every night?
Data Decentralization
The strongest advantage to decentralizing data is that it gives ownership of the
data to the group that creates and maintains it. Users also have complete control
of the data and can prevent anyone else from even knowing that it exists. For data
that does not need to be shared, this control presents no problems. However, scat-
tered control of data can interfere with other users when many people need access
to the data. An example of complete decentralizationincluding data, hardware,
and personnelis displayed in Figure 13.14.
Reality Bytes: Perfection is Probably Impossible even in the Cloud
One of the big risks of centralization is that if something goes wrong, individuals
have little ability to recover data. Instead, the centralized provider is responsible
for securing the data and providing backups. The cloud and virtual computers make
this process easier, but few technologies are perfect. In February 2011, as many as
150,000 Gmail users faced the loss of years of data stored on Googles servers. One
noted that I logged in and my account also looks like a brand-new Gmail account!
10 years of emails (17000 of them) are gone. Google engineers worked for several
days to restore the data. Ultimately, most of the data was restoredfrom old-style
tape drives. The bottom line is that users should try to keep backup copies of their
data. If it is not possible to store data on client systems, at least move copies of criti-
cal data to different online providers.
Adapted from Laurie Segall, Google Nukes Thousands of Gmail Accounts, CNN
Online, February 28, 2011.
Data replication is sometimes used to provide the advantages of decentral-
ized dataand still provide companywide access. With replication, the database
is copied to multiple servers throughout the company. Users work on their local
copies, which provide fast access to the data. The changes are copied to the other
servers at regular intervals, so everyone has access to the latest data. This tech-
nique is often used with groupware products to distribute spreadsheets and word-
processed documents.
Personnel
When most users think about decentralization, they often forget about the infor-
mation systems personnel. Traditionally, the MIS roles have been performed by
centralized MIS staffs. However, the increased decentralization of hardware in-
creases pressures decentralize the personnel by having them report directly to user
departments.
Centralization
Most of the advantages of a centralized MIS staff accrue to the MIS workers. For
example, MIS workers often feel more comfortable with other MIS specialists.
Centralization creates a group of homogeneous workers who share the same edu-
cation and experiences. Moving MIS workers to user departments places them in
a minority position.
One implication of this situation is seen by looking at the career path of an MIS
worker. In a centralized environment, workers are typically hired as programmers.
They eventually become systems analysts. Some move on to become team or proj-
ect leaders, and a few can look forward to becoming managers of IS departments
and perhaps a CIO someday. If programmers are moved to user departments (say,
MIS personnel are members
of user departments
Marketing
Finance
Accounting
Figure 13.14
Complete decentralization. Each department maintains its own hardware, software,
and data. Corporate standards and the network enable workers to utilize data across
the company. MIS personnel are members of the user departments and support tasks
within that department.
human resources), what career path do they have? Maybe they could become team
leader or manager of the HRM department, but they would be competing with
HRM specialists for those positions.
Centralization also makes it easier for the company to provide additional train-
ing to MIS staffers. Because hardware and software changes occur constantly,
MIS employees need to continually learn new tools and techniques. If they are all
located in a central facility, it is easy to set up training classes and informal meet-
ings to discuss new technologies.
Centralization also gives the frm the ability to hire MIS specialists. If there
are 50 positions available, two or three can be set aside for workers specializing
in areas such as database administration or local area networks. If all workers are
distributed to user areas, the individual departments will be less willing to pay for
specialists.
Finally, when the entire MIS staff is centralized, it is easier to see how much
MIS is costing the frm. If the MIS functions are dispersed to user departments,
they may be performed on a part-time basis by various employees. It is diffcult
to control the costs and evaluate alternatives when you do not know how much is
being spent.
Decentralization
The primary advantage to decentralized MIS staffng is that the support is closer
to the users. As a result, they receive faster responses to questions and problems.
More important, as the MIS staffers spend more time with the users, they gain a
better understanding of the problems facing the users department. Communica-
tion improves and problems are easier to identify and correct. These are powerful
advantages to the individual departments and have the potential to create much
better development and use of information systems.
Te Help Desk
One issue with decentralized MIS support is that it can be expensive to place MIS
personnel in each department. Many companies compromise by creating a help
desk that is staffed by MIS employees who specialize in helping business manag-
ers. When business managers have questions, workers at the help desk provide an-
swers. Typical problems involve personal computers, networks, and access to cor-
porate databases. One advantage for business managers is that they do not have to
search for answersthey simply call one number. This system can also cut costs
and ensure consistent support. The knowledge of the support workers is easily
shared throughout the company. It is also easier to train and evaluate the workers.
To provide more decentralized support, some companies are using their net-
works to provide more detailed help to business departments. They set up a spe-
cial program in the background on each personal computer. When someone calls
for help, the microcomputer specialist can see the users screen and take control of
the users machine. This method simplifes communication between the user and
the specialist, making it easier to solve problems and make changes immediately.
Of course, it also raises several security issues, because the help desk personnel
could monitor any machine at any time.
Recentralization with Cloud Computing
How can Internet technologies be used internally to centralize
data but still support decentralized user access? No organization is
completely centralized or completely decentralized. The true challenge is to create
a system that matches the needs of the organization. Networks are a critical part of
the solution. With reliable high-speed networks, data can be stored anywhere. But
most organizations do not yet have high-speed networks everywhere. In particu-
lar, connections to offces in other cities or nations can be expensive and relatively
slow. Consequently, bandwidth is a crucial factor in deciding centralization issues.
Cloud computing and private clouds or Web-based intranets provide a solution to
many of these problems.
Networks
Web technologies are particularly good at handling low-bandwidth connections
to users. Web browsers are relatively effcient at displaying data. Many business
pages contain only basic data and graphs, which can be easily and quickly sent to
managers. Streaming media technologies can be used to send more complex data,
such as speeches, to many users at the same timeeven when the managers are
connected through dial-up lines.
The capabilities of the Internet browser have led frms to consider a new ap-
proach to organizing the MIS resources. The thin-client approach, illustrated in
Figure 13.15, uses a relatively simple computer to run a Web browser that is re-
sponsible for displaying data and getting input from the user. This approach recen-
tralizes many of the MIS functions. All of the data and most of the applications
reside on centralized servers. The use of Web standards simplifes many decisions.
Users can choose almost any type of client hardware, including laptops, tablets,
and PDAs. As long as the system runs a browser, it can access the corporate data.
Of course, some client computers will have more capabilities than others.
The browser client is becoming a user interface device, with responsibility for
displaying data and translating input to a standard form. This approach simpli-
fes the development of applications and provides more fexibility to users and
organizations. For example, some users or entire organizations might stick with
standard desktops for years to save money. Others might move to wireless-based
Figure 13.15
Thin client. The thin client device is only responsible for displaying data and
translating user input. The data and the application software reside on the central
servers.
Server
Data and applications
Thin client
Browser-based access
User interface
Technology Toolbox: Managing Projects
Problem: You need to track workers and schedules on a project.
Tools: Project management software such as Microsoft Project has becoming in-
creasingly sophisticated.
The tools emphasize Gantt charts, but the primary strength is that all of the informa-
tion is stored in a database. The tools quickly provide different views of the data,
including Gantt, network, and critical path charts, calendars, and worker schedules.
The frst step in any project is to identify the goal. Do you need to be fnished
by a specifed date, or do you have a starting date and need to estimate the fnish?
The next step is the most important: break the project into multiple tasks. Tasks can
be grouped and subdivided into even smaller activities for large projects. Each task
should have a clear starting and ending point. For each task, estimate the amount of
time required to complete the task. Later, you can record when tasks begin and track
their progress. Once the basic tasks have been defned, you need to specify the rela-
tionships: Which tasks cannot be started until a prior task has fnished? For example,
you cannot begin integration testing of software until after it has been written.
You can also assign resources to each task. Generally, you will assign people or
outside contractors to particular segments. If technology or other hard resources are
going to be a limiting factor, you should include them as well. Microsoft Project
has a system to automatically level the project if resources are over allocated. For
example, if you are scheduled to do three daylong tasks on one day, it will push the
tasks forward to provide a better estimate of when the project will be completed.
Most development projects involve multiple workers or contractors performing
many tasks. Each task can be dependent on many other tasks. The Gantt chart high-
lights these dependencies. As the project progresses, you can add completion data
to each task. If a task is delayed, the system will automatically push back the future
tasks and show the new completion date.
Microsoft Project can also be integrated with a Web server, enabling each partici-
pant to check the progress and share calendars. The Web server acts as a groupware
tool and can e-mail each person and record their comments.
Quick Quiz:
1. What advantages are provided by storing the project information in a DBMS?
2. Why is estimating development time one of the most diffcult activities?
tablets or Web-enabled cell phones that rely on voice input instead of keyboards.
The key point is that the choice of the user device should no longer matter to the
application developer. Regardless of the user device, the back-end databases are
the same, the Web servers are the same, and the applications are the same. Note
that currently, it does take a middle-ware component to strip down Web sites so
that they are small enough for todays cell phones, but that limitation is likely to
change in the near future. The simplicity of this approach is that it recentralizes
the primary items that gain from centralization: the main business applications
and the shared data. Users are free to use whatever devices they prefer and to load
additional software on their computers. A key beneft of the thin-client approach is
that the clients can be built from relatively simple hardware and software, reduc-
ing the cost and improving the reliability of the clients. With fewer problems, user
support becomes easier and cheaper.
As cloud-based computing software expands and improves, it will become
easier to move more applications to cellphones and tablets. For example, Google
and Microsoft both have Web-based personal productivity suites. Other compa-
nies provide installable apps that can be used to open and edit standard Word and
Excel fles on tablets.
One challenge is that large data transfers require high-speed connections. For
example, bulk data transfers from one division to another will suffer if sent over
dial-up lines. Since high-bandwidth Internet connections are expensive, compa-
nies will cluster their servers in a few key locations. As shown in Figure 13.16,
these server locations will be connected with high-speed lines, and everyone else
will use lower-priced basic services to connect over the Internet.
Figure 13.16
Intranet networks. Server locations are connected by high-bandwidth networks to
replicate data. Individual users obtain data using standard Internet connections.
The network personnel still have to decide where to locate the servers and
which data should be stored on each server. Many systems use data replication,
where each server location holds the same data. Data changes are exchanged on
a regular basis, often overnight, so each server has a relatively up-to-date copy of
the data. The MIS department is responsible for maintaining the servers and the
networks.
Hardware
Companies have several choices for the central server hardware. But the software
environment is more important than the specifc hardware. The servers need to run
software that generates the Web data while interacting with the database manage-
ment system. Several companies offer competing technologies for these services
that run on diverse hardware platforms. The main issues are (1) cost, (2) scal-
ability of the servers so that the system can be expanded without interfering with
existing operations, and (3) reliability, maintainability, and backups to ensure the
systems can remain operational at all times.
Data
Because of the challenges of running and securing servers, most companies lean
toward centralizing the data. Certainly the basic fnancial data is consolidated
in a data warehouse. This approach works well for managers retrieving data for
analysis. The main problems arise in terms of creating or modifying the data. The
workers who create and analyze the data to produce more useful information need
more sophisticated tools. They also need greater access to the data, compared with
users who simply view the data.
Reality Bytes: How Do You Know if Your IT Department is Working?
Shortly after Obama took offce in 2009, the administration began planning to
reduce the number of government data centerslargely to reduce costs. First, it had
to fgure out how many federal data centers existed. In 2011, the administration an-
nounced plans to shut down 137 of the 2,094 centers it identifed. The plan ultimately
targeted 800 centers to be closed by 2015. The plan estimates it will save $3 billion a
year. The government spends about $450 million a year just on electricity to run the
centers. Because of overbuilding, the White House estimated only 27 percent of cen-
ter capacity was utilizedfar lower than average values for private-sector data cen-
ters. In mid-April, while Obama was at a fundraiser in Chicago, a microphone was
left operational and some off-the-cuff remarks were widely reported. Mostly, Obama
complained about government IT support, calling it horrible, across the board, in-
cluding the Pentagon, Homeland Security, and federal agencies. Mark Knoller, a re-
porter for CBS who heard the comments tweeted that Obama said he thought that
as president, hed have some cool phones and stuff in the Oval Offce, but now he
says we cant get our phones to work! More publicly, Vivek Kundra, the federal
CIO, has been critical of federal IT, noting that projects too often cost more than
they should, take longer than necessary to deploy, and fdeliver solutions that do not
meet our business needs.
Adapted from Damian Paletta, Government to Pull Plug on 137 Data Centers, The
Wall Street Journal, April 27, 2011; and Patrick Thibodeau, Obama Unvarnished:
Government IT is Horrible, Computerworld, April 15, 2011.
Giving users, even managers, the ability to create new data scares most MIS
people. The security challenges are much greater when users need to add and
change data. It is more diffcult to control access and ensure that only authorized
people can make the changes. Plus, if something goes wrong, the IS employee is
the one who will be blamed. In the old days of simple transaction data, it was rela-
tively straightforward to set up controls and procedures for the daily operations.
And it was reasonably simple to keep transaction logs of all the changes so errors
could be corrected. But in todays environment, teams of workers perform the
analyses and information creation, so team members need access to work in prog-
ress. For instance, the fnancial budgeting team uses the marketing and production
forecasts to generate estimates of future cash fows.
From the standpoint of data creation, the cloud approach requires two steps
beyond traditional systems: (1) managers need tools that will create the fnal data
and reports in a format suitable for the browsers, and (2) managers need an easy
method to securely transfer information to the servers.
Better data creation tools have been created in the past few years, but in many
cases they are still hard to use and can require specialized training. Likewise, sev-
eral systems have been created to simplify transferring data to cloud-based serv-
ers. Some are relatively easy to use. As always, the challenge lies in providing se-
curity so that the transfer process is easy for authorized managers but impossible
(or exceedingly diffcult) for unauthorized users. Both of these conditions require
that managers have more powerful tools and often higher-speed data connections.
Managers will also need more support and technical assistance. Hence, portions
of the MIS organization must be decentralized to handle these issues.
Confict Management
Why is the MIS department involved in so many conficts? How
do you solve them? The answer to the frst question should be relatively easy
if you reread the sections on centralization and decentralization. Trying to resolve
the issues is considerably more diffcult because each organization is different. An
answer that is often popular is to fre the CIO. But that rarely solves the underly-
ing problems.
Reality Bytes: Minimizing Complexity with Thin Clients
The Co-operative Group runs a chain of grocery stores in Manchester, England. The
company was in the process of building new corporate offces and asked its IT tech-
nical architect, Ian Cawson, to establish a new PC policy before the building opened
in 2012. He determined there was not enough time to conduct a full business analy-
sis, so he selected a thin-client approach using Wyse hardware for 2,500 of the of-
fces 2,750 users. All of the PCs are created as virtual desktops in the central server.
He also plans to roll out the system to the 19,000 users in offces worldwide in three
to fve years. Overall, Cawson notes that the approach does not really save costs. In-
stead, it makes it easier to manage things like upgrades and patches. Users will also
be able to access their desktops from portable devices, such as the Apple iPad.
Adapted from Robert L. Mitchell, Grocer Goes with Thin Clients, Computerworld,
May 3, 2011.
Centralization and decentralization are often causes of conficts. Managers
want the fexibility to respond to the changing environment. They often are will-
ing to implement technology that helps them directly. But they rarely want to pay
for the technology, and they generally dislike having to pay for infrastructure
technology. The CIO and the MIS department provide and support IT, but they are
usually squeezed by costs. Software and employee costs increase. Although hard-
ware costs are declining, it simply means that organizations buy more hardware,
and installing more hardware means more work. Leading edge MIS departments
automate as many of the tasks as possible, from software installation to support to
network monitoring. One way MIS departments try to reduce costs is to standard-
ize as much of the technology as possible. The company purchases one machine
and confgures it the same for all users. Users are forbidden from customizing
or adding any software or hardware. While this approach makes life easier for
MIS, it does restrict the technology and applications available to the company.
What is the goal of the company: To make life easier for MIS or to use technol-
ogy to improve the business? At what cost? The issues are further complicated
because no one really knows if the MIS department is effcient or wasting money.
Hence, frms have an incentive to look at outsourcingto provide a market price
comparison.
The conficts get even worse as MIS becomes more focused on security. Many
restrictions have been imposed in the name of security. Yes, security is impor-
tant, but remember that security requires a trade-off. For example, it is possible to
protect a document so that it is 100 percent secure: all you have to do is com-
pletely destroy the document (and any copies). This action guarantees that no one
can steal or read the document. But it also ensures that you cannot use the docu-
ment. By defnition, security creates this trade off, which pretty much guarantees
there will be conficts between the central security managers and the rest of the or-
ganization. Keep in mind that computer security people lose their jobs if someone
breaks into the computers and steals data and money. They rarely get fred if their
Reality Bytes: Outsourcing Politics at Infosys Technologies
Many politicians have found it useful to pound the drum about outsourcing frms tak-
ing away American jobs. As with most issues of international trade, the politicians
are ignorant of the economics, such as reduced costs and higher profts for American
frms. But, in August 2010, they substantially increased the fees for skilled employ-
ment visasmaking it more expensive for Indian frms to place their employees in
U.S. jobs. In 2011, U.S. agencies began investigating one of them (Infosys Technolo-
gies Ltd.) for possible abuse of the temporary business visa program. Essentially, the
Indian frms send their employees to U.S. businesses to handle tasks such as business
analysis. Indian frms reportedly earn about 60 percent of their total $50 billion in
annual revenue from U.S. frms. Infosys was accused by a former employee of using
short-term B1 visas to bring in employees to work on a long-term basis. Long-term
employment requires an H-1B visa, which are more tightly controlled. In the past,
the annual limit for H-1B visas was often reached within a few weeks of the start of
the annual time period.
Adapted from Magha Bahree and Amol Sharma, U.S. Moves from Rhetoric to Ac-
tion on Visas, India Realtime, May 25, 2011.
security controls force users to take 20 more steps to fnish a job. Consequently,
security administrators tend to lean on the tighter side of securityeven when it
creates additional problems and work for users.
The answer to resolving these conficts is to understand that they will arise and
to build a mechanism to resolve the disputes. You could escalate all disputes to
the level of the CIO or the board of directors, but that would waste an incredible
amount of time. It is far better to establish a neutral oversight committee to evalu-
ate MIS progress and arbitrate conficts. Just make sure the committee is chaired
by a business leader and not an MIS manager.
Summary
Managing an MIS organization is diffcult. Even as a business manager, working
with MIS and choosing the proper role for technology can be challenging. Ideally,
the goals of the MIS department should be aligned with the overall business goals.
But with issues of centralization, cost control, and security staring you in the face,
conficts can easily arise. One of the more diffcult problems facing MIS depart-
ments and company executives is the confict between centralization and decen-
tralization. These issues were involved in many decisions during the last 5 to 10
years, from politics to corporate organizations, to the way in which MIS fts into
the organization. Although there is no single answer that applies to all situations,
there has been a distinct trend during the last few years toward decentralization. In
larger organizations, this propensity has been hampered by the highly centralized
organizations and computer systems that have been in place for the last 30 years.
Decentralization of MIS can occur in any of four areas: hardware, software,
data, and MIS personnel. Economics is driving the decentralization of hardware,
because of tremendous price performance values in personal computers. The chal-
lenge is to accommodate this decentralization without losing the benefts of cen-
tralization. One option would be a completely decentralized information system,
where each user and department is responsible for its own information. Today,
the Internet standards provide new technologies to gain the benefts of both cen-
tralization and decentralization. Applications running on Web servers can retrieve
centralized data to be displayed and modifed using thin-client browsers. The goal
is to gain the economies of scale and improved control and ease of sharing of-
fered by centralized servers, yet provide users with the individual tools needed to
perform their jobs. The simpler client hardware and software platforms offer the
promise of less user support.
Managing servers and networks, as well as building applications, can be diff-
cult tasks for many companies. It is hard to fnd and reward good IS workers, and
continually solving technical problems takes time away from the daily business
tasks. So, many organizations have chosen to outsource various IS functions
from development to maintenance to development and operation of the servers.
Outsourcing provides a short-term increase in cash for the company, access to
computer specialists, and the ability to concentrate on the companys primary
business. However, frms requiring specialized talent, high security and control,
high levels of recent technology, new state-of-the-art information technology, or
complex market structures should avoid outsourcing and retain in-house manage-
ment of the information function.
Key Words
Web Site References
Job Boards
Careerbuilder www.careerbuilder.com
Computerworld www.computerworld.com/topic/10/
Careers
Dice www.dice.com
Jobs.com www.jobs.com
Kforce.com www.kforce.com
Monster www.monster.com
Studentjobs.gov www.usajobs.gov/studentjobs
U.S.Government www.usajobs.gov
Wall Street Journal careers.wsj.com
A Managers View
It is diffcult to manage and evaluate an MIS organization. With multiple
roles and many different types of employees, IT departments can be expen-
sive. You can try to use outsourcing and contractors to reduce costs. Ulti-
mately, you have to decide on the strategic role of MIS. Is cost reduction
your primary goal, or are you going to use technology to improve the busi-
ness and gain a competitive advantage? Centralization and decentralization
are key issues in managing information systems. Many conficts arise when
the IS departments are not aligned with the business practices. New Web-
based technologies offer new methods of maintaining the cost advantages
of centralization while still providing decentralized user access and control.
advocacy role
centralization
certifcations
chargeback
data administrator
data integrity
database administrator
decentralization
offshoring
outsourcing
replication
service level agreement (SLA)
standards
systems analyst
thin client
total cost of ownership (TCO)
Review Questions
1. What are the main issues in managing an MIS department?
2. What are the basic roles of the MIS department?
3. What types of MIS jobs are available?
4. How do you determine which aspects of MIS to outsource and which to keep
in-house?
5. Is it true that students should avoid IT careers because all of the jobs are
moving to other countries?
6. Even if a company has an ERP system and buys most of its software, what
aspects of MIS does it need to keep?
7. What are the advantages of centralizing computer hardware, software, and
data? What are the advantages of decentralization?
8. Why do different types of organizations require different levels of IT
centralization?
9. How do Web-based operations solve the centralization/decentralization
problems? Do they make some problems worse?
10. Why do conficts arise with MIS and what structures should be created to
help resolve them?
Exercises
1. Using salary surveys and local advertisements, fnd typical salaries for
various MIS jobs in your area.
2. Interview a worker at a large company (perhaps a friend or relative)
and ask them to evaluate the MIS department. Comment on the degree
of centralization. Does the organization use chargeback to pay for MIS
services?
3. Make a list of symptoms you expect to see in a company that is too
decentralized. That is, company users are free to choose any hardware and
software, and databases are maintained by each department. Data is shared
through reports that are printed in each department and forwarded to other
departments on paper. There is no central MIS staff and no CIO. Treat it as
a company that started small using personal computers and grew but did not
come up with a centralized information system approach.
4. As a manager in a small company, you do not have an in-house MIS
development team. You need a special data-analysis program created that
will run as a browser-based service on your Web site to help other managers
analyze sales data. The mathematical analysis is fairly complex, but you have
formulas for all of the calculations. Find a company or person that could
handle the development for you.
5. Find data on the number of computer science and MIS majors at your school
for the past few years and identify any trends. How might these trends affect
the job market?
6. Obtain a sample of at least 20 job ads for programmers. For example,
randomly sample ads from Dice or Monster. Summarize the programming
languages and skills in demand. Identify the types of companies hiring
workers. In particular, are they general business organizations, or are they
MIS development specialists?
7. Using the various salary surveys, compare salaries for small versus large
businesses in your area. Briefy explain the differences. If possible, talk with
employees at small and large businesses to see how the jobs might differ.
8. Using the Web, fnd at least three job descriptions for MIS workers in
network support. Summarize the main tasks of the job and explain how they
support the company.
9. Browse the online contract-worker site vworker.com and examine at least
fve project descriptions. Summarize the projects and comment on the level
of detail provided and the suggested values.
10. Choose a case in this book and explain the level of centralization within the
management of the company.
Technology Toolbox
11. Assuming that you work as a manager, create a set of mailbox folders to
handle your expected mail. Create the rules that will move incoming mail to
the appropriate folder.
12. Assuming that you are in charge of a sales division for a large company,
create an e-mail rule to handle messages from customers. You have a couple
of key clients that you have known for years, and they have a pattern of
sending messages. For example, they send personal messages a few times a
year, they send questions about price changes monthly, and once in a while
they send complaints about late deliveries or quality concerns. Sometimes
they send a new order directly to you, which you have to forward to the sales
staff.
13. Create a project management analysis for starting a new company. Identify
the major tasks and their dependencies. If possible, create the project in
Microsoft Project (you can get a free demonstration copy). If it is not
available, at least draw a Gantt chart by hand.
14. Create the Gantt chart for the development exercise using Microsoft Project.
Assign resources at 100 percent as indicated and use resource leveling to
determine the time it will take to complete the project.
15. Using the data in the previous exercise, explore ways to complete the project
earlier. Note that you can consider hiring more people, but there is a limit
adding more people to the project means you have to add more managers
and increase some of the development times to compensate for the additional
overhead.
Task Length Depends on Resources
1 Feasibility Statement 5 days
2 Get hardware list and costs 1 day Analyst
3 Count forms and reports 1 day Analyst
4 Estimate development time 1 day Analyst
5 Get benefts from user 1 day Analyst
6 Create statement 1 day 2, 3, 4, 5 Analyst
7 Management Approval 1 day 1
8 Analysis 17 days 7
9 Interview users 7 days Analyst
10 Evaluate competition 3 days Analyst
11 Search for existing software 3 days Analyst
12 Evaluate options 4 days 9, 10, 11 Analyst
13 Management Approval 1 day 8
14 Design 15 days 13
15 Design and create database 2 days Analyst
16 Build forms 8 days 15 Programmer
17 Create reports 4 days 15 Programmer
18 Design application 3 days Programmer
19 User approval 1 day 14
20 Management approval 1 day 19
21 Implementation 10 days 20
22 Purchase hardware 2 days Analyst
23 Transfer data 3 days 22 Programmer
24 Integration test 4 days 23 Programmer
25 Train users 1 day Trainer
26 Write procedures 1 day Analyst
27 Transfer operations 1 day 24 Analyst, Programmer
28 Review 1 day Analyst, Programmer
Teamwork
16. Interview computer users and managers in a local frm (or your university)
and determine the degree of decentralization in their information system
organization. Talk to several users and see whether their perceptions
agree. Are they receiving all of the advantages of centralization and
decentralization? If not, how could the system be modifed to gain these
benefts without signifcantly increasing the drawbacks? Be sure to analyze
hardware, software, data, and personnel.
17. Have each person select one country and fnd the average cost for
programmers. Identify technology and telecommunication capabilities
available. Identify social and cultural factors (such as education) that
could affect programming abilities. Combine the data from each person
and identify a nation in which you would want to establish an offshore
outsourcing facility.
18. Have each person research a separate case in the textbook and identify the
degree of MIS centralization within the organization. Compare the results
along with the sizes of the organizations. Identify any patterns.
19. Assume that you want to install a new wireless network for a company with
300 employees in one location. Find at least three companies that could
handle the installation and confguration.
20. Assuming that the team is a company that needs to standardize its
technology, select a single smart phone that would be used by each person
and choose a standardized list of applications that would be available. If you
had to use this corporate phone, would you give up your personal phone?
21. The team wants to start a company that provides business advice that uses
the Web and smart phone applications. For example, it might specialize
in answering marketing questions and analyzing sales data. Assuming
the company grows to 50 main employees, describe the IT systems and
organization that will be needed.
22. From a management perspective, explain if Rolling Thunder Bicycles is a
centralized or decentralized company.
23. How should the company handle typical information system tasks such
as backing up data, creating employee accounts, maintaining hardware,
selecting new hardware and software, and so on?
24. The company wants to move the entire application to a Web-based system to
support both internal operations and allow customers to place orders online.
Find a company that can develop the software. Find a company that can host
the application. If possible, estimate the costs.
25. Assume users are complaining about lack of support from the MIS
department. How can you improve MIS responsiveness? How can you do it
without substantially raising costs?
Additional Reading
Arnett, Kirk P. and C.R. Litecky, Career Path Development for the Most Wanted
Skills in the MIS Job Market, Journal of Systems Management, February
1994, pp. 6-10. [Job skills.]
Arnold, David and Fred Niederman, The global IT workforce, Communications
of the ACM, July 2001, vol. 44, no. 7. [A special section on global issues in
IT management.
Bringing I.T. Back Home, CIO Magazine, March 1, 2003. [Failure rate of 78
percent on outsourcing projects.]
Fryer, Bronwyn, Diffcult at best, Computerworld, January 4, 1999, p. 38.
[High demand for staff with ERP skills.]
Managing Unruly Desktop Computers Costs Businesses Dearly, The Wall
Street Journal, February 16, 1995, p. A1. [Maintenance costs of personal
computers.]
McWilliams, Gary, After Landing Huge Navy Pact, EDS Finds Its In Over Its
Head, The Wall Street Journal, April 6, 2004. [EDS loses $1.6 billion in
trying to build a network for the Navy.]
York, Thomas, Shift in IT roles ahead: changes in business and technology will
alter IT careers, Infoworld, January 18, 1999, p. 75. [Predicting the future of
IT jobs is hard, but useful.]
Cases: Te Energy Industry
Te Industry
How do you control a huge organization geographically dispersed around the
world? How do you choose and implement information technology to help work-
ers perform tasks from simple communication to advanced nonlinear analysis of
seismic data? Then what happens when you merge two of these huge companies?
Check the petroleum industry for different answers to these questions. The BP/
Amoco, Exxon/Mobil, Chevron/Texaco, and Conoco/Phillips mergers of the late
1990s and early 2000s represented giant combinations. The mergers resulted in
lower administrative costs. But they also made it more important to combine data
and standardize processes across the new organizations.
0
100
200
300
400
500
600
19941995199619971998199920002001200220032004200520062007200820092010
B
i
l
l
i
o
n

$
Annual Revenue
ExxonMobil
Royal Dutch Petroleum
ChevronTexaco
0.00
0.02
0.04
0.06
0.08
0.10
0.12
19941995199619971998199920002001200220032004200520062007200820092010
R
a
t
i
o
ExxonMobil
ChevronTexaco
The industry is also facing an eventual shortage of expertise, as the few in-
dustry experts retire. Many of the frms are turning to technology to leverage the
knowledge of these workers. Communication technologies and high-resolution
digital cameras make it possible for experts to diagnose problems remotely, re-
ducing travel time. A few companies are experimenting with expert systems and
knowledge management to store the accumulated knowledge of the experts.
Exploration
The energy sector contains a variety of companies and some interesting interre-
lationships and partnerships. The upstream process of oil exploration is notori-
ously variable. The exploration companies perform enormous data collection and
analysis operations around the world. When oil prices are high, their services are
in high demand. To spread the risk most of the exploration companies are inde-
pendent and provide services to all of the major petroleum companies. But that
means rivals end up sharing data on some projects. Analytical tools that have been
refned and improved over the years can provide a competitive advantage to a
large company, so they are cautious about sharing techniques.
The exploration companiessuch as Amerada Hess, Anadarko, and Schlum-
bergerwork around the world. They invest billions of dollars in both mechani-
cal and information technology. According to the Energy Department, the total
cost of fnding and producing a barrel of oil dropped from $15 in 1977 to $5 in
2001primarily due to improved technology (Ricadela 2002). Randall Notting-
ham, an analyst at Strategy Analytics, says that the oil industry spends $9 billion
a year on IT, not counting the oil feld technology and robotics (Ricadela 2002).
The oil exploration side of the business generates petabyes of data. All major
oil felds and potential felds around the globe have been examined with seismic
sensors. Producing felds generate even more data collection. Amerada Hess has
100 terabytes of data in its Houston center alone. All of it is refned and analyzed
through complex geological models. These models are proprietary to the various
research frms. They require huge amounts of computational power. Mike Heag-
ney, Suns global energy manager says that oil companies are probably second to
NASA in data volumes. The systems still arent fast enough; they consume pretty
much anything we put out (Ricadela 2002).
To process that much data, Sun sells 106-unit server farms. IBM sells its mon-
ster symmetric processor systems as well as giant Linux cluster grids. John Sher-
man, executive vice president of marketing and systems for Landmark (a division
of Halliburton), observes that what were trying to do in essence is look down
in the earth. These [systems] are MRIs on supersteroids (Weil 2002). Amerada
Hess also employs a 200-node Dell cluster to run the huge analytical jobs. The
PC-based systems cost about 10 percent of the price of the large IBM systems. A
few companies perform the analysis as outsourcersbut the analytical tools are
shared across the industry, reducing the competitive value.
Refneries
Refneries are some of the most complex chemical and mechanical systems in
existence. Much of the systems are monitored and controlled through remote sen-
sors coupled to computer systems. But the systems still need to be monitored by
humans. It is critical to catch and repair a component before it fails. If something
small fails, it can easily wipe out other components down the line.
Houston-based SAT corporation created a portable electronic system that man-
agers can use to diagnose problems. Each component is given an RFID tag. Work-
ers carry handheld computers that read the tag and provide device-specifc data
to check. For example, workers might be told to check the temperature, pressure,
and vibration on a pump. The handheld expert system then provides instructions
on what to do if problems exist. Bill Johnson, reliability manager at Lyondell-
Citgo Refning LP, notes that this thing will prompt some action. It allows us to
identify problems earlier and do better troubleshooting when we identify those
problems (Thibodeau 2004). The data is eventually uploaded to the main com-
puters to check for trends over time.
Because oil is so important to the current U.S. economy, the government and
the frms are concerned about security from terrorists. Along with enhanced physi-
cal security measures, the companies are trying to create some type of network to
help them share data and identify trends and threats. Protecting the IT infrastruc-
ture used to fnd and produce oil is also important. The main oil and gas compa-
nies have created the Energy Information Sharing and Analysis Center (ISAC).
Users or outsiders can post warnings on the system. These warnings are analyzed
and then threat notices are sent to the member companies.
The process is not that simple for the oil companies themselves. Mark Evans,
CIO at Tesoro Petroleum Inc., notes that it is hard to retrieve information from
the Supervisory Control and Data Acquisition (SCDA) systems that run the op-
erations at most companies. For a long time, weve been unable to share that
information within our own company. Thats really the frst step (Meehan 2002).
Ultimately, the companies want to share security tips and best-practices informa-
tion as well as notices of immediate threats.
Meehan, Michael, Energy Firms Move To Thwart Cyberattacks,
Ricadela, Aaron, Pay Dirt, Information Week, March 18, 2002.
Thibodeau, Patrick, SAT Corp.s Handhelds Help Refneries Quickly Spot
Problems, Computerworld, June 7, 2004.
Weil, Nancy, IBM, Landmark Deal Takes Linux To Oil Industry,
Case: ExxonMobil
ExxonMobil (ticker: XOM) is the largest of the newly combined oil companies.
With over $370 billion in revenue and more than 83,000 employees in 2010, it is
the second largest company in the United States. The company is involved in all
aspects of oil production and marketing, from exploration to refning, chemicals,
and distribution and marketing. The company has several proprietary software
tools to help fnd and analyze deposits, including its Stellar basin-modeling soft-
ware (www.exxonmobil.com).
The oil glut of the late 1990s and early 2000s drove crude oil prices down to
$10 a barrel, a number that quickly became nostalgic after the $40 and $50 a bar-
rel prices in 2004 during the Iraq war, followed by the $90-100-120 prices later
in the decade. The unusual aspect to the price surge is that the oil companies have
not dramatically expanded their exploration and production. In the past, compa-
nies would instantly respond to price signals, and the eventual increase in produc-
tion would lead to a drop in prices. In 2003, Exxons spending on exploration and
production rose by 15 percent, but was fat through 2004 (Warren 2004a). One of
the big questions is whether oil prices will remain high. The frms must forecast
the price of oil for the near future.
With businesses in 200 countries, ExxonMobile relies on a suite of software
representing best practices to reduce the complexity of its information systems.
Patricia C. Hewlett, vice president of global information technology, said the IT
organization has developed an upstream suitcase, or collection of tools that can
be installed when the company moves into a new market. The system includes the
ERP system, tools to monitor equipment, track personnel, and manage work per-
mits. Its a suite of standard computing applications we can use anywhere in the
world, she said. She also noted that the standardized package enabled the com-
pany to reduce staffng by about 15,000 employees after the merger. Exxon is one
of the worlds largest users of SAP, and it took considerable effort to install the
system. But Hewlett observed that it helped improve the quality and timeliness
of data, and weve gotten even greater benefts than forecast. In terms of tech-
nology, the company does experiment with some leading edge ideas, but it sticks
with standard, well-tested applications for systems that directly affect operations
and productivity (Mitchell 2006a).
With its huge size, global reach, and thousands of employees, ExxonMobile
wanted to create an identity-management system to simplify user logins and better
control access rights. But, in 2006 when the company began testing various prod-
ucts, Ms. Hewlett found that they could not scale up to the needs of ExxonMobile.
She noted that available products could handle a small number of static roles but
were not well suited to managing dynamic, attribute-base roles. The company
also found that even though tools such as Microsofts Active Directory could track
users and roles, none of the companys applications were designed for rolesso
they all needed to be rewritten. The project was put on hold until some of the
problems could be resolved (Mitchell 2006b).
Merger
Exxons takeover of Mobil was a major milestone both for the industry and for the
IT department. SAP was chosen as the main fnancial IT platform. Both compa-
nies had been running SAP software, but they were separate global systems and
required changes so that the fnal system was using a single set of accounts and
defnitions. Plus, Mobil had only begun its conversion to SAP a few months be-
fore the takeover. Since the takeover occurred shortly before 2000, both IT depart-
ments were also busy working on Y2K updates (King 1998).
To promote competition, the FTC required the two frms to divest 2,431 gas
stations in the United States. Mobil had more than 3,500 stations confgured to use
its Speedpass RFID system. Waving the key-ring device in front of the pump trig-
gers the system to access the customers credit record and authorize the charge via
satellite. The newly joined IT system also had to install the system in the merged
stores (Hamblen 1999).
In 2000, the company announced that it was implementing mySAP for all of
its employees. mySAP is a Web-based platform that provides links into the com-
panys main databases. A key goal was to integrate all of the data and provide
a consistent interface. Suzanne McCarron, an ExxonMobil spokesperson, notes
that by consolidating and upgrading our systems, we will streamline our business
processes, lower information system support costs and provide access to common
consistent dataall of which will result in overall cost savings, rapid information
technology project implementations and improved performance (Songini 2000).
The system provides access to supply chain management, plant maintenance, hu-
man resource, and accounting tools. It also contains industry-specifc oil and gas
applications. In particular, the specifc functions include dispatch planning and
optimization and inventory management. In terms of profts, the higher prices for
exploration and production added to Exxons profts. However, the increase in
price at the pumps caused consumers to cut back on driving, reducing sales vol-
ume (Cummins and Warren 2004).
The oil business is closely tied to politics in many countrieswhich consider-
ably increases the risks. In 2003, Exxon was in talks to invest in Russian oil felds
through Russian oil baron Mikhail Khodorkovsky. Exxon would love to gain a
foothold in the Russian oil production industry. Unfortunately for Exxon, Khodor-
kovsky was arrested and imprisoned before any deals could be reached. Russian
president Putin charged Khodorkovsky with failing to pay taxes and had him ar-
rested. Many observers believe Putin was more concerned about Khodorkovskys
political ambitions (White, Whalen, and Warren 2004).
Finding new oil felds or even mapping the extent of existing felds is a diffcult
problemparticularly when the felds are in hostile environments such as the ar-
tic or deep sea. The oil companies have relied on seismic surveys and drilling for
years, and many experts have decided that no new easy oil reserves exist. Faced
with relatively low prices and stiff competition, many oil frms cut back on re-
search and development in the late 1990s and early 2000s. Harry Longwell, an ex-
ecutive vice president at ExxonMobile took a different perspective and launched a
program in 1996 to try new ideas and fnd new ways to search for oil. The concept
of using electromagnetic systems to search for oil has been around for years, but
no one had been able to make it work. Funded by ExxonMobile, Dr. Len Srnka
found a way to search for oil deposits in deep-sea areaslocations that cost mil-
lions of dollars to drop a well. The technology makes it easier to fnd the specifc
location of oil deposits, providing an accurate target for drilling (Warren 2004b).
In 2006, the organization had more than 2 petabytes of operational data online.
Most of the data involves operations, including refnery and production facilities.
The company wants to build a knowledge management system to reduce duplica-
tion, index everything, and make it available for decision-makers (Mitchell 2006).
Questions
1. What benefts did ExxonMobil gain by centralizing on SAPs ERP software?
2. How does ExxonMobil use technology to reduce costs?
3. What information technology problems can ExxonMobil expect if it tries to
expand into Russia?
Additional Reading
Cummins, Chip and Susan Warren, High Prices Help Exxon Mobil, Shell, The
Hamblen, Matt, Merged Exxon Mobil Faces IT Issues, Computerworld,
December 1, 1999.
King, Julia and Kim S. Nash, Exxon/Mobil Sets Up Mega SAP Project,
Mitchell, Robert, Exxon Mobil: Focus on Flexibility, Computerworld, October
30, 2006.
Mitchell, Robert, Stepping Into Identity Management, Computerworld,
November 20, 2006.
Songini, Marc L., Exxon Mobil Adopts Mysap.Com as Its Primary Backbone,
Warren, Susan, Oil companies Curb Their Spending, The Wall Street Journal,
June 1, 2004.
Warren, Susan, Exxon Mobil Bets On New Technology In the Hunt for Oil, The
Wall Street Journal, August 17, 2004.
White, Gregory L., Jeanne Whalen, and Susan Warren, A Global Journal Report;
Tough Drill: For Wests Oil Giants, Vast Fields In Russia Prove Hard to Tap,
Case: Royal Dutch Petroleum (Shell)
Royal Dutch Petroleum (ticker: RD) is the main holding company for Shell Oil. It
is headquartered in the Netherlands. The company did not participate as heavily in
the late 20
th
century mergers. Still, in 2010, it had sales of about $368 billion and
97,000 employees. Note that the number of employees had been increasing, com-
pared to the declining number of employees at ExxonMobile, but it did drop from
2009 to 2010. The company experienced a major public relations setback in 2004
when it was forced to restate its fnancial data for 2002 and 2003. The change
was driven by an admission from Shell offcials that they had overstated the com-
panys oil and gas reserves by 22 percent. Four senior executives were fred as
a result of the misstated information (Wang 2004). The company has fallen be-
hind the other major producers in developing reserve felds. In 2004, Shell began
pushing more money into exploration. Malcolm Brinded, head of exploration and
production, said that he was going to focus on high-margin projects in the UK and
the United States. The goal is to re-establish the competitiveness of the portfolio
(Cummins and Warren 2004).
Shell spends a considerable amount of money on hiring temporary workers, as
much as $100 million a year. Most of the departments used a homemade collec-
tion of paperwork to hire, track, and pay the workers. The company was working
with 20 different organizations that supplied contract labor and needed to cut its
costs. Shell reduced its outside service providers to four and then installed soft-
ware from IQNavigator. The system automates most of the processes needed to
hire, track, and pay the temporary workers. By consolidating the information,
Shell can negotiate longer-term and volume discounts with the main suppliers. To
cut costs even more, the outside suppliers are the ones who pay for the software,
through a 5 to 8 percent assessment for each hire (Hoffman 2003).
For several years in the mid-2000s, Shell faced problems with identifying the
status of its oil reserves. In 2004, the company downgraded the book value of its
reserves by 20 percent, leading to a major restructuring and the departure of top
executives. After being sued by investors complaining about misleading informa-
tion, Shell settled for $350 million (Hanney 2007).
Outsourcing and Standards
Since Shell is a smaller player in the market, it has found it necessary to partner
with other companies to reduce costs and expand its capabilities. In particular,
the company partnered with Chevron and Schlumberger Ltd. to defne a vendor-
neutral suite of applications for petroleum companies. Known as OpenSpirit, the
technology should make it possible to integrate applications from multiple ven-
dors into one framework (Ohlson 2000).
The OpenSpirit system is specifcally designed to transform data from a variety
of common sources, including seismic systems. It can scan databases, handle 2-D
and 3-D projects, and maintain everything in a GIS database. The system sup-
ports multiple languages and platforms, including Java and C++ on Windows, So-
laris (Sun), and Linux. It also contains connectors to ArcView (for GIS) and Excel
(www.openspirit.com).
Shell has also turned to outside vendors to provide additional expertise in inte-
grating data. In 2001, the company chose IBM to confgure and set up three new
data centers located in Houston, The Hague (headquarters), and Kuala Lumpur.
Shells general manager for IT projects, Alan Matula, said that we were look-
ing for a trusted technology partner to help us achieve aggressive TCO [total
cost of ownership] targets in our MegaCentre project. It is one of the most im-
portant IT initiatives in Shells history (Vijayan 2001). Shell also worked with
IBM to develop a high-performance Linux-based cluster to analyze seismic data.
In 2004, Shell went even further and negotiated an agreement to outsource most
IT functions to India-based Wipro and IBM. The company is trying to reduce its
9,000-employee IT workforce by 30 percent by 2006 in an effort to cut about
$850 million a year (McDougall 2004).
Shell also realized that its communication network was fragmented across
business units, according to Rob van Zwieteren, the telecommunications man-
ager. The goal is to save $50 million over three years by consolidating all commu-
nications into a single network infrastructure. The master contract is handled by
Cable & Wireless (Cope 2001).
In early 2008, Shell signed a $4 billion outsourcing contract with AT&T, EDS,
and T-Systems. In the initial fve-year deal, almost 3,000 IT workers were trans-
ferred to the service providers. Each of the three partners would specialize, so
minimal overlap existed. AT&T handled networking and telecommunications, T-
Systems controlled hosting and storage, and EDS was responsible for end-user
computing and integration of the infrastructure (Chapman 2008). The IT orga-
nization of 8,000 employees and contractors handles IT tasks for 150,000 users
worldwide.
Unfortunately, contract employees can be more diffcult to control than tradi-
tional full-time employees. In October 2008, Shell learned that someone had used
Shell employment data to fle fake unemployment claims. Without naming the
worker, offcials noticed that had pulled records from the master employment da-
tabase and used the data to fle false claims to obtain unemployment insurance
benefts (McMillan 2008).
Solving problems in a huge company with experts scattered around the globe can
be a challenge. Shells Arjan van Unnik notes that what we had was a commu-
nity of expatriates who might link up when they encountered a problem. We had
knowledge management, but not that much (King 2001). To improve communi-
cation and sharing, Shell implemented a $1.5 million project using off-the-shelf
collaboration software in 1999. The system evolved into 13 Web-based communi-
ties used by more than 10,000 employees. By sharing technical data and providing
the opportunity for employees to ask questions, the system is estimated to have
provided $200 million in benefts in less than two years (King 2001).
One challenge with KM projects is that the name has garnered a negative rep-
utation. Several large projects were created that tried to capture wide levels of
knowledge across an entire organization. Projects that large and broad tended to
fail, giving KM a bad name. Yet the concept of KM is still importantparticu-
larly in a geographically diverse company, and in an industry facing retiring ex-
perts. Consequently, the KM label was discarded on the Shell project, and it was
renamed new ways of working. The group found that the system actually had
to direct user discussions. It was not enough to just store and retrieve knowledge.
People need more guidance. In particular, when the system splintered into more
than 100 communities, the IT group had to redefne the groups and educate people
to reduce the number of communities down to 12 specifc communities (Kontzer
2003).
Questions
1. What problems has Royal Dutch experienced because of decentralization?
2. How is Royal Dutch using information technology to improve
communications in its decentralized environment?
3. How is Royal Dutchs focus on reducing IT costs and centralizing services at
three data centers going to solve its problems?
Additional Reading
Chapman, Siobhan, Shell Signs $4B Multisupplier Outsourcing Deal,
Cope, James, Shell To Set Up $250M Global Data Pipeline, Computerworld,
October 15, 2001.
Cummins, Chip and Susan Warren, High Prices Help Exxon Mobil, Shell, The
Hanney, Brian, Shell Pays $350m in Reserves Fiasco, Accountancy Magazine,
May 2007.
Hoffman, Thomas, Contingent Workforce: Managing the Temporary Players,
King, Julia, Shell Strikes Knowledge Gold, Computerworld, July 16, 2001.
Kontzer, Tony, The Need to Know, Information Week, August 18, 2003.
McDougall, Paul, Shell Objects To Reported $1B Outsourcing Price Tag,
McMillan, Robert, Shell Fingers IT Contractor in Theft of Employee Data,
Ohlson, Kathleen, Chevron, Shell and Schlumberger Team on Energy Software
Venture, Computerworld, October 9, 2000.
Vijayan, Jaikumar, Shell, IBM Agree To $100M E-Business Applications Deal,
Wang, Michael, Shell Names Peter Voser as New Finance Chief, The Wall
Case: ChevronTexaco
Chevron merged with Texaco at the end of 2001. The new frm (ticker: CVX) had
sales of over $120 billion in 2003 with over 60,000 employees. It is headquar-
tered in San Francisco. The merger created some issues with the ongoing Internet
explorations by both companies. But, the dot-com crash at about the same time
made those issues irrelevant (King 2000).
One of the more interesting aspects of the merger was the consolidation of their
telecommunication networks. Because both companies had operations around the
world, they had both contracted with AT&T to provide international telecommuni-
cation services. The merger provided the opportunity to consolidate the networks
to reduce costs. In terms of fnancial systems, Chevron was using SAP software,
while Texaco had an older custom system built over the years. Texaco replaced
the system with SAP prior to the merger. (Collett 1999).
Partly because of the merger, partly because of the expanded use of the Inter-
net, from 1999 to 2001, Chevrons Internet demands increased by 200 percent a
year for those three years. The company upgraded routers and servers to handle
the new demands (Maselli 2001).
Managing desktops in a company scattered around the world is diffcult. Dave
Clementz, the CIO, standardized the company on a common network backbone.
He also rolled out Windows XP early in 2002 to ensure that everyone had the
same platform. For some cases, he installed thin-client terminals. The applica-
tions for the thin clients are stored on a central server, making it easier to update
and troubleshoot the applications. The company is also planning to move to Web-
based applications. Employees who need only simple tools, like e-mail, could run
everything through a thin-client browser (Maselli 2001). IT directors at both com-
panies were leaning to Web-based applications to provide more centralized man-
agement and reduced cost of applications.
In 2003, the combined ChevronTexaco signed an agreement to work with IBM
and BearingPoint for supply chain management, procurement, and outsourcing.
IBM was a lead contractor on merging the SAP systems of the two companies.
IBM was also negotiating to provide outsource facilities and hosted services (Mc-
Dougall 2003). ChevronTexaco CEO David OReilly in 2003 that combining IT
operations from the two companies has saved the company $2.2 billiontwice
the amount originally projected. When the two companies began the merger, they
realized early on that they needed outside help (Evans 2003).
Big companies with multiple locations also generate issues with security. It is
diffcult enough to track IT assets. (Who is currently assigned to which comput-
er?) It is also diffcult to control user access to systems and applications. Chevron
purchased eProvision Day One software from Business Layers to manage these
processes. The system uses Lightweight Directory Access Protocol (LDAP) to
track the user groups. It stores account data and access rights for voice mail, e-
mail, and even cell phones. The system consolidates data from dozens of older
databases used throughout the companyoften including Excel spreadsheets.
Centralizing the account data makes it easier to see and control user access rights.
It also makes it easy to remove all permissions when an employee leaves the com-
pany (Verton 2001).
Chevron uses a high-technology Cooperation Center at its San Ardo, CA oil
feld to help managers visualize construction, maintenance, and drilling opera-
tions. Jim Crompton, IT adviser for Chevrons operations observed that people
come out of their functional view and begin to see the same picture. The work
they do is so interdependent. If you drill a new well and it produces 100 barrels of
oil, you have to have the pipeline, you have to have the processing capability, you
have to have everything thats downstream of that well or you end up having to
suspend operations in one place until it all works together (Chabrow 2006).
In total, Chevron Corp. adds 2 terabytes of data a day to its systems, but Gary
Masada, corporate CIO sees the data as an opportunity, noting its an issue of,
you have this information; how are you going to search it and use it? In many
cases, Mr. Masada prefers large projects because they are better integratedlead-
ing to greater benefts. For instance, the Global Information Link connects 50,000
desktops at 1,800 locations and was projected to save the company $50 million a
year (Anthes 2006).
In 2008, Louie Ehrlich was appointed CIO of Chevron. In an interview, he talks
about how he handled the transition. His frst action was to go around and talk
with the 60 leading executives at the company and ask them three basic questions:
What was important to you in your business, whats working well with IT at
Chevron and whats not? Later, he held lunch meetings with typical employees
and asked them similar questions (Mitchell 2009). He noted that the high number
of mergers had created duplication and an complex IT environment. One of his
primary goals was to simplify, reduce the number of vendors and reduce the num-
ber of applications.
In 2011, Chevron was interested in mobile computing and tested using tab-
lets such as the Apple iPad. Peter Breunig, general manager of technology man-
agement and architecture noted that In 1990, it was the PC desktop with Excel.
Now you can give them mobile platforms that are more powerful and more com-
pact. Thats like having calculators on steroids, with tie-ins back to corporate data
stores. Thats an opportunity, in my mind (Hamblen 2011).
Questions
1. What benefts does ChevronTexaco gain with thin-client and Web-based
applications?
2. Why is ChevronTexaco using LDAP to consolidate its user accounts instead
of standardizing the underlying systems?
3. What did ChevronTexaco gain by consolidating its international
telecommunication systems? Were there any other options?
Additional Reading
Anthes, Gary, Chevron: Where Size Is Opportunity, Computerworld, October
30, 2006.
Chabrow, Eric, Oil Companies Turn to IT to Shave Costs, Boost Effciency,
Collett, Stacy, Chevron/Texaco Merger Could Recast AT&T Role,
Evans, Bob, Business Technology: Sometimes Letting Go Can Really Pay Off,
Hamblen, Matt, Chevron and TD Bank Try Tablets, Computerworld, February
24, 2011.
King, Julia, IT Implications Linger for Chevron/Texaco Merger,
Maselli, Jennifer, Chevron Bolsters IT Infrastructure to Accommodate Growth,
Information Week, September 17, 2001.
McDougall, Paul, ChevronTexaco Prefers IBM For Services, Information
Week, May 22, 2003.
Mitchell, Robert L., The Grill: Louie Ehrlich, Computerworld, April 13, 2009.
Verton, Dan, Chevron Adds Software To Control Network Access,
Case: Exploration
NuTec Energy Services is a specialty company that provides data and data visu-
alization services for oil exploration. In particular, the Houston-based company
takes raw seismic data and converts it into 3-D images. The company focuses on
the Outer Continental Shelf (OCS) in the Gulf of Mexico. Customers purchase
data by the OCS block which is one square kilometer. The analyses use from 2
to 20 terabytes of data per project. Although proftable, the company was being
squeezed between customers who wanted lower prices and increased computa-
tional and storage costs. NuTec solved its cost problems by moving to Linux-
based servers and installing a dual-controller storage area network (SAN). The
open-source system reduced costs by 84 percent. The challenge is that the servers
have to deliver the massive data quickly to the 220 or more servers performing
the computations. The system requires a high-speed network to move the data off
the storage devices to the requesting computers. To solve the problem, NuTec in-
stalled two subsystems. One system serves the highly nonlinear analysis comput-
ers that perform intensive computations but pull data in bulk. Data for that system
is stored on an EMC Clariion FC4700 linked to four servers running GFS. It feeds
130 Linux computational servers. The second analytical system performs a sim-
pler calculation, but draws huge amounts of data quickly. Data for that time imag-
ing system is stored on a MetaStor E4600 array from LSI, routed through McData
6140 SAN switches to 220 computational servers (Scheier 2004).
Anadarko Petroleum in Woodlands, Texas, faced similar problems with data
storage. The company needed to consolidate the seismic 2-D and 3-D data used
by its engineers, geologist, and geophysicists. The company installed network-
attached storage (NAS) devices from Network Appliance. The new system uses
EMC Symmetrix boxes to hold 110 terabytes of data. The conversion took 20
months. But CIO Morris Helbach notes that an exploration and production com-
pany lives and dies by the way it acquires, manages and provides access to data
(Songini 2003). In 2006, Anadarko expanded its market position by acquiring two
large competitors, Kerr-McGee and Western Gas Resources, making Anadarko
the largest independent oil and gas producer in the United States (Weekly Corpo-
rate Growth Report 2006).
British Petroleum (BP) faced a slightly different problem with similar data. The
third-largest oil company has hundreds of leases in the Gulf of Mexico. But it does
not have the staff to analyze all of the data. Steve Decatur, staff development de-
ployment leader in the Houston offce, observed that it would take us four years
with our current manpower to get all these properties analyzed (Bryce 2001).
The company found an innovative solution. It created a Web site and invited free-
lancers to analyze the data. Freelance analysts who develop suitable drilling plans
get $50,000. If the wells produce oil, they receive a cash bonus. If BP does not
drill, the analysts can take the plans to another company. The company initially
provided data on fve Gulf tracts. Based on the quality of responses, it eventually
added 18 more to the site (Bryce 2001).
Having the information is only the frst step. Employees also need to be able
to fnd the information needed when they have a question. Knowledge Manage-
ment tools can help provide access to data. Intec is a Houston-based engineering
frm that provides support and project management to the oil industry around the
globe. Beginning as a small frm, the engineers were collaborative and shared in-
formation on 3x5 fle cards. By 2002, the company was too big to use paper-based
information. A team purchased KM software from AskMe and built a library-
based system to store the companys knowledge and engineering expertise. Now,
engineers can go online, ask a question, and get results and sample documents
from the library in minutes (Melymuka, 2003).
Oil felds are often in remote locationsand exploratory drilling operations
tend to be even farther away from civilization. Yet communicating among the
workers on site and with managers and coworkers around the world is still im-
portant. A startup company (Texas Energy Network) built a network that runs on
the long-term evolution (LTE) phone system to provide high-speed access in out-
of-the-way places. Instead of trying to run cables and fber optic lines, the LTE
system uses wireless modems. The company hopes to cover several miles with a
centralized tower (Lawson 2010).
Questions
1. What are the benefts to outsourcing analysis of oil feld data? What are the
drawbacks?
2. What information technologies are important in the oil exploration industry?
Additional Reading
Bryce, Robert, BP Drills Web, eWeek, July 9, 2001.
Lawson, Stephen, LTE Demo Will Test Broadband for Oil Fields,
Melymuka, Kathleen, Knowledge Management Helps Intec be Smarter by the
Hour, Computerworld, June 23, 2003.
Scheier, Robert L., Seismic Data Firm Boosts Throughput, Cuts Costs With
Super-Scalable SAN, Computerworld, January 6, 2004.
Songini, Marc L., Profle: Joan Dunn, Computerworld, January 6, 2003.
Weekly Corporate Growth Report, Anadarko Petroleum to Purchase Kerr-
McGee Corporation for 2.77 Times Revenue, July 3, 2006.
adopted by rivals?
sectors?
916
How does your company affect the rest of the world? What infuence does the outside
world have on your company?
How does information technology affect individuals? As a manager and a company, do
you treat individuals the way you expect to be treated by other companies?
How does technology affect jobs? If computers do more of the work, what jobs are left
for people?
How does technology change the relationship between businesses and consumers?
Can information technology change education?
How does technology affect different areas of society?
Can information technology improve governments?
Do criminals know how to use computers?
How do your actions affect society? Is it possible to follow the laws and still be wrong?
What major laws affect technology and the use of computers?
What risks are created through using cloud computing?
Chapter Outline
Information Management
and Society
14
Chapter
Introduction, 917
Individuals, 918
Privacy, 919
Privacy Laws and Rules, 930
Anonymity, 931
Jobs, 932
Loss of Jobs, 932
Physical Disabilities, 934
Telecommuting, 934
Business: Vendors and Consumers, 936
Intellectual Property, 936
Digital Rights Management, 940
Balance of Power, 942
Education and Training, 944
Social Interactions, 945
Social Group Legitimacy, 945
Access to Technology, 948
e-Mail Freedom, 949
Liability and Control of Data, 949
Government, 950
Government Representatives and
Agencies, 950
Democracy and Participation, 950
Voting, 951
Information Warfare, 953
Rise of the World-State?, 955
World Government Cases, 956
Crime, 959
Police Powers, 959
Freedom of Speech, 960
Gambling, 961
Responsibility and Ethics, 962
Users, 962
Programmers and Developers, 963
Companies, 964
Governments, 964
Some Computer-Related Laws, 965
Property Rights, 965
Privacy, 968
Information Era Crimes, 971
Data in Multiple Countries, 973
Threats to Shared Servers, 974
Subcontractors, 975
Summary, 976
Key Words, 977
Exercises, 979
Cases: Health Care, 983
917 Chapter 14: Information Management and Society
Sutter Health
Introduction
How does your company affect the rest of the world? What in-
fuence does the outside world have on your company? Why
should you care? Try an easier question: How much do your customers care about
privacy? When you use information technology to help your company, it means
you are collecting and analyzing data on customers and employees. As shown in
Figure 14.1, your company lives within an environment. Companies infuence the
world through relationships with customers, employees, and other companies. In
turn, your organization is affected by events in the world ranging from govern-
ment laws to education and public opinion. When you make business decisions,
you need to think about these interaction effects. Even if you cannot change the
world yourself, you should be aware of the effects of your choices so that you are
ready to deal with the consequences.
If nothing else, history has shown that technological change is inevitable. Com-
petitive economics virtually guarantees that the search for new products, new
manufacturing techniques, and other ways to gain competitive advantage will
continue. Changes in technology often affect society. Technology can change in-
dividuals, jobs, education, governments, and social interactions. As components
of society, each group has rights and responsibilities to others, such as a right to
privacy and obligations regarding ethics.
Technology effects on individuals can be benefcial or detrimental. Often a
change in technology helps one set of individuals and harms another group. Typi-
cal problems include loss of privacy, depersonalization, and changing incentives
or motivations. Advantages include lower prices and better products and service.
The effect on jobs is hard to predict, but most observers conclude that workers
will require more education and training. Most authorities think that increases in
technology in the past generally led to an increase in the number of jobs. Now,
however, many of the new jobs require higher levels of education, and the work-
ers displaced by technology rarely have the qualifcations needed for the new jobs.
How do information systems affect society? From one perspective, health care or-
ganizations are simply another business. Yet, because of their costs, importance to
our daily lives, and widespread governmental involvement, health care has a spe-
cial role in society. Not surprisingly, most physicians and other health care work-
ers receive minimal training in information systems and are not comfortable with
it. Consequently, information systems for medical care are being introduced slowly.
On the other hand, information systems have the ability to reduce errors as well as
costs. In particular, health care organizations like Sutter Health are working to im-
plement electronic drug ordering and dispensing systems. Sutter is also working to
implement paperless medical offces and already has 400 member physicians using
a completely electronic systemincluding X-rays and prescriptions. The group is
also working with telemedicineparticularly with intensive care units (ICUs) to re-
duce the costs of physicians and provide better care to remote locations. Of course,
security and privacy issues become major areas of concern with medical information
systems.
Technology also has an effect on crime. Technology creates new crimes, new
ways to commit crimes, and new ways to catch criminals.
In addition to the increased demand, technology has provided new teaching
methods. Although there is considerable debate over the costs and benefts of
technology in education, there is usually a place for technology, even if only as
a specialized technique. However, most educators know that technology cannot
succeed by itself. Technology can isolate or it can build connections, but it re-
quires people and applications to improve learning.
Governments attempt to control these impacts of technology by creating laws,
but laws often bring their own problems. Also, in times of rapid change, laws rare-
ly keep up with the changes in technology. Governments are also directly affected
by improved communication facilities. For example, technology makes it possible
for governments to better understand the needs of the citizens and provide more
avenues for communication.
Technology can alter any number of social interactions. Social groups can gain
or lose power, and types or methods of criminals are altered. Furthermore, society
can become dependent on technology, which is not necessarily bad, but it causes
problems if the technology is removed or substantially altered.
Individuals
How does information technology affect individuals? As a man-
ager and a company, do you treat individuals the way you ex-
pect to be treated by other companies? Information technology plays
an important role in the lives of most individuals. Many jobs are directly involved
in the collection, processing, and evaluation of data. Performance of many work-
ers is continually monitored by computers. As consumers, virtually our entire lives
are recorded and analyzed. Governments maintain massive fles on all public as-
pects of our lives. Increasingly, these public fles are accessible to anyone via the
Web and a few dollars. Privacy is a delicate and controversial issue. Citizens must
work together to live in a society, which requires giving up some elements of pri-
Figure 14.1
Information management and society. Every organization and individual exist in
a social environment. Changes in the frm and changes in technology affect the
environment. Changes in the environment can affect the frm. An understanding of
these interactions will make you a better manager.
Operations
Tactics
Strategy
Government
Consumers
Employees
Culture
Businesses
Privacy
Education
Company
vacy. Businesses and governments often need to identify customers and employ-
ees to perform basic functions. Yet history reveals that individuals can be threat-
ened or coerced if some people or organizations collect too much information.
Although data has been collected on citizens for many years, recent improve-
ments in technology raise greater concerns about privacy. As computer capabili-
ties increase, it becomes possible to collect, integrate, and analyze the huge vol-
ume of data. Using publicly available data, it is possible to collect an amazing
amount of data on any person.
Privacy
As Figure 14.2 indicates, companies, governments, and employers collect data
about many aspects of our lives. Most of the modern marketing efforts includ-
ing data mining and building customer relationships require information about
customers. Marketing and sales can be improved by maintaining databases of
consumer information and tracking sales and preferences at the customer level.
Combining government statistics and data from market research frms with geo-
graphical data can provide a precise picture of consumer demands. It also might
represent an invasion of privacy for individuals. With databases available even to
small companies, it is easy to acquire basic data on any individual. For instance,
phone numbers and addresses are readily available online. Data collected by gov-
ernmental agencies such as voter registration and property records can be pur-
chased from several online sources. More comprehensive commercial databases
are available from specialty marketing companies. Few laws exist that limit the
use of personal data.
It is easy to obtain lists from universities, clubs and social organizations, maga-
zine subscriptions, and mail-order frms. Statistical data can be purchased from
the U.S. government. Although most U.S. agencies are forbidden to release spe-
cifc individual observations until 50 years after the collection date, statistical av-
Trends
The industrial revolution in the late 18th century caused many changes to
society. Before the revolution, workers were predominantly employed as
craftsmen, farmers, or lesser-skilled laborers. Mechanization brought stan-
dardization and assembly lines, for which jobs were reduced to simple, re-
petitive tasks.
As transportation improved, people moved from farms to cities, and cities
spread to suburbs. Communication systems improved and linked the popu-
lations back together. Better product distribution mechanisms changed the
way products are sold. Companies (such as Sears, through its catalogs) be-
gan to distribute products nationally instead of relying on small local stores.
National and international markets developed with every change in the com-
munication and transportation systems.
These changes were so strong that philosophers and writers began to
take note of how technological changes can affect society. From the bleak
pictures painted by Dickens, Marx, and Orwell, to the fantastic voyages of
Verne, Heinlein, and Asimov, we can read thousands of opinions and predic-
tions about how technology might affect the political, economic, and social
environments.
erages can be highly accurate. By combining the statistical averages with your
address, your actual income might be estimated to within a few thousand dollars.
Because most people prefer to maintain their privacy, companies have an ethi-
cal (and sometimes legal) obligation to respect their wishes. Individuals can al-
ways ask companies not to distribute personal data. Companies should give con-
sumers the option of protecting personal data by building the option into their
databases and informing consumers whenever companies collect data.
Consumer Privacy
As shown in Figure 14.3, a tremendous amount of data is collected on consum-
ers. In the early years, consumer activists primarily worried about government
data collection. Governments had computers and the authority to force citizens to
provide data. Today, businesses can easily collect, obtain, and integrate almost the
same level of data available to government agencies. In fact, government agencies
have begun using commercial databases in some cases. Credit card and credit bu-
reau data are the two most detailed sources of consumer data.
Consumers have little control over the collection of personal data. But it is
interesting how cheaply people will give up their privacy. Grocery store loyalty
cards collect a tremendous amount of personal purchase data. Customers routinely
sign up for the cards in exchange for a tiny discount on prices. The purchase data
is sold to marketing companies and manufacturers to track sales and the success
of marketing campaigns.
Governments
Employers
Businesses
Figure 14.2
Privacy. Businesses, employers, and governments all have reasons to collect data
on individuals. The challenge lies in balancing these needs with the privacy of the
individuals. Historically, there have been many instances of abuse and fraud when
this balance is not achieved.
A signifcant question remains as to whether consumers really care about their
privacy. The loyalty card data and lack of concern over fnancial records indicate
that many people do not care about privacy. Yet over 50 percent of U.S. house-
holds signed up for the national do-not-call list to stop telemarketing calls. Per-
haps the conclusion is that customers do not mind having data collected, but they
do not like being solicited directly.
On the other hand, at times, consumers seem more concerned about online data
collection. The technology for Web sites did not initially consider the demands of
e-commerce. It was originally intended to simply display pages independently
every request for a page is independent of any other request. For e-commerce,
the Web server needs to track information about the person requesting a page.
For instance, a shopping cart system needs to store items selected by the cus-
tomer. Similarly, any site using security needs to track the user through a series
of pagesotherwise it would force the user to log in for every new page. These
problems were solved with the creation of magic cookies. A Web cookie is a
small text fle that the server asks the browser to store on the users computer.
As shown in Figure 14.4, whenever the browser requests another page from that
server, it returns the cookie fle containing an identifcation number. Hence, the
server knows which user is returning. This use of cookies is common and rela-
tively benign. Yes, the cookie could be used to track visitors, but presumably the
visitor is purchasing items and already willingly provides detailed information to
complete the transaction.
Figure 14.3
Privacy. Many businesses and government agencies collect data. With increasing
computer capabilities, it is becoming possible to collect and correlate this data to
track many aspects of individual lives.
credit cards
loans & licenses
financial
permits
census
transportation
financial
regulatory
employment
environmental
subscriptions
education
purchases
phone
criminal record
complaints
finger prints
medical
records
grocery store
scanner data
Blogs
Social networks
organizations
Figure 14.4
Web cookies. Cookies are used to keep track of the user across page requests. Each
time the user PC requests a page, it returns a small text fle (cookie) containing an
identifcation number.
Web server
User PC
time
Request page.
Send page
and cookie.
Display page,
store cookie.
Find page.
Request new
page and
send cookie.
Use cookie to
identify user.
Send
customized
page.
Reality Bytes: Medical Privacy
In early 2011, U.S. Representative Gabrielle Giffords and several others were shot in
one of the worst public shooting sprees in the United States. Less than a week later,
three employees of the Tucson University Medical Center were fred for accessing
the medical records of some of the victims without permission. A nurse working
under contract was also fred. The UMC noted that it had installed technology to
track unwarranted data access and the three were fred in accordance with UMCs
zero-tolerance policy on patient privacy violations. Although most medical workers
are careful about patient privacy, this example is not an isolated incident. In April
2009, a Kaiser Permanent hospital near Los Angeles fred 15 workers for accessing
medical records of a patient. In 2008, the University of California revealed that over
13 years, as many as 165 medical personnel, including physicians, had improperly
accessed records about celebrities.
Adapted from Jaikumar Vijayan, Three Fired for Accessing Records of Tucson
Shooting Victims, Computerworld, January 13, 2011.
Figure 14.5 shows a more troublesome use of cookies. In 2000, it was revealed
that Doubleclick.com (now owned by Google), the leading Web advertisement-
placing frm was using cookies as a third party to track page visits by millions of
people. Leading Web sites register with Doubleclick to carry advertising. Compa-
nies wishing to advertise on the Web create the ad and pay Doubleclick to carry it
on its servers. The original Web site includes a link to Doubleclick software that
delivers the ads and records page views so that the site owners can be paid the cor-
rect fee. However, Doubleclick also includes a cookie that is sent to the visiting
PC with each ad placement. Anytime the user visits a site that deals with Dou-
bleclick, the identifying cookie, date/time, and site visited are stored on Double-
clicks servers. Web users were understandably upset when Doubleclick attempted
to market this data collectionparticularly when the company wanted to tie the
online identities to real-world names and addresses. In effect, third-party cook-
ies enable the company to provide information on every site visited by users. In
the fall of 2010, the Wall Street Journal ran a series of article examining Web
sites, cookies, and other tracking devices. Researchers found hundreds of sites
tracking usersmany of them placing multiple tracking items on user computers.
To prevent this loss of privacy, browsers enable you to turn off cookiesbut
then you will not be able to use many secure sites, such as those run by banks.
Todays browsers offer more control over cookiesin particular, you can refuse
to accept third-party cookies such as those placed by Doubleclick. Finding the
option is not always easy, but it is one important step to protect your privacy. In
2011, Microsoft and other browser vendors began implementing new ideas for
specifying privacy levels in the browser. The main problem is that no Web sites or
companies are confgured to handle the new methods. And the new methods are
not standardized, so it will likely take time before usable privacy controls arrive.
Useful Web site
User PC
Useful Web Page
Text and graphics
[Advertisements]
National ad Web site
Doubleclick.com
Link to ads
Requested
page
Ads, and
cookie
Request
page
Hidden
prior
cookie
Figure 14.5
Misuse of Web cookies. Doubleclick.com as a third party places cookies onto visitor
PCs. Every time the visitor sees an ad delivered by Doubleclick, Doubleclick records
the user, the date/time, and the site visited. Doubleclick attempted to market this huge
database.
Wireless technologies offer even more methods for tracking people. Did you
know that over 50 percent of emergency calls are made on cell phones? The feder-
al e-911 law requires cell phone operators to provide location data on cell phones.
Manufacturers have embedded GPS locator chips within cell phones. Triangula-
tion and signal strength are also used as backup methods. While this data is useful
for emergencies, it could also be used for commercial purposes. The company and
Web site Loopt has a system where you and your friends can sign up, so that you
can open a map to see where your friends are located. This site is used by other
applications to enable tracking of your friends. Other sites such as foursquare en-
courage people to check-in to stores to indicate their location to friends, and the
rest of the world.
In the realm of autos, GMs On-Star system has been around for years and it
tracks the location of your car at all timesfor a monthly fee. In 2011, insurance
companies in some states began offering discounts to customers who install driv-
ing monitors in their carsnot only tracking location but also how you drive. As
an individual, do you care if you are tracked? Keep in mind that any stored data
can be retrieved by the police or by lawyers in a court case. This data has played
a role for both prosecution and defense attorneys in some high-profle cases. But,
if you are a completely honest person, perhaps the tracking systems could be
goodthey might be useful for proving your innocence if you are falsely accused
of somethingby the police or by your girlfriend/boyfriend.
Imagine the commercial opportunities of broadcasting messages to consumer
cell phones as they walk by your store. Now, ask yourself whether you want to be
continually interrupted as you walk through the mall, and whether you want stores
to be able to track that you walked by the store. On the other hand, the same tech-
nology could be used to broadcast emergency messages to any cell phone within a
danger zone (fre, tsunami, terrorist attack, and so on).
Employee Privacy
Computers have created other problems with respect to individual privacy. They
are sometimes used to monitor employees. Computers can automatically track all
of the work done by each person. Some employers post this data on public bul-
Reality Bytes: Google v. Spain
International laws on privacy are often stronger than those in the United States.
For example, the Spanish government has an Agency of Data Protection. The exis-
tence of the agency alone is importantit handles citizen complaints about the way
their personal data is handled on the Internet. And, the agency has taken a relatively
strong position with Google and its search system. The agency claims that Google
must delete links to any Web sites that contain information that might compromise
an individuals right to privacy. For example, whenever citizens fnd embarrassing
information about themselves on the Web, they can petition to have Google remove
any links to that data. In a legal case before the Spanish courts, Google argues that
the privacy agency does not force news agencies to remove any of the content. Other
countries target the original news source rather than the search engines.
Adapted From David Roman, Google Contests Spains Privacy Laws, The Wall
letin boards to encourage employees to work harder. Some software available for
local area networks enables managers to see exactly what every employee is do-
ingwithout the employees knowing they are being watched. Some employers
read their employees electronic-mail messages. Currently, all of these activities
are legal in the United States.
Many companies use electronic badges, which employees use to unlock doors.
The systems are run by a centralized computer that can be programmed to allow
access to specifc people only during certain hours. These systems enable employ-
ers to track the daily movements of all employeesincluding the amount of time
spent in the restroom.
Courts have repeatedly held that property owned by the employer is complete-
ly within its control. Hence, employers have the right to impose any controls or
monitoring they wish; as long as they do not violate other laws, such as the dis-
crimination laws.
Reality Bytes: Hard to Claim Benevolence When You Take Away The Internet
In the Arab Spring of 2011, unrest in North African nations led to political changes
in many countries. Tunisia was one of the frst nations to remove its longtime leader.
Egyptian citizens quickly began to followwith protests and talk of overthrowing
long-time leader Mubarek. Mubarek, and many other leaders, believed that the Inter-
net and cell phones were providing communication tools to dissidents. So, by Janu-
ary 28, 2011, the Egyptian government shut down virtually all networks in Egypt.
James Cowie, CTO at Internet monitoring frm Renesys noted that This is on a
differnet level entirely. Theres no cutting off the fnger to save the patient here. This
really is the Armageddon approach. In a developed nation like Egypt, the Internet is
not just a tool for the wealthyit has become part of the way business is conducted.
And, Internet connections are relatively robust. About the only way to shut down ev-
erything was for an offcial to call all of the ISPs and demand they cease operations.
But, shutting down the entire local Internet affects the entire country. That message
alone would likely encourage the rest of the population to question the control and
intentions of the leaders. A few days later, Mubarek resigned. Internet access and cell
phone networks were eventually restored. Iran is another nation that is going even
further to control the Internet. The leaders a building a new national Internet that
would be completely under the control of the government. Reza Begheri Asl, direc-
tor of the Iranian telecommunication ministrys research institute said that almost 60
percent of the homes and businesses are on the new network and within two years
it would cover the entire country. Some reports say the regime plans to roll out a
new operating system to replace Microsoft Windows. Abdolmajid Riazi, when he
was deputy director of communication technology in the ministry of telecommunic-
aitons, said that It will instead empower Iran and protect its society from cultural
invasion and threats. Top offcials have said that they consider Western culture and
ideas to be a major threat. Cuba currently runs a dual-Internet system, one for the
government (and tourists) and a local public one with limited access. The U.S. State
Department funds the development of tools to help bypass Internet censorship.
Adapted from Jaikumar Vijayan, Egypt's 'Net Blockage an 'Armageddon Ap-
proach', Computerworld, January 28, 2011; and Christopher Rhoads and Farnaz
Fassihi, Iran Vows to Unplug Internet, The Wall Street Journal, May 28, 2011.
It is easy to question why employers might feel the need to monitor their em-
ployees. At some point, you have to trust your employees to do their jobs. If an
employee has so little work to do that he or she can waste time using the Internet
for personal use, the bigger problem is money the company is wasting on an
unneeded employee. On the other hand, most major fnancial losses come from
insiders such as employees and consultants. Furthermore, many companies have
been burned by criminals when the companies were not careful enough in hiring
and monitoring. Other industries have specifc requirements that almost mandate
they monitor employees. For instance, the brokerage industry has many regula-
tions dealing with customer contacts, so it routinely monitors employee e-mail
messages and phone calls to ensure rules are followed.
Government Privacy
Privacy from government agencies and their employees can be a touchy issue. As
citizens, we agree to cooperate with government agencies to improve all of our
lives. And to function properly, government agencies can require detailed personal
data. For example, as shown in Figure 14.6, most governments collect health data,
police records, driving records, international travel, and detailed fnancial data for
taxes. Recipients of federal programs, such as farm subsidies, must report busi-
ness data. Many people are also required to complete census surveys collecting
detailed information about their lives. In the United States, much of this data is
protected and can be shared or released only under specifc conditions. But there
have been several cases of government employees illegally browsing through re-
cords for their neighbors or even selling data. In 1991, 18 people were accused of
selling Social Security information, including six government employees (Gov-
ernment Computer News, January 6, 1992, p. 58).
In the United States, few laws or regulations control the use of data held by pri-
vate organizations. However, several federal laws control the use of data collected
by government agencies. For example, federal agencies are restricted from shar-
ing databases except in specifc situations. In most cases the FBI cannot access the
IRS data without special permits. In terms of collection and use of data by private
Reality Bytes: Should Public Data be Private?
Many people are unaware of the amount of public data collectedor aware that it
is public. Basic items such as birth, death, and marriage records are recorded pub-
liclytypically at county offces. This data was originally considered to be public so
that people could verify the identity of others. In essence, you register with the local
government, and it vouches for your identity. Property and some debt records are re-
corded publicly so that ownership can be verifed. Criminal data is public so citizens
can recognize lawbreakers. Originally, the fles were on paper, and local newspapers
would report on the items of interest to the local citizens. Then companies started
retrieving the data and putting it online. Leading to cases where a person with a
speeding ticket can Google his or her name and see not only the ticket and fne but
personal information such as drivers license number and date of birth. The rules
about what can be done with public data vary by state.
Adapted from Elizabeth Garone, Help, Google Knows About My Speeding Ticket,
companies, few restrictions exist. Contrary to popular belief, there is no right to
privacy specifed in federal law. However, an element of privacy is contained in a
few scattered federal laws and some state laws, and in some Supreme Court inter-
pretations. For example, one federal law prohibits movie rental stores (and librar-
ies) from disclosing lists of items rented by individuals. But this law was largely
voided by the Patriot Act of 2001.
As everyone was reminded on September 11, 2001, the fip side to privacy is
the need for governments to identify and track individuals to prevent crimes and
terrorism. As an open society, the United States has chosen to lean toward individ-
ual rights and privacy; but many people have suggested that more control and less
privacy would make it easier to stop potential terrorists and criminals. Because of
the capabilities of modern information systems and networks, it is now possible
to build powerful systems that identify and track individuals within the nation and
around the world. Some people have suggested that the United States should es-
tablish national identity cards, as used in many European nations. A single, unifed
database would make it easier to track individual actions.
People tend to be split on the issues of government privacy. Some hate the fact
that they have to provide data. Others wonder what the problem is: if you tell your
friends how much money you make, why not tell the government? In one sense,
public information keeps everyone honest. And for many people, it probably does
not matter if various government agencies collect personal data. No one really
cares about the personal details of the housewife in Peoria. On the other hand,
Personal
nancial data
Company
nancial data
Political
negotiations
Health
data
Travel
data
Figure 14.6
Government and privacy. Spying on ordinary people is not an issue. Spying on
business and political leaders or journalists can cause problems. Collecting data on
targeted individuals such as dissidents or minorities can stife innovation.
some people within governments have abused their positions in the past. Consider
the tales of J. Edgar Hoover, longtime head of the FBI. He was obsessed with col-
lecting data on people and built fles on tens of thousands of people. Ostensibly he
was attempting to remove subversives and was a leading cause of the McCarthy
anticommunism hearings in the 1950s. He also collected thousands of secret fles
on politicians, journalists, and business leaders. He used these fles to harass and
blackmail leaders. Even if a modern-day data collector is not as blatant as Hoover,
and even if modern politicians have fewer moral problems, there is still an im-
portant risk. What if a politician tries to spy on or interfere with political negotia-
tions? CIA records released in 2007 also reveal the extent of then-president John-
sons use of the CIA to illegally spy on U.S. citizens in the 1960s and 1970s. He
used the communist threat to justify spying on college students, journalists, and
others participating in the anti-war movement. The same terrorist threat could
be invoked whenever some politician or police agent decides he or she wants to
spy on people. Only now, the spying could easily obtain almost any information
about your lifeby tracking your cell phone location, your purchases, your car,
the material you download and messages you send, your location at work (through
employee badge data), in addition to tracking you directly with advanced tracking
devices, satellites, and sensors. But, terrorists still walk free.
Some local governments, particularly in the U.S. and England, have become so
frustrated with crime that they have installed video cameras on every corner and
in every police car. By digitizing the video feeds, automated tools can watch and
listen to many events simultaneously. Any potential crimes can be fagged and
brought to the attention of a human guard. Beyond the deterrence effect, the data
Reality Bytes: Google Tracks Police Requests
As a leading portal to the Internet, Google is used by everyone. And, in a twist on
the old adage You are what you eat, the digital version has become: You are what
you Google. If someone looked at all of your searches, they could identify what you
are working on, planning, or perhaps even thinking. So, the police are increasingly
getting Court warrants to obtain Google search data. Agents from other countries
place additional requests on Google. In 2010, Google set up an online site to display
the frequency of these requests so citizens could see the increasing importance. In
the frst half of 2010 alone, Google counted more than 4,200 requests in the United
States (the country with the largest number of requests). As mobile computing be-
comes mainstream, police also turn to the cell-phone providers, not just to obtain
calling data, but also e-mail and text messages. In 2007, Verizon reported to Con-
gress that it receives 90,000 requests a year. In 2009, Facebook told Newsweek that
it received 10 to 20 subpoenas and court orders a day. The U.S. Justice Department
has argued in a Colorado Court that it should be able to access e-mails without a
search warrant. Does it matter? Ryan Calo, director of the consumer privacy project
at the Center for Internet & Society at Stanford Law School notes that When your
job is to protect us by fghting and prosecuting crime, you want every tool available.
No one thinks D.O.J. and other investigative agencies are sitting there twisting their
mustache trying to violate civil liberties. Theyre trying to do their job.
Adapted from Miguel Helft and Claire Cain Miller, 1986 Privacy Law is Outrun by
the Web, The New York Times, January 9, 2011.
records are also useful for identifying and prosecuting criminals. But, what will
happen when face recognition technology improves and an automated system can
track every person every day? Try reading some science fction novels, such as
Colossus, for a few ideas.
Protecting Your Privacy
Despite the shortage of laws, you can take several actions to protect your privacy
and restrict access to personal data. First, it is your responsibility to direct em-
ployers and companies you deal with to not distribute your personal data. You can
also ask them why they need personal data and whether it is optional or required.
In particular, all federal agencies are required to explain why they need data from
you and the purposes for which it will be used. You can also write to direct-mar-
keting associations and fle a request that your name not be included in general
mailings or unsolicited phone calls. By using variations of your name or address,
such as changing your middle initial, you can keep track of which organizations
are selling personal data. In some cases, you can refuse to give out personal data
(such as a Social Security or taxpayer identifcation number). If a private com-
pany insists, simply stop doing business with it. In a world where frms increas-
ingly rely on a single number for identifcation, it is important that you protect that
number.
With most government agencies and with banks, creditors, and credit-reporting
agencies, you have the ability to check any data that refers to you. You have the
right to disagree with any inaccurate data and request that it be changed. You can
also fle letters of explanation that are reported with the original data. In 1994,
Congress updated the Fair Credit Reporting Act of 1970. The new version requires
credit bureaus to verify disputed information within 30 days or delete it. Business-
es that provide data to the credit agencies would also be required to investigate
claims of incorrect information. The bill also limits who can have access to the
data stored by the credit agencies and controls how it can be used in direct market-
ing campaigns. In 1994, according to the Associated Press, the bureaus processed
450 million fles, selling 1.5 million records a day and handling almost 2 billion
pieces of data every month.
Reality Bytes: Opt Out Lists
Stop telemarketing phone calls:
www.donotcall.gov
Stop some junk mail:
Mail Preference Service
Direct Marketing Association
PO Box 643
Carmel, NY 10512
Stop credit agencies from selling your address to credit card companies:
Credit Bureau Screen Service
888-567-8688
Privacy Laws and Rules
The United States has few laws regarding privacy, although a few states do of-
fer some stronger protections. On the federal level, the Bork Bill states that vid-
eo rental stores and libraries cannot release their rental data. It was passed by
Congress when some over-zealous reporters obtained the video rental records
of a judge nominee (Bork). The 1974 Family Educational Privacy Act prohibits
schools from releasing grade data without permission from the student. The Priva-
cy Act of 1994 placed some minimal limits on the sales of state and local drivers
license data. The Privacy Act of 1974 limits what data can be collected and shared
by federal agencies. However, various rules, interpretations, and practices have
created enough loopholes to circumvent most of the original provisions.
In terms of fnancial data, various laws give consumers the ability to obtain
their credit records once a year and the right to dispute items in the report and
have the dispute resolved within 30 days. In 2001, a federal rule took effect that
was initiated by President Clinton to provide some control over the use of medical
data. Health care providers already complained about the high cost of implement-
ing the provisions, and some of the provisions were withdrawn when George W.
Bush became president. Nominally, the rules state that transfer of data (particu-
larly prescription drug data) requires permission from the patient. However, most
healthcare organizations require customers to sign waivers enabling them to share
data.
In contrast to the United States, the European Union has stronger consumer
privacy laws. Most EU nations have adopted the European commissions 1995
Data Protection Directive. Since 1978, France has had its own strict Data Protec-
tion Act. The laws basically state that personal data can be collected only with the
users permission, the user must be clearly told how the data will be used, and the
Reality Bytes: Pineda v. Williams-Sonoma and Her ZIP Code
A common thread in marketing is to identify your customers. One simple approach is
to ask them for a ZIP Code when customers check out. Most people assume the data
is needed by the credit card company, but it is not. It is just used to track customers.
Detailed Census data can then be paired with the ZIP codes to estimate income and
other statistics about customers. Yet, some customers feel pressured when asked for
personal data. In California, it is no longer legal for retailers to ask for personal data.
A 1990 state law prohibits merchants from recording personal identifcation informa-
tion; but merchants commonly ignored itat least in terms of the ZIP code. In 2009,
Jessica Pineda purchased items at a Williams-Sonoma store which asked for her ZIP
code. The merchant then used her credit-card number and ZIP Code to retrieve her
home address and add her to the catalog list. She began receiving catalogs and other
mail from the company. It is likely that the vendor also sold her information to other
companies. In 2011, the California Supreme Court ruled that a ZIP Code constitutes
personal information and it is illegal for merchants to ask for it. A ZIP Code might
seem innocent, but with huge databases, it seems that ZIP Code, gender, and birth-
date are suffcient to identify most people.
Adapted from Scott Thurm, California Court Says Stores Cant Ask for ZIP Codes,
user has the right to see and to change any personal data. The laws have an ad-
ditional important condition: personal consumer data cannot be moved to a nation
with lesser privacy controlsnotably the United States. The United States has
negotiated a loose safe harbor provision, so that companies can bring European
consumer data to the United States if the companies formally agree to abide by the
EU directives and also agree not to resell the data. These provisions make it more
expensive to collect data in Europesometimes beyond the price of small busi-
nesses. For example, in the United States, it is relatively easy to purchase e-mail
lists of potential customers for a few hundred dollars. In Europe, these lists would
generally be illegal to use, since the customer did not agree to the unsolicited use
of his or her address.
Anonymity
Anonymity is the fip side of the privacy question. Until recently, it has been dif-
fcult or impossible to provide anonymous access to the Internet. Using advanced
encryption, some frms now offer people the ability to use the Internet without
revealing any data about themselves. Remember that the Internet works by as-
signing IP addresses to all computers. Any server that you visit saves your IP ad-
dress in log fles. It can be somewhat diffcult to match an IP address to a spe-
cifc individual, but it provides a general location. If necessary, a court order can
be obtained to track an IP address to a specifc individual. However, Figure 14.7
shows that a proxy server can hide your IP address. Sites that you visit see only
the address of the intermediate server. The server also intercepts all cookies and
Figure 14.7
Anonymity server. By passing requests through a proxy server, the Web site server
sees the IP address of the proxy server, not the original source. However, if the
anonymity server is monitored in real time or keeps logs, it is possible to trace back
to the original computer.
Web Server
Anonymity Server
Proxy server
Temp IP
Address
The Web server, and
investigators, see the
IP address of the
anonymity server, not
the real user.
If the anonymity
server is monitored in
real time, or if it keeps
records, it is possible
to trace back to the
original source.
other fles sent by some servers. However, if the anonymity server keeps logs or if
it is monitored in real time as data is being transferred, it is easy to trace sessions
back to the original source. The primary U.S. wire-tapping law, Communication
Assistance for Law Enforcement Act of 1994 (CALEA), was interpreted in 2006
to require that all ISPs must provide the ability to invisibly tap Internet connec-
tions and provide real-time feeds to a law enforcement agency when ordered by
a court. Because of these laws, truly anonymous servers cannot exist within the
United States, but a few are run by various people around the world. It is not
clear whether they can be monitored in real time or are subject to court-ordered
retrieval of data.
The ultimate question that you, as an important member of society, must answer
is whether anonymity should be allowed, or how it should be controlled. Certainly
it can be used to improve privacy. People who have a stronger belief in personal
privacy might seek out and use anonymity serversothers will decide that pri-
vacy is not an issue. But what about drug dealers, terrorists, child pornographers,
and other illegal activities that society wishes to stop? Or what about anonymous
harassment? Someone could use the technology to harass and intimidate people
on the Internet. Perhaps society should not allow anonymity? On the fip side,
who makes that decision? If some nation chooses to ban dissenting viewpoints, or
if a government whistleblower needs to protect a career or a life, anonymous sites
can be valuable tools to increase information and open discussions.
Jobs
How does technology affect jobs? If computers do more of the
work, what jobs are left for people? Technology can affect jobs in other
ways as well. It opens up the world to people with physical disabilities. By remov-
ing location as an issue, networks make it possible to work on jobs from almost
anywhere in the world. You can telecommute or consult around the world without
leaving your home.
Loss of Jobs
There is no question that technology causes some workers to lose their jobs. In
the 19th century, Luddites reacted to textile automation by destroying machines.
Information technology is no exception. Norbert Weiner, a computer pioneer in
the 1940s, predicted a major depression would result from computers replacing
workers. Despite these predictions, during the last 100 years technology has in-
creased the number of jobs and raised the standard of living for most workers.
Since the introduction of computers in the 1950s, the worlds economies have
grown and incomes have increased. However, individual workers can lose jobs in
the short run. Even in the long run, lower-skilled workers experience greater diff-
culty in fnding new jobs. Compare the automated shipyards of Singapore to those
in the United States. In Singapore, one man using computer screens and a joystick
moves hundreds of containerized cargoes without ever leaving his offce. In the
United States, each crane requires a crew of four workers, including one just to
identify shipments and destinations that are handled by computer in Singapore. In
Europe, the Dutch port in Rotterdam cut employment in half by installing robotic
cranes and automated transfer vehicles.
The point is that some jobs will disappear, but others will take their place. In
the shipyard example, more technical expertise will be needed to program and
repair the equipment. Figure 14.8 shows the changes in jobs for the next few years
that are anticipated by the Bureau of Labor Statistics. Although they do not make
the top-ten list in 2010, a few of the fastest growing jobs are in computer technol-
ogy. But due to outsourcing and cost issues, the list is shorter now than it was a
couple of years ago.
Most economic experts believe that technology increases the total number of
jobs. New technology creates demand for people to design it, manufacturing frms
to produce it, and people to maintain and repair it. Computer hardware also cre-
ates demand for software programmers. More important, technology can cause
the economy to grow, creating more jobs in all sectors. By most indications, new
jobs created by technology tend to be higher paying, physically safer, and less
repetitive than those replaced by technology. Information technology can also re-
duce product prices, raising the standard of living by enabling people to buy more
goods. On the other hand, technology typically causes some workers to lose their
jobs. Unfortunately, many of these displaced workers cannot be retrained for the
new jobs created by the technology. Similarly, the new jobs might pay less money,
have lower status, or might have less desirable work environments. As computers
and software become increasingly powerful and intelligent, it encroaches on even
more jobs. The U.S. recession starting in 2008 resulted in many companies reduc-
ing the number of workersparticularly in middle management.
Governments have created several programs to provide benefts of money,
retraining, and relocation to workers who lose their jobs. Managers need to un-
derstand the effects on employees when new technology is introduced. Many
corporations provide ongoing educational payments and training classes to help
workers improve their skills. Others provide out placement services to help unem-
ployed workers in their job search.
Figure 14.8
Future jobs. Today there is no guarantee that your job will continue to exist. Demand
for specialists changes constantly. Jobs that are well defned and require little
innovation or thought can usually be performed easily by computers.
U.S. Bureau of Labor Statistics. http://www.bls.gov/emp/
-60 -40 -20 0 20 40 60 80
Occupational Change 2010
As individuals, we need to remember that changing technology can eliminate
virtually any job. One of the best plans is to continue your education and learn
new skills. Remember that technology continually changes. Some of the skills
you learn today will be obsolete in a couple of years. We must all continually
learn new skills and adapt to changes. Applying these skills in your current job
adds experience that will help you fnd a new job. It also benefts your current em-
ployer and might help you keep your job or stay with the company if new technol-
ogy makes your current job obsolete.
The concept of continually acquiring new skills sounds straightforward. How-
ever, many times you will have to choose among multiple technologies. Guessing
wrong can lead you to invest time and money in a technology or skill that fades
away. As you become more involved with technology, you will increasingly fnd
it necessary to predict the future. Identifying trends and deciphering fact from
rumor are important skills to learn.
Physical Disabilities
Technology offers many possibilities to provide jobs for workers with physical
disabilities. In fact, in 1992, the U.S. Congress passed the Americans with Dis-
abilities Act, stating that companies are not allowed to discriminate against dis-
abled employees. Common uses of technology include the use of scanners and
speech synthesizers for visually impaired workers; voice input devices and graph-
ics displays for workers who cannot use keyboards; and telecommuting for those
who work from home. In 2001, the U.S. government began requiring that all soft-
ware it purchases must be accessible to users with disabilities. Since the federal
government employs hundreds of thousands of workers, this order should encour-
age all software providers to improve their software.
Most Windows-based software contains features to facilitate usage by people
with various physical challenges. In some cases, additional accessibility tools can
be downloaded or purchased to provide more features. Speech recognition pack-
ages are useful for many applications. Sometimes adaptive devices are needed to
provide alternative ways to enter data and obtain the results.
Web sites still present accessibility problems, particularly for those with visual
impairments. Many sites rely on color and graphics, which are diffcult for the ac-
cessibility tools to interpret. These issues are being discussed by many vendors.
Check Microsofts accessibility site for more details.
Telecommuting
The fact that about 70 percent of U.S. jobs are service-based raises interesting
possibilities for workers. Many services like accounting, legal advice, education,
insurance, investments, data analysis, computer programming, and consulting are
not tied to a physical location. As a service provider, you could be located any-
where and still perform your jobas long as you have the appropriate telecom-
munications system. As communication improves to include video links and faster
document transfer, even more jobs can be performed from remote locations.
Some companies are experimenting with home-based workers, especially in
cities such as Los Angeles and New York with long commute times. Some work-
ers like the concept; others try it for a few months and return to a traditional work-
place job. Several advantages and complications arise from the perspective of the
worker, the frm, and society.
If a substantial number of workers choose to work from home, the frm gains
two main advantages: (1) decreased costs through smaller offces, and (2) fex-
ibility in hiring additional workers on a contract basis. Some people have pre-
dicted that companies might also gain from increased use of part-time workers,
thus avoiding the cost of insurance and other benefts. The greatest complication
to the frm is evaluating and managing employees. Without daily personal contact,
including conversations, it is harder to spot problems and make informal sugges-
tions as corrections.
To the worker, the most obvious beneft lies in reducing the time and expense
of commuting to work. The biggest drawback lies in the loss of personal contact
and daily ritual of a typical work schedule. Depending on your home environ-
ment, there can be substantially more interruptions and distractions at home. It is
also more diffcult to get away from your job. Working from home on a fexible
schedule requires strong motivation and organization. Before you choose to work
at home, talk to someone with experience.
A few frms have experimented with intermediate telecommuting options. As
indicated in Figure 14.9, the frm leases smaller offces in city suburbs and work-
ers operate from these satellite offces instead of one central location. The offces
are linked by high-capacity telecommunication lines. Workers keep a traditional
offce environment but cut their commuting costs. Businesses maintain traditional
management control but do not save as much money.
Advantages
Decreased overhead.
Flexibility in part -time workers.
Disadvantages
Harder to evaluate workers.
Harder to manage workers.
The Firm
Advantages
Reduced commuting costs.
Flexible schedule.
Disadvantages
Loss of personal contacts.
Distractions.
Employees
Suburban
work centers
Figure 14.9
Telecommuting. In the simplest form of telecommuting, individual workers connect
to offce computers from their homes. An intermediate method has been used to
avoid the problems of distractions and the cost of creating a home offce. Workers
report to satellite centers in their suburban neighborhood. Workers retain a structured
environment but reduce their travel time.
A few people have speculated about the effects on society if there is a large
shift to telecommuting. At this point, there is not much evidence to support any of
the hypotheses, but many of them focus on negative aspects. People could become
isolated. Jobs could become highly competitive and short-term. Firms could list
projects on the network and workers would compete for every job. Workers would
essentially become independent contractors and bear the responsibilities and costs
of insurance, retirement, and other benefts, with little or no job security. They
would also have no loyalty to any particular frm. Firms could become loose coali-
tions of workers and teams that are constantly changing, with little control over
future directions. It is hard to predict what will really happen, but by understand-
ing the negative effects, they become easier to avoid.
Business: Vendors and Consumers
How does technology change the relationship between busi-
nesses and consumers? Business consists of transactions. Changes in the
way transactions are handled can alter society. In particular, as digital content be-
comes more important, can the existing laws created decades ago still be applied?
And if the laws are replaced, will they affect the balance of power in the relation-
ship between vendors and consumers?
Intellectual Property
Intellectual property is the general term to describe ownership of ideas (patents)
and creative expressions (copyrights). For many years, there was a solid distinc-
tion between the two: ideas involving physical items (such as machines) could be
patented. A patent essentially grants a monopoly to an inventor for a fxed period
of time (originally 17 years in the United States but now 20 years). During that
time, no other company can introduce a similar deviceeven if the second creator
did not use any knowledge from the frst inventor. A copyright is created for other
creative workstraditionally writing and music. It protects the specifc article
from being copied and grants the owner the sole right to create derivative works
(such as a sequel). But it does not prevent others from creating similar works.
For example, one person could write a story about space explorers. Someone else
could also write a story about space explorers, and it would not be an infringement
on the frst story. If the second story used the same characters and plot, it might
be an infringement, but it might not, depending on the interpretation of the courts.
In the mid-1990s, the U.S. patent offce began granting patents for nonphysical
itemsspecifcally for business ideas. Patents were supposed to be granted only
for nontrivial ideas, but for a while, the patent offce got carried away and forgot
that patents are only supposed to be granted for nontrivial and nonobvious in-
ventions. For instance, it granted a business process patent to Amazon.com for
one-click checkout of Web sales; so no other Web site was allowed to offer a
checkout system with a single click without paying royalties to Amazon.
The goal of patents and copyrights is to encourage creativity by offering pro-
tected rewards to innovators. Remember from economics that without a barrier to
entry, any frm that makes a proft will attract competitors. Patents and copyrights
are designed to be barriers to entry for a limited time. But the laws were written
in decades when the goal was to protect companies from other companies. For in-
stance, before computers, only a large company would be able to copy a book and
reprint it. The laws made it clear that this action was illegal, and the injured party
could easily fnd and sue the single violator for damages. Several exemptions to
the copyright law were specifcally created to support important noninfringing
uses that are considered valuable to society. For example, educational institutions
can make limited copies of items for discussion and research.
The basic laws made sense when large companies were the primary threats to
copying products and content. Only large print shops, music frms, and competi-
tors had the money, tools, and distribution networks to become a serious threat.
For example, the music industry did not consider cassette tape copies made by
individuals as a serious threat.
Digital content changed most of the underlying assumptions of the intellectual
property discussion. First, it is easy for anyone to make perfect copies. Second,
it is equally easy for everyone to distribute those copiesat virtually no cost.
Instead of a large competitor, now the threat is millions of your own customers.
Some of these issues are cultural and economic. For example, some industry-
sponsored reports indicate that software piracy in Southeast Asia is huge: over 90
percent of software in use is copied. Nations such as Vietnam do not have the tra-
dition of paying for creative works, and do not have much money to pay for them.
The most famous case of these copyright issues involved the company called
Napster. Napster was a pure Internet frm that ran a Web site to make it easy for
consumers to fnd and share digital music fles. In an attempt to stay within the
copyright laws, Napster did not store any fles and did not charge for its services.
Instead, it was simply a giant directory. Individuals searching for specifc songs
went to the Napster site, found a fellow enthusiast with a desired fle, and copied
the fle from the others machine. Napster lost the ensuing lawsuit from the mu-
sic industry. But the battle is far from over. Napster made it easy by providing a
single target to sue. What if there is no central company directing the copying?
One such frm, LimeWire, was shut down in 2011 for violating copyright laws by
providing software for sharing music.
Reality Bytes: Cignet and Massachusetts General Fined for HIPAA Violations
The U.S. Department of Health and Human Services (HHS) is in charge of enforcing
the provisions of the Health Insurance Portability and Accountability Act (HIPAA).
Two major features of the rules are that patients can receive copies of their health re-
cords and that medical institutions must protect the privacy of patient data. In 2011,
HHS completed enforcement actions against two companies based on separate viola-
tions of those provisions. The insurance company Cignet had failed to provide re-
cords to patients in a timely manner. The company also failed to cooperate with in-
vestigations and to produce records to HHS. Cignet paid a $1.3 million penalty. Mas-
sachusetts General Hospital faced a different problem. In March 2009, an employee
of the Hospital accidentally left health documents for 192 patients on a subway train.
The Hospital ended up paying HHS $1 million in fnes. The two actions represent the
frst fnes imposed due to HIPAA. Perhaps the actions will help convince businesses
to protect customer data. But much remains to be done. A report from the accounting
frm Kaufman, Rossin & Co. revealed that in 2010 businesses somehow gave up or
lost health data on about 5 million patients.
Adapted From Jaikumar Vijayan, HIPAA Privacy Actions Seen as Warning, The
Wall Street Journal, February 25, 2011.
Technology Toolbox: Privacy
Problem: How do you improve privacy on the Internet?
Tools: Most current browsers have tools for improving privacy by reducing the data
exchanged with Web sites. But, browser vendors tend to hide these tools because
most receive large amounts of advertising money that might be disrupted. Also,
stronger privacy sometimes makes Web sites unusable or harder to use.
Microsoft Internet Explorer (IE) has a couple of privacy controls. The simplest
one is to control the use of cookies. In particular, you should disable third-party
cookies. These are installed and used almost exclusively by advertising companies
such as Double-Click (Google) to track which sites you visit. To disable them, use
the main menu (press the Alt key) Tools/Internet Options/Privacy tab, then select the
Advanced button. First-party cookies should be set to Accept, and third-party cook-
ies set to Block. If you have problems with some sites, you might set it to Prompt,
but that will generate warning messages on almost any site you browse. IE9 and
above has a new option under the Privacy tab to Never allow websites to request
your physical location, that you might want to select. If you are going to browse a
Web site for research or where you want to ensure absolutely minimal tracking and
minimal history, IE provides the InPrivate browsing mode. Start IE then click the
small new tab icon on the tabs row. Find and click the link for InPrivate Browsing
which will open a new browser window. In this mode, minimal information is ex-
changed or stored on your computer.
The Google Chrome browser has similar settings under different names. To block
third-party cookies, click the Settings icon and choose the Under the Hood tab. In
the Privacy section, click the Content Settings button. In the Cookies section, check
the option to Ignore exceptions and block third-party cookies. Chrome also has a
mode that does not save any cookies, temporary fles, or history. Click the Settings
icon and pick the New Incognito Window option.
Firefox and IE have a new option to Tell Web sites I do not want to be tracked.
The problem with this option is that it is not yet a standard and none of the Web sites
actually support it. Firefox has a private browsing option. Press the Alt key to acti-
vate the menu, choose Tools/Start Private browsing.
At the server level, it is also useful to opt out of some Web site tracking. A big one
is to opt out of Google: http://www.google.com/privacy/ads/, which includes getting
rid of Double-Click tracking.
Quick Quiz:
1. Can you prevent Web sites from collecting your personal data?
2. What do you gain by blocking third-party cookies?
3. Why would you not want to use Private or Incognito browsing all the time?
In response to these problems, content providers began encrypting their content
to make it more diffcult to share. Cable and satellite TV providers were among
the frst to scramble their signals to make it more diffcult for people to steal the
signals. In response, several companies started selling and advertising black boxes
that would descramble the signals, making it possible for anyone to steal the TV
signals. Technically, it was illegal for people to steal the signal, but it was eco-
nomically diffcult for companies to stop the theft.
To solve this problem and help protect IP rights, the United States passed the
Digital Millennium Copyright Act (DMCA). One of the most important changes
in the act is a provision that makes it illegal to circumvent any copyright protec-
tion scheme. For example, this provision made it illegal for companies to advertise
and sell TV descramblers. The law was generally accepted in terms of hardware
issues, but problems and complaints quickly arose when circumvention focuses
on software and information.
In existing cases, the DMCA provision has been interpreted to mean that any
discussion of how to circumvent protection is illegal. The frst case pursued under
this provision relates to DVD movie disks and the DeCSS program. DVD fles are
encrypted so that they can be played only on specifc machines with authorized
software. The fles are only weakly protected. Direct from the manufacturer, the
fles can be copied to other computers, but can only be viewed with the special
software. Since that software did not originally exist for some computer systems
(notably Linux), a few experts found a way to defeat the encryption. They posted
this method (known as DeCSS) on the Internet. The movie studios promptly sued
every Web site that carried the program for violating the DMCA. Many people are
concerned that these actions violate the spirit of free speech and open discussion.
Of course, whether or not the movie industry wins the case is almost irrelevant.
One key factor of the Internet is that it is impossible to destroy knowledge once
it has been created. Several Web sites in foreign nations that do not support the
DMCA carry the information.
Reality Bytes: Dirty Politics Goes Online
It is unlikely that anyone likes political campaigns, but they are a necessary com-
ponent of a democratic society. On the other hand, political campaigns that become
dirty are even more annoying. Apparently, political consultants are now working
with technology practitioners to generate online attacks. Someone hacked and re-
leased thousands of e-mails from HBGaary Federal, a security consulting frm. The
messages show that the company pitched ideas to lawyers at Hunton & Williams, a
D.C. law and lobbying frm. The ideas included create fake social network accounts
to mislead people, scanning Facebook and other accounts to fnd negative informa-
tion about opponents, and planting false documents to undermine the credibility of
activists. Workers from HBGary Federal, Berico, and Palantir suggested many ways
to spread disinformation online. It is not clear whether the actions would be illegal or
if any were actually undertaken. But, given the challenges of tracking actions online,
it is only a matter of time before dirty tricks become more common.
Adapted from Dan Eggen, Hacked e-Mails Show Web is an Increasingly Useful
Tool in Dirty-Tricks Campaigns, The Washington Post, March 4, 2011.
Computer software, music and digital books face similar issues. As more con-
tent moves online, creators and publishers searched for a stronger solution that
will prevent customers from copying and sharing digital data.
Digital Rights Management
Digital rights management (DRM) tools use encryption and rules embedded in
the data to control how purchasers can use and transfer data. Figure 14.10 shows
the basic concept. Several companies provide DRM tools, leading to different ap-
proaches, but many of them follow a similar process that is pushed heavily by Mi-
crosoft. When an item is purchased, software on the purchasers computer creates
an identifer for that machine and sends it to the server. The server encrypts the
data, embeds a set of rules, and often embeds a user ID. If the content publisher
fnds an illegal copy of a fle online, it can be traced back to the original purchaser.
To date, no one has tested this concept in court to see if it could be held against the
original purchaser. As a simple defense, the purchaser could argue that a virus or
Trojan Horse copied the fle invisibly.
For DRM to work the software that plays or displays the fle must know how
to decrypt the fle and it must follow the embedded rules. Even if the fle is trans-
ferred to a different computer, the software will not play the fle if the rules are not
met. Most systems accomplish this task by requiring that the fle be played with a
specifc software package. For example, originally fles purchased through Apple
could be played only with the iTunes software, and fles protected with Microsoft
DRM could be played only with Microsoft media software. The process becomes
Figure 14.10
Digital rights management. When digital content (music, books, movies, and so on)
is purchased, it is encrypted with a set of rules that tie it to the purchasers computer.
Software on other computers checks the rules to decide whether the data can run on
other machines. HDTV goes a step further and verifes the security of the hardware
before sending data.
Content
Server
Purchase item
Encrypted item
Tied to computer
Usage rules
Embedded user ID
Attempt to play (or
share) on different
computer and player
software examines the
rules/rights.
Hardware devices validate entire chain to
ensure no insecure devices will capture
the signal to convert it. HDCP handles
these issues with HD TV.
more complicated if the data needs to be compatible with multiple devices created
by different manufacturers. HDTV is the primary example. You can buy a Blu-
Ray disk created by several companies, play them on several different players and
connect the player to any television. To protect the content, all of the manufac-
turers had to agree on the high-bandwidth digital content protection (HDCP)
format.
Effective DRM is diffcult to create. For example, music eventually is convert-
ed to analog form and played through speakers. It is relatively easy to convert it
back to digital form that is no longer protectedbut some quality might be lost
in the process. HDTV took the process a few steps further. The high-defnition
players examine the entire collection of devices, including the connecting cables,
to ensure that all of the devices support HDCP. If not, it plays the video in a lower-
quality picture. Some early devices even required the video output to be directly
connected to the display device (TV).
Many people look at DRM as a challengesomething to defeat simply because
it is there. In theory, any DRM scheme can be defeatedgiven enough time and
effort. Programmers have full access to the computer so they can trace every step
in the process. Technically, DMCA makes these efforts, and publishing it, illegal
in the United States; but it will not stop the work.
The fip side of copyrights is the argument that content should be free. But
who would take the time to create useful content if it is free to everyone? Stephen
King, the noted author, ran an experiment in the late 1990s. He wrote a novella
and began writing a novel in installments. He distributed them through his Web
site. He attracted hundreds of thousands of visitors and free downloads. He ulti-
mately stopped the projects because of lack of revenue (it was more proftable for
him to devote his time to paying projects).
At this time, DRM is a highly contentious issue. It makes life more diffcult for
honest consumers and limits what they can do with the data they purchased. For
example, imagine the problems that can arise if all of your music is locked to one
computer and that computer dies. Even if you have backups, you will need special
permission to transfer them to a new device. Many companies establish rules to
support using the content on three or four computers, but that rule will not cover
a lifetime of upgrades. In 2007, Steven Jobs, the CEO of Apple, attempted to con-
vince the music industry to allow him to drop DRM conditions on the songs his
company sells.
By 2011, most digital music was sold without copy protection (DRM). Does
that mean consumers became more honest? Certainly, it is now easier to purchase
music online, particularly individual songs. Some prices have dropped with com-
petition. But, the music industry continues to report declining revenues from all
forms of music sales. (For some interesting charts, see http://www.businessinsid-
er.com/these-charts-explain-the-real-death-of-the-music-industry-2011-2.)
DRM presents an interesting dilemma to businesses that sell digital content.
The fear is that if you do not include DRM, everyone will copy your content and
share it with others, reducing the number of sales. But, if you do use DRM, you
make life more diffcult for legitimate users who pay for the product. For example,
as of 2011 most book publishers sell DRM-protected books for electronic readers.
Effectively, the DRM locks the book to specifc devices. Customers run the risk of
losing their existing books if they switch to a different device, or even if a device
breaks or is lost. In the meantime, the DRM methods were generally broken, with
methods available online, within a few months. Additionally, almost anyone could
use a scanner to turn a print copy of a book into an unprotected digital copy in
less than an hour. Similar issues face publishers of music, movies, and software.
Music moved away from DRM largely because of the decision to avoid locking
music to specifc devices. Movies continue to cling to DRM. Software varies, with
big publishers such as Microsoft and Adobe leading the way with DRM (their
software uses Internet connections to limit the number of installations.) Clearly,
it is within the rights of the copyright owners (publishers and authors) to choose
to implement DRM. Also, in the U.S. the DMCA makes it illegal to circumvent
DRM protections (although this provision is largely untested in court). But, from
a business perspective, managers need to think about exactly what is gained and
lost through DRM.
Balance of Power
Some of the issues with intellectual property arise because of questions of bal-
ance of power between the creator, the publisher, the retailer, and the consumers.
For instance, the Internet, with its digital content, provides the opportunity for
authors and creators to circumvent the traditional publishers. Currently, authors
receive only a small portion of the list price of an item. Retailers, distributors, and
publishers take the majority of the money. In a world that requires distribution of
physical items, these are the costs artists must pay to reach consumers. In a digital
world, anyone can sell products directly to consumers. Of course, it will be sim-
pler and more cost-effective when consumers adopt a small-payments mechanism.
The point is that the few large publishers in each industry have a strong interest in
Figure 14.11
Price discrimination. By selling directly to individual customers one at a time, it
becomes possible to fnd and charge the maximum amount each is willing to pay.
This action shifts profts/benefts from consumers to producers.
Q
P
D
S
P*
No price discrimination.
Price paid by everyone.
Person 1
P1
Person 2
P2
Charge each person
the most he or she is
willing to pay.
maintaining control over the distribution system, so most proposals have catered
to these large frms. For example, the DRM systems keep the role of the retailers
and the publishers, so the costs to consumers are likely to remain high.
All of these issues are challenging with multiple perspectives. Society needs to
protect and encourage innovation because it creates new products and moves the
economy and society forward. But, in these cases, technology also makes it easier
for individuals to copy and distribute protected works, and it is economically in-
feasible for an innovator to sue millions of people over patent or copyright viola-
tions. The issues also revolve around important economic concepts. Digital con-
tent essentially has zero cost to copy and distribute (but high fxed costs to create).
As shown in Figure 14.11, another interesting economic issue is that some peo-
ple are willing to pay higher prices for the content than others are. In a typical
market, the price is determined by supply and demand so that the market clears
with no shortages or surpluses. Effectively, everyone pays the same price (or close
to it). People who were willing to pay more money for a latest release got a deal
(or consumer surplus) because they paid the same fxed price. But, in an online
digital world, it is possible that the customer will purchase directly from the pro-
ducer. Every sale is separate, and individuals might not even know how much oth-
ers are payingparticularly if you want to buy a newly released item. In this situ-
ation, it is possible to set up a system where each customer pays his or her highest
priceperhaps through an ongoing series of auctions. For example, if you want
to be the frst person to download a specifc song, you will pay more. If you are
willing to wait a few days or weeks, you can pay a lower price. Or, more sophisti-
cated methods could be created. In the end, each person pays a different price, but
one that each is willing and able to pay. The balance of power is changed and the
producer captures more of the consumers money. Is this situation good or bad?
The answer depends heavily on your personal beliefs, but it is certainly different.
When Amazon tried an experiment several years ago, (the company charged dif-
Reality Bytes: Poker is Not Gambling?
U.S. law makes it illegal to gamble by wire which includes Internet gambling. On-
line casinos and games of chance were banned shortly after the Web became popular.
Federal law prohibits credit card companies from transferring money to online casi-
nos. But, for several years, online poker sites fourishedeven in America. Credit
card companies did not transfer money to them, but a few banks handled special ac-
counts for them. In April 2011, Federal authorities cracked down on the online poker
sites. Initially, eleven people were arrested, including three founders of the largest
poker sites. Federal agents fled restraining orders against 76 bank accounts in 14
nations; alleging at least $3 billion in money-laundering penalties. PokerScout, a site
that tracks playing, estimated 1.8 million U.S. players wagered $16 billion in 2010.
About one-third of the money deposited by players ended up going to the online
companies as the rake or house percentage. Prosecutors also seized fve of the In-
ternet domains used by the poker companies, shutting down the services and putting
up warning notices. In their defense, the arrested founders argued that the sites are
not illegal because poker is not gambling.
Adapted from Alexandra Berzon and Chad Bray, Eleven Charged in Federal Crack-
down on Online-Poker Companies, The Wall Street Journal, April 16, 2011.
ferent prices for the same book in different cities), customers were unhappy when
they found out some people paid a lower price. Apple negotiated with the studios
to be able to charge different prices for each song. Popular, newly-released songs
cost more than older songs. It is not perfect discrimination, but someday new pric-
ing mechanisms might be introduced that add even more levels of prices.
Education and Training
Can information technology change education? For hundreds of
years, the principles and techniques of education have changed only slightly. As
new technologies are introduced, people have often declared that the world of
education would change markedly. Yet, few technologies have had a lasting im-
pact on education. Television is a classic example. Although movies and news
reports are sometimes used for teaching purposes, the role of television in formal
education is minimal. However, it is used for informal education and for training,
especially with the availability of videotapes for teaching specifc tasks. And, as
bandwidth has increased, video communication is growing as a way to connect
with students in distant locations.
One of the drawbacks to video education is the lack of interaction and feed-
back. Multimedia tools that combine video, sound, and computer interaction rep-
resent one attempt to surmount this limitation. However, three basic problems
arise when applying technology to education. First, technology is often expensive,
especially compared with traditional methods. Second, it is time consuming to
create lessons that generally are diffcult to change. Third, there is little conclusive
evidence that the techniques are equal to or superior to existing techniques. Espe-
cially in light of the frst two problems, it is diffcult to test the new technologies.
In many cases, by the time prices have fallen and lessons are created, an even
newer technology emerges.
Figure 14.12
Information technology in education. The technology has the potential to change
education, particularly in terms of individualized attention, course management,
and distance. But it is expensive and time consuming to provide the infrastructure
and create new applications. Nontraditional areas such as continuing professional
education (CPE), employer training, and the military have found several benefts in
the technologies.
Can technology improve education?
Computer-assisted instruction to provide individual attention
Course management
Distance learning
Do people want more technology in education?
Teachers
Students
Employers
Are the answers different for lifelong learning?
Professionals
Employers
Military
Despite these obstacles, technological innovations are often used for special-
ized teaching purposes. For instance, interactive and multimedia computer tools
can be used to provide more in-depth study for advanced students or to handle
repetitive drills for those students needing extra work. Increasingly available two-
way video links are used to connect teachers and students in remote locations.
The main questions regarding technology in education are summarized in Fig-
ure 14.12. Note that nontraditional areas have been faster to adopt the technolo-
gies, for example, business training classespartly to reduce the cost of hiring
instructors and partly because the lessons are available to workers at any time and
can be studied at whatever speed the student desires.
The Internet is increasingly being pushed as a means to expand the reach of
higher education. Several universities offer individual courses over the Internet.
The early examples often consisted of simple e-mail-based systems where stu-
dents worked on their own and occasionally sent messages to the instructor. A
few organizations offer complete programs over the Internet. Improving Internet
speeds have made new communication methods feasibleincluding interactive
voice and some two-way video communication.
The real key to online education is to use all of the power of the technology
to develop entirely new applications. Communication is only one aspect of the
Internet. Building more intelligence into the applications to create entirely new
procedures will lead to more useful tools. Researchers have worked for years to
develop computer-assisted instruction tools that will provide individualized atten-
tion to each student. While some individual products have been successful, these
tools require considerable creativity and effort to create.
Social Interactions
How does technology affect different areas of society? As any
good science fction book illustrates, advances in technology can alter society in
many different ways. Sometimes the responses are hard to predict, and they often
occur gradually, so it can be diffcult to spot patterns. At the moment, four patterns
appear to be important: social group power, equal access to technology, e-mail
freedom, and liability and control over data.
Social Group Legitimacy
One interesting feature of technology is that it has substantially lowered commu-
nication costs, including the costs of producing and distributing information to
large public groups. For example, desktop publishing systems enable groups to
create professional-quality documents at low cost. Furthermore, video production
facilities are easily affordable by any group, and access to mass markets is pro-
vided free through public-access channels on cable television. Web sites can be
created by anyone. These technologies enable small groups to reach a wider audi-
ence for minimal cost.
The only catch is that with growing professionalism of small-group produc-
tions, it becomes harder to distinguish fact from fction, and it is harder for the
public to tell the difference between mainstream, professional commentary and
radical extremists. For example, do you believe stories that are printed in The
New York Times? What about stories printed in supermarket tabloids that sport
titles such as Space Alien Eats Movie Star? Now consider the Internet and run
some searches on medical questions. You will fnd hundreds of Web sites and
comments. Which ones do you believe? Web sites present the strongest challenge
ever to trust and reliability issues. Literally anyone can create a site and say any-
thing. Nonsensical comments will be found by the search engines and displayed
along with accurate statements. Throw in Twitter and your cynicism meter should
go off the scale. Consider the example Web sites in Figure 14.13 and see if you
can determine which one to believe.
This issue has some interesting effects. For example, in several instances, dis-
gruntled customers have created sites criticizing companies. If you search for a
particular company, you are likely to encounter several of these sites. The Web
makes it easy for people to criticize anyoneand the entire world can see the
results. Of course, traditional defamation laws still apply, but in situations where
there is an element of truth, companies will fnd it diffcult to stop these activities.
The same issues can be applied to television broadcasts, except that for the mo-
ment, the high costs of broadcasts restrict this option to a few participants. With
his War of the Worlds broadcast, Orson Welles shocked many listeners because
they had come to accept radio broadcasts as fact. With existing technology, it is
possible to create realistic-looking fctional broadcasts. It is not even necessary to
resort to tricks such as hidden explosive charges. It is possible to create computer-
generated images that exceed the quality of broadcast signals, so they appear to
be realistic. Advertisers have made heavy use of these techniques. Every time you
watch a commercial, you should remind yourself that a portion of what you are
seeing is probably a computer-generated image. Now, imagine what would hap-
pen if an extremist organization used this same technology to create newscasts
with altered pictures.
Figure 14.13
A test of cynicism. Which Web site do you believe? Why? Would it help to know that
the one on the left is from Johns Hopkins medical center, and the one on the right
from a site called arthrituscure.org? With information technology, anyone can create
a Web site. It can be diffcult to determine the truth. Of course, in many cases, truth
may be only shades of gray, and there seldom are any right answers. All consumers
must learn to challenge everything.
Technology Toolbox: Working in a Global Environment
Problem: How do you deal with multiple languages and currencies?
Tools: Foreign currency and language translation tools are available.
Global business presents several
challenges. Some, such as cultural
issues, are too hard to handle with
technology. Others, such as for-
eign exchange conversion and even
some language translation, can be
handled by automated systems.
However, before you use these sys-
tems, you need to understand their
features and limitations.
Items sold in one nation are usually priced in the national currency. Products in
Europe might be priced in the national currency or in euros. Currencies exchanges
are handled by banks or by specialty foreign exchange (FX) brokers. Due to interna-
tional trade, large banks buy and sell currencies and record the going exchange rates.
Currently, the United States uses a foating exchange rate, where the rate changes
constantly in response to trade and interest rate differentials. Other nations, such as
China or the member European Union countries, fx their exchange rates so that they
change only when major economic conditions force a revaluation. Since the rates
change, you need to convert currencies on a specifc date. The other challenge with
exchange rates is that the commonly cited rates are from large interbank transac-
tions. Most people do not get to use those rates and have to pay additional fees to
banks or brokers. For example, many credit card transactions use the interbank rates
and then tack on 2 percent as a fee. Physically exchanging currencies at a border or
airport kiosk will cost you even more in fees.
Oanda is probably the most powerful online FX converter (www.oanda.com/con-
vert). It automatically converts between any two currencies (from164) on a specifed
date, using interbank or typical credit card rates. You can use the system to convert
prices on your Web site for customers in other countries.
Language translation is another challenge you need to solve to compete in world
markets. Ultimately, you need to have Web sites (and product documentation) trans-
lated by human experts. However, if you happen to receive a note or see a Web site
in a foreign language, or if you simply want to know the meaning of a few words,
you can use the automated online translators. Several free translation systems are on
the Web. The babel.altavista.com site is one that has been around for several years.
Some experimental translation software is becoming good enough that it matches
human translators. However, for now you are still better off hiring a person to trans-
late Web sites and documents that will be read by customers. The online sites are
useful for quick, approximate translations into your language. In most cases, you
should be able to understand the gist of the document, even if it contains errors or
poorly worded phrases. An interesting way to test the sites is to enter a phrase in
your language (say, English); translate it to a second language (say, Spanish); and
then ask the system to translate that phrase back (to English). See how far off it is
from the original phrase.
Quick Quiz:
1. What cautionary messages do global Web sites use when converting currencies?
2. If you cannot afford a human translator, is it better to leave your Web site in
English, or to use a machine translation?
Currency 1 U.S. Dollar Equals
Euro (EUR) 0.69022
Japanese yen ( JPY) 80.840
British pound (GBP) 0.62269
Australian dollar (AUD) 0.92384
Mexican peso (MXN) 11.7305
Access to Technology
Picture a world in which fnancial instruments are traded electronically, goods are
purchased through computer-television systems, libraries are converted to elec-
tronic media, and businesses require suppliers to exchange data over computer
links. Large portions of the United States and Europe are getting closer to this
scenario every day. Now, what happens to the individuals in poorer nations who
can barely afford to eat, much less invest in personal and national information sys-
tems? If the means of production are based on technology and certain groups do
not have access, the gap between those who have access and those who do not (the
digital divide) will widen. Although some groups will be content to live without
technology, some will become upset at the imbalance.
Figure 14.14 shows that access to the Internet requires more than simple PCs
and software. Individuals need access to telecommunication lines. More impor-
tant, nations need high-bandwidth connections to other nations. The fgure shows
the currently installed bandwidth between major world regions. Note the major
connections between the United States and Europe or Asia, and the relatively
small connections to Latin America and Africa. It takes time and money to install
new fber-optic connections across long distances. Telecommunication frms are
reluctant to incur these high fxed costs until a region has enough paying custom-
ers to cover the costs with long-term usage.
Some companies have worked to give others access to technology. A few re-
cycle older computers to libraries and citizen centers. On the international front,
businesses can donate older personal computers to organizations for shipment to
Figure 14.14
International bandwidth. Access to the Internet requires more than computers; it
requires high-speed communication lines between nations and continents. As noted
by www.telegeography.com, most connections are between the United States and
Europe. Numbers are estimates. See/pay TeleGeography.com for better numbers.
7,900 Gbps
4,400 Gbps
619 Gbps
2,700 Gbps
7,600 Gbps
other countries. After three to fve years, the technology is often out of date in the
United States, but even old technology is better than nothing in some countries.
The program one laptop per child has attempted to provide low-cost computers
to schools in developing nations, but they do not run standard software. And, little
work has been done to expand the bandwidth.
Wireless connections offer enormous potential to nations that cannot afford to
install fber-optic connections across their countryside. Wireless is particularly
convenient and substantially cheaper to install in high-density areas.
Still, before you begin to believe that the rest of the world is poor and bereft
of Internet access, consider that several other nations have superior Internet access
than the United States. Also, note that even in developing nations, big cities, gov-
ernments, and universities have computers and Internet connections. These facili-
ties make it possible for most people to obtain access to the Interneteven if they
do not have access at home.
e-Mail Freedom
Some organizations have observed an interesting feature when they frst replaced
paper mail with electronic-mail systems. The frst people to use the technology are
generally younger, more likely to take risks, and bolder than the typical employee.
If the top management levels accept and respond to electronic messages, they are
likely to get a different perspective on the organization, its problems, and potential
solutions. E-mail systems provide a means for employees (and sometimes cus-
tomers) at the lower levels to bypass the hierarchy of middle management. A few
directed comments and words of encouragement can enhance this effect, which is
useful if managers are searching for new approaches to solving problems.
Similarly, customers who spend considerable time on social networks and
Twitter can provide useful feedback about your company. Many frms monitor
these networks to identify potential problems and provide additional support to
highly-connected customers.
Liability and Control of Data
Virtually all of our legal structures and interpretations were created before the
advent of a computerized society. Although federal and state governments have
passed a few laws specifcally to address problems with computer interaction,
most legal systems still rely on laws and defnitions created for a paper-based
world. Sometimes these concepts do not ft well in a computerized environment.
For example, how would you classify the operator of a Web site? Is that person
a publisher of information, like a newspaper? Or is the operator merely a vendor
offering disk space to anonymous writers? In particular, are the owners of Web
sites responsible for the content of messages posted on their systems? To date,
the court systems have tended to make the decision based on whether the own-
ers exercise editorial control. In 1995, the New York supreme court ruled that
Prodigy could be sued for libel. An anonymous writer posted a message that was
highly critical of the fnancial status of a certain frm. The frm claimed that the
comments were false and sued Prodigy for publishing false information. Since
its inception, Prodigy maintained a policy of forbidding people to post profane
messages. The Prodigy staff used software to scan messages. The court noted
that these actions constituted editorial control, so Prodigy could be treated as
any other publisher of information (like a newspaper). These concepts were later
clarifed into law. Now, Web sites that do not exercise control over the content
are merely distribution channels (like booksellers) and cannot be held liable for
the content. However, many Web hosting companies place restrictions on content
(such as pornography) and will remove a site that is reported to violate its policies.
Government
Can information technology improve governments? Following
the expansion of the Internet, the concept of e-government became popular. In
some ways, government agencies are similar to businesses. Most federal and state
agencies now provide data and communications via the Internet. Some are quite
sophisticated.
Government Representatives and Agencies
Governments can be slow to adopt new technologies. Typically, government agen-
cies have limited budgets, long procurement cycles, and requirements for special
allocations to acquire new technology. They tend to have smaller IS staffs, who
also receive less pay than their counterparts in private business. Moreover, gov-
ernment projects tend to be large and involve thousands of people, which makes
them expensive, harder to create, and more diffcult to implement.
In the United States, the federal government has moved many data sources to
the Internet. Almost all federal data is available in computer form. Many agencies
are positioning themselves as providers of economic data to facilitate business de-
cisions. Fedstats (www/fedstats.gov) is one of the best starting points for fnding
data produced by federal agencies. Even municipal governments are beginning to
post notices and data on the Internet. Most government agencies are still nervous
about electronic commerce. One of the main problems they face is the inability
to positively identify consumers over the Internet. Of course, government agen-
cies operate on government time. Little has been done to reduce the time it takes
to release government data. For example, data reports from the 2000 census were
released over a fve-year time period, and data from the 2010 census began slowly
trickling out in 2011. Data from many agencies is months or years out of date
when it is released. Furthermore, many economic statistics are revised over time,
so preliminary numbers you see one month may be replaced with different values
several months later. Nonetheless, the government agencies are important sources
for many types of data. Government data is important for its accuracy.
Politicians campaigning for offce also use technology. But, due to spam, you
cannot send direct e-mail to them. Most have Web sites that enable you to enter
questions or sign up for marketing spam from the politicians offce. For many
years government offcials have used databases to track letters and comments, so-
licit contributions, and tailor speeches to specifc audiences. Politicians still rely
on television to create images, but Web sites are commonly used to provide de-
tailed position papers and background information that is too long to be covered
in depth by traditional media.
Politicians signifcantly use technology in election campaigns. They use data-
bases to pinpoint donors. More aggressively, they use databases to identify voters
who will cast votes for a specifc candidate, then use reminders and rides to the
polls to ensure those voters get to the voting booth.
Democracy and Participation
The U.S. Constitution and its amendments clearly recognize that democracy re-
quires the participation of the citizens. And participation requires that citizens be
informedhence the importance of the press. Information is required to produce
knowledge, which can lead to wisdom and better decisions. More important, it
is not always clear exactly what information will be useful later. The Internet is
a powerful source of information. Of course, distinguishing fact from fction is
critical. Yet today it is still possible for a nation to control the content available
within its borders. China maintains its hold by owning and controlling all routers
that connect to the Internet. Ultimately, it may become impossible for a nation to
control all information. Between the massive data fows, encryption, automated
document translation, and wireless capabilities, it will become increasingly dif-
fcult to control data.
These points were emphasized in the 2011 political upheavals in northern Af-
rica (Tunisia, Egypt, Libya, and so on).Cell phones and the Internet gave citizens
the ability to communicate and coordinate their actions. Dictators Mubarak and
Gadhaf actually shut down Internet connections and cell phone networks in an
attempt to restrict the ability of their citizens to communicate. (Side note to fu-
ture dictators: It is really hard to convince people you are a benevolent dictator
when you take away cell phones and the Internet.) Eventually, people found ways
to communicate. In Libya, enterprising workers built a completely separate cell
phone network.
Voting
With the fasco of the 2000 U.S. election, people began to realize the deplorable
status of existing voting systems. The level of mistakes due to machine, user, and
counter error is unacceptable in a modern society. Several people have mentioned
the possibility of creating electronic voting systems to provide faster and more ac-
curate tallies of votes. But many challenges exist as shown in Figure 14.15. Sev-
eral experts have testifed before Congress that they do not believe current tech-
nology is capable of surmounting all of the problems. But ultimately, the question
comes down to whether a superior system can be developed, even though it may
Prevent fraud by voters (identify voters).
Prevent fraud by counters.
Prevent fraud by application programmers.
Prevent fraud by operating system programmers.
Prevent attacks on servers.
Prevent attacks on clients.
Prevent loss of data.
Provide ability to recount ballots.
Ensure anonymity of votes.
Provide access to all voters.
Prevent denial of service attacks.
Prevent user interface errors.
Identify and let voters correct data entry errors.
Improve on existing 1 in 6,0000 to 1 in 10,000 error rates.
Figure 14.15
Electronic voting requirements. Electronic voting sounds convenient and easy to
set upuntil you look at the detailed requirements. Many avenues for fraud exist.
Additionally, complex systems are hard to create and susceptible to errors.
not be perfect, and whether it can prevent major problems. There is a long history
of building and revising voting machines in an attempt to minimize fraud and
abuse. But existing machines still miscount an average of one in 6,000 to 10,000
ballots. The other serious drawback to existing systems highlighted in the 2000
election was the usability issue, where thousands of ballots were disqualifed and
thousands more counted incorrectly because people did not understand them.
A few people have suggested that it would be nice to implement a voting sys-
tem that works as easily as ATMs or even using their own PCs to connect over
the Internet. But electronic voting has two main complications over traditional
electronic commerce. First, it is critical to authenticate each voter. Current e-com-
merce handles this step with credit cardswhich are not available to all voters
and not secure enough to use as a public voting identifer. (What would stop a
business from assuming your identity?) Second, the votes have to be auditable,
but anonymous so that no votes can be traced back to an individual. This second
condition is even trickier if you are concerned about vote selling. Ideally, voters
should not be able to show their fnal vote to anyone else. If they can, it opens the
possibility of buying votes. Currently, there is little incentive to buy votes because
there is no way to prove how someone voted, so no way to enforce the agreement.
Voting from your home over the Internet might take years to develop, largely
because of the challenge of protecting the client computers and denial of service
attacks. Security experts can protect the servers and data transmissions can be
protected through encryption. But how can a government ensure the security of a
PC in your house? Given the level of viruses, hoaxes, and false statements on Web
sites, it would seem to be relatively easy to attack millions of home computers to
control an election.
On the other hand, society has an additional critical objective in designing a
new voting system: the need to make it easier for people to vote to increase par-
ticipation rates. To combat this problem, several states have implemented paper-
based ballots shipped to each persons home. Ultimately the point is that no sys-
tem is perfect, so the question quickly becomes whether an electronic vote system
is better than the existing methods, and whether it is possible to prevent signifcant
fraud. In a test of electronic voting systems in the Georgia 2001 election, almost
all said the system was easy to use and over 94 percent said the entire state should
move to the electronic system.
Reality Bytes: Where Have You Been?
In 2011, researchers looking at cell phones noticed that Apple iPhones and Google
Android phones routinely tracked the cell phone location and stored the data in an
unprotected fle on the phone. Both Apple and Google later admitted that the phones
regularly sent the location data to their corporate computers. Both companies appar-
ently used the data to build a location database. Apple and Google are not alone. A
Wall Street Journal examination of 101 popular apps found that 47 of them sent loca-
tion information to app vendors and other companies. Once the data is collected, no
rules exist, so it can be sold to anyone else.
Adapted from Julia Angwin and Jennifer Valentino-Devries, Apple Receives iPhone
Location Data, The Wall Street Journal, April 21, 2011.
The issues of electronic voting systems are constantly debated in the press.
Computer scientists state that even electronic voting booths can never be trusted,
let alone voting over the Internet. Yet, many citizens actually would prefer the
convenience of voting over the Internet, so there is pressure to develop a workable
solution. Some interesting cryptographic methods exist that have the ability to
make it possible. Securely implementing these technologies will take time, and re-
quire some interesting adaptations. One powerful trick (homomorphic encryption)
is the ability to encrypt votes so that it is possible to record the vote totalwithout
ever observing or decrypting the individual votes.
Information Warfare
As frms and entire economies become more dependent on information systems,
the underlying infrastructure becomes critical to the nation. Think about how the
information society will work in 10 years or so. Communication, including tele-
phones, will be based on Internet protocols. B2B e-commerce will take place over
the Internet, with automated agents placing orders and handling most transactions.
Private and government services will be provided through Web sites. Web ser-
vices will be offered through interlinked sites.
Now, imagine what happens if some nation or group decides to attack this in-
formation system. Inexperienced people using software scripts found on various
Internet sites have already attacked individual companies. A few major attacks
Reality Bytes: Movie Theaters v. DirecTV
Video is an interesting case in technology. Largely because of the bandwidth issue,
video was one of the last entertainment categories to be digitized. It is also hugely
expensive to create movies and entire business models evolved to handle the fnanc-
ing, production, and distribution of video, with many of the steps controlled by a few
large frms. The one thing that most of the industry agrees on is that movies have a
shelf life or immediacy. People will generally pay higher prices for newly released
movies, but a considerable amount of back-end money exists for older movies that
can be sold to a wider audience at lower prices. In 2011, DirectTV signed a deal with
some studios that allows the satellite company to rent movies to customers for about
$30only two months after they debut at the theater. Traditionally, other offerings
provided movies four months after release for a price of about $6. Theater owners
and several directors (including James Cameron) expressed displeasure with the plan
saying it was a distribution model that cannibalizes theatrical ticket sales. Studios
counter that most movies make their money within a couple weeks of release and
that the need to make up revenue lost to declining DVD sales and changing con-
sumer habits. But director Todd Phillips noted that Knowing a flm will be available
at home so quickly removes the sense of urgency people feel to go see the flm in
theaters. Directors want to see the large screen and social atmosphere of theaters
remain as the primary market for movies. Theater owners are concerned about go-
ing out of business. But, ultimately consumers will make the decisions on when they
want to see movies and how much they are willing to pay.
Adapted from Michelle Kung and Ethan Smith, Filmmakers Pan DirecTV Plan,
have been launched against the Internet DNS infrastructure. These denial-of-ser-
vice attacks can be mounted by anyone. If an experienced, dedicated group of ex-
perts attacked a nation, they could stop service to huge segments of the economy.
This threat is one aspect of the Internet that scares many agencies.
The United States and other national defense departments have begun planning
for information warfare (IW)in terms of both potential defenses and attacks.
The ultimate objective of information warfare is to control the information avail-
able to the other side so that you can encourage them to take certain actions. This
defnition includes the ability to intercept communications, as well as to provide
new data that will be accepted as valid. IW goes way beyond hacking into a sys-
tem or destroying enemy computers and networks. In many ways, IW has been a
part of war and conquest for centuries. The increasing use of computers, both in
the military and in economies, has made IW more important. IW has existed in
many respects from the early centuries of warfare. Some aspects became promi-
nent in World War II, such as code breaking, the Navajo code talkers, the use of
the BBC to send coded signals, and misinformation. Misinformation and control
of the press (domestic and worldwide) have become key aspects to IW, particu-
larly given the worldwide reach of CNN.
Some U.S. reports indicate that the Chinese military is attempting to develop
viruses that can be inserted into foreign networks to disrupt the fow of data or
provide false information. In 2010, an interesting virus Stuxnet was released
on the world. Analysts who studied the virus determined that it was specifcal-
ly targeted to attack Iranian nuclear facilities and was probably created by some
government-funded agency (probably not in the U.S.). Some reports out of Iran
indicated that it was successful in causing at least temporary problems at several
facilities. Information attacks can be targeted against military or civilian objec-
tives. Military uses of information warfare are common today. One of the frst
steps the U.S. air force takes is to disable the enemys air defense systems to gain
control over the enemy airspace.
Civilian attacks are still new, but the potential is huge. The military goal would
be to destroy the economic ability of a nation to build and deploy weapons, but an
RealityBytes:ChinasStateInternetInformationOffce
China has developed one of the strongest controls over the Internet of any nation.
Particularly given the population and size of the Internet, managing and controlling
policy can be a complex task. In 2011, China created a new central agency to han-
dle all aspects of the Internet known as the State Internet Information Offce. It is
not clear how the many existing agencies will ft into the new organization, but it
is in charge of online content management, supervision of online gaming, video
and publications, promotion of news sites, and government propaganda. The Offce
would also have the authority to investigate and punish people and companies who
violate the rules. It will also be in charge of the telecommunications companies that
provide Internet access. At least 14 other agencies have a hand in controlling the
Internet in China, but the new offce makes it clear that the Chinese government will
continue to enforce limits on access to the Internet.
Adapted from Michael Wines, China Establishes New Internet Regulator, The New
York Times, May 4, 2011.
attack would also destroy the underlying infrastructure. The Internet was original-
ly designed to survive military attacks through decentralization and the ability to
route around broken links. However, as e-commerce has evolved, several vulner-
abilities have been created that would enable governments or terrorists to disrupt
major sections of the Internet by attacking some critical points.
In general, the same security controls that businesses use to protect systems
on a daily basis are important to defend against international attacks. Ultimately,
many aspects of the Internet infrastructure need to be improved to prevent attacks
by terrorists, since the underlying components were not designed with security in
mind. Several Internet committees are working on these new standards.
Rise of the World-State?
In ancient history (literally), communities of people formed into city-states to
share common resources and provide a common defense. Because communica-
tion was limited and transportation costs were high, the city-states were largely
self-suffcient. However, merchants traveled among cities to barter products that
were only available in some locations. Over many years, transportation and com-
munication costs declined, giving rise to nation-states. Through various battles
and political arrangements, people accepted the role of the national governments,
although some issues are still being fought.
Reality Bytes: Google v. National India and Kazakhstan
As a leading Internet company, Google encounters many issues around the world.
The company has taken the lead in many issuestrying to convince nations to re-
main open and allow everyone access to content on the Internet. In 2011, Indian
regulators passed a set of rules that imposed restrictions on Internet content. Among
other things, the regulations created by the Ministry of Communications and Infor-
mation Technology require Web sites to remove objectionable content and to respond
to government requests to remove content within 36 hours. Google was primarily
concerned about being held liable for indexing content that was hosted by other sites.
But a key issue in any Web site is whether the site is required to monitor posted
content to look for possible violations of national laws. In the U.S. the laws do not
require monitoring but do require some content to be removed when it is brought to
the attention of the Web site publisher. In a more extreme situation, the Kazakhstan
government in May 2011 ordered all Kazakh domains (a top-level domain of .kz) to
route traffc only to servers in Kazakhstan. Kazakhstan is a former Soviet republic
that is governed by a long-term president in an increasingly authoritarian regime. In
response, Google suspended operations on its internal Kazakhstan servers and routed
all search requests to the generic Google.com servers. Bill Coughran, a Senior Vice
President at Google, noted that We fnd ourselves in a diffcult situation. Creating
borders on the Web raises important questions for us, not only about network eff-
ciency but also about user privacy and free expression.
Adapted from Amol Sharma, Google Raised Objections to Draft India Internet
Rules, The Wall Street Journal, May 10, 2011; and Alan Cullison, Google Redi-
rects Kazakhstan Traffc, The Wall Street Journal, June 8, 2011.
For years, many writers have suggested that increasing international trade, de-
clining transportation costs, and improving global communication will eventually
lead to a world-state. National governments might still exist, but commerce would
be more regional and global, and world laws would be more important than na-
tional policies. The rise of the European Union and other free trade areas (such as
NAFTA) are sometimes seen as forerunners of this world.
International e-commerce provides some support for this hypothesis. In an en-
vironment where digital data and services can be transferred instantly around the
world, it is easy to see the irrelevance of individual national laws. With encryption
and a wireless (satellite) connection, how can a national government impose rules
or taxes on the digital transfer? If a serious digital monetary system is developed
and accepted, how will a nation impose its independent economic policies? Some
of these issues are being addressed today by global political organizations. Na-
tions are slowly learning to cooperate and create common procedures and laws.
On the other hand, a world-state would be a massively complex system that
would undoubtedly be politically unstable. There are still many regional tensions
and periodic fghts over physical resources. It would take many years of pros-
perity and economic growth before nations were willing to accept a truly global
government. However, in the meantime, many issues will need to be negotiated in
a global setting because they are beyond the control of any national government.
Some international organizations facilitate these discussions, but most are some-
what cumbersome.
World Government Cases
Some of the issues in world governments are best seen through cases. Figure
14.16 shows one of the frst major international cases. France has a national law
that makes it illegal to trade (buy and sell) Nazi memorabilia. Based on their ex-
periences in WWII, this law probably makes sense for France. Yahoo is a U.S.
Figure 14.16
France v. Yahoo. France charged Yahoo with violating French law by allowing the
sale of Nazi memorabilia online. A French court convicted and fned Yahoo, although
Yahoo had no physical presence in France. A U.S. court later refused to grant France
the ability to collect the money from Yahoo using U.S. courts.
Yahoo:
Auction site
French law:
Illegal to trade
Nazi memorabilia
Auction: Sell Nazi
Buy
Should a nation
be able to set
law for the rest
of the world?
based company that started as one of the leading search engines. Yahoo also had
an online auction system at one time. The problem is that anyone in the world
could use the auction system, so someone might offer to sell or buy Nazi trinkets
through the systemand conceivably, a French citizen might be the buyer or sell-
er. France apparently asked Yahoo to block all trading in Nazi memorabilia, but
Yahoo decided it was too hard or expensive to do that just for France. So, France
prosecutors charged and tried Yahoo in French courts for violating their laws. Un-
surprisingly, the French prosecutors wonthe only surprise is that Yahoo both-
ered to show up. The courts imposed large fnes on Yahoo. However, Yahoo had
no physical facilities in France, and later, U.S. courts ruled that France had no au-
thority to use American courts to extract the money from Yahoo. Of course, Yahoo
would never be able to establish offces in France in the future without addressing
the problems. The real issue of the case is whether one country can impose its
laws on other nations, simply because they are connected through the Internet.
Now, consider the case of Microsoft v. Fujian Dongbai Group, which is a de-
partment store in China. Microsoft (and the Business Software Alliance) charged
the company with violating copyright laws by using copies of Offce and Win-
dows on its computers to run its business. In April 2011, Microsoft prevailed and
the Chinese company agreed to pay about $138,000 to Microsoft (see the Wall
Street Journal note on April 21, 2011). How is this case different from the Ya-
hoo case? The most important difference is that for several years, the U.S. gov-
ernment has diplomatically been pressuring nations, including China, to create
Figure 14.17
America v. WikiLeaks. WikiLeaks published several documents obtained from the
U.S. government, including diplomatic cables marked secret. The U.S. government
took several steps to harass or shut down WikiLeaks. Ultimately, WikiLeaks moved
everything out of the U.S. and continues to run its servers at www.wikileaks.ch.
1. Video from Iraq on alleged
U.S. abuses.
2. U.S. military documents on
Iraq and Afghanistan wars.
3. U.S. diplomatic cables or
State Department e-mails,
many classified secret.
Uploads to WikiLeaks
Amazon
PayPal
EveryDNS
1. After cable postings, WikiLeaks servers
are attackedrumors mention
government.
2. WikiLeaks moved to Amazon servers for
bandwidth and capacity.
3. Government officials contacted Amazon.
Amazon kicked out WikiLeaks for
violating terms of service.
4. Government officials contacted PayPal
for taking donations. PayPal dropped
WikiLeaks.
5. EveryDNS dropped registration for
WikiLeaks.
intellectual property laws (including copyrights). China now has a copyright law,
and Microsofts case was presented to Chinese courts. So, can a country (the U.S.)
impose its laws on other countries? Not directly, but perhaps it can convince oth-
ers that the law is necessary and each nation can implement similar laws. This
situation is a classic case of international cooperation and diplomacy. Over time,
several laws have been presented for adoption around the world. Copyright is a
big one, but computer crime laws including hacking and anti-terrorism are also
important.
Figure 14.17 presents another recent case that is more ambiguous. WikiLeaks
is a site and European company that is dedicated to publishing information pro-
vided by insiders to reveal corruption or other problems within organizations. In
2010, a person within the U.S. military (possibly Bradley Manning, a network
technician), sent three major items to WikiLeaks: (1) Combat video from Iraq
showing probable attacks by U.S. soldiers on civilians; (2) Classifed govern-
ment documents regarding the wars in Iraq and Afghanistan; and (3) Thousands of
U.S. State Department diplomatic cables (e-mails), many of which were classifed
as secret. This latter group of fles appears to have angered President Obama
and the U.S. State department. Although some of the documents were simulta-
neously published by traditional newspapers (such as The Wall Street Jour-
nal, The Guardian, and El Pas), most of the anger appeared to be directed
at WikiLeaks. (See news accounts beginning in the Fall of 2010 for many details
and rumors.) WikiLeaks quickly fell under attack from people trying to shut down
the site with denial of service attacks. WikiLeaks claimed the attacks were coordi-
nated by the U.S. government, but evidence has not been provided to support that
claim. To handle the attacks by providing increased bandwidth and server capac-
ity, WikiLeaks purchased space on Amazons cloud system. Apparently, U.S. poli-
ticians contacted Amazon and asked that the site be removed. Amazon complied
immediatelystating that WikiLeaks was in violation of the terms of service
agreement. Another U.S. based company, EveryDNS, was the DNS registrar for
the main WikiLeaks.org name. That company also dropped WikiLeaks as a cus-
tomer, claiming that the DoS attacks were causing problems with their operations.
Reality Bytes: Cut Our Network and Well Bomb You
In 2011, the Pentagon announced that some acts of cyberwar will be treated as any
other type of attack on the United States. If electronic attacks on computers or the In-
ternet cause death, damage, or high-level disruptions equivalent to a traditional mili-
tary attack, then the U.S. will have grounds to treat the attack as an act of war and
respond with use of force. Or, as an unidentifed military offcial put it If you shut
down our power grid, maybe we will put a missile down one of your smokestacks.
These new principles are based on the Pentagons interpretation of the Laws of
Armed Confict. The Pentagon still needs to determine which acts or levels would
trigger a response and what level of response might be applicable to each case. And
the bigger problem is going to be tracing the actual source of any attack. Still, na-
tions need to begin thinking about and negotiating methods of identifying, stopping,
and perhaps retaliating against cyber attacks.
Adapted from Siobhan Gorman and Julian E. Barnes, Cyber Combat: Act of War,
The Wall Street Journal, May 31, 2011.
This step meant that no one could fnd WikiLeaks.org by entering just the name.
WikiLeaks raised some of its money through donations to a German company that
used PayPal to transfer funds. Again, it appears that American politicians pres-
sured PayPal to drop the WikiLeaks company, and PayPal complied immediately.
WikiLeaks still survivesin part because a Canadian company (EasyDNS) is
handling the domain registration for the WikiLeaks.ch name. But, the case raises
important questions about the role of governments in controlling various elements
of the Internet. The WikiLeaks case is particularly interesting because it involves
the U.S. government and a complete lack of due process or appeal. But, because
the frms involved are individual companies, they are free to set their own poli-
cies. It is not clear what actions the U.S. government would have taken if compa-
nies such as Amazon or PayPal had declined to drop the WikiLeak sites.
Many subtler international and ethical issues exist across the Internet. For in-
stance, China owns and controls all of the Internet routers in that country. It rou-
tinely blocks many sites and has the ability to track online activities of its citizens.
As a sovereign nation, China is free to perform all of these tasks. But, some people
have criticized American companies such as Google and Cisco for creating tools
that enable China to implement these policies. The situation raises the question of
whether companies should provide capabilities requested by large customers. A
pure business approach would argue that anything that makes more money for the
company is good. But, are there features or tools that companies should stay away
from? The answer is probably yes, but who makes these decisions? Do these
decisions need to be made within individual frms, individual nations, or with the
help of a global forum? The Internet is still young and many questions remain to
be answered.
Crime
Do criminals know how to use computers? Crime has many aspects
both in the Internet/information world and in the real world. Security issues
related to protecting information systems and Web sites are discussed in detail in
Chapter 5. The issues in this chapter refer to questions of how governments can
combat crime in society. Criminals today have access to the same technologies as
everyone else. Drug dealers and weapons merchants use encrypted spreadsheets
to track their sales. Terrorists use encrypted e-mail to transfer information. Con
artists use the Internet to steal money from victims. Entirely new forms of ha-
rassment and stalking have been created with chat rooms, e-mail, cell phones,
and other electronic communication systems. Most people want the government
to protect them from these many forms of crime. The complication is that the elec-
tronic tools make it more diffcult for police to work. So you as a citizen need to
identify the trade-offs you are willing to accept.
Police Powers
For years, politicians have used the threat of crime to argue for granting increased
powers to police agencies. Interception and decryption of communications (wire-
tapping) is a classic example. The United States passed the Communications
Assistance for Law Enforcement Act (CALEA) in 1994. It has taken effect and
requires that when requested by the government, any telecommunication com-
pany must route any communication that passes through its facilities to an off-site
U.S. government facility. The FBI similarly created the Carnivore (DCS-10000)
system to monitor and record all Internet communications of a targeted person.
However, the FBI gave up on its technology and beginning in 2007 now requires
all ISPs of any size to purchase and install equipment to collect and forward to a
law enforcement agency all transmissions to or from a specifc client. VoIP is the
primary reason for the law, to enable law enforcement agencies to intercept phone
calls placed over the Internet, but it also applies to data. In practice, if you run
even a small wireless access point that gives public users access to the Internet,
you must install technology to capture these transactions and forward them to the
police on demand. On a global scale, the National Security Agency (NSA), in
cooperation with other national partners, routinely captures and monitors interna-
tional communications. Under federal mandate, wireless providers are phasing in
locator systems designed to route emergency crews to callers who use cell phones.
Of course, these same locator systems could be used by police to monitor the loca-
tions of suspected criminals.
Three questions must always be addressed with each new technology: (1) Is
the technology effective or are there other ways to accomplish the same result?
(2) How can society control the use of the technology and is it worth the loss of
privacy? (3) Who is going to pay for the technology? The technology press con-
tains many stories of abuses of power and informationincluding those by IRS
agents and state and local police agencies. The police can also tell stories of how
the criminals use modern technology to thwart investigations, and how additional
police powers can be used to reduce crime.
Freedom of Speech
As constitutional scholars have long known, freedom of speech is a diffcult con-
cept. In practice, many limits are placed on individual speech to protect society.
The classic example: you are not allowed to yell Fire! in a crowded theater
(when there is no fre) because the result is dangerous to many people. Similarly,
Reality Bytes: Do Not Call, We Mean It
Spam, or unwanted commercial messages and phone calls are a global problem. And
when most of the phones are mobile and people pay for minutes of usage, the calls
are beyond annoying. Most nations have followed the lead of the U.S. and estab-
lished do-not-call registries. The problem is that many, including the U.S., have min-
imal enforcement provisions. India, the number two country behind China, in terms
of mobile phones learned the problem the hard way. Although a do-not-call registry
was established in 2007, it was widely ignoredactually, the list was often used as
a base list for calling. Until the Finance Minister Pranab Mukherjee received a tele-
marking call while in Parliament in 2010. A new law adds progressively higher fnes
on telemarketers, and requires commercial messages to be tagged with key numbers
or headers. More importantly, the cell phone carriers will be responsible for increas-
ing fnes for each infraction; starting at $2,200 for the frst offense and quacking in-
creasing to $22,000 for the third or more offenses. Wireless carriers are complaining
about the cost of installing hardware and software to comply with the law. And the
largest provider, Bharti Airtel Ltd, announced it would stop providing text service to
telemarketers.
Adapted from Megha Bahree, India Adds Teeth to Do Not Call, The Wall Street
there are restrictions on speech on the Internet. A big element is that you can-
not defame or harass others. While this statement seems obvious, what happens
when people sign up with an anonymity server? They could then use free e-mail
services, chat rooms, and Web sites to attack other people or companies.
The fip side of this situation is the issue of how to control these problems.
Should a police agency have the ability to routinely break the anonymity server
to identify all people? But that raises the question of what constitutes defamation
and harassment? It is legal for a person to report truthful information about a com-
pany or an individual, but sometimes marginal in whistleblower situations. But
what if the person being criticized is a public offcial and uses the police power to
retaliate and harass the original person? Of course that action would be illegal as
well, but how do you prevent it?
The main thing to remember is that there are many sides to all of these discus-
sions. Also remember that many people have strong personal preferences on each
side, and debates are often flled with emotional and unsubstantiated claims. In the
coming years, these topics will become increasingly important. It is critical that
you form an educated opinion and make sure that your voice is heard.
Gambling
Gambling on the Internet is a multibillion dollar, worldwide business. In many
countries, gambling is routine and available at many locations. In the U.S., gam-
bling is largely discouragedexcept in some states or cities, and in cases where it
is used to raise money for governments, and you can probably fnd a dozen other
exceptions. The Federal Wire Act makes it illegal to gamble by wire. It was origi-
nally created to prevent gambling institutions from spreading through telephone
and telegraph connections. Each state also has its own gambling laws, making
it challenging to determine the status of gambling in any location. However, if
it crosses a state line, the Federal law can be applied and enforced. In the ear-
ly 2000s, the U.S. government used the law to effectively force basic gambling
Web sites out of the United States. Of course, several international sites exist and
provide gambling to anyone on the Internet. To prevent U.S. citizens from using
these sites, The Unlawful Internet Gambling Enforcement Act of 2006 enabled the
government to go after credit card companies and banks and prevent Americans
from using credit card payments to gamble online. Consequently, it is diffcult for
Americans to gamble online because it is diffcult to make payments and create
accounts.
Partly due to the absence of general online gambling, and partly due to the tele-
vised national tournaments, online poker became a large business in the United
States. At least until April 2011, when the U.S. government shut down several
U.S. based poker sites, confscated their bank accounts, took control of the do-
main names, and charged several owners of the sites with violating the Federal
Wire Act (which includes jail time and monetary penalties if convicted). The ini-
tial defense by the owners was that poker is not gambling, but that argument could
be diffcult to win. However, even if the poker sites return, you should avoid them.
By 2011, many poker-bots, or automated systems were capable of beating even
the best human poker players. And, if the bots could coordinate their activities
within one game, there would be no way to win against them.
The era of online poker playing is probably over. But, how long will it be before
something else attempts to take its place? Does the U.S. need a better defnition of
gambling? Or, should gambling simply be expanded and opened up to any site?
The evils of gambling are often espoused as a problem, but are U.S. citizens more
susceptible to gambling addiction than other nations? Of course, online gambling
could be a bigger problem for some people. Money lost online might not seem as
real as money lost in hand. Again, these are questions that you have to answer as
citizens through voting.
Responsibility and Ethics
How do your actions affect society? Is it possible to follow the
laws and still be wrong? In any society, but particularly in one with open
information, ethics and morality are important. Laws do not always keep up with
changes in society. Think about small towns a century ago. A few basic laws ex-
isted, but people generally behaved responsibly, in part because if you gained a
negative reputation, people would not trade with you. In an Internet world where
people can write almost anything about you for others to fnd, it is possible that
your reputation and honesty will become even more important. For example,
eBay uses a ratings system to formalize these concepts. But even without worry-
ing about your future prospects, you should strive to create an honest world.
Users
Computer users have certain responsibilities in terms of computer security and
privacy. First, they have an obligation to obey the laws that pertain to computers.
The U.S. government and some states, along with other nations, have laws against
computer crimes. Most other traditional laws also apply to computer crimes. One
law that has received much attention is the copyright law. European and U.S.
copyright laws prohibit the copying of software, music, books, movies and other
content except for necessary backup. It is the responsibility of users to keep up
Reality Bytes: Telecommuting Tax Problems
The old world rules that are dependent on physical location cause problems when
they encounters new digital technologies. Telecommuting seems like a beneft to em-
ployees and businesses. Even for cities, telecommuting should be able to reduce the
demands on services such as transportation, highways, and parking. But, as cities
face losses in tax revenue, politicians become more creative in seeking ways to make
money. In March 2010, the Tax Court of New Jersey ruled that TeleBright Software
Corp., a company with its offces in Maryland, was actually doing business in New
Jerseysimply because one of its employees lived in New Jersey and used telecom-
munication to work for the frm. Why does anyone care? Because if a company is
declared to have a business presence in another state, it also has to pay income and
sales taxes on any sales in that state. A survey conducted in April 2011 revealed that
35 states/regions (including D.C and New York City) said telecommuting would cre-
ate a nexus allowing the state to impose income taxes on the company who em-
ploys that worker. In other cases, cities (particularly New York City) have declared
that telecommuting means the worker is still subject to income taxes at the location
of the company. So, a worker living in Kansas and working for a frm in NYC would
have to pay income taxes to NYC (and Kansas).
Adapted from Barbara Haislip, The Hidden Cost of Letting Workers Telecommute,
with the changes in the laws and to abide by them. In the last few years, publish-
ers have increased their efforts to stop illegal copying of software, called software
piracy.
Although it might seem to be trivial, making illegal copies of software (or vid-
eos, books, or other copyrighted works) can cause several problems. First, it takes
money away from the legal owners of the software, which reduces their incen-
tive to create new products. Second, you run the risk of hurting your employer. If
employees illegally copy company-purchased software, the owners of the copy-
right can sue the employer. Third, copying software provides an illegal advantage
over your competitors. A small design frm might decide to copy a $20,000 CAD
design system instead of buying it. Consequently, honest frms are hurt because
the original frm will be able to make lower bids on jobs because their costs are
lower. Fourth, as an individual, you have a reputation to defend. If your friends,
colleagues, or employers learn that you are willing to break the law by copying
software, they can easily believe that you are willing to break other laws.
Users of computer systems also have an obligation as part of computer ethics
to customers and clients. Most information in computer databases is confdential.
It should not be revealed to anyone except authorized employees. Some nations
have laws to protect this privacy. If you fnd a company violating these laws, it is
your responsibility to question the practice.
Users have an obligation to use the information provided by computer applica-
tions appropriately. When a user sets up calculations in a spreadsheet, the user
must realize that those calculations might be wrong. The calculations must be test-
ed and the information produced should always be checked for reasonableness.
You should not believe information solely because it comes from a computer. All
data should be verifed.
Programmers and Developers
Programmers would never get jobs if they could not be trusted. This trust is one of
the most crucial requirements to being a programmer. As a programmer or devel-
oper, not only do you have to be honest, but you must also avoid any appearance
of dishonesty. For example, practical jokes involving security violations can be
dangerous to your career.
Programmers have more responsibilities than many other employees. Software
is used in many critical areas. If a programmer attempts a job that is beyond his
or her capabilities, crucial errors can be introduced. For example, consider what
might happen if an underqualifed person took a job programming medical life-
support systems. If he or she made a mistake, a person might die. Although mis-
takes can be made by anyone, they are more likely to arise when a programmer
attempts too diffcult a job.
Along the same lines, programmers have an obligation to test everything they
do. It also means that companies have the responsibility to provide adequate time
for programmers to perform the tests. The important step is to identify compo-
nents that are critical and to build in safeguards.
There have been enormous increases in the demand for software in the last de-
cade. At the same time, new tools allow programmers to create much more com-
plex applications. But our ability to create this new software has far outstripped
our ability to ensure that it is error-free. Even commercial programs, such as word
processors and spreadsheets, still have errors that can cause problems. In spite of
the best efforts of conscientious, talented people, software used appropriately can
produce erroneous information.
Liability for erroneous information produced by software has not been fully
established yet. Laws and court decisions during the next few years should settle
many aspects of who is responsible when software makes mistakes or fails. A
related issue is the extent to which the user is responsible for correctly entering
information needed by the program and for using the information produced by the
program appropriately.
Companies
Every company has obligations to society, customers, employees, and business
partners. In terms of society, a frm must obey all relevant laws. For customers,
frms must ensure privacy of data. That means companies will collect only the
data that they truly need. The data must be safeguarded so that only those who
need it for their job have access. If customer information is sold or distributed
for other purposes, customers should be notifed. Consumers must be allowed to
remove their names from any distribution lists.
For employees, a company must provide training and monitoring (compliance
programs) to ensure they understand the laws and are following them. Firms must
provide suffcient funds to allow the employees to meet their individual respon-
sibilities. Companies must provide enough time and money to test software ad-
equately. Firms have an obligation to allow their employees a certain amount of
privacy. For instance, companies have no reason to routinely monitor and read
employees electronic mail messages.
Companies are required to abide by all partnership agreements. In terms of
computers, they must safeguard all data acquired from partners. They must not
use the data in a manner that would injure the frms involved.
Governments
Federal, state, and local governments have obligations to establish laws that pro-
vide a means for those unfairly injured to allow them to gain compensation from
those who did the damage. Until the 1980s, relatively few laws at any level were
specifcally directed at computer usage. Instead, laws intended for other purposes
were stretched to cover computer crimes. Frequently, citing mail fraud laws was
the only recourse. Some criminals were not convicted because the crime was con-
sidered victimless by the jury, or the injured corporation declined to prosecute.
Starting in the mid-1980s, the federal government and nearly every state passed
new laws concerning computer crime. In 1986, the Computer Fraud and Abuse
Act and the Electronic Communications Privacy Act were enacted. The Comput-
er Fraud and Abuse Act makes it a federal crime to alter, copy, view or damage
data stored in computers subject to federal law. The law provides fnes of up to
$100,000 and up to 20 years in prison. The Computer Abuse Amendments Act of
1994 expanded the original defnitions to include transmission of harmful code
such as viruses. It distinguishes between actions taken with reckless disregard
for potential damages (misdemeanor) and intentionally harmful acts (felony). It
also modifed the language so that crimes causing damages of more than $1,000 or
involving medical records are within federal jurisdiction. Furthermore, it placed
controls on states in terms of selling drivers license records.
Most states have enacted similar laws for the few computers that might not be
subject to federal law. European countries have been ahead of the United States in
developing legislation to deal with computer crime.
In terms of enforcement, most federal, state, and local agencies have few, if
any, offcers devoted to solving computer crimes. In fact, many software piracy
cases have been pursued by U.S. Secret Service agents. One complication is that
most law enforcement agencies lack proper training in computer usage and inves-
tigation of computer crimes.
Some Computer-Related Laws
What major laws affect technology and the use of computers?
Laws form the foundation of society. They provide the structure that enables busi-
nesses to exist. As society changes, the laws must also be changed. Hence, as the
use of computers grows, we can expect to see more laws governing their use.
Existing laws will be extended and new ones created. To date, computer laws
have been concerned with three primary areas: property rights, privacy, and crime.
These areas overlap, and they cannot cover all possible issues. As information
technology and robotics become entwined into all our activities, virtually any law
can be applied or interpreted to the situation.
Laws continually change and new interpretations and applications regularly
arise. You will generally need a lawyer to help you understand and apply the cur-
rent laws. This short appendix can provide you with only a limited background.
You can fnd additional information in many places on the Web. This information
will help you identify problems and generally enable you to obey the laws. How-
ever, a lawyer is still the best source of informationparticularly if you anticipate
problems or conficts.
Property Rights
A property right gives you ownership and control over an object. While the term
originated with physical property, the more important issues now involve intellec-
tual property. If you write a book, a song, or a computer program, you should be
able to receive money from sales of that item. Copyright, patent, trademark, and
trade secret laws provide defnitions of ownership and control transfer of these
rights. They provide you with the opportunity to prevent others from using or
stealing your work. Each of the four types of property-rights laws applies to dif-
ferent material.
Copyrights are used for books, songs, and computer software. The laws ap-
ply to the specifc item, such as a book. You cannot copyright a general concept.
For example, you can obtain a copyright for a specifc word processing applica-
tion. But other people are free to write similar applications, as long as they do
not utilize your specifc code or text. Copyrights generally last for 50 years after
the death of the writer. In the case of a work produced by a group of workers in a
company, the copyright lasts for 75 years after the publication of the work. After
that time, the work falls into the public domain, where anyone can use or copy
it with no restraints. The times vary by country and have been changed several
times, so always check on current values.
Patents were originally designed for mechanical devices, although today you
can receive a patent for any device that is innovative and useful. For many years,
computer software applications could not receive patents because laws of na-
ture including mathematical algorithms were specifcally excluded. In the last
few years, the U.S. Patent Offce has changed this interpretation and now grants
patents for computer software. A U.S. patent right exists for 20 years from the
date the application was fled. The strength of a patent is that it prevents other peo-
ple from creating a similar product, even if they do not directly copy your work.
Consequently, a patent is much more diffcult to obtain than a copyright.
Trademarks are used to create a unique name. Once you fnd and trademark a
name (or logo), no one else can use that name without your permission. It is rela-
tively easy to obtain a trademark, except that you must fnd a name that no one
else has already chosen. You can begin your search at the U.S. Patent and Trade-
mark Offce: http:/www/uspto.gov.
Trade secret laws provide you with the ability to seek damages if someone
steals your secret information. The catch is that you are responsible for protecting
the information. The laws are generally used to enforce a nondisclosure agree-
ment (NDA). If a company wants to work with another company or a consultant,
it is a good idea to have the outsiders sign an NDA, in which they agree not to re-
veal any information you share. If you forget to have them sign an NDA and they
release your secret information, you will have few options. It is your responsi-
bility to protect the data.
These four basic protections have different purposes and different strengths and
weaknesses. Copyrights and trademarks are relatively easy and inexpensive to ob-
tain. You simply fll out a form, submit the material, pay a fee, and wait a few
months for the agency to process the request. Actually, a copyright exists as soon
as you create the material. You do not need to fle the registration form. However,
there are some legal and monetary advantages to registering the copyright. Pat-
ents require considerable documentation and a formal review to identify prior and
related patents and to determine the legitimacy of the innovation. They usually
require the help of a specialized law frm, take at least a year to obtain, and will
probably cost about $30,000 in legal and processing fees. Trade secret protec-
tion requires no registration with the government, but requires you to create and
enforce a security policy to ensure that your information is adequately protected.
In a digital age, copyright law is the most challenging to apply and to enforce.
The frst question is identifying ownership. Who owns a particular item? If you
write a book on your own time with your own resources, then generally you own
the rights. If you write a computer program for your employer as part of your job,
the employer owns the copyright. Interestingly, if you are an outside contractor
and create a program for a company, it is more likely that you own the copyright,
unless you agree to transfer the rights.
There is an interesting exception to copyright law: mere collections of data can-
not be copyrighted. Consider the example of Feist Publications v. Rural Tele-
phone Service [499 U.S. 340 (1991)]. Feist wanted to publish a telephone direc-
tory, but Rural would not provide the data. So Feist copied much of the data from
Rurals printed directory. The U.S. Supreme Court eventually ruled that Feists ac-
tion was not a copyright infringement because the directory contained only data,
which is not suffciently original to obtain a copyright. Now consider the case of
ProCD, Inc. v. Zeidenberg [86 F3d 1447 (7th Cir. 1996)]. ProCD collected and
published a CD-based list of phone numbers and addresses, which they generally
obtained from printed phone directories. Zeidenberg purchased a copy of the CDs
and transferred them to his Web site. He then charged people to access the site.
ProCD sued for violating the copyright laws. Based on the Feist case, Zeidenberg
was found innocent of copyright infringement. However, he was guilty of violat-
ing the shrink-wrap license agreement that came with the CDs. Note that the data
collection argument probably applies to most data collected by federal and state
agencies.
Copyright protection gives you the ability to stop others from profting from
your work. There are a few minor exceptionssuch as parody, excerpting short
quotations, and educational fair use, which allows educational institutions very
limited provisions to make a limited number of copies for teaching purposes. A
more interesting, unanticipated exception involves money. Consider the 1994 case
of U.S. v. LaMacchia, who was a student running a bulletin board system on
university computers. He routinely placed commercial software on the site and
allowed people to download (steal) the software for their own use. The catch is
that he did not charge access to the system and made no money from the process.
Without this proft motive, the court ruled that LaMacchia could not be convicted
on charges of criminal violation of the copyright laws. Of course, the commercial
software vendors could sue him on civil grounds, but unless he was an unusually
wealthy student, there would be little gain. On the other hand, the university could
throw him out for violating university policy. Congress has modifed the copyright
provisions to cover this situation, so now anyone who violates copyright laws can
be criminally charged, fned, and potentially jailed.
Copying becomes a more serious problem every day. As more works are cre-
ated and distributed in digital form, it becomes more diffcult to protect them.
Even though you might have a legal right to prevent copying, it becomes increas-
ingly diffcult to prevent the distribution of your work, particularly if individual
ethics are weak. For example, say that you write a story and sell it through your
Web site. Once the frst few people have read the story, they could copy it and e-
mail it to their friends. What are you going to do? Arrest and sue your customers
who frst read the story? On the other hand, if a publisher took your story, printed
it, and sold it, you clearly have the legal authority and monetary incentive to seek
compensation. Consider a similar example. You build a Web site and create some
interesting graphics and sound effects. Over time, other people routinely down-
load your objects and use them on their own sites. Have they violated copyright
laws? Can you stop them? Can you even fnd them? Would it be economically
worthwhile to pursue them?
It is unlikely that individual motivations and ethics will improve. That is, de-
spite the laws, many people will still copy anything they can (software, art, text,
photos, video clips, and so on). Whatever technology might be applied, it is un-
likely to be economically feasible to pursue them. Yet without incentive, why
should you create and distribute new works? One possible outcome is that large,
expensive content will disappear. Why should you write and distribute an entire
book in one piece, when most people would steal it instead of paying $20 a copy?
Instead, you could sell the book a section at a time, for a few cents per section. By
releasing the sections over time, people would have to pay to receive the most re-
cent (and organized) sections. Yes, some people might wait and have a friend pay
for the section and e-mail it, but it is a question of economics. If the price is low
enough, more people will opt to get the data earlier and directly from the source.
The federal white paper (Intellectual Property and the National Information
Infrastructure) contains an extended discussion of copyright issues and possible
federal solutions. It is available online from the Information Infrastructure Task
Force (IITF) Web site. You should also read Pamela Samuelsons criticism of the
white paper proposal, which points out that the discussion strongly favors copy-
right holders as opposed to the public, particularly since the primary author (Bruce
Lehman) was a lobbyist for the copyright industry.
Privacy
Privacy is an intriguing concept. Humans are a social group: we can accomplish
far more by living in communities and sharing our talents. Yet individuals have a
desire to keep some things private. More to the point, we have a desire to control
what information we wish to share. For example, you might not want everyone to
know exactly how old you are or how many times you were sick last year, but it
is okay if your mother knows these things, and possibly essential that your doctor
knows them.
Society has a vested interest in knowing some things about you and your life.
For example, communities need to know how much you paid for your car and
your house so they can fairly assess taxes. Society needs to track criminal behav-
ior to help identify antisocial people who might harm others. Medical researchers
need to track diseases to identify trends, causes, and potential solutions.
Businesses have an incentive to obtain considerable amounts of data on groups
and individuals. And individuals have an incentive to provide some information
to businesses. Whenever you make a purchase, you need information, and busi-
nesses are generally happy to provide you that information. The problem is how
do you fnd the business or company that best matches your needs? Conversely,
how can a company identify its potential customers? With no information, com-
panies might resort to mass e-mail (spam) that clogs networks and irritates people
who have no use for the services advertised.
The catch is that we do need to share information about ourselves, with gov-
ernment agencies, with researchers in various disciplines, and with businesses.
Yet there is no reason that everyone in the world should be able to obtain every
detail of our lives. The diffculty lies in determining where to draw this line. It is
Freedom of Information Act
Family Educational Rights and Privacy Act
Fair Credit Reporting Act
Privacy Act of 1974
Privacy Protection Act of 1980
Electronic Communications Privacy Act of 1986
Video Privacy Act of 1988
Drivers Privacy Protection Act of 1994
Communications Assistance for Law Enforcement Act of 1994
Health Insurance Portability and Accountability Act of 1996
Childrens Online Privacy Protection Act of 1998
Identity Theft and Assumption Deterrence Act of 1998
Graham-Leach-Bliley Act of 1999
U.S. Patriot Act (antiterrorism) of 2001
CAN-SPAM Act of 2003
Fair and Accurate Credit Transactions Act of 2003 (FACTA)
Figure 14.18
Privacy laws. Only a few specialized laws exist to protect privacy in the United
States. Some, like the Patriot Act, have actually removed earlier privacy protections.
further complicated by the fact that every person (and social group) has different
preferences.
First, it is important to realize that there is no constitutionally defned right to
privacy, especially with respect to data. A couple of Supreme Court rules have
interpreted a right to privacy within the constitutional freedoms. But these rights
apply only to governmental intrusion. A few laws have been enacted in the United
States to provide minimal restrictions on the use and sharing of personal data. Fig-
ure 14.18 lists the most notable laws. Most are related to fnancial data and credit
reporting agencies.
The Freedom of Information Act generally provides people with the ability to
obtain information held by governmental agencies. There are limits for national
security and on the release of data relating to individual personal data. For ex-
ample, you cannot ask the IRS for personal information about your neighbor.
The most important feature of the Family Educational Rights and Privacy Act
is that it limits the release of educational data. Institutions can release basic in-
formation such as the names of students (commonly sold to businesses), but they
cannot release grades without the students express written permission.
The primary purpose of the Electronic Communications Privacy Act was to ex-
tend traditional wiretap provisions to electronic communication, which includes
cellular phone and e-mail transmissions. Essentially, the law makes it illegal for
individuals to intercept these conversations, and requires law enforcement agen-
cies to obtain court permission to intercept and record the conversations. On the
other hand, it is specifcally legal for an individual to record his or her transmis-
sions (although a few states limit this right). Consequently, employers generally
have the legal right (since they own the equipment) to monitor most communica-
tions by employees. Note that there may be some exceptions and an honest em-
ployer will always notify employees frst.
The Fair Credit Reporting Act primarily gives consumers the right to inspect
credit recordsand it gives them the right to correct errors. The Drivers Privacy
Act limits the use and release of state motor vehicle data. Its primary purpose was
to prevent release of specifc data to individual requesters. However, it has gener-
ous exceptions for insurance companies, research, and business use. The Video
Privacy Act was created to limit the release of rental records from video stores and
libraries.
The Privacy Protection Act of 1980 is primarily concerned with law enforce-
ment investigations. It provides some defnitions for when police searches are le-
gitimate and when they are an invasion of privacy. The act predates the advances
in information technology, so it is generally silent on the issue of privacy in terms
of electronic data.
On the other hand, the Privacy Act of 1974 deals more directly with the collec-
tion and dissemination of information by the federal government. It specifcally
limits the collection of data by an agency to information that is relevant to its
work. It provides citizens with the ability to examine and contest the data. The act
initially limited agencies from sharing and matching data with other agencies, but
most of these restraints have been removed by subsequent amendments. For ex-
ample, the postal service is generally not permitted to disclose data on individual
addresses. However, it does release data to a few large commercial service bu-
reaus. Companies can submit address lists to these bureaus for correction of their
mailing lists.
The Communications Assistance for Law Enforcement Act (CALEA) requires
telecommunications frms to pay for wiretap facilities for police to listen to con-
versations. In 2004, the FTC began discussions to expand the coverage to nontra-
ditional communication providers, such as ISPs.
The Health Insurance Portability and Accountability Act (HIPAA) is used to
limit sharing of medical information. Many health care organizations now ask you
to sign forms that give them the authorization to share the data. Since it is unlikely
that consumers have the ability to refuse to sign or modify these preprinted agree-
ments, the overall effectiveness is minimal.
The Childrens Online Privacy Protection Act is much stronger, and if you run a
Web site, you need to be aware of its provisions. As long as you make it clear that
you are collecting data only from adults, the law does not apply. However, if you
do collect data from children, you must be careful to minimize the personal data
collected and generally have to obtain verifable parental consent.
The 1998 Identity Theft and Assumption Deterrence Act prohibits knowingly
transfer[ring] or use[ing], without lawful authority, a means of identifcation of
another person with the intent to commit, or to aid or abet, any unlawful activity
that constitutes a violation of Federal law, or that constitutes a felony under any
applicable State or local law. That is, it specifcally makes identity-theft illegal,
and targets the person who steals the identity, even if that person does not use the
stolen identity.
The Graham-Leach-Bliley Act of 1999 primarily deregulated some fnancial
services. In exchange, it imposed some trivial privacy clauses. In particular, it re-
quires fnancial institutions to notify customers that they have the right to opt out
of (1) selling their names to other companies and (2) marketing requests from the
institution. Institutions reportedly spent hundreds of millions of dollars sending
notices to customers, but many feel they deliberately made the process obscure
and few consumers replied to the mass mailings. Consequently, businesses are
basically free to continue using consumer data in any manner they want.
The U.S. Patriot Act was not directly concerned with privacy. However, it ef-
fectively repeals almost all governmental restraints. Someday it might result in
some interesting lawsuits. One objective was to remove restraints to federal shar-
ing of data.
The CAN-SPAM Act is a halfhearted attempt to reduce the fraud involved with
most unsolicited commercial e-mail (spam). Although it is a relatively weak law,
you need to make sure that your messages conform to its provisions. For example,
you need to include a physical address in each unsolicited message. The message
must state that it is an advertisement. You cannot include false return addresses or
message headers. And you must provide a working opt-out system. It also imposes
limits on how you collect e-mail addresses. Individuals cannot sue violators, but
ISPs do have the authority to sue for substantial sums of money. The most power-
ful aspect of the law is that it applies to the sender of the message and to the frm
being advertised. To be safe, if you send unsolicited e-mail messages, be sure to
send them yourself from a verifed list. Do not purchase lists, and do not use third
parties to send messages on your behalf.
The Fair and Accurate Credit Transactions Act of 2003 (FACTA) primarily
added defnitions and a few features to the Fair Credit Reporting Act. A couple of
useful points from the FACTA of 2003 is that credit agencies must provide free
copies of credit reports to customers once a year (http://www.annualcreditreport.
com), and that merchant receipts cannot contain more than the last fve digits of a
customers credit card number.
The bottom line is that this piecemeal approach to privacy means that it is dif-
fcult for consumers to determine their rights and for businesses to identify their
responsibilities. Consequently, except for the few specifc limitations (e.g., credit
and educational records), most businesses are free to collect and share informa-
tion. On the other hand, you can improve relationships with customers by always
asking them for permission to use and share personal data. Keep in mind that
states have their own laws that apply to transactions and companies within that
state. For example, California prohibits merchants from requiring personal infor-
mation from customers when they make a purchase. An interesting decision in
2011 pointed out that even asking for the customers ZIP codea common mar-
keting practicewas in violation of the state law.
Information Era Crimes
As commerce moves to digital form, existing crime laws need to be extended and
new ones need to be created. The biggest concerns are fraud, theft, and destruction
of property. To understand the complications, consider what happens if someone
steals your car. Why is that bad? Largely because you no longer have the use of
the car. Now, what if someone steals your companys marketing plan? Why is that
bad? You still have the use of the plan. Similarly, what if someone deleted your
computerized customer database? Assuming that you are smart enough to keep a
backup, what have you lost? The point of these questions is to show you that our
traditional views on crime might not apply to crime related to information. Fur-
thermore, computers create the prospect of new types of crime. For instance, what
happens if someone writes a program that prevents you from obtaining access to
your fnancial records? The alleged criminal did not steal or destroy anything, so
what crime has been committed?
The Computer Fraud and Abuse Act of 1986 provides answers to many of the
questions regarding crime in the digital realm. In particular, it outlaws (1) access
to computers without authorization; (2) damage to computers, networks, data, and
so on; (3) actions that lead to denial of service; and (4) interference with medical
care. Interestingly, the act charged the U.S. Secret Service with enforcement.
Enforcement of the act has been challenging. It has been diffcult to fnd quali-
fed law enforcement personnel, including prosecutors. Besides, many businesses
are reluctant to prosecute cases because they do not want competitors or share-
holders to learn the details. On the other hand, sometimes companies and the Se-
cret Service are too enthusiastic in their pursuit of alleged criminals. For example,
one of the frst cases supported by the Electronic Frontier Foundation (EFF) in-
volved a (bulletin board system) BBS that supplied a document obtained from the
telephone company that detailed information about the 911 system. The phone
company complained that the document was stolen and that hackers might use it
to break into its system. The Secret Service confscated the BBS computer equip-
ment and arrested the teenage owner. In court, with the help of the EFF, it was
shown that the document could be purchased from the phone company for a few
dollars.
Examining crime historically, the same problems exist in preventing more tra-
ditional crime and enforcing the laws. In the United States, it was the introduc-
tion of the FBI and their professional investigative techniques that improved the
detection and enforcement of various crimes. In the digital arena, until society
gains more experience and improved training of police, attorneys, and judges, it
will face the same problems of weak laws, diffculty in prosecution, and variable
enforcement.
The Digital Millennium Copyright Act (DMCA) of 1998 changed some copy-
right provisions to synchronize the U.S. laws with the European laws. It also in-
cluded a unique controversial provision that makes it a federal crime to create or
to distribute devices that circumvent copy protection schemes. Part of its original
purpose was to prevent people from advertising and selling black boxes to decode
scrambled satellite TV signals. Many people believe that these provisions are too
strict and that they infringe on the free speech rights in the Constitution. For in-
stance, some researchers have been threatened with prosecution under the DMCA
if they attempted to publish their work. The problem with copyright laws is that
they can provide only limited legal protection. To enforce these laws, a copyright
holder generally has to prosecute violators. But as the record industry was aware
in the Napster case, it is virtually impossible to fnd everyone who copies a song
even more impossible to take them all to court. So, property owners are searching
for ways to prevent casual theft. The problem is that in theory, it is impossible to
completely prevent the copying of a digital work. So, portions of the DMCA are
required to make it diffcult for people to sell circumvention technology. By mak-
ing it more diffcult for people to copy a work, the laws essentially raise the cost
of stealing. But there are fne lines between protecting copyright holders, protect-
ing consumers rights to use a work, and protecting everyones right to study new
ideas. It will take time and discussion to draw these lines.
Driven by the September 11 attack on the World Trade Center in New York, the
U.S.A. Patriot Act (antiterrorism bill) of 2001 provides considerable new pow-
ers to federal, state, and local police agencies. Some of these provisions reduce
privacy by making it easier for police agencies to monitor conversations, inter-
cept e-mail and Internet messages, and detain people without cause. Law enforce-
ment agencies are asking for even more fexibility to investigate people. These
provisions do have some justifable uses, and there are times when enforcement
agencies have to jump through too many hoops to perform their jobs effectively.
However, as J. Edgar Hoover proved, the challenge lies in preventing abuse of the
laws, particularly preventing people from using them as political tools.
Cloud Computing
What risks are created through using cloud computing? Cloud
computing raises some interesting questionsparticularly from a legal perspec-
tive. A private cloudowned and operated by your own company in one or two
locations is less of an issue. As long as you control the location and operation
of the servers, a private cloud is no different from any other set of servers. The
problems begin to arise with public cloudswhen you rent processors and data
storage space on servers run by other companies. Three basic problems arise with
storing data and processing public clouds: (1) Data is transferred across multiple
national boundaries and subject to laws in multiple nations; (2) Shared servers
might be at risk if police confscate or search the computers because of a second
company, or if attackers initiate a denial of service attack against another com-
pany on the server; (3) With multiple levels of contractors and subcontractors, you
must ensure everyone respects the security and privacy of your data.
Data in Multiple Countries
Figure 14.19 shows the frst issue. One of the strengths of cloud computing is that
multiple servers in different locations are used to hold the databoth as backup
and to provide faster access across the world. These are useful technical attributes.
But, from societys perspective, the data is crossing multiple boundaries and is po-
tentially subject to laws in multiple nations. For instance, what if some of the cus-
tomer data is transferred to servers in Europe? Does that data become subject to
European privacy lawswhich are much stricter than those in the United States?
But, do you even know where the data is currently stored? What if you are storing
data that is politically sensitive? Can one nation confscate or block your data?
These questions are not hypothetical. In 2010, the Web site WikiLeaks (not a
U.S. company) began releasing data that it obtained from sensitive U.S. diplo-
matic e-mails. The U.S. government was unhappy about the release of the data
which they claimed was stolen (and then given to WikiLeaks). At one point, the
WikiLeaks servers were subjected to a distributed denial of service attack and they
struggled to remain online. The company paid Amazon to host their content on the
Amazon cloud serverswith the hope that the huge bandwidth and distributed
service would mitigate the effect of the attacks. Technically, the process worked.
But, high-level politicians in the U.S. government contacted Amazon (a U.S. com-
pany) and urged them to stop hosting WikiLeaks content. Amazon complied and
removed the content. The U.S. government also pressured the U.S. company that
was providing DNS services to WikiLeaks and told them to drop the listingso
no one could go directly to the Web site. The WikiLeaks site is currently available
through a Swiss registration: www.wikileaks.ch. (This information is available in
many news reports from late 2010 and early 2011.) Regardless of which side you
might take in the WikiLeaks discussion, the business point is that national gov-
Figure 14.19
National location issues in the cloud. If your data is transferred across multiple
borders and stored in multiple countries. What national rules need to be followed?
What countries might seek to control or limit your use of the data?
ernments can exert huge control over cloud service providersdepending on the
location. Political, security, and tax issues could vary depending on the location
of your data. So, you might have to write contracts to control exactly where the
cloud provider will store your data. However, in a related situation, Research in
Motion, the Canadian company that sells the Blackberry cell phone and service
encountered problems in 2010 and 2011. Several nations insisted that the com-
pany run e-mail servers within their countriesso national investigators could
confscate or tap the servers if necessary to control communications by citizens
within their countries.
If you want some idea of government involvement or interference in the In-
ternet, check out Googles statistics (http://www.google.com/transparencyreport/
governmentrequests/). Interestingly, the United States and Brazil are at the top
of the list for countries asking Google for data and for requests to remove links.
In 2011, Freedom House also examined national limits placed on Web traffc and
digital media (Sanja Kelly and Sarah Cook, Freedom on the Net 2011).
Treats to Shared Servers
As shown in Figure 14.20, a second potential problem with cloud computing is
that the servers and Internet connections are ultimately shared with other com-
panies. The cloud provider makes money by building huge server farms and then
leasing out capacity to many customers. In many cases, the companies lease vir-
tual machines, and any physical server can run many virtual machines.
Figure 14.20
Complications with shared servers. When police investigate a second company, they
might collect all of the data on a server or even confscate the physical server. These
actions might include your innocent data and disrupt your services. Similarly, if
attackers food a companys site, it might knock down your site as well.
Company 1 (you) Company 2
Police
The problem with sharing the physical machine is that it is not always easy
to separate the companies. At least a few cases have already arisen where police
investigators have confscated entire physical servers to gather evidence against a
company using the server. The problem is that taking down the entire server af-
fects other companies. Even if the police only need to collect data, they are likely
to collect all of the dataincluding yours. Similarly, an attacker or group of peo-
ple might become upset at a particular company and launch a denial of service
attack against that company. If that company uses the same cloud service as you
do, it is likely that your server and Internet connection will be disrupted as well.
Stronger cloud providers like Amazon might have enough bandwidth and server
capacity to handle the increased load, but it is still a potential risk.
Subcontractors
Figure 14.21 shows a more subtle problem with cloud computing services. When
you sign a contract with a cloud service provider, you deal with the top level, but
many sublevels can exist below that one. For example, it is possible that each
server location is run by a separate contractor. Employees from several contrac-
tors might have access to the servers, and ultimately the data, at any location.
For instance, network and security specialists, database administrators, and server
operators might all be hired by different companies. Throw in electricians, clean-
ing companies, and security guards and multiply by the number of countries
where the servers are located. Potentially dozens of companies might have access
Your company
Main contractor
Subcontractor
Subcontractor
Hired workers
Figure 14.21
Controlling subcontractors. You can sign a detailed contract with a cloud provider,
but that company might subcontract services to several companies in different
countries. Workers at many levels might have access to servers that host your data.
to the physical computers and data in different locations. Hopefully, all of them
are trained and honest and know how to protect your data. But, what if you are
required to guarantee that the data is accurate and protected? For example, you
might want to store employee or healthcare data. Realistically, can you meet that
guarantee when you do not know all of the levels of people who might have ac-
cess? Encryption might help, but only if the encryption keys and all decryption are
handled by local servers.
For most cases, it is possible to handle these three types of problems, so they
are not intended to suggest cloud computing should be avoided. Cloud computing
can provide some signifcant benefts. But, you need to be aware of these potential
problems before you commit to a cloud service provider so you can ask the right
questions.
Summary
Technological change and increasingly aggressive use of information systems by
businesses have several consequences. Technology affects individuals, their jobs,
educational systems, governments, and society as a whole. Businesses have to be
careful to protect the privacy of consumers and workers. Security provisions, dis-
closure policies, and audits are used to ensure that data is used only for authorized
purposes. To ensure accuracy, it is crucial to allow customers (and workers) to
examine relevant data and make changes.
Technology is generally believed to increase the total number of jobs avail-
able. However, the workers displaced by the introduction of technology are rarely
qualifed for the new jobs. Businesses and governments need to provide retraining
and relocation to help those workers who lose their jobs. Sometimes technology
allows physically disabled people to work in jobs they might not otherwise be
able to perform.
Improved communication networks, huge databases, and multimedia tools pro-
vide possibilities for education and training in the public and business sectors.
However, because of high development costs, technology tends to be used for
specialized training.
Governments have long been involved in data collection, and technology en-
ables them to work more effciently. Of course, many political observers would
argue that perhaps governments should not be too effcient. For example, it would
be diffcult for businesses to operate in an environment where the laws were
changed every day. Technology also has the potential to improve communication
between citizens and their representatives.
Technology and society produce other interactions. One feature is that lower
prices, improved capabilities, and ease of use have made improved communica-
tion available to virtually any size groupproviding a wider audience for small
extremist groups. The new technologies also offer the ability to alter pictures,
sound, and video, making it diffcult to determine the difference between fact and
fction. Another important social issue is providing access to technology for every-
one. It would be easy to create a world or nation consisting of haves and have-
nots in terms of access to information. Those with information would be able to
grow and earn more money, while those lacking the data continually lose ground.
Increasing dependence on technology brings with it new threats to the security
of the frm. Managers need to recognize and evaluate these threats and understand
some of the techniques used to minimize them. The most common threats come
from inside the company, in terms of workers, consultants, and business partner-
ships. These threats are diffcult to control, because frms have to trust these in-
dividuals to do their jobs. Training, oversight, audits, and separation of duties are
common means to minimize threats. Depending on the communication systems
used, there are threats from outsiders and viruses that can access computers with
modems, over networks, or by intercepting communications. Dial-back modems,
access controls, encryption, and antivirus software are common techniques to
combat these threats.
Working in todays business environment means more than just doing your
job. Each individual and frm has ethical obligations to consumers, workers, other
companies, and society. In addition to obeying the laws, it is important for work-
ers and companies to remember that the data in information systems refers to real
people. The lives of people can be adversely affected by inaccurate data, poorly
designed information systems, or abuse of the information.
Key Words
A Managers View
As a manager, you need to understand how businesses, technology, and so-
ciety interact. Dealing with changes in privacy and security threats will be-
come increasingly important to managing a company. Evaluating changes in
society will also give you an advantage in the marketplace; it is important
to know your customers. As a citizen, you need to be aware of the negative
and positive effects of technology. In particular, changes in technology often
lead to changes in political power and control. As a manager and a citizen,
you are obligated to make ethical decisions and to understand the conse-
quences of your actions.
computer ethics
cookies
copyright
digital divide
digital rights management (DRM)
high-bandwidth digital con-
tent protection (HDCP)
information warfare (IW)
intellectual property (IP)
nondisclosure agreement (NDA)
patent
privacy
software piracy
Web Site References
Technology and Society
ACM/society www.acm.org/usacm
Center for democracy and technology www.cdt.org
Center for information technology and
society
www.cits.ucsb.edu
Computer professionals for Social
Responsibility
www.cpsr.org
Internet Society www.isoc.org
Privacy
Electronic Frontier Foundation www.ef.com
Electronic Privacy Information Center www.epic.org
FTC advisory committee www.ftc.gov/acoas
FTC privacy and security business.ftc.gov/privacy-and-security
Platform for privacy preferences www.w3.org/P3P
Privacy, ACM www.acm.org/usacm/privacy
Privacy International www.privacyinternational.org
Privacy Rights www.privacyrights.org
Review Questions
1. Do employees need to worry about the data collected by their employers?
2. If everyone is identifed by some biometric measure, will that cause more
dehumanization? Will it reduce individual privacy?
3. Do you think increasing use of computers causes a loss of jobs? What about
in the past or in the future?
4. What are the personal benefts to telecommuting? Why would people choose
to return to commuting jobs after trying telecommuting?
5. Do computers and digital content change the balance of power relationship
between consumers and businesses? Should consumers have a right to make
personal (backup) copies of digital works?
6. How does information technology add legitimacy to fringe groups?
7. Do you think state, local, and federal governments are making effcient use of
computers? Will citizens ever be able to vote online?
8. In what ways have computers affected society and organizations? Will these
patterns continue? Are there other important patterns that might arise?
9. Should governments be granted more powers to monitor and investigate
people and transactions on the Internet?
10. What are the ethical responsibilities of users in terms of information
systems?
11. Do we need additional privacy laws in the United States? What provisions
would you add?
12. As a business manager running a Web-based company, which laws and rules
do you need to pay careful attention to?
13. What is information warfare and what controls or oversight should be placed
on it?
14. List the primary U.S. laws related to computer crime and describe each in
one sentence.
15. If data is stored in an international Web cloud, what potential problems exist
for companies? Do we need new world laws to cover these situations?
16. If you write a blog and publish it on a Web site, who owns it? Can other users
copy it? What if you are paid by your company to write the blog?
17. Should governments have teams to create computer viruses that can be sent
to other countries?
Exercises
1. Research the tools (hardware and software) available for a new employee of
yours who is blind. List the sources, capabilities, and costs.
2. Should people be allowed to use the Internet anonymously? Should ISPs be
required to pay for hardware and software that can track individual usage in
case of a lawsuit or criminal charge? Is it possible to prevent anonymous use
of the Internet?
3. Do you think governmental agencies should share data about citizens? For
example, should the FBI be able to access IRS records to locate suspected
criminals? Should the FBI be allowed to access fles from state and local
governments? For instance, should all arrest records be automatically
relayed to a central database? Should medical records be accessible to law
enforcement agencies? Say that it is technically possible for the FBI to
build a national database that contains DNA records for all citizens. If all
medical records (from accidents, blood tests, and medical treatment) were
computerized and automatically forwarded to the FBI, the agents could easily
locate virtually any criminal.
4. Some remaining Federal laws limit the ability to create huge, integrated
collections of personal data. Some agencies, including the FBI, have turned
to buying this data from private companies (e.g., ChoicePoint). Should
government agencies be allowed to circumvent laws by purchasing data on
individuals from private agencies?
5. Research the issues involved in electronic voting. What problems need to be
overcome? What technologies could be useful? Does an electronic voting
system have to be perfect, or simply better than the existing manual system?
6. Should vendors be allowed to charge different prices for online products,
or should everyone pay the same price? Answer the question both from the
perspective of the consumer and as a vendor or artist.
7. Should consumers be able to sue software companies for security failures or
other problems with the software? What limits these lawsuits now?
8. What aspects of education would you prefer to have online or automated?
What elements would you prefer to keep in person?
9. Find at least fve news sites on the Web. Evaluate them in terms of (1) style/
presentation, (2) accuracy, (3) believability, and (4) balanced news.
10. Identify which privacy and computer crime laws might apply to the following
situations:
a. Someone intentionally downloads a program from a Web site which is then
used to run a denial of service attack on government computers.
b. A hospital employee sells a celebritys health report to a newspaper.
c. You download and distribute copies to your friends of a software program
that captures digital video and audio streams and converts them to
unprotected fles.
d. A Chinese government agent intercepts your PayPal data and uses it to buy
electronics items.
e. Homeland Security asks your ISP to provide copies of all of your Skype
conversations.
f. You send marketing e-mail to customers without including your business
address.
11. Should Internet gambling be legal in the United States?
12. Should all consumers be allowed to pay the same price for identical items
purchased on the Web?
Technology Toolbox
13. Find at least two translation sites and test them with sample text. If you read
the second language, comment on the results. Translate the text back to the
original language and comment on the quality.
14. Find at least two foreign exchange sites and convert $100 (USD) into a
different currency. How much does your credit card company charge for
currency exchanges?
15. Why would you not want to use Private or Incognito browsing all the
time?
16. Research the current status of the do not track option to see if there is any
progress towards it becoming a standard.
17. Even with cookies blocked and Private browsing, explain how your
Internet activities can still be tracked.
Teamwork
18. Assume that you are selling a new release for popular music. Create a silent
auction and have everyone write down the price they are willing to pay for
the music. Add up the numbers to get the total revenue you would obtain.
Now, look up an average price for a similar item. Assume that each person
who was willing to pay at least that amount would actually buy the item, and
the others would not. Count the number of items sold and multiply by the
fxed average price to get revenue. Compare the two values for total revenue.
19. Have each team member select a developing nation. Research the
information technology available in that country. How do people get access
to the Internet? What percentage of the people have used the Internet?
Combine the results and create a list of options that might be used by other
nations to improve Internet access.
20. Split the team into two groups to participate in a debate. The proposition is
that programmers and developers should be licensed. One group should fnd
evidence and arguments to support the proposition, the other to defeat it. If
possible, conduct an actual debate. Otherwise, outline your arguments and
compare them in writing.
21. Examine the arguments against electronic voting. Divide the arguments
among the team members and have each person research existing
technologies and proposals. Identify the methods used to avoid or minimize
the stated problem. Combine the results and write a proposal defending the
use of electronic voting.
22. Interview or survey at least 30 people, not students in the class and be sure
to include a range of demographics including older people. Ask them what
they would think about a national ID number and how it might affect them.
Ask them what they think about the potential benefts and how it would be
different from the existing system.
23. Split the team into three groups and have each group choose one nation. Find
at least one computer-crime or privacy law for that country. Note: It helps
if at least one person in the group can read Web documents in the countrys
language. Combine the results and summarize which crimes are most
commonly outlawed.
24. Research a couple of current patent violation cases involving IT and argue
why patent laws and procedures should be changed.
25. What privacy problems might exist at Rolling Thunder? What rules or
procedures should be enacted to avoid problems?
26. Your boss says that with the decline in sales, it would be wise to cut costs and
suggests that you could buy only a single copy of some of the offce software
and install it on multiple machines. What do you do?
27. The management at Rolling Thunder is thinking about trying to get a patent
for an online process of confguring and ordering a custom bicycle. Search
the patent records to see if anyone already has a similar patent, and estimate
the probability of obtaining such a patent.
28. Rolling Thunder Bicycles wants to begin sales to Europe. It is currently
run completely in the United States. The company will add a Web site and
establish separate contacts with bike stores in Europe. What effect will these
actions have on the information system?
Additional Reading
Arkin, William and Robert Windrem, The U.S.-China information War,
MSNBC, December 11, 2001. http://www.msnbc.com/news/607031.asp.
[Description of some aspects of the U.S. information warfare preparations.]
Collett, Stacy, 5 Legal Gotchas in the Cloud, Computerworld (whitepaper),
April 18, 2011. [Good summary of legal issues in cloud computing in
terms of where data is located and multiple contractors.] http://www.
computerworld.com/s/article/355454/Legal_Risks_in_the_Cloud
Government Computer News, January 6, 1992.
Grosso, Andrew, The individual in the new age, Communications of the ACM,
July 2001 44(7), 17-20. [A readable legal perspective on individual versus
society.]
Jones, Douglas W. Problems with voting systems and the applicable standards,
May 22, 2001. http://www.house.gov/science/full/may22/jones.htm [Issues
on electronic voting systems.]
Kelly, Sanja and Sarah Cook, Freedom on the Net 2011: A Global Assessment
of Internet and Digital Media, Freedom House, 2011. http://www.
freedomhouse.org/uploads/fotn/2011/FOTN2011.pdf [A summary of national
restrictions and censorship on Internet traffc.]
Machalaba, Daniel. U.S. Ports Are Losing the Battle To Keep Up With Overseas
Trade, The Wall Street Journal, July 9, 2001. [Effect of automation on jobs at
the loading docks.]
Mercuri, Rebecca T. Scoping Identity Theft, Communications of the ACM,
May 2006, 49(5), p. 17. [Data and laws on identity theft.]
Stoll, Clifford, The Cuckoos Egg: Tracking a Spy Through a Maze of Computer
Espionage, New York: Doubleday, 1989. [Fascinating story of a spy
searching US networks.]
Cases: Health Care
Te Industry
Health care makes up a substantial portion of expenditures in the United States.
In 2004, health care expenditures were projected to be $1.8 trillion, a whopping
15.3 percent of gross domestic product. As the population ages, particularly the
boomers, the U.S. government estimates that these values will rise to $3.4 tril-
lion or 18.4 percent of GDP by 2013 (Brewin 2004). It is also a highly complex
industry because the ultimate consumer, the patient, rarely bears the direct costs
of the services. In 2000 (the most recent federal study), private medical insurance
covered about 41 percent of total health care expenses, Medicare and Medicaid
combined paid about 31 percent, and individuals paid 19 percent of the costs out
of their own pockets (AHRQ 2003). The Census Bureau maintains a history of
health expenditures and annually updates the projections. The 2008 actual data
were $2.3 trillion (16.2 percent of GDP), with a projection for 2015 of $3.4 tril-
lion. The tables also separate private and public expenditures and breakdowns for
hospitalization, physician, drug, and nursing home costs (http://www.census.gov/
compendia/statab/cats/health_nutrition/health_expenditures.html, Table 130). In
2008, about 53 percent of the costs were paid by private funds, 35 percent federal,
and12 percent state. Almost 12 percent of the costs were paid out of pocket by
individuals, versus private insurance.
Federal Involvement in Care
Some people have heard about mistakes made in operating roomswhere the sur-
geon operated on the wrong knee or arm. Some patients have taken to writing on
their limbs before surgeryjust to make sure everyone knows which leg or arm
to work on. Not as many people are aware of the problems that arise with drugs.
0
1
2
3
4
5
6
7
8
9
10
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
2
0
0
0
2
0
0
1
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
2
0
0
7
2
0
0
8
2
0
0
9
2
0
1
0
B
i
l
l
i
o
n

$
Annual Revenue
Sutter Health
Partners Healthcare
CareGroup
Physicians sometimes prescribe the wrong drug. Nurses occasionally deliver the
wrong drug or incorrect dose to a patient. The federal government has stepped
in to reduce this problem. The Food and Drug Administration (FDA) in 2003 is-
sued a ruling to take effect in 2004 for new drugs and 2006 for existing drugs. All
individual doses of drugs will be required to have bar codes. Hospitals will then
have to implement bedside bar code readers that match patients and drugs before
giving the drugs. Any errors will be fagged by the system. Hospitals are spending
millions of dollars to add the new systems. In terms of those surgeries, new rul-
ings require hospitals and physicians to adopt a marking system and mark every
operating site on the body while the patient is conscious, to reduce mistakes. Ap-
parently, they could not fgure out a way to bar code the body parts.
In 2004, President George Bush stated that he wanted the entire health care
system to move to electronic records. Within 10 years, hospitals are supposed to
have a system that allows electronic sharing of medical data (Brewin 2004). This
push goes far beyond individual hospitals. It means that the entire health care and
health informatics industry has to agree on standards and has to come up with a
way to identify patients and transfer data securely.
The industry has been trying for several years to devise a health care informa-
tion system that would work for the entire industry. Tommy Thompson, U.S. Sec-
retary of Health and Human Services, had asked the industry to derive a blueprint
for an electronic health-record system. In late 2003, the initial plan was rejected
by the industry because it was overly complex. The original design was cumber-
some, yet still did not address all of the potential issues. Part of the problem is that
it tried to focus on a detailed level and include everything from medical records
to billing to patient history. A bigger problem is that several proprietary systems
already exist, and vendors are concerned that a government-designated system
would be incompatible. At this point, there is not even a framework or structure
for defning the overall approach (Landro, 2003).
The federal government has also taken an interest in improving health care
by reducing errors through bad information. Panni Kanyuk, a senior Datamoni-
tor health-care analyst noted that information technology is an important tool in
improving patient quality of care, and were seeing this resonate in the market.
Most health-care providers surveyed responded that they anticipated increasing IT
spending by at least 10 percent per year (McGee 2004).
Several medical care providers have adopted electronic systems for recording
physician drug orders. The systems have the ability to reduce errors and improve
communication. However, they still require physicians, pharmacists, and patients
to be conscientious and observant. A study at one hospital identify 22 ways that
medication errors were facilitated by a computerized physician order entry sys-
tem. Although handwriting errors are no longer an issue, other problems can arise
with information errors and usability problems. Ross Koppel of the Center for
Clinical Epidemiology and Biostatistics at the University Pennsylvania School of
Medicine observed the largest problem is that the system asks house staff to twist
the software like a pretzel rather than the software corresponding to the way the
work is done (McGee 2005).
Information technology has the potential to manage data and reduce costs and
time, but the medical world is highly fragmented. Many physicians work in their
own offces, and hospitals are often independent or connected into loose networks.
Throw in a few dozen insurance companies, and various state and federal organi-
zations, and even simple communication becomes challenging. A study by a re-
searcher at Sutter Health in California found that even with HER systems, physi-
cian costs are high because of the problems dealing with multiple health plans.
Annual costs for even a small clinical practice were estimated to be $51,221 for
non-clinical personnel time and $34,052 for physician timefor each physician
in the clinic. The total costs of $85,273 represented 10 percent of total revenue for
a practice (Hardy 2009).
Privacy
Privacy has become a more important issue. The Health Insurance Portability and
Accountability Act of 1996 (HIPAA) has fnally been given some teeth. Any or-
ganization that handles medical data has to comply with rules that prevent them
from releasing medical data. The rules cover accidental releases, such as over-
heard conversations, and security breaches.
Actually, President Bush watered down the original medical privacy rules
created by President Bill Clinton. The original rules required mandatory patient
consent to disclose data, even for treatment and payment. The new rules simply
require that patients be notifed of the privacy policies (CNN Online 2002). The
main argument was that getting patient signatures might slow down the treatment
process. Although privacy advocates were upset, the real-world effect was prob-
ably minimal. Most health care providers were simply requiring patients to sign a
form that waives most privacy anyway.
The law has been useful at encouraging health care providers to tighten up the
access to medical records. For example, hospitals cannot release or use patient
data for charity solicitations. The billing system has to be separate from the medi-
cal system, so that a billing clerk could see that a patient was billed for a test but
not see the test results (Tarkan 2003).
In 2010, the U.S. Congress passed a healthcare law that was not exactly popu-
larit was challenged in court on several grounds, and will probably eventually
end up before the U.S. Supreme Court, or be overwritten when Obama leaves
offce. The law is known as the Patient Protection and Affordable Care Act. One
of its main, and controversial, elements is that it requires everyone to buy health
insurance. One of the problems often cited by reform proponents is that uninsured
people cost the system billions of dollars a year because hospitals and doctors are
ethically required to treat sick and injured people even if they have no insurance.
And people with no insurance tend to avoid paying for routine car and wait until
they have to go to the emergency rooms. The goal of the law was to help reduce
the projected federal defcit due to rising Medicare and Medicaid costs (Moore
2011). Although, it is diffcult to see how costs can be reduced without deliberate
increases in supply.
Additional Reading
AHRQ, Statistical Brief #27: National Health Care Expenses in the U.S.
Community Population, 2000. November 2003. Agency for Health care
Research and Quality, Rockville, MD. http://www.meps.ahrq.gov/papers/
st27/stat27.htm
Brewin, Bob, Health Care IT Plans Get a Renewed Push, Computerworld, May
3, 2004.
CNN Online, Change in the Air for Medical Privacy Rules, April 16, 2002.
Hardy, Kyle, Study Shows Physicians Using IT Still Experience High Cost in
Billing and Insurance, Healthcare IT News, May 14, 2009.
Landro, Laura, Plans for Health-Record System Need Work, Industry Players
Say, The Wall Street Journal, September 15, 2003.
McGee, Marianne Kolbasuk, Health-Care Providers Pump More Dollars into
I.T. Information Week, June 7, 2004.
McGee, Marianne Kolbasuk, Computerized Systems Can Cause New Medical
Mistakes, Information Week, March 9, 2005.
Moore, Stephen, ObamaCare Doesnt Add Up, The Wall Street Journal, June
29, 2011.
Tarkan, Laurie, A Privacy Laws Unintended Results, The New York Times,
June 3, 2003.
Case: Sutter Health
Sutter Health is one of the largest health care systems in northern California. In
2002, the not-for-proft organization had almost 40,000 employees and patient
services revenues of almost $4 billion. Its 26-member hospitals with 5,773 beds
recorded almost 240,000 discharges in 2002 and over 2.5 million outpatient vis-
its (www.sutterhealth.com). Like any hospital group, Sutter works to improve the
health of its community. And it struggles to balance costs, medical technologies,
and information technology.
Quality Care
Sutter is working to build the new systems to implement the FDA drug bar code
rules and improve the quality of drug prescriptions. It is deploying 6,000 PCs on
mobile carts. They will be connected by Wi-Fi wireless networks because it would
be too expensive to install wires in every hospital (Cuneo 2004). One advantage
to the new system is that drug manufacturers will be required to put bar codes on
individual doses. They have used them on bulk shipments in the past, but that still
required hospitals to repackage and remark all of the drugs. John Hummel, CIO of
Sutter Health, observed that those codes will save the company about $2 million a
year (Brewin 2004).
Sutter recognized that a key part of the bar code point-of-care (BPOC) tech-
nology is that it had to be combined with a computerized physician order entry
(CPOE) system. When the nurse scans patient and drug codes, the system has to
verify the original drug order. That means the physicians need to enter the drug
orders electronically. Sutter chose to implement a high-end BPOC system that
also checks for patient allergies and asks nurses to double-check drugs if they
have similar names or similar appearance to other drugs. Hummel involved nurses
early in the selection process to ensure the system would be easy to use and that
they would willingly adopt it. After implementation, nurses indicated a 42 percent
improvement in satisfaction and a 64 percent improvement in perception of sys-
tem effciency (Johnson et al. 2004). Furthermore, the real-time dispensing data
feeds directly into the charging system, so patient bills are more accurate.
One of the early challenges that Hummel faced was the need to integrate data
from multiple medical systems. In 1999, he installed an interface engine from
Century Analysis, now a division of Sybase. It serves as a hub and transfers data
across systems. For example, the data from the Siemens Medical Health Services
that handles picture archives can be transferred to or accessed from the Lawson
ERP system. In 2003, he upgraded the engine to an eGate intregration platform
from SeeBeyond Technology Corporation. The new system utilizes XML to trans-
fer data quickly between even more systems. His ultimate goal is to provide a
single view of all patient data. The company is using Identity Hub software, an
artifcial-intelligence tool from Initiate Systems, that uses statistical techniques to
match historical data. It is critical that physicians accurately identify patients to
retrieve the proper records. Hummel observes that if I go to the doctor, he needs
to know the difference between John Hummel and John C. Hummel (McGee
September 2003).
As of 2004, more than 1,000 Sutter physicians store patient data electronically,
and more than 400 of them have totally paperless offces. Even X-rays and pre-
scriptions are stored digitally. The goal is to have data for all 4.5 million Sutter
patients integrated in 2006. The system will make it easier for patients and physi-
cians to use the health servicesregardless of location.
Hummels group is also building a virtual intensive care unit (ICU), where each
hospital will have telecommunication links with real-time videoconferencing to
an ICU physician. The ICU staff will be able to remotely monitor patients and
collaborate with on-site staff. With the eICU system, one ICU physician and nurse
monitor dozens of patients at hospitals that cannot afford full-time intensive care
units. Keeping a physician online full-time provides additional supervision. Nurs-
es do not have to worry about fnding a doctor or surgeon for unnecessary cases.
Dr. Daniel Ikeda, director of the system, notes that when Im in the eICU, Im
a lifeguard. I use the technology to look for troubling trends, before they become
serious complications. A critically ill patient can turn sour in a matter of minutes.
(Kolbasuk 2003).
The electronic ICU has been successful, based on its ability to prevent deaths
and on the increasing usage found by physicians. The system was created for about
$25 million, but by 2006 had already saved 425 sepsis-related patient deaths, and
cases of ventilator-associated pneumonia fell form 37 in 2005 to 8 in 2006 (Hoff-
man 2007)
Sutter Health is also rolling out Internet services to physicians and patients. By
2005, all Sutter physicians should have Web access, where they can download
charts, check lab results, or order prescriptions. On the patient side, customers can
use the Internet to communicate with doctors, view their records, order reflls, or
get additional instructions. The system is being used by 15,000 patients in Palo
Alto (Cuneo 2004).
Sutter Health is using more sophisticated decision support and data analysis
tools to help reduce errors and problems throughout the health care process. For
example, a data warehouse tracks injuries that occur during childbirth, as well as
the frequency of induced labor. The system is being used to reduce the instances
of maternal tissue tears (McGee October 2003).
The California Pacifc Medical Center in San Francisco is an affliate of Sutter
Health. The Center uses an automated Site of Care (SOC) system to enter data
electronically at the bedside. The system saves time and money by reducing tran-
scription costs. It also makes it easy to conduct studies and analyze the data. Re-
ports can be generated quickly in response to physician queries. After a decade of
use, the system contains substantial information and knowledge that can be used
to improve patient care (Parker 2002).
Pushed in part by the federal government to improve medical care by reducing
errors, Sutter Health is adding more support for electronic records keeping. Hum-
mel noted that my IT employees arent making widgets, theyre saving peoples
lives, by giving physicians electronic access to information and medical records
(McGee 2004). Hummel also reported that California providers were likely to
implement complete electronic records in 4-5 years, ahead of the 2014 timetable
set by President Bush. Perhaps fueled by the importance of the technology and
the enthusiasm of Mr. Hummel, Sutter Health was ranked 18 in Computerworlds
2007 100 Best Places to Work in IT survey.
Simply storing health records electronically does not solve all problems. Hospi-
tals and physicians also need interoperable electronic health care records. Parsing
the jargon, a key strength of electronic records is the ability to share the data with
other providers. The U.S. Department of Health and Human Services is working
to defne standards to support interoperability within the federal government. The
agency will provide some software to small and midsize physician practices at
low cost (Havenstein 2005b). Sutter Health is building a system to connect its 26
hospitals and more than 5,000 physicians. CIO Hummel observed that much of
the data is stored in systems purchased from Epic Systems Corp, but that Sutters
interface department builds over 800 interfaces a year to integrate all our ven-
dors (Gilhooly 2005).
In 2010, Sutter Health partnered with iTriage to offer a smartphone app that
provides basic medical information including lists of 300 symptoms, 1,000 dis-
eases, and 350 medical procedures. The system also taps a database of every doc-
tor, hospital, urgent care clinic, and pharmacy in the nationalong with directions
to health care providers in any community (Merrill 2010).
In 2011, Sutter Health began offering its electronic health record (EHR) system
(Sutter Community Connect) to regional physicians in independent practices. Ty-
ing more physicians into their system makes it easier to integrate patient data. Jeff
Burnich, MD and senior vice president of the Sutter Medical Network noted To
truly reinvent care, we see an imperative to connect as many physicians across our
network as possible by extending an HER option to community doctors in inde-
pendent practice (Healthcare IT News 2011). The system already connects more
than 12,000 caregivers and almost 450,000 patients use Sutters online services.
Managing the Technology
John Hummels IT department consists of almost 1,000 employees, with a budget
of 3.9 percent of the hospitals net revenue, spending $105 million for operations
(Cuneo November 2003). He still manages to send birthday and anniversary cards
to each employee every year. But the department is busy, scheduling 750 large-
scale projects for 2004 alone (Cuneo December 2003). The state of California
essentially mandated the huge number of projects. In 1994, the state passed a law
requiring all hospitals to be earthquake proof by 2008. Sutter Health has planned
$5 billion in capital improvements. One of the big tasks for Hummel is building a
new data center to consolidate all of the servers into one secure, safe location. By
2004, Sutter had completely replaced fve hospitals and is planning to rebuild six
more. Building the hospitals from scratch means that information systems can be
built into them from the beginning (Brewin and Thibodeau 2004).
The hospital company operates in more than 100 communities and needs a net-
work to connect all of the facilities. In 2003, the network team installed high-
speed fber-optic links to the new data center. The 50-micron multimode fber ca-
ble can carry 10 gigabits per second. Chris Kennedy, network engineer for Sutter
Health, notes that we didnt want to impede our LAN system and therefore chose
10 Gbps multimode fber. Most of our applications at the desk are pushing 100
Mbps though. With the new fber backbone, we will be ready to easily push one
gigabit from our desks without compromising our network. Category 6 cable was
run to desktops in anticipation of the need to handle gigabit speeds (Oliver 2003).
Sutter Health is also spending money to improve privacy and protect data
partly driven by the HIPAA regulations. In addition to the common security and
privacy controls, the IT department has a team of a half-dozen white-hat hack-
ers to continually test the system to look for problems before real hackers can
fnd them (McGee September 2003).
Like several other hospitals, Sutter has turned to wireless networks to provide
data services to physicians and nurses as they move throughout the hospital. To-
day, most patients are give bar-code bracelets. Nurses use a portable scanner to
identify the patient, retrieve drug medication information and have the computer
match the records to ensure the right patient is given the correct drug at the proper
time. Sutter uses Citrix remote terminal software to improve security. Physicians
and nurses basically download a screenshot that provides data, but is not stored
locally, so the medical provider sees only a small snapshot of the data, not the
entire patient record. The wireless network is also secured against eavesdroppers
(Havenstein 2005a).
Costs
In 2002, Sutter Health raised its prices to Health Net insurance by 25 percent. Af-
ter a public battle fought in newspaper ads, 20,000 Health Net members jumped
to other insurers to avoid the battle. Ultimately, the two companies negotiated 15
percent annual increases.
In 2003, Sutter Health encountered a more serious public relations problem.
Driven by ever-rising prices for health care, several large employers began ex-
amining prices at many of the large health care providers in California. More-
over, federal regulators have been examining overpricing claims against many
hospitals. In 2003, the Service Employees International Union brought signifcant
pricing data to the attention of several organizations, including CalPers, the large
retirement investment agency. They found that fve of the Sutter Health hospitals
had inpatient charges that were up to 53 percent higher than the national average.
Five of its hospitals were also being investigated by the federal Centers for Medi-
care and Medicaid Services for possible overbilling. In addition, nine of the Sutter
Health hospitals had total proft margins exceeding 10 percent in 2001, compared
to a state average of 3.5 percent on not-for-proft hospitals (Benko 2003). In early
2004, CalPers dropped 13 of the Sutter Health hospitals from its insurance cov-
erage. It dropped 25 hospitals from other frms. A spokesman said We wanted
to send a message to these providers that their costs are over the top. CalPers
provides coverage for 1.2 million state employees, retirees, and their families
(Wojcik 2004).
In early 2009, Sutter cut 121 jobs from its IT department in Rancho Cordova
to reduce costs. The cut represents about 7 percent of its IT workforce. By that
time Sutter had rolled out its electronic health records system to fve of its eight
physician organizations and one hospital. The company put rollouts on hold to
additional hospitals (Monegain 2009). In 2011, Sutter Health signed a contract
with outsourcer Affliated Computer Services (ACS) to run the MIDAS+ health
care analytic tools. The DataVision system pulls from existing clinical data and
provides reports and analyses, including comparisons with similar activities from
healthcare organizations across the nation. Krystin Dozier, Vice President of Clin-
ical Effectiveness at Sutter Health noted that MIDAS+ DataVision provides us
with the comprehensive solution we need to easily view performance across our
enterprise. We are particularly excited about SmartReport which shows us, in a
volume relative, DRG-weighted priority, those clinical areas and quality issues
where we can track issues and target improvement efforts (MIDAS+ 2011).
Questions
1. What obstacles will Sutter Health face to implement a completely digital
health care information system by the end of 2006?
2. Why is Sutter using a data gateway to transfer information across machines
instead of standardizing the underlying systems?
3. If Sutter is so advanced in its use of technology, why are its hospitals so
expensive?
Additional Reading
Benko, Laura B., Price Check, Modern Health care, April 21, 2003, Vol 33
(16), pp. 6-8.
Brewin, Bob, Sidebar: FDA Mandate Could Have $7B IT Price Tag,
Brewin, Bob and Patrick Thibodeau, Earthquake Law Pushes Hospitals To
Spend Big On IT, Comptuerworld, February 16, 2004.
Cuneo, Eileen Colkin, Uptick In Care, Information Week, November 3, 2003.
Cuneo, Eileen Colkin, A Project for Every Problem, Information Week,
December 22, 2003.
Cuneo, Eileen Colkin, Mobile Care, Information Week, March 1, 2004.
Gilhooly, Kym, Rx for Better Health Care, Computerworld, January 31, 2005.
Havenstein, Heather, Wireless Leaders & Laggards: Health Care,
Havenstein, Heather, Medical Software from Feds Could Beneft Big Health
Care, Computerworld, August 8, 2005.
Healthcare IT News, Sutter Health to Spend $50M to Help Docs with EHRs,
April 5, 2011.
Hoffman, Thomas, Saving Lives Via Video At Sutter Healths eICU,
Johnson, Van R., John Hummel, Terance Kinninger, and Russell F. Lewis,
Immediate Steps Toward Patient Safety, Health care Financial
Management, February 2004, vol 58(2).
McGee, Marianne Kolbasuk, Mission Critical, Information Week, May 19,
2003.
McGee, Marianne Kolbasuk, Health Care & Medical: Tech Innovation Keeps
The Doctor In, Information Week, September 22, 2003.
McGee, Marianne Kolbasuk, Putting a Clamp On Medical Mishaps,
McGee, Marianne Kolbasuk, Health Care & Medical: E-Health Revives Health-
Care IT, Information Week, September 20, 2004.
Merrill, Molly, Sutter Health Uses Smart App to Boost Access to Care,
Healthcare IT News, August 31, 2010.
MIDAS+ Press Release, ACS Awarded Contracts With Sutter Health and
Memorial Hermann Healthcare System, April 7, 2011.
Monegain, Bernie, Sutter Health Cuts 121 IT Jobs, Healthcare IT News, May
19, 2009.
Oliver, Carol Everett, A Cure for Bandwidth Bottlenecks, Communications
News, September 2003, vol 40(9), p. 26.
Parker, Anne, Tracking a Decade of CIS, Health Management Technology,
April 2002, Vol 23(4), pp. 28-30.
Weber, Joseph and John Cady, The New Power Play in Health Care, Business
Week, January 28, 2002.
Wojcik, Joanne, Citing High Costs, Calpers Removes 38 Hospitals From Blue
Shield Network, Business Insurance, May 24, 2004, vol 38 (21), p. 4.
www.sutterhealth.com
Case: Beth Israel Deaconess Medical Center
Beth Israel Deaconess Medical Center (BIDMC) in Boston is a teaching affliate
of Harvard Medical School. It is a not-for-proft hospital that is part of the Care-
Group Health System. The hospital has 534 beds and a Level 1 Trauma Center. It
is a major biomedical research university (bidmc.harvard.edu). Like all hospitals,
BIDMC has worked to install information technology to help patients, physicians,
and nurses.
Emergency rooms in a major city are always hectic. The ER at BIDMC treats
60,000 patients a year (an average of 164 per day). Triage is the standard medi-
cal practice of identifying the most severe cases and treating those frst (if the
treatment can reasonably be expected to succeed). But with new patients arriv-
ing constantly, and nurses and physicians rotating among cases, it can be hard to
keep track of the current situation. In the old days, hospitals used white boards
to list major issuesbut that sacrifces patient privacy and can lead to errors if
the board is not updated or erased. At BIDMC, three doctors devised a new solu-
tion: an electronic dashboard that consists of a four-foot wireless plasma dis-
play. Patients are color-coded for severity (red for serious) and by gender (pink
or blue). The entire ER was rebuilt with wireless technology in 2002. When a
patient arrives, clerks enter registration data into a laptop, and the pertinent data
is transferred to the plasma display identifying them by their initials. Unlike other
hospitals, patients are immediately moved to beds. The wireless system enables
clerks to come to the patients. Dr. John Halamka, CIO of CareGroup notes that
if you think of a traditional emergency department, you walk in and immediately
youre sitting at a triage desk. Maybe youre in pain, maybe you cant make it to
the desk. Well, thats nuts. We put you in a bed, and then the registration people
come to you and take your information (Ewalt 2001).
Physicians treating patients enter orders and diagnoses into wireless laptops,
with notations transferred to the display. When a procedure is completed, such
as X-rays, the corresponding display element (XR) turns green. The physicians
laptops also connect to the hospitals primary information system, so they can
retrieve medical histories. In addition to improving care, the system has made the
ER team more effcient. Nearby facilities averaged 450 hours in a six-month pe-
riod where they had to turn patients away. BIDMC was overloaded only 40 hours
in the same time period (MSNBC 2003).
The hospital has created wireless access in some other parts of the hospital. In
some wards, patients are even given laptops so that they can check their e-mail or
surf the Web. Halamka observes that unless Im in critical condition, I need to
access the outside world. People are there for a long time, so we give them PCs.
To improve privacy and security, the hospital encrypts all wireless transmissions
and requires that all wireless devices be registered before being granted access to
the network (Ewalt 2001).
In terms of basic data and operations, BIDMC was an early adopter of comput-
erized drug cabinets to monitor inventory levels throughout the hospital. The cabi-
nets have a built-in PC board running a Sybase database and a fat panel display.
Located throughout the hospital, they are tied to the central pharmacy to signal
when an item needs to be reflled and to provide data for patient billing (Whiting
1999). With the new federal drug bar code regulations, the drug cabinet capabili-
ties will probably not be needed.
Patient Care
As the U.S. population gets more comfortable with online interactions, it is only
natural that patients want to be able to e-mail their physicians and obtain advice
or renew prescriptions. Many physicians have resisted this technology. Some have
claimed they are worried about privacy and liability issues. A few more cynical
observers have noted that physicians do not usually get paid for these communi-
cations. Either way, few physicians embrace online interactions with patients. In
2004, Blue Cross Blue Shield of Massachusetts began a pilot study with several
health care organizations, including 200 physicians at BIDMC. Blue Cross pays
doctors $19 for each Web visit, and patients kick in a $5 co-payment. BIDMC an-
ticipates participation by about 250 patients with perhaps two e-visits each for the
frst year. Contacts and billing are handled through a secure site by RelayHealth
(McGee 2004).
Within the hospital, BIDMC is moving to electronic records. It is using a Web-
based order-entry system for prescriptions, lab tests, and supplies. The system
includes reminders for physicians and nurses and can electronically notify them
when lab results come back. Massachusetts requires that all medical data be stored
for 30 years, including images such as MRI, ultrasound, and X-ray scans. In ad-
dition to meeting state requirements, the system database can be used for data
mining (McGee September 2003). Physicians can access stored images almost
instantly through the network. Ronald Mitchell, CareGroups director of radia-
tion information systems, notes that in the operating rooms, were installing dual
high-resolution fat-panel monitors so that surgeons can view the images prior to
and during procedures (McGee October 2003).
Network Disaster
One of the challenges to an electronic medical system is that it has to keep run-
ning24 hours a day with no interruptions. In November 2002, the network at
Beth Israel Deaconess crashed and had to be completely rebuilt in a matter of
days. In the meantime, physicians and staff had to resort to paper records that had
not been used for years. Dr. Halamka, the CIO, widely reported the problems he
encountered, to show other hospitals how to improve their networks.
On Wednesday, November 13, 2002, Halamka noticed that the network was
sluggish and taking too long to send and receive e-mail. He talked with the net-
work team, and they had already noticed the problem. It appeared to be com-
ing from a surge in one of the switches. They had experienced these spikes be-
fore, but happened to have a consultant on-site looking into the problem with that
switch. To help identify the problem, network engineers began shutting down vir-
tual LAN (VLAN) segments. That action was a mistake, because it forced the
switches to recalculate the traffc distributions, and all data traffc ground to a halt
while the switches continually reconfgured. They quickly turned everything back
on, but the network was still sluggish. Around 9 P.M., the engineers spotted the
problem: a loop in the spanning tree protocol. When data arrives at a switch, the
switch computes the shortest path and directs the message to the destination. The
problem was that the spanning tree could only look out to seven hops. Once data
travels beyond seven jumps, it can lose its way and get redirected to the begin-
ningcreating a loop. On Wednesday, a researcher had loaded several gigabytes
of data into a fle-sharing application, and it looped, clogging the network. The
network team took standard steps to cut links and reduce the probability of loops
and went home for the night.
The next morning, as usage ramped up, the network slowed to a crawl again.
The team tried other options with no success. The network was beginning to cause
problems for the physicians and patients. One physician was monitoring a critical
patient and needed several lab reports to help spot the problem. But it was taking
fve hours to get lab reports completed. Fortunately, the patient survived. At 3:50
P.M., the hospital closed its emergency room for four hours.
At 4:00 P.M., Halamka called Cisco, their network provider. Cisco triggered its
Customer Assurance Program (CAP) where the company commits every resource
possible to solving the problem. A nearby team from Chelmsford moved in and set
up a command center. Their frst problem was that the network was so slow they
could not get status information from the switches. They fnally found some an-
cient 28.8Kbps modems to use to bypass the network and found the problem at 9
P.M.. The image archive system was 10 hops away from the closest core network
switch. Huge volumes of data were being abandoned because the spanning tree
system could only go out to seven hops. The problem was that the network had
been cobbled together since 1996 one piece at a time, using outdated switching
technology. The team decided to upgrade the backbone to the image system with
a Cisco 6509 router/switch. The router element provides more sophisticated com-
munications by constantly evaluating bandwidth and rerouting traffc as needed.
Shortly after 9 P.M., Cisco loaded a 6509 onto a commercial fight from San Jose
to Boston. Working through the night, the CAP team rebuilt the image network, a
task that originally took six months.
The next morning, when the load increased again, the network still began to
crash. By 10:00 A.M., Halamka decided to shut down the entire network and re-
vert to paper. Most of the medical staff had already given up on the system any-
way. Employees cranked up the copiers to generate blank forms. Some interns and
physicians had never used paper prescription forms before and had to be trained.
Runners were used to carry the paper and communicate orders.
By Saturday morning, the system was still down. The engineers decided the
entire network was outdated. At 5 A.M., three additional Cisco engineers arrived
from Raleigh. At 8 A.M., two more 6509 routers arrived by plane from Cisco. The
team of 100 people spent the day building a new network. By Saturday night, the
new core network was in place. But it took most of the night and Sunday to debug
small glitches, such as a dead network card and out-of-date frmware. On Mon-
day morning, the network was fnally stable. At noon, Halamka declared the crisis
over (Berinato 2003).
The main lesson from the disaster: networks have to be evaluated on a con-
tinual basis. You cannot just plug new items in and expect everything to work
correctly. You need to have an overall architecture that supports the entire system.
And you have to be willing to rebuild and replace core equipment as new tech-
nologies are introduced.
Moving Forward
After rebuilding the network and SQL Server databases with highly-redundant
systems, the IT system has succeed since 2004 with better than 99.9 percent up-
time (Halamka 2007). With the main network problems solved, Beth Israel Dea-
coness Medical Center continued to expand its use of information technology. In
2004, the hospital, along with the parent company CareGroup, Inc., began install-
ing RFID tags that can be read by hardware from PanGo Networks, Inc. that ties
into the wireless Wi-Fi network at the hospital. The hospital is using the tags to
track equipmentthe PanGo software provides a map of the location of expensive
equipment. Halamka noted that the two-campus hospital has millions of dollars of
expensive equipment and loses almost $400,000 of equipment a year, because in
the course of normal care, it gets misplaced (Rosencrance 2004). The system can
also be used to monitor the location of medical workers to help locate them in an
emergency, or even to locate patients in the future.
Beth Israel Deaconess is one of 31 hospitals in 10 large cities tapped by the
Centers for Disease Control and Prevention to provide automated data feeds from
its emergency rooms. The CDC is concerned about pandemics and terrorist threats
and uses the real-time data collection to help spot early trends. Barry Rhodes,
associate director for technology and informatics in the Division of Emergency
Preparedness and Response at the CDC observed that the amount and rate of data
streams to CDC is really unprecedented [compared with] what we have done in
the past (Havenstein 2006).
Beth Israel Deaconess and a few other medical providers have begun making
patient data available directly to patients via Web sites. Although many are wary
of potential privacy issues, the practice has the ability to keep patients more in-
volved, and it gives them the ability to spot and correct errors in the database
(McGee 2007).
Beth Israel was one of the frst big hospitals to adopt the Apple iPad as a useful
device for interacting with physicians. CIO John Halamka was even featured in an
Apple video. He noted that Sometimes doctors are overwhelmed with data. What
we have tried to do on the iPad is give doctors at the point of care the tools they
need at the exact moment the doctor can make a difference. The doctors can use
the devices to retrieve data and even show images to the patients (Merrill 2011).
Questions
1. How does the emergency room system at BIDMC protect patient privacy?
2. Why have physicians been so slow to adopt online and e-mail
communications from patients?
3. Why did the BIDMC network get so bad and fail? Why was it not fxed
earlier?
Additional Reading
Berinato, Scott, All Systems Down, CIO, February 25, 2003.
Ewalt, David M., CareGroups Wireless Hospital, Information Week,
September 17, 2001.
Halamka, John, Conservation of Aggravation, Computerworld, January 15,
2007.
Havenstein, Heather, CDC Upgrading IT to Gather Data From Hundreds of
Hospitals, Computerworld, February 13, 2006.
McGee, Marianne Kolbasuk, Health Care & Medical: Tech Innovation Keeps
The Doctor In, Information Week, September 22, 2003.
McGee, Marianne Kolbasuk, Putting A Clamp On Medical Mishaps,
McGee, Marianne Kolbasuk, E-Visits Begin To Pay Off For Physicians,
McGee, Marianne Kolbasuk, Doctors Debate Giving Patients Online Access To
Health Data, Information Week, May 26, 2007.
Merrill, Molly, iPad 2 Looks Even Better for Doctors, Healthcare IT News,
April 8, 2011.
MSNBC, Next Frontiers: Using Technology To Get Ahead In Business,
February 3, 2003.
Rosencrance, Linda, Boston Hospital Will Track Assets with Wireless System,
Whiting, Rick and Bruce Caldwell, Data Capture Grows Wider, Information
Case: Partners Health care System
Partners Health care System, Inc. is an association of hospitals in the Boston area,
including Brigham and Womens Hospital and Massachusetts General Hospital.
The not-for-proft organization had revenues of $4.6 billion in 2003. Across the
organization, the physicians and staff see 11,000 patients a day (annual report on
the company Web site www.partners.org).
Telemedicine
With support from engineers at MIT, the hospitals in Partners were pioneers in
the use of telemedicine. The system was created to expand the reach of the orga-
nization and provide quality care to patients in outlying areas. The system origi-
nally ran on ISDN phone lines with videoconferencing equipment. In 2000, the
organization turned to an Internet-based system using Microsoft Windows 2000
Advanced Server. Dr. Joseph Kvedar, director of the Telemedicine Program, notes
that with the help of Windows 2000 Advanced Server, we will be able to extend
the reach of our providers around the globe and to take the considerable body of
knowledge within our organization and make it available virtually anytime, any-
where (Microsoft 2000). The system is integrated into Microsofts Internet In-
formation Server (IIS) to exchange data and information as well as provide video
streaming. The unit estimates that there are 20,000 potential users in one target
group alone, and the system might eventually reach hundreds of thousands of us-
ers. Scalability and network load balancing were critical factors in upgrading the
system. To access the system, users need only a Web browser and an Internet con-
nection, as opposed to proprietary commercial lines that were needed with the old
system (Microsoft 2000).
Telemedicine offers the possibility of providing detailed expert services to
many new areas. However, some serious obstacles remain. Notably, insurance
companies are reluctant to pay for the services. In part, they are concerned about
fraud and billing abuse. Despite the obstacles, Kvedar says that Partners has seen
e-visits increase by 25 percent a year (Kolbasuk 2004). Some home-health nurses
are turning to digital cameras and camera-equipped cell phones to treat basic skin
problems. They take photos of skin wounds on diabetic patients and transfer them
via the Internet to wound specialists. The specialists can examine the patterns us-
ing the digital history fles. They can also check progress on many patients in one
day. Partners is equipping up to 180 nurses with digital cameras or high-resolution
cell phone cameras to provide care for 2,800 at-home patients (Kolbasuk 2004).
Medicine revolves around a tremendous amount of knowledge. Some of it is gen-
erated by research, some by best practices, some by experience. In the mid-1990s,
medical researchers were concerned about the high error rates at Brigham and
Womens and Massachusetts General hospitals. Sometimes simple things were
causing huge problemssuch as physicians not knowing a patients allergies or
forgetting that two drugs had bad interaction effects. So, they built a knowledge
management system to assist physicians prescribing drugs. Doctors enter orders
into the computer and it examines patient data, test results, other drugs, and diag-
nostic information to evaluate the drug choice. It then makes recommendations.
John Glaser, CIO at Partners, notes that serious medication errors have been re-
duced by 55 percent, and about 400 times a day, a physician changes his mind
on an order based on the computer (Melymuka 2002). Although that is only 3
percent of the total drug orders, it can save lives. Because the system provides
only concrete advice that is not debatable, physician acceptance has been fairly
goodeven though entering the data can add 30 minutes a day to their workload.
One physician even thanked Glaser: I just want to tell you our system has saved
my ass a couple of times (Melymuka 2002).
Unfortunately, Glaser notes that Only about 3 percent of hospitals have sys-
tems like this. Thats because its hard, but also because the ROI is fuzzy and
messy (Melymuka 2002). The low adoption rate is one of the reasons for the
push by President Bush to force hospitals to move to electronic records and track
drugs with bar codes. The Center for Information Technology Leadership backed
by Partners notes that if the industry can standardize on electronic records, hospi-
tals and insurers could save $86 billion a year (Brewin 2004).
Technology Management
One problem with technology that is rarely discussed outside of IT departments
is that individual users often want new projects or special treatment. Special re-
quests can be important and they can be useful, but individual users usually want
their pet projects to move to the top of the list. Particularly in healthcare, it is
easy for some to argue that their projects can save lives. But, if the IT department
spends all of its time responding to special requests, no time will remain for the
big projects. Partners Healthcare experienced these problemsparticularly since
the department did not have a means to track all of the special requests. Mary Fin-
lay, deputy CIO, said the team counted over a 100 ways special requests arrived
and that they were marginalizing resources by spreading them across the special
requests plus the major initiatives. To solve the problem, the IT department add-
ed a tracking system to handle evaluation and approvals of nonscheduled projects
(Artunian 2005).
John Glaser, CIO of Partners Healthcare, realized a major problem was brew-
ing in July 2004. The electronic medical record (EMR) system serving 6,000 phy-
sicians and nurses had been suffering from minor problems since the start of 2004,
but in July and early August, the system appeared to be melting downwith fre-
quent outages and slowdowns. Partners, Bostons largest hospital group, had been
using EMR for 15 yearslargely adopted by two big hospitals: Massachusetts
General and Brigham and Womens hospitals. The system continued to degrade
in August and the IT staff did not have solid answers for the causes. Mr. Glaser
brought in IBM consultants, but also spent considerable time defecting and ab-
sorbing criticisms from physicians. He recognized that he had to be visible and
talk with the doctors, noting they are angry and upset, and they want to yell at
someone, and it has to be you. You have to roll with it. You have to resist the
temptation to fght back. Eventually, his team found the root cause of the prob-
lem: failure to upgrade an older operating system that could not handle the load.
It required three months to upgrade the system and test everything, but the staff
added additional servers to reduce the outages. By the end of December, the new
system was in place and working well (Worthen 2005).
Within the general shortage of IT workers in 2010, the issue of hiring for hospi-
tal IT is an even greater problem. For example, hospitals located in smaller com-
munities would have problems attracting top talent. Located in Boston, Partners
does not have that problem, but Sue Schade, CIO at Brigham and Womens Hos-
pital, part of Partners, noted that recruiting was problematic because the hospital
runs its own custom systemnot one of the standard systems in the industry such
as Epic or Meditech. So, it is not possible to fnd people with experience, and they
have to be trained specifcally for the custom system (Monegain December 2010).
New Technologies
In 2005, Partners Healthcare created a pilot project to equip home health care
nurses with camera-equipped cell phones. The nurses use the phones to take pic-
tures of wounds and skin lesions and send them immediately to specialist nurses
trained in enterostomal therapy (wound care). Doug McClure, corporate manager
for telemedicine technology solutions, noted that there are only so many patients
these wound-care nurses would be able to see in a dayand theres only so many
of these specialty nurses available. The technology will enable them to work on
cases more effciently. The initial cell phone cameras had limited resolution (1
megapixel), but the organization is looking for higher-resolution phones as the
technology improves (McGee 2005). The hospital is also working with Ambient
Devices to help remind chronically-ill patients to take their medicine. When pa-
tients open a smart pill box, an electronic message is sent to the hospital. The hos-
pital computer then remotely turns a special light at the patients house from red to
green (McGee 2006).
In 2010, Brighams was one of the hospitals in 15 communities to receive a
share of $220 million from the federal government to serve as models for wide-
spread use of healthcare information technology. Jonathan Teich, MD, who is an
assistant professor at Harvard and works as an attending physician in emergency
medicine at Brigham and Womens Hospital among other jobs noted that We
can read all about how to do it, and we can read books and guidance, but I think
that providers really want to see examples of how its working somewhere else,
someplace like them (Monegain June 2010). The cities became known as Beacon
Communities, and each selected specifc and measureable goals for improvement.
In 2011, Partners was one of several hospitals to adopt IBMs business analytic
technology to examine effectiveness and potential safety issues of pharmaceuti-
cals across a large population of users (Healthcare IT News 2011). The Netezza
data warehouse tool will allow researchers to analyze data from millions of de-
identifed patient records. Examining the massive number of cases across time
might reveal interactions or other effects that would be diffcult to spot in smaller
studies.
Questions
1. Why did it require a federal law for hospitals to adopt bar code systems for
drug prescriptions and delivery in hospitals?
2. What are the drawbacks to telemedicine?
3. What will it take for telemedicine to be used more often?
Additional Reading
Artunian, Judy, Over the Transom: Dealing With Just-this-once, ple-e-e-ase! IT
Project Requests, Computerworld, November 28, 2005.
Brewin, Bob, Health Care IT Plans Get a Renewed Push, Computerworld, May
3, 2004.
Healthcare IT News, Vendor Notebook: DiagnosisOne Launches New
Informatics Platform, May 6, 2011.
Kolbasuk McGee, Marianne, E-Health On The Horizon, Information Week,
May 17, 2004.
McGee, Marianne Kolbasuk, Digital Images Unleashed, Information Week,
July 11, 2005.
McGee, Marianne Kolbasuk, Dude, Its Time to Take Your Medicine,
Melymuka, Kathleen, Knowledge Management Helps Cut Errors by Half,
Microsoft, Partners Health care System, Inc., February 17, 2000.
Monegain, Bernie, Health IT By Example, Healthcare IT News, June 2, 2010.
Monegain, Bernie, IT Staffng Troubles Ahead for Hospitals, Healthcare IT
News, December 1, 2010.
Worthen, Ben, Glaser Faces the Music, CIO Magazine, February 1, 2005.
adopted by rivals?
sectors?
1000 Glossary
Glossary
10Base-T A system of connecting computers
on a LAN using twisted-pair cable. The
method relies on compression to increase raw
transfer rates to 10 megabits per second.
24 -7 Operation of an application or database
24 hours a day, 7 days a week. Because the
database can never be shut down, performing
maintenance is a challenge.
Access speed A measure of disk drive speed.
Loosely, the time it takes a disk drive to move
to a particular piece of data.
Accounting journal Raw fnancial transaction
data are collected by the ac counting department
and stored in a journal. Mod ern accounting
requires the use of a double-entry system to
ensure accurate data.
Activity-based costing (ABC) ABC allo-
cates costs by examining a detailed break down
of the pro duction activities. The cost of each
process is computed for each different product.
The detail provides a better picture of the
production cost for each item.
Advanced encryption standard (AES) A
new U.S. standard for single-key encryption.
Approved in 2001 by the government to replace
DES and triple DES. With 128 bit keys, it is
substantially more diffcult to break; but still
very fast to encrypt and decrypt.
Advocacy role Someone in MIS, usually
the chief information offcer, who bears
responsibility for exploring and presenting
new applications and uses of MIS within the
company.
Agent An object-oriented program designed
for networks that is written to perform specifc
tasks in response to user re quests. Agents
are designed to automatically communicate
with other agents to search for data and make
decisions.
American National Standards Institute
(ANSI) An organization responsible for
defning many standards, including several
useful information technology standards.
American Standard Code for Information
Interchange (ASCII) American standard
code for information interchange. A common
method of numbering characters so that they
can be processed. For instance, the letter
A is number 65. It is slowly being replaced
by the ANSI character set table and the use
of international code pages that can display
foreign characters.
Angel investor An individual who provides
a limited amount of funding to start-up frms.
Unlike a partner, the investor is rarely involved
in management. The amount of funding is
generally small--$25,000 to $100,000.
Antitrust laws A variety of laws that make
it illegal to use monopoly power. Some basic
(economic) actions to achieve a competitive
advantage are illegal. Strategic plans must be
evaluated carefully to avoid violating these
laws.
Application generator A software tool
for developers that helps build software
applications. It is usually associated with a
DBMS but several standalone tools exist to
help create new forms and reports.
Application service provider (ASP) A
specialized Internet frm that provides an
individual application to other businesses. For
example, a reservation system can be run by an
ASP to provide services to other companies.
Artifcialintelligence(AI)An attempt to
build machines that can think like humans.
Techniques evolved from this research help
solve more complex problems. Useful
techniques include expert systems, neural
networks, massively parallel computers, and
robotics.
Aspect ratio Used to defne the ratio of width
to height in display screens. Standard defnition
TV was typically 6:4 (such as 640 x 480).
HDTV is generally 16:9 (such as 1920 x 1080).
Movies sometimes use a 1.85:1 ratio which is
slightly wider than the HDTV 1.77:1.
Assumptions Models are simplifcations of
real life, so they require assumptions about
various events or conditions.
Asynchronous Transfer Mode (ATM) A
packet-based network system that uses high-
speed transmission lines (150 megabits
and over) and routers to maximize network
effciency and throughput.
Attributes Descriptions of an object or entity.
For example, a customer object would at least
1001 Glossary
have attributes for name, phone number, and
address.
Auction In an e-commerce context, a Web-
based system where individuals bid for items.
Useful when you do not know the exact value
of an item or have only a few items to sell. The
auction site helps handle payments but charges
a percentage fee.
Audit trail The ability to trace any transaction
back to its source. In accounting, transaction
values are accumulated on the general ledger
and used to create reports. An audit trail is a set
of marks or records to point back to the original
transaction.
Augmented reality A display of a real-world
scene with computer-generated data added
to it. The computer data often adds names,
descriptions, or drawings. For instance, a phone
camera might be used to display a scene and
tags are added to identify points of interest.
Authentication The ability to verify the
source of a message. Dual-key systems are a
useful technique. The sender uses a private key
to encrypt the message. The recipient applies
the senders public key. If the decrypted
message is readable, it had to have come from
the alleged sender, because the keys always
work in pairs.
Backbone A high-speed communication line
that links multiple subnetworks. It is usually a
fber-optic line.
Backward chaining In an expert system,
the user enters a conclusion and asks to see
whether the rules support that conclusion.
Bandwidth Traditionally, the amount of
frequency allocated to a communication
channel, such as the portion of the spectrum
allocated to a single radio or television station.
But, the term is commonly used to indicate
the basic capacity or transmission speed of
a communication channel. For example, 20
megabits per second instead of 5 megahertz.
Bandwidth hogs On a shared network, a
small percentage of users will transmit vastly
more data than the average user. ISPs often
defne pricing methods to penalize heavy users
to ensure that more bandwidth or capacity is
available for all users.
Barriers to entry Anything that makes
it more diffcult for new frms to enter an
industry. Several possibilities would violate
antitrust laws. An acceptable barrier is the
increased use of information systems, which
raises the cost of entering an industry because a
rival would have to spend additional money on
information technology.
Beginners All-purpose Symbolic
Instruction Code (Basic) An early computer
programming language designed to be easy to
program and to teach. Visual Basic is a current
version for Windows programming.
Benchmark A set of routines or actions
used to evaluate computer performance. By
performing the same basic tasks on several
machines, you can compare their relative
speeds. Benchmarks are especially useful
when the machines use different processors and
different input and output devices.
Best practices Methods that are known
to work for solving specifc problems.
Most problems, including those in software
development, have multiple solutions. Best
practices are a collection of techniques for
solving problems that have been tested and
avoid common mistakes and problems.
BETWEEN A portion of a SQL statement
used to specify a lower and upper bound in a
WHERE clause. Commonly used for dates,
such as OrderDate BETWEEN 01-Jan-2008
AND 31-Dec-2008.
Bill of materials Used in manufacturing, it
is a list of components used to manufacture a
fnished product. In an ERP system, data from
it is often used to trigger inventory deductions
and to add the fnished product to inventory.
Bill presentation and payment Web-based
software that automatically displays bills and
invoices for customers. The payment side
accepts various forms of payment including
credit cards and electronic checks. Generally
run as a Web service.
Binary data A collection of ones and zeros
called bits. Computer processors operate
only on binary data. All data forms are frst
converted to binary.
Biometrics A feld of study that is trying to
determine how to identify people based on
biological characteristics. The most common
devices are fngerprint and handprint readers.
Bit The smallest unit of data in a computer.
1002 Glossary
All data is converted to bits or binary data.
Each bit can be in one of two states: on or off.
Bits are generally aggregated into collections
called a byte.
Bitmap image A method of storing images.
The picture is converted to individual dots that
are stored as bits. Once a picture is stored in
bitmap form, it is diffcult to resize. However,
bitmaps are good for displaying photographic
images with subtle color shading.
Blog Web log. Say it fast and you can hear the
abbreviation. A special type of Web site with
software that makes it easy for a user to enter
comments. Typically used as a daily journal.
Bluetooth A short-range wireless network
technology invented by IBM. It is most
commonly used for cell-phone devices such as
headsets. The data transmits a 1 mbps up to 32
feet. It is one of the few network protocols to
automatically encrypt the data. Newer versions
offer higher data rates.
Board of directors A group of people paid
to oversee and evaluate the decisions of the
company. Technically the CEO reports to the
board of directors, but they are charged more
with reviewing the CEOs decisions. Most
boards have the authority to remove a CEO, but
many board members are selected by the CEO.
Boolean search Searching for data by
using the logic operators AND, OR, and
NOT conditions in a WHERE statement; for
example, fnd a list of customers where city =
Detroit and age > 50 and do not own a car.
Bottom-up development An approach to
designing and building systems in which
workers build system components to solve each
problem as it arises. Eventually the pieces are
combined to create an integrated system. The
method relies on standards and controls to
facilitate cooperation and integration. See also
top-down development.
Brainstorming A group technique in which
each individual is asked to come up with
possible suggestions to a problem. Any ideas
are useful, regardless of how wild they are.
Even fanciful ideas could stimulate someone
else to improve it or to explore a related area.
Break (report) A report that organizes output
by sections that are based on the data values.
Common business examples include reports
by customer or employee, where data for each
person is displayed in a group.
Break footer The section of a break or group
report that displays subtotals for the data within
the group. See also break.
Break header The section of a break or group
report that displays the column headings for the
data within the group. See also break.
Broadcasts A technique of transmitting
messages using radio, micro, or infrared waves.
Broadcast messages are sent to all devices in
a certain area. Others in the vicinity can also
receive the messages.
Browser A software tool that converts World
Wide Web data into a graphical page with
hypertext links. Using standard (HTML)
commands, companies can offer data and
additional links to users. Users simply click
on individual words and pictures to retrieve
additional data and move to other network sites.
Brute force An attack on encrypted data that
attempts to use every possible key. Can be
stopped by using very long keys. For example,
using a key or password of only three letters
means there are only 26*26*26=17,576
possible values. Even a slow computer can test
all combinations in a few seconds.
Bulletin board system (BBS) Similar to a
typical bulletin board, except that people access
it from computers. The BBS enables users to
store com ments, pictures, and fles for other
people to retrieve. Bulletin boards are usually
orga nized by topics and can be searched for
spe cifc phrases or comments. They are a
useful way to disseminate informa tion that is of
interest to many different people.
Bus Most computers have special slots called
a bus to provide high-speed connections to
other devices. Various manufacturers make
boards that ft into these slots. The processor
can exchange data with these other devices, but
performance is sometimes constrained by the
design of the bus.
Bus network A network organizing scheme in
which each computer is attached to a common
transmission medium. Protocols are needed to
determine when a machine can transmit and to
recover from collisions.
Business process management (BPM) Also
see workfow software. The concept that
1003 Glossary
business actions have to be performed in a
specifc sequence. Managing the process entails
fnding effciencies through automating or
reordering. For example, purchasing expensive
items requires discussions and approvals by a
variety of managers.
Business to business (B2B) Business-
to-business electronic commerce; sales by
suppliers to other businesses over the Internet;
often long-term relationships. See B2C and
EDI.
Business to consumer (B2C) Business-to-
consumer electronic commerce; purchases by
individual consumers similar to traditional
mail-order systems, but conducted on secure
Web sites over the Internet.
Byte A collection of bits. Traditionally, 8 bits
make up one byte. From binary arithmetic, an
8-bit byte can hold 2 to the 8th power, or 256,
possible numbers. In many systems a byte is
used to hold one character.
C A powerful programming language that is
fexible and creates effcient code. A language
commonly used to build complex applications
and to create commercial software products.
C++ An object-oriented extension of the C
programming language. It is commonly used
to build commercial software. It produces
effcient code and supports the development of
reusable objects.
Cable modem An Internet connection device
that translates local area network protocols to
run over a television cable line. It can provide
transmission speeds around 1.5 Mbps. But the
communication line is shared with other users.
Cache A buffer between the processor and a
slow er device such as a print er, disk drive, or
mem ory chips. The cache generally consists
of high-speed memory. Data is transferred in
bulk to the cache. It is then pulled out as it is
needed, freeing up the processor to work on
other jobs instead of waiting for the slower
device to fnish.
CAN-SPAM Act The U.S. Act that makes
it illegal to send commercial e-mail messages
to people who do not want to receive them.
For business, the key is that it makes e-mail
messages legal, as long as all of the rules are
followed.
Capability maturity model integration
(CMMI) A system designed at the Carnegie
Mellon Software Engineering Institute to
help organizations improve their software
development processes. A key element is to
work toward a formal development model that
is measurable and is continually upgraded. The
CMMI system is an upgrade of the older CMM
process.
Carrier-Sense, Multiple-Ac cess/ Collision
Detection (CSMA/CD) A communications
protocol that determines how computers
will behave on a shared-medium network.
Ethernet protocols rely on CSMA/CD. Other
alternatives are Token Ring and packet
switching.
Case-based reasoning An expert system
approach that records information in the form
of situations and cases. Users search for cases
similar to their current problem and adapt the
original solution.
Catalog management system A software
tool that holds product descriptions, images,
and prices to simplify changing and uploading
data to a Web site. It makes it easier to track
thousands of products to ensure the Web site
data is correct.
CD-ROM Compact disk-read only memory.
Data is stored and retrieved with a laser. A
special machine is required to create data on
a CD-ROM. Used to hold data that does not
change very often. Useful for multimedia
applications because a disk can hold about 650
megabytes of data. The format used to store
music CDs.
Centralization A business scheme for
performing most operations and making
management decisions from one location in
an organization. MIS organization can be
examined in four areas: hardware, software,
data, and personnel. See also decentralization.
Certifcateauthority(CA)Dual-key
encryption and authentication require that the
public key be published and available to others.
A certifcate authority is an organization that
validates the owners identity, issues the keys,
and runs the public directory. Almost anyone
can run the software to be a CA, but others
must trust that host.
CertifcationsVendors provide exams to test
workers in their specifc technologies and offer
a certifcate so that potential employers can
1004 Glossary
be sure that job applicants possess a defned
level of knowledge. Common certifcations
include Cisco (networks) and Microsoft (server
administration and development). The industry
constantly argues over whether certifcations
have value.
Change agents Objects or peo ple who
cause or facilitate changes. Sometimes the
change agent might be a new employee
who brings fresh ideas; other times
change can be mandated by top-level
management. Sometimes an outside event
such as a competitor or a hurri cane forces an
organization to change.
Change drivers Concepts or products that
have altered the way businesses operate.
Classic examples include bar code scan ners
in retail stores, handheld miniterminals or
notebooks by delivery frms and sales peo-
ple, and reservation systems by travel and
entertainment industries.
Charge-back system A scheme for charging
other internal departments for services. For
example, some frms charge departments a
fee based on how often they use the central
computer. The goal is to ration a limited
resource by avoiding free use and to provide
a lever for user departments to hold MIS
accountable.
Chart of accounts A listing of all the
accounts and subaccounts in the general ledger.
It must be defned ahead of time for each
business.
Check in A step in version control systems.
When a user is fnished making changes to a
fle, the user checks in the fle to the repository
to make it fully available to other users. The
user must frst check out the fle.
Check out A step in version control systems.
A user checks out a fle or document to
indicate that changes will be made. To prevent
concurrency problems, the document is usually
locked so that others cannot make changes at
the same time. When fnished, the user checks
in the fle.
Chiefexecutiveoffcer(CEO)The head of
a company. The person ultimately responsible
for setting the direction and policies of the frm.
Usually the CEO is also the chairperson of the
board of directors.
Chiefinformationoffcer(CIO)The person
who is in charge of the MIS organization within
a frm, charged with overseeing operations,
setting MIS priorities, and being a top-level
advocate for MIS. Also develops and supports
strategy for the frm.
Circular reference In a spreadsheet, a set of
cells that eventually refer to each other. In the
simplest example, cell A1 would use values
stored in cell A2, but cell A2 uses the value
stored in A1. This technique is sometimes used
to create an iterative solution to a model.
Classes Base descriptions of objects.
Technically, classes describe generic attributes
and methods. Objects are a specifc instance of
a class.
Click-through rate Used in Web advertising,
the percentage of people viewing an online ad
who actually click it to see the details on the
advertised product or service. By 2000, the
average click-through rates had declined to less
than 1 percent. But it is not necessarily a good
measure of advertising effectiveness.
Client-server network A network
confguration in which a few machines are
used as fle servers and the others (clients) are
independent workstations. Shared data is frst
sent to a fle server where it can be examined or
transferred by another client.
Client-server organization A method of
organizing the MIS function so that some
operations are centralized while others are
decentralized. The client-server model
separates all of the components into two
categories: servers or clients. The functions
associated with the server tend to be
centralized, whereas the client components and
tasks are dispersed among the users.
Clip art Artwork created and sold to be used
by nonartists. Hundreds of collections are
available of people, places, buildings, and other
objects. Clip art images are often used to create
presentations and illustrate reports.
Clipboard The method used to transfer
data between software packages in windows-
oriented operating environments. All objects
that are cut or copied are placed onto the
clipboard, ready to be pasted to another
location or another package. Clipboard
viewers exist to show the current contents of
the clipboard. Some software systems allow
1005 Glossary
a clipboard to hold several cuttings. Many
automatically delete the older cutskeeping
only the most recent.
Clipper chip An encryption method created
by the U.S. top-secret National Security
Agency (NSA). It uses a secret algorithm
to encrypt and decrypt digital messages. It
was particularly designed for digital voice
communication. Its key feature is the use
of two escrow keys assigned to each chip.
If the police decide they want to listen to a
conversation between two suspects, they can
get a court order, collect the escrow keys, and
instantly decrypt the call.
Closed loop A system or piece of computer
code in which every step in a control
mechanism is contained inside the system
and does not utilize external input. See also
feedback.
Closed system A system that is entirely self-
contained and does not respond to changes
in the environment. Most closed systems
eventually fail due to entropy.
Cloud computing The process of running the
main part of an application on servers on the
Internet. The servers are generally scalable and
provide redundancy. Cloud services available
to the public are usually charged on some type
of per-use fee basis so frms can buy the level
of computing.
CMYK Cyan-Magenta-Yellow-Key (black).
A color model used in the printing world to
precisely defne colors. Colors can be expressed
by specifying the percentage needed of each of
the primary colors. See also RGB.
Coaxial cable A cable used to transmit data.
Cable television is a widespread application.
The inner cable is surrounded by a plastic
insulator, which is surrounded by a wire mesh
conductor and an outer casing. The wire mesh
insulates the internal signal wire from external
interference.
Cold site A facility that can be leased from a
disaster backup specialist. A cold site contains
power and telecommunication lines but no
computer. In the event of a disaster, a company
calls the computer vendor and begs for the frst
available machine to be sent to the cold site.
Collision In networks, a collision arises when
two computers attempt to broadcast messages
at the same time. The network protocols need
to identify the situation and determine which
machine will go frst.
Co-location Installing your computer or
network equipment in anothers facilities
to obtain access to high-speed data
communication lines. You pay a fee for use
of the facilities, power, cooling, and network
usage.
Column A vertical part of a table that holds
data for one attribute of an entity in a database
or spreadsheet. For example, a table to
describe automobiles will have columns for
make, model, and color.
Command-line interface A method of
controlling the computer by typing commands.
The user must generally memorize specifc
commands. Older machines still use them
because GUI systems require too much
overhead. Some people prefer command
lines, because it is faster to type one or two
commands than to manipulate an image on the
screen.
Commerce server A software system that
runs an e-commerce Web server. It handles the
product catalog, searching, a shopping cart, and
the payment mechanism. Several vendors sell
versions to be run on your own server, or you
can lease space on a hosting company.
Commercial off-the-shelf software (COTS)
Purchased software for building applications.
Relatively popular because it is faster than
building from scratch.
Common Business-Oriented Language
(COBOL) An early programming language
designed to handle typical transaction-
processing tasks. Its death has been predicted
for years, but it is hard to throw away billions
of lines of code.
Common Object Request Broker
Architecture (CORBA) A model largely
developed in the UNIX community that will
enable objects to communicate with each other
across networks. In particular, it is designed
to enable users to combine different data types
from various software vendors into a single
compound document. The data could reside on
any server on the network.
Competitive advantage Something that
makes your company better or stronger than
1006 Glossary
your rivals. Examples include lower costs,
higher quality, strong ties to loyal customers,
and control over distribution channels.
Composite key In defning a database table,
each table must have a primary key. When
the primary key consists of more than one
column, it is referred to as a composite key.
The business relationship between the multiple
columns are many-to-many.
Compound document A document that
incorporates different types of data: text,
graphics, sound, and video. The different
objects might be transmitted across a network
to be included in a fnal document.
Computer-aided design (CAD) Programs
that are used to create engineering drawings.
CAD programs make it easy to modify
drawings. They also make it easier to keep
track of material specifcations. They can
perform spatial and engineering estimates
on the designs, such as surface or volume
calculations.
Computer-aided software engineering
(CASE ) Computer programs that are
designed to support the analysis and
development of computer systems. They make
it easier to create, store, and share diagrams and
data defnitions. Some versions even generate
code. There are two cate gories of CASE tools:
software develop ment and maintenance of
existing sys tems.
Computer-integrated manufacturing
(CIM) Using a computer to control most of
the production equipment in a manufacturing
environment. The computer can monitor the
production statistics. It is also used to set
individual machine controls.
Computer ethics The concept that all of us
have an obligation with respect to data. For
example, managers have a responsibility to
customers to protect personal data, to collect
only data that is truly needed, and to give
customers the ability to correct errors in
personal data.
Computer information system (CIS) See
management information system (MIS).
Composite key In relational databases, a
key that consists of more than one column.
The columns are combined to yield a unique
primary key.
Concurrency A situation that arises when
applica tions attempt to modify the same piece
of data at the same time. If two people are
allowed to make changes to the same piece
of data, the com puter system must control the
order in which it processes the two re quests.
Mixing the two tasks will result in the wrong
data being stored in the comput er.
Content management system Changing text
and images on a Web site can be challenging,
particularly with thousands of pages and
hundreds of contributors. Contributors can
write changes in simple text format to a content
management system which then formats and
uploads the data to the Web site automatically.
Context diagram The top level of a data fow
diagram that acts as a title page and displays
the bound aries of the system and displays the
external entities that interact with the system.
Continuous quality improvement The
concept that any process can be improved by
continually evaluating the system and making
adjustments and refnements. The concept is
also applied to service processes, but relies on a
measurable performance objective.
Converge The ability of an iterative model to
stabilize on a fxed solution. The alternative is
that values continually increase and never reach
a solution.
Cookies Small text fles that a Web server
sends to client computers. When the user
returns to a site, the browser automatically
returns the cookie fle. Servers use them to keep
track of transactionsso they know when the
same user has returned. Marketers have used
them to track individual users on the Web.
Copyright A legal ownership right granted to
the creators of intellectual property. All works
are automatically copyrighted. Registering with
the copyright offce is not required but grants
additional protection to the owner.
Critical success factors A limited number
of concrete goals that must be met for the
organization to be successful. Identifying
these key factors helps determine the strategic
directions and highlights the areas that can
beneft from improved information systems.
Customer relationship management (CRM)
A system for tracking and integrating all
customer data. Salespeople, managers, and
1007 Glossary
clerks all have access to the same data, so
everyone has the same consolidated view of all
customer interactions.
Cut, copy, paste A common mechanism used
to transfer and link data between different
software packages. The data to be transferred
is marked. When it is cut or copied, it is
placed on the clipboard. Switching to the
second package, the object is pasted into the
appropriate location. Dynamic and static links
are specifed through options in the paste
special menu. With the cut option, the original
object is deleted. With copy, the original is
unchanged.
Data Consists of factual elements (or opinions
or comments) that describe some object or
event. Data can be thought of as raw numbers
or text.
Data administrator MIS manager who
is charged with overseeing all of the data
defnitions and data standards for the company
to ensure that applications can share data
throughout the company.
Data dictionary Contains all of the infor-
mation to explain the terms used to defne a
system. Often includes report descriptions,
business rules, and security considerations.
Data encryption standard (DES) An
older method of encrypting data that was
commonly used by fnancial institutions. With
current computer capabilities that can break
a DES-encrypted message, DES is no longer
considered a secure encryption system.
Datafowdiagram(DFD)A diagramming
technique used to analyze and design systems.
It shows how a system is divided into
subsystems and highlights the fow of data
be tween the processes and subsystems. It
displays processes, external entities, fles, data
fows, and control fows.
Data independence Separating programs
from their data defnition and storage. The
main advan tage is that it is possible to change
the data without having to change the pro-
grams.
Data integrity (1) A concept that implies
data is as accurate as possible. It means the
database contains few errors. (2) Keep ing data
accurate and correct as it is gath ered and stored
in the comput er system.
Data mart A small version of a data
warehouse. A database designed to hold concise
collections of data for retrieval and analysis by
managers.
Data mining An automated system that
examines data for patterns and relationships. It
is partly based on statistics, but also searches
for more specifc associations. The results are
not always applicable to other situations.
Data mirroring The ultimate backup
technique where all data that is stored on one
machine is automatically transferred and stored
on a second computer. Useful to prevent loss of
data and recover from disastersparticularly
when the second computer is located many
miles away.
Data store A fle or place where data is
stored. In a realistic setting, a data store could
be a computer fle, a fle cabinet, or even a
reference book.
Data types To humans, there are four basic
types of data: text and numbers, images, sound,
and video. Each data type must be converted to
binary form for computer processing.
Data warehouse A single consolidation point
for enterprise data from diverse production
systems. The data is typically stored in one
large fle server or a central computer. Because
legacy systems are diffcult to replace, some
data is copied into a data warehouse, where it is
available for management queries and analysis.
Database A collection of related data that can
be retrieved easily and pro cessed by computers;
a collection of data tables.
Database administrator (DBA) (1) A person
appoint ed to manage the data bases for the frm.
The DBA needs to know the technical details
of the DBMS and the computer system. The
DBA also needs to under stand the business
operations of the frm. (2) A management
person in the MIS department charged with
defning and maintaining the corporate
databases. Maintaining data integrity is a key
component of the job.
Database management system (DBMS)
Software that defnes a database, stores the
data, supports a query language, produces
reports, and creates data-entry screens.
Decentralization Moving the major
operations and decisions out to lower levels
1008 Glossary
within the frm. In MIS, decentralization has
largely been led by the declining cost and
improved capabilities of personal computers.
See also centralization.
Decision biases Without models and careful
analysis, decisions made by people tend to be
biased. There are several biases in each of
the four systems categories: data acquisition,
processing, output, and feedback.
Decision process The steps required to make
a decision. It includes problem identifcation,
research, specifcation of choices, and the
fnal selection. Midlevel managers are often
involved in the initial stages and affect the
outcome, even though they may not make the
fnal decision.
Decision support system (DSS) System to
use data collected by transac tion-process ing
systems to evaluate business models and assist
manag ers in making tactical decisions. There
are three major components: data collection,
analysis of models, and presentation.
Decision tree A graphical representation of
logic rules. Each possible answer to a question
or situation leads to a new branch of the tree.
Default value A value that is auto matically
displayed by the computer. Users can often
override the default by deleting the old value
and entering a new one. The goal is to choose a
value that will almost always be entered, so the
user can skip that item.
Dehumanization Some people feel that
technology isolates people and decreases
our contact with other members of society.
Treating people as identifcation numbers and
summary statistics can lead managers to forget
the human consequences of their decisions.
Denial of Service (DoS) Preventing legitimate
users access to systems and networks. A
common Internet trick is to force thousands
of zombie computers to food a server with
millions of meaningless messagespreventing
anyone else from using the system.
Descriptive model A model that is defned
in words and perhaps pictures. Relationships
between objects and variables tend to be
subjective. Useful for an initial understanding
of a system but diffcult to evaluate by
computer.
Desktop publishing (DTP) The art of
creating professional documents with personal
computers and small laser printers. Beyond
basic word processing, DTP software provides
controls to standardize pages, improve the page
layout, and establish styles.
Detail section The section in a report that is
repeated for every row in the associated tables.
It is often used for itemized values, whereas
group and page footers are used for subtotals.
Device drivers Small software modules that
provide the interface from an operating system
to a hardware device. Manufacturers improve
and rewrite their device drives, so you should
periodically update your system to obtain the
newer drivers.
Diagnostic situations Spotting problems,
searching for the cause, and implementing
corrections. Examples include responding
to exception reports to identify problems and
potential solutions, and determining why the
latest marketing approach did not perform as
well as expected.
Dial-back modem A special modem placed
on a central computer. When a user attempts
to log in, the dial-back modem breaks the
connection and calls back a predefned
phone number. Its use minimizes the threat
of outsiders gaining access to the central
computer.
Digital cash An electronic version of money
that is provided and verifed by a trusted third
party. It consists of an encrypted number for
a specifed value that can only be used one
time. It provides for verifable and anonymous
purchases using networks.
DigitalcertifcatePart of an authentication
mechanism used with dual-key encryption.
Companies that host servers need to encrypt
transactions over the Internet. They purchase
a digital certifcate from a certifcate authority
and install it on the Web server. The client
browser recognizes the certifcate key and
encrypts the data.
Digital dashboard A visual presentation
of broad measures of current activity in an
organization. The data is generally displayed
as gauges, and the system must be customized
for each organization. As part of an executive
information system, managers can drill down to
get more data.
1009 Glossary
Digital divide The distance between those
individuals or nations who have network
capabilities and those who do not. Despite
declining costs, many people and many nations
cannot afford the hardware and software. If a
large portion of the economy moves online,
it could alienate those who cannot afford the
network connection.
Digital rights management (DRM) A
combination of encryption and Internet
validation for protecting vendor copyrights to
prevent unauthorized copying of digital content
(software, music, books, movies, and so on).
Digital signature Any electronic signature
technology that verifes the user. U.S. law now
recognizes digital signatures as equivalent to
handwritten ones. The most secure system
is to obtain a digital certifcate from a public
company that verifes each persons identity.
But the IRS accepts a simple PIN issued by the
agency as a digital signature.
Digital subscriber line (DSL) A special
phone service connection available to
customers within 3 miles of the phone
companys switch. It provides about 1 Mbps
transmission speed for Internet connections.
Digital video/versatile disk (DVD) A digital
format primarily used for storing video and
movies. However, it can also hold audio and
traditional computer data. One side of the disk
can hold over 3 gigabytes of data.
Direct sequence spread spectrum (DSSS)
A network transmission protocol commonly
used for wireless connections. It subdivides the
allocated frequency to send multiple packets
at the same time. Communication packets can
shift frequencies at each time slot. By making
more effcient use of the spectrum, more data
can be transmitted.
Disintermediation In an e-commerce context,
using a Web-based system to skip over sections
of the production chain, such as manufacturers
selling directly to consumers. The approach
can give the manufacturer a higher percentage
of the sale price, but risks alienating retailers,
resulting in lost sales.
Distribution center (DC) A central point in
a supply chain where incoming bulk goods are
split and merged into multiple shipments to the
fnal destination. For example, a truckload of
bread would be unloaded and individual boxes
placed on other trucks, along with other food
items for distribution to a grocery store.
Distribution channel The layers of
distributors in between the manufacturer and
the fnal customer. If a producer can gain
control over this means of getting the product
to the consumers, the producer can prevent new
rivals from entering the industry. Improved
communication systems offer the possibility
of eroding control over some distribution
channels.
Diverge The property of an iterative model
where successive computations keep leading to
larger values (in magnitude). The model never
reaches a stable solution. Generally due to
insuffcient or incorrect feedback mechanisms.
Documentation Descriptions of a system, its
components, the data, and records of changes
made to the system.
Domain name system (DNS) A set of
computers on the Internet that converts
mnemonic names into numeric Internet
addresses. The names are easier for humans
to remember, but the computers rely on the
numeric addresses.
Download To transfer fles from a remote
computer to a local computer (usually a
personal computer). See also upload.
Drill down To use an information system to
get increasingly detailed data about a company.
In an enterprise information system, the ability
to look at overall company data, and then select
breakdowns by regions, departments, or smaller
levels.
Dot-com Abbreviation given to the many
Internet frms formed in the late 1990s because
their Internet names ended with the .com suffx.
For a couple of years, having a dot-com name
was prestigious and attracted funding. When
hundreds of these frms failed in 2000 and
2001, they became known as dot-bombs.
Dots per inch (dpi) A measure of the
resolution of devices including printers and
displays. Higher values representing more dots
per inch provide more detailed images and text.
Some people use the term pixels (ppi) instead
of dots.
Drill down The action in a data analysis
package or executive information system where
the user clicks a link to obtain more detail
1010 Glossary
about a specifc situation. See also roll up.
Dual-key encryption A method of encrypting
a message that requires two keys: one to
encrypt and one to decrypt. One of the keys is
a public key that is available to anyone. The
other key is private and must never be revealed
to other people. RSA is a popular dual-key
encryption system. Dual-key systems can also
be used to authenticate the users.
Dynamic data exchange An early method
of linking data from multiple sources with
the Windows operating system. The software
packages literally send messages to other
software packages, which enables them to
combine and update data. See also dynamic
integration as well as Object Linking and
Embedding (OLE).
Dynamic host control protocol (DHCP) The
standard Internet method for assigning Internet
addresses to a computer. A DHCP server is
given a database with a range of IP addresses
and it assigns an unused number to a computer
when it requests one on startup. The process
means that IP addresses on specifc computers
can change over time unless the computer is
allocated a static address in the database.
Dynamic integration A means of linking
data from multiple documents. One compound
document (or container) can hold data objects
created by other software. As the original data
is changed, it is automatically updated in the
container document. See also static integration.
e-Business Electronic business. The process
of conducting any type of business over the
Internet. It includes all forms of e-commerce
and m-commerce, as well as internal processes
and Web services.
e-Commerce (EC) Electronic commerce. The
process of selling items over the Internet. The
most familiar form is business-to-consumer,
but it includes business-to-business and auction
sites like eBay.
e-Discovery Electronic discovery. In legal
cases, a request for all electronic data
particularly e-mail communications, but might
also include accounting or other data from any
computer system. Companies establish policies
to defne storage life and destruction policies to
ensure old data is deleted automatically from
systems.
E-mail Electronic mail, or messages that
are transmitted from one computer user to
another. Networks transfer messages between
the computers. Users can send or retrieve
messages at any time. The computer holds the
message until the recipient checks in.
EBCDIC: Extended Binary Coded Decimal
Interchange Code A method of numbering
characters so that they can be processed by
machines. Used exclusively by large IBM and
compatible computers. See also ASCII.
Electronic data interchange (EDI)
Exchanging transaction data with entities
outside the control of your frm. Private
connections can be established directly between
two frms. Public networks are also being
formed where one provider collects data and
routes it to the appropriate client.
Encryption A method of modifying the
original information according to some code,
so that it can be read only if the user knows the
decryption key. It is used to safely transmit
data between computers.
End-user development Managers and
workers are to develop their own small
systems using database management systems,
spreadsheets, and other high-level tools.
Enterprise network A network that connects
multiple subnetworks across an entire frm.
Often, the networks use different protocols and
different computer types, which complicates
transmitting messages.
Enterprise resource planning (ERP) An
integrated computer system running on top of
a DBMS. It is designed to collect and organize
data from all operations in an organization.
Existing systems are strong in accounting,
purchasing, and HRM.
Entrepreneurship The act of creating
and organizing a business. Generally, an
entrepreneur takes the risks to create a new
business in search of a proft.
Ergonomics The study of how machines can
be made to ft humans better. One of the main
conclusions of this research in the computer
area is that individuals need to be able to
adjust input (and output) devices to their own
preferences.
Escrow key In an encryption system, it is a
special key that can be used by government
1011 Glossary
offcials to decrypt a secret conversation. The
Clipper chip uses escrow keys.
Ethernet A network communications protocol
that specifes how machines will exchange
data. It uses a broadcast system in which
one machine transmits its message on the
communication medium. The other machines
listen for messages directed to them.
Ethics The concept that various elements of
society have obligations to the others. In IT, it
focuses on the roles of users, developers, and
vendors.
Event-driven approach A user-interface
approach where the user controls the sequence
or operations and the software responds to
these events. Events can range from a simple
key-press to a voice command. Modern,
window-based software does not follow a
sequential process. Instead, actions by users
generate events. The programs
respond to these events and alter data or offer
additional choices. Typical events include
mouse clicks pointing to items on the screen,
keystrokes, changes to values, or transmissions
from other systems.
Exabyte A count of the number of bytes one
step above petabyte and one below zettabyte.
The new IEC defnition uses exbibyte for
binary data and exabyte only for decimal data
which is 10
18
.
Exbibyte The new IEC defnition for binary-
based counting instead of exabyte. The binary
version is 2
60
. Technically 2 raised to the 60th
power or 1024*1024*1024*1024*1024*1024
(6 times).
Exception report Report that is triggered
by some event to signify a condition that is
unusual and needs to be handled immediately.
Executive information system (EIS) A
type of decision support system that collects,
analyzes, and presents data in a format that
is easy to use by top executives. To achieve
this objective, the EIS is based on a model of
the entire company. In most cases the model
is pre sented graphically and the executives
retrieve information by pointing to objects on
the screen.
Exhaustive testing Testing every possible
combination of inputs to search for errors.
Generally not a feasible option, so most
computer systems will always contain errors.
Expert system (ES) System with the goal of
helping a novice achieve the same results as
an expert. They can handle ill-structured and
missing data. Current expert systems can be
applied only to narrowly defned problems.
Diagnostic problems are common applications
for expert systems.
Expert system shell A program that provides
a way to collect data, enter rules, talk to users,
present results, and evaluate the rules for an
expert system.
Export An older method of exchanging
data among various software packages. One
package exports the data by storing it in a
format that can be read by other software.
Object Linking and Embedding is a more
powerful way to exchange data.
Extensible business reporting language
(XBRL) A specifc XML style for reporting
fnancial data in a standard way. Predefned
tags are used to mark the fnancial data to make
it easier for computers to extract and compare
data from diverse companies.
Extensible markup language (XML) A tag-
based notation system that is used to assign
names and structure to data. It was mainly
designed for transferring data among diverse
systems.
External agents Entities that are outside
the direct control of your company. Typical
external agents are customers, suppliers, rivals,
and governments. Competitive advantages can
be found by producing better-quality items or
services at a lower cost than your rivals. Also,
many frms have strengthened their positions
by building closer ties with their suppliers and
customers.
External entity Objects outside the boundary
of a system that communicate with the
system. Common business examples include
suppliers, customers, government agencies, and
management.
Extraction, transformation, and loading
(ETL) The process in data warehouses that
involves taking data from existing systems,
cleaning it up, and moving it into the data
warehouse.
Extranet A network confgured to give certain
outsiders, usually customers and suppliers,
1012 Glossary
limited access to data using Web-based
systems.
Extreme programming (XP) A new version
of development loosely based on prototyping.
Pairs of developers rapidly build and
simultaneously test applications. The goal is to
build releases and then modify them to meet
the changing needs of the users.
Facsimile (Fax) A combination scanner,
transmitter, and receiver that digitizes an image,
compresses it, and transmits it over phone lines
to another facsimile machine.
Fault tolerance The ability of a computer or
a system to continue functioning properly even
if some of the components fail. Fault-tolerant
machines rely on duplication of subsystems
with continuous monitoring and automatic
maintenance calls.
Feasibility study A quick examination of
the problems, goals, and expect ed costs of a
proposed system. The objective is to determine
whether the problem can reasonably be solved
with a computer system.
Feedback Well-designed systems have
controls that monitor how well they meet their
goals. The information measuring the goals
and providing control to the system is known as
feedback.
Fiber optic cable A thin glass or plastic
cable that is internally refective. It carries a
light wave for extended distances and around
corners.
File server Computer on a network that is
used to hold data and program fles for users
to share. To be effective, it should use a
multitasking operating system.
File transfer protocol (FTP) A standard
method of transferring fles on the Internet. If
you control a computer, you can give other
users access to specifc fles on your computer
without having to provide an account and
password for every possible user.
Firewall A small, fast network computer
device that examines every packet entering a
company. Rules or flters can be created that
will reject certain packets that are known to be
dangerous to the network.
First mover In a model of rivalry, the frm
that takes the initial action. Sometimes the frst
mover gets a beneft by setting the strategy
and the market. But the costs are often higher
because the technology is newer. Games such
as chess recognize that the frst mover has a
slight beneft.
Five Forces model Michael Porters model
used to search for competitive advantage. The
Five Forces are: rivals, customers, suppliers,
potential competitors, and substitute products.
Floating point operations per second
(FLOPS) The number of mathematical
calculations a processor can perform in one
second. Typically measured in millions (mega-
FLOPS) or billions (giga-FLOPS). Bigger
numbers represent faster processors.
Flow chart An old pictorial method for
describing the logic of a computer program. It
has largely been replaced by pseudocode.
Font size An important characteristic of text
is its size. Size of type is typically measured in
points. For reference, a capital letter in a 72-
point font will be approximately 1 inch high.
Forward chaining In an expert system, the
ES traces your rules from the data entry to a
recommendation. Forward chain ing is used to
display questions, perform calcula tions, and
apply rules.
Frame A related set of information that
humans group together. Sometimes groupings
can be arbitrary. A concept used in discussing
AI applications and human cognition.
Frame relay A network communication
system that uses variable-length packets. It
is useful for high-speed, large bursts of data.
It is being used for long-distance network
communications.
Franchise A means of organizing companies.
Independent operators pay a franchise fee to
use the company name. They receive training
and beneft from the name and advertising of
the parent company. They purchase supplies
from the parent company and follow the
franchise rules.
Frequency division multiplexing (FDM)
Supporting multiple communications at the
same time by assigning a specifc frequency
range to each participant. For example,
television and radio stations are assigned
specifc frequency ranges to avoid collisions.
1013 Glossary
Front-end processor A simple com mu ni-
cations device for large central computers that
ac cepted all of the terminal wires and then
assigned each user to an open communi cations
port on the computer. This device decreased
the number of physical access ports required on
the computer.
Full duplex A method of transferring data,
usually over phone lines, so that data is
transmitted in both directions simultaneously.
In terms of speaker phones, it means that
people on both ends of a call can talk at the
same time. With half duplex, the initial speaker
blocks others from talking.
Functions See methods.
Fuzzy logic A way of presenting and
analyzing logic problems that is designed to
handle subjective descriptions (e.g., hot and
cold).
General ledger A collection of accounts that
break fnancial data into specifc categories.
Common categories include accounts
receivable, accounts payable, inventory, and
cash.
Geographic information system (GIS) De-
signed to identify and display relationships
among business data and locations. Used to
display geographical relationships. Also used
to plot delivery routes and create maps.
Gibibyte The IEC defnition for billion in
binary base 2 (2
30
). It replaces the term gigabyte
which now is to be used for decimal billion.
Gigabyte Approximately 1 billion bytes of
data. Technically, 2 raised to the 30th power or
1024*1024*1024 (3 times). It is one step above
megabyte and one below terabyte..
Global positioning system (GPS) A system
of 24 satellites created by the U.S. Department
of Defense. The civilian receivers will identify
a location to within about a few feet. Used for
navigation, track vehicles, and plotting delivery
routes.
Graphical user interface (GUI) A system
that is based on a graphics screen instead of
simple text. Users perform tasks by clicking
a mouse button on or manipulating objects
on the screen. For example, copies are made
by dragging an item from one location on the
screen to another. Pronounced as gooey.
Grid computing A system that networks
multiple computers so that they cooperatively
process the designated tasks, effectively
functioning as a single computer.
Group breaks Reports are often broken
into subsections so that data in each section is
grouped together by some common feature.
For example, a sales report might group
items by department, with subtotals for each
department.
Group decision support system (GDSS) A
type of groupware that is designed to facilitate
meetings and help groups reach a decision.
Each participant uses a networked computer
to enter ideas and comments. Votes can be
recorded and analyzed instantly. Comments
and discussion are automatically saved for
further study.
Groupware Software designed to assist
teams of workers. There are four basic
types: communication, workfow, meeting,
and scheduling. The most common is
communication software that supports
messages, bulletin boards, and data fle
transfers and sharing.
Hacker Primarily used to indicate a person
who devotes a great deal of time trying to break
into computer systems.
Hardware The physical equipment used in
computing.
High-bandwidth digital content protection
(HDCP) The digital rights management
technology created by the movie and television
industries to make it more diffcult for people
to copy high-defnition TV signals. To play
protected videos all of your equipment will
need to support the HDCP standard.
High-DefnitionTelevision(HDTV)
Transmission of television signals in digital
form. It provides clearer reception. It also
supports encrypted transmissions so that
broadcasters can control who receives the
images. HDTV also supports compression,
so that more data (better pictures or more
channels) can be transmitted in the same
frequency space.
Hot links See dynamic integration.
Hot site A facility that can be leased from a
disaster backup specialist. A hot site contains
all the power, telecommunication facilities,
1014 Glossary
and computers necessary to run a company. In
the event of a disaster, a company collects its
backup data tapes, notifes workers, and moves
operations to the hot site.
Hub A network device used to connect
several computers to a network. Commonly
used in a twisted-pair LAN. A cable runs from
each computers NIC to the hub. The hub is
often connected to a router.
Hypertext markup language (HTML) The
standard formatting system used to display
pages on the Internet. Special tags (commands
inside angle braces, e.g., <HTML>) provide
formatting capabilities. Several software
packages automatically store text in this format,
so users do not have to memorize the tags.
Icon A small picture on a computer screen
that is used to represent some object or indicate
a command. A classic example is the trash can
used to delete fles on the Apple Macintosh.
Image A graphic representation that can be
described by its resolution and the number of
colors. They can be stored as bit-mapped or
vector images.
Import An older method of exchanging
data among various software packages. Most
software (e.g., a database management system)
can export or store data in a text fle format.
Another software package (e.g., a spreadsheet)
can import or retrieve this data. Object Linking
and Embedding is a more powerful way to
exchange data.
Inference engine Within an expert system,
the inference engine applies new observations
to the knowledge base and analyzes the rules to
reach a conclusion.
Information Data that has been processed,
organized, and integrated to provide insight.
The distinction between data and information is
that information carries meaning and is used to
make decisions.
Information center An MIS group
responsible for supporting end users. It
typically provides a help desk to answer
questions, programmers who provide access
to corporate databases, training classes, and
network support people to install and maintain
networks.
Information rights management (IRM) A
system to control exactly what each group
can do with digital data, including documents,
music, and video fles. A good IRM system
can prevent a document from being read by
outsiders, even if the document is somehow
shipped outside the companys computers.
Information system A collection of
hardware, software, data, and people designed
to collect, process, and distribute data
throughout an organization.
Information technology (IT) The hardware
and software used to create an information
system. Sometimes used as an abbreviation for
management information systems.
Information threats There are two classes of
threats to information: (1) physical, in the form
of disasters; and (2) logical, which consists
of unauthorized disclosure, unauthorized
modifcation, and unauthorized withholding of
data. The primary source of danger lies with
insiders: employees, ex-employees, partners,
or consultants.
Information warfare (IW) The use of
information in a confict setting. It includes
protecting your own information, providing
misinformation to the enemy, and monitoring
and disrupting the enemys information.
Inheritance Creation or derivation of objects
from other object classes. Each derived class
inherits the attributes and methods of the prior
class. For example, a savings account object
can be derived from an account object. The
savings account object will automatically have
the same attributes and methods. Attributes
and methods specifc to the savings account can
be added.
Initial public offering (IPO) The step when
frms frst sell stock to the public. A method of
raising additional funds and a major step for
most start-up frms.
Input devices People do not deal very well
with binary data, so all data forms must be
converted into binary form for the computer.
Input devicesfor example, keyboards,
microphones, and bar code readersmake the
conversion.
Input-Process-Output A shorthand
description of a subsystem. Each subsystem
receives inputs and performs some process.
The output is passed to another subsystem.
Instant Messaging (IM) A two-way
1015 Glossary
electronic communication in real time. Short
comments that you type are immediately
displayed on the recipients screen. It generally
requires that both parties run the same software.
Integrated data The practice of combining
data from different sources to make a decision.
Data can come from different departments
throughout the business, and it can come in
many different forms. Networks, groupware,
and products that support dynamic linking are
all useful tools to integrate data to make better
decisions.
Integrated Services Digital Network
(ISDN) A set of services, and a transmission
and control system, offered by telephone
companies. It uses complete digital
transmission of signals to improve transmission
speed and quality.
Intellectual property As defned by
copyright laws, the concept that property such
as music, books, software, and movies can be
protected. The laws clearly defne the owners
of the property and specify that the owners
can establish any type of copy protections they
desire.
Internet A collection of computers loosely
connected to exchange information worldwide.
Owners of the computers make fles and
information available to other users. Common
tools on the Internet include e-mail, ftp, telnet,
and the World Wide Web.
Internet Assigned Numbers Authority
(IANA) The Internet committee that is
responsible for allocating IP address segments.
It generally focuses on technical issues,
manages the DNS root zone and coordinates
the global numbering system.
Internet Corporation for Assigned Names
and Numbers (ICANN) The Internet
committee founded in 1998 that coordinates the
Internet naming system. It is largely a political
committee, formed within the United States,
but with international membership, it discusses
new top level domains, internationalization
issues and so on.
Internet Engineering Task Force (IETF)
The primary technical committee that defnes
Internet standards. It is responsible for
standards such as router protocols and HTML.
Internet service provider (ISP) A private
company that provides connections to the
Internet. Individuals pay a fee to the ISP. The
ISP pays a fee to a higher-level provider (e.g.,
NSP) to pass all communications onto the
Internet.
Intranet A network within an organization
that utilizes standard Internet protocols and
services. Essentially, this includes Web sites
that are accessible only for internal use.
Intrusion detection system (IDS) A software
tool containing sensors and a set of rules that
monitors network traffc looking for attackers.
Snort is a commonly-used tool for IDS because
it is freely available from the open-source
community.
Intrusion prevention system (IPS) Similar
to an IDS but more active. It uses software to
monitor network activity and a set of rules to
indicate when an attack occurs. It then uses
router and system commands to shut down or
delay access to attackers.
Iterative solution Building a model and
evaluating it until the parameter values
converge to a fxed solution. Sometimes an
iterative model will diverge and never reach
an acceptable solution. See also circular
reference.
Intrusion detection system (IDS) A
combination of hardware and software that
monitors packets and operations on the network
and computers. It watches for suspicious
patterns that might indicate an attack.
Internet Protocol version 6 (IPv6) A set
of standards that defne how raw data is
transmitted on the Internet and how machines
are addressed. Version 6 contains several
improvements to the older version 4. For
example, version 6 supports 128-bit addresses
compared with 32 bits in version 4. It will take
several years for people to move to version 6.
Joint application design (JAD) A method
to reduce design time by putting everyone
in development sessions until the system
is designed. Users, managers, and systems
analysts participate in a series of intense
meetings to design the inputs (data and screens)
and outputs (reports) needed by the new
system.
Just-in-time (JIT) inventory A production
system that relies on suppliers delivering
1016 Glossary
components just as they are needed in
production, instead of relying on inventory
stocks. JIT requires close communication
between manufacturers and suppliers.
Kerberos A security system created at MIT
that enables systems to have a single sign-on.
Users log into the Kerberos server and other
systems can validate the users identity from
that server. Much simpler than requiring users
to log in multiple times. Named after the
hound that guards the gates of Hades (spelled
Cerberus in Latin).
Kilobyte Approximately one thousand bytes
of data. Technically it is 2 to the tenth, or 1024.
The next step up is megabyte.
Knowledge A higher level of understanding,
including rules, patterns, and decisions.
Knowledge-based systems are built to
automatically analyze data, identify patterns,
and recommend decisions.
Knowledge base Within an expert system, the
knowledge base consists of basic data and a set
of rules.
Knowledge engineer A person who helps
build an expert system by organizing the data,
devising the rules, and entering the criteria into
the expert system shell, trained to deal with
experts to derive the rules needed to create an
expert system. The engineer also converts the
data and rules into the format needed by the
expert system.
Knowledge Management (KM) A system
that stores information in the context of a set
of decisions. It contains cross-references and
search methods to make it easy for workers to
understand how and why decisions were made.
Last mile The connection from an ISP to
individual households and businesses. In many
cases, the most diffcult connection to make
because of the cost and monopoly control.
Most households are limited to a few choices:
telephone and cable TV. Some technologies
exist to run communications over power lines.
The other option is wireless.
Latency The delay between initiating an
action and seeing a result. In communications,
it is the delay between sending a message and
receiving a reply. Often a problem with satellite
connections because the signal must travel huge
distances.
Legacy system Information systems that were
created over several years and are now crucial
to operating the company. They probably
use older technology, and the software is
diffcult to modify. However, replacing them is
diffcult and likely to interfere with day-to-day
operations. Any changes or new systems must
be able to work with the older components.
LIKE An SQL command used within a
WHERE clause to search for patterns in text.
Two pattern-matching characters are used. A
% (* in Access) matches any characters. An _
(? in Access) matches exactly one character.
For example, WHERE LastName LIKE Jo*,
matches any last name beginning with those
two letters.
Limited liability company (LLC) A legal
variation of organizing a company. It protects
the owners with the same separation of funds
offered to corporations, but because it does not
allow it to issue stock, the record keeping is
somewhat easier.
Local area network (LAN) A collection of
personal computers within a small geographical
area, connected by a network. All of the
components are owned or controlled by one
company.
Magnetic hard drives Magnetic hard drives
(or disk drives) consist of rigid platters that
store data with magnetic particles. Data is
accessed by spinning the platters and moving a
drive head across the platters to access various
tracks.
Magnetic ink character recognition (MICR)
A special typeface printed with ink containing
magnetic ink. It can be read rapidly and
reliably by computers. Banks are the primary
users of MICR. Checks are imprinted with
MICR routing numbers. MICR readers are
more accurate than straight OCR because
they pick up a stronger signal from magnetic
particles in the ink.
MailfltersPrograms that automatically
read e-mail and sort the messages according
to whatever criteria the manager prefers. Junk
mail can be discarded automatically.
Malware A generic term used to describe
software that does nasty things. It includes
viruses, Trojan Horses, spyware, and so on.
Management informa tion system (MIS)
1017 Glossary
An MIS consists of fve related compo nents:
hardware, software, people, procedures, and
databases. The goal of management informa-
tion systems is to enable managers to make
better decisions by providing quality informa-
tion.
Manufacturing Resource Planning (MRP
II) An integrated approach to manufacturing.
Beginning with the desired production levels,
we work backward to determine the processing
time, materials, and labor needed at each step.
These results generate schedules and inventory
needs. Sometimes known as a demand-pull
system.
Market basket analysis A data mining
technique pioneered to see if two items are
commonly purchased at the same time. Can
also be used to identify any pairs of items that
are associated with each other.
Mass customization The ability to modify the
production line often enough to produce more
variations of the main product. The goal is to
cover virtually all of the niche markets.
Materials requirements planning (MRP)
An early production system, where at each
stage of production, we evaluate the usage of
materials to determine the optimal inventory
levels.
Mathematical model A model that is defned
by mathematical equations. This format is easy
to use for forecasts and for simulation analyses
on the computer. Be careful not to confuse
precision with accuracy. A model might
forecast some value with great precision (e.g.,
15.9371), but the accuracy could be quite less
(e.g., actual values between 12 and 18).
Mebibyte (MiB) An International
Electrotechnical Commission (IEC)
replacement for the term megabyte. Mebibyte
is base 2 (2
20
), megabyte is for base 10 (10
6
).
Media For transmissions, the means of
connecting computers in a network. Common
methods include twisted-pair and coaxial cable;
fber-optic lines; and radio, micro, and infrared
waves.
Media access control (MAC) The network
protocol that governs how data bits are
sent across a connection medium. Almost
always implemented in a LAN card. It is
most commonly noticed when you need to
control security based on physical cardsin
which case you need the MAC address that is
uniquely assigned to every network interface
card.
Megabyte Loosely, 1 million bytes of data.
Technically, it is 1,048,576 bytes of data, which
is 2 raised to the 20th power or 1024*1024. The
next step up is gigabyte.
MegafopsMillions of foating-point
operations per second. A measure of the
processor speed, it counts the number of
common arithmetical operations that can be
performed in one second.
Megahertz One million cycles per second, a
measure of the clock chip in a computer, which
establishes how fast a processor can operate.
Menu tree A graphical depiction of the menu
choices available to users in a system.
Metadata Describes the source data, and
the transformation and integration steps, and
defnes the way the database or data warehouse
is organized.
Methods Descriptions of actions that an
object can perform. For example, an employee
object could be hired, promoted, or released.
Each of these functions would necessitate
changes in the employee attributes and in other
objects. The methods carry out these changes.
Microsecond One-millionth of a second.
Few computer components are measured in
microseconds, but some electrical devices
and controllers operate in that range. One
microsecond compared to one second is the
same as comparing one second to 11.6 days.
Million instructions per second (MIPS) A
measure of computer processor speed. Higher
numbers represent a faster processor. However,
different brands of processors use different
instruction sets, so numbers are not always
comparable.
Millisecond One-thousandth of a second.
Disk drives and some other input and output
devices perform operations measured in
milliseconds. One millisecond compared to
one second is the same as comparing 1 second
to 16.7 minutes.
Mirror drive A backup system where data is
automatically written to a second disk drive.
If the primary drive fails, operations can be
1018 Glossary
switched instantaneously to the mirror drive.
Model A simplifed, abstract representation of
some real-world system. Some models can be
written as mathe mati cal equa tions or graphs;
others are sub jec tive de scriptions. Models
help managers visualize physical objects and
business processes. Information systems help
you build models, evaluate them, and organize
and display the output.
Modem Modulator-demodulator. A device
that converts computer signals into sounds that
can be transmitted (and received) across phone
lines.
Morphing Digital conversion of one image
into another. The term is an abbreviation of
metamorphosis. True morphing is done with
digital video sequences, where the computer
modifes each frame until the image converts to
a new form.
Motherboard The main board in a computer
that contains sockets for the process and
RAM. It also contains an interface bus so that
interface cards can be added to the system.
Multimedia The combination of the four
basic data types: text, sound, video, and
images (animation). In its broadest defnition,
multimedia encompasses virtually any
combination of data types. Today, it typically
refers to the use of sound, text, and video clips
in digitized form that are controlled by the
computer user.
Multitasking A feature of operating systems
that enables you to run more than one task
or application at the same time. Technically,
they do not run at exactly the same time. The
processor divides its time and works on several
tasks at once.
Musical Instrument Data Interchange
(MIDI) A collection of standards that defne
how musical instruments communicate with
each other. Sounds are stored by musical
notation and are re-created by synthesizers that
play the notes.
Nanosecond One-billionth of a second.
Computer processors and memory chips
operate at times measured in nanoseconds. One
nanosecond compared to 1 second is the same
as comparing 1 second to 31.7 years.
Natural language A human language used for
communication with other humans, as opposed
to a computer programming language or some
other artifcial language created for limited
communication.
Near-feldcommunication(NFC) A very
short range wireless communication method
useful for touchless payments because the
short range makes it diffcult to intercept
the transmission or misidentify the sender.
Typically a range of a few centimeters with a
relatively low bandwidth. Similar to RFID, but
with a shorter range.
Network A set of items connected together. In
MIS, it is typically a connection of computers.
And social networks are connections of people.
Network address translation (NAT) A
network confguration where internal
computers use non-routable addresses (usually
in the 10.0.0.0 range). When connecting to
devices on the Internet, the boundary router
temporarily assigns a real IP address and then
directs the incoming messages to the original
computer by changing the address within the
packets.
Network attached storage (NAS) A disk
drive unit that stands alone and is connected
to the high-speed local area network instead of
directly to a single computer. Similar to a SAN
but it uses standard network connections and
is accessible to any computer attached to the
network.
Network effect The concept that a network
becomes more valuable and useful as the
number of participants increases. In social
networks, larger networks attract more people,
enabling them to grow even larger.
Network interface card (NIC) The
communication card that plugs into a computer
and attaches to the network communication
medium. It translates computer commands into
network messages and server commands.
Network operating system (NOS) A special
operating system installed on a fle server,
with portions loaded to the client machines.
It enables the machines to communicate and
share fles.
Network service provider (NSP) A high-
level Internet service provider offering
connections to ISPs. The NSP leases high-
speed, high-capacity lines to handle the
communication traffc from hundreds of ISPs.
1019 Glossary
Neural network A collection of artifcial
neurons loosely designed to mimic the way the
human brain operates. Especially useful for
tasks that involve pattern recognition.
Neuron The fundamental cell of human brains
and nerves. Each of these cells is rela tively
simple, but there are approximately 100 million
of them.
Newsgroups A set of electronic bulletin
boards available on the Internet. Postings are
continuously circulated around the network as
people add comments.
Nondisclosure agreement (NDA) A written
agreement where the signer agrees to keep
certain information confdential and not tell
anyone. Commonly used by startup companies
to keep basic technology, general operating
practices, and marketing plans secret.
Normalization A set of rules for creating
tables in a relational database. The primary
rules are that there can be no repeating
elements and every nonkey column must
depend on the whole key and nothing but the
key. Roughly, it means that each table should
refer to only one object or con cept.
Numbers One of the basic data types, similar
to text on input and output. Attributes include
precision and a scaling factor that defnes the
true size or dimension of the number.
Object A software description of some entity.
It consists of attributes that describe the object,
and functions (or methods) that describe the
actions that can be taken by the object. Objects
are generally related to other objects through an
object hierarchy.
Object hierarchy Objects are defned from
other base objects. The new objects inherit the
properties and functions of the prior objects.
Object Linking and Embedding (OLE) A
standard created by Microsoft for its Windows
operating system to create compound
documents and dynamically link data objects
from multiple software packages. You begin
with a compound document or container that
holds data from other software packages.
These data objects can be edited directly
(embedded). Most OLE software also supports
dynamic linking.
Object orientation An approach to systems
and programming that classifes data as various
objects. Objects have attributes or properties
that can be set by the programmer or by users.
Objects also have methods or functions that
defne the actions they can take. Objects can be
defned from other objects, so most are derived
from the four basic data types.
Object-oriented DBMS A database system
specifcally created to hold custom objects.
Generally supports developer-defned data
types and hierarchical relationships.
Object-oriented design The ultimate goal
of the ob ject-oriented approach is to build a
set of reus able objects and pro cedures. The
idea is that eventu ally, it should be possible to
create new sys tems or modify old ones simply
by plugging in a new module or modifying an
existing object.
Object-oriented programming (OOP) The
process of writing software using sets of
extensible objects. Programmers frst create
objects that encapsulate internal data structures
with software methods. New objects can be
created by inheriting properties and methods
from more generic classes. A goal of OOP was
to encourage reuse of objects to reduce the time
it takes to create new applications.
Offshoring The practice of sending jobs to an
outside contractor located in a different country.
One-to-many relationship Some object
or task that can be repeated. For instance, a
customer can place many orders. In database
normalization, we search for one-to-many
relationships and split them into two tables.
Online analytical processing (OLAP) A
computer system designed to help managers
retrieve and analyze data. The systems are
optimized to rapidly integrate and retrieve data.
The storage system is generally incompatible
with transaction processing, so it is stored in a
data warehouse.
Open operating system An operating system
that is supposed to be vendor neutral. It should
run on hardware from several different vendors.
When a buyer upgrades to a new machine, the
operating system and software should function
the same as before.
Online transaction processing (OLTP) A
computer system designed to handle daily
transactions. It is optimized to record and
protect multiple transactions. Because it is
1020 Glossary
generally not compatible with managerial
retrieval of data, data is extracted from these
systems into a data warehouse.
Open source development A method of
creating software where the source code
is released to the public and anyone can
contribute to the project by writing sections of
the code. Usually one person takes the lead to
control the integration changes and planning
for new releases. The Linux operating system
initiated by Linus Torvalds is a common
example.
Open system An open system learns by
altering itself as the environment changes.
Operating system A basic collection of
software that handles jobs common to all
users and programmers. It is responsible for
connecting the hardware devices, such as
terminals, disk drives, and printers. It also
provides the environment for other software,
as well as the user interface that affects how
people use the machine.
Operations level Day-to-day operations and
decisions. In a manufac turing frm, machine
settings, worker schedules, and maintenance
re quire ments would rep resent management
decisions at the opera tions level. Infor ma tion
systems are used at this level to collect data and
perform well-defned com puta tions.
Optical character recognition (OCR)
The ability to convert images of characters
(bitmaps) into computer text that can be stored,
searched, and edited. Software examines a
picture and looks for text. The software checks
each line, deciphers one character at a time, and
stores the result as text.
Optimization The use of models to search for
the best solutions: minimizing costs, improving
effciency, or increasing profts.
Output devices Data stored in binary form
on the computer must be converted to a format
people understand. Output devicesfor
example, display screens, printers, and
synthesizersmake the conversion.
Outsourcing The act of transferring
ownership or management of MIS resources
(hardware, software and personnel) to an
outside MIS specialist.
Packets Network messages are split into
packets for transmission. Each packet contains
a destination and source address as well as a
portion of the message.
Packet switching network A
communications protocol in which each
message is placed into smaller packets. These
packets contain a destination and source
address. The packets are switched (or routed)
to the appropriate computer. With high-speed
switches, this protocol offers speeds in excess
of 150 megabits per second.
Page footer Data that are placed at the bottom
of each page in a report. Common items
include page totals and page numbers.
Page header Data that is placed at the top of
every page in a report. Common items include
the report title, date, and column labels.
Parallel processing Using several processors
in the same computer. Each processor can be
assigned different tasks, or jobs can be split into
separate pieces and given to each processor.
There are a few massively parallel machines
that utilize several thousand processors.
Parameter Variables in a model that can be
controlled or set by managers. They are used
to examine different situations or to tailor the
model to ft a specifc problem.
Patent Legal protection for products (and
sometimes business processes). It grants the
owner sole right to sell or create modifcations
of the product for 20 years. No one can create
the same product unless approved by the patent
owner.
Peer-to-peer communication A method of
sharing data and information directly with
colleagues and peers, instead of transferring
data through a shared central server.
Peer-to-peer network A network
confguration in which each machine is
considered to be an equal. Messages and
data are shared directly between individual
computers. Each machine continuously
operates as both a client and a server.
Personal digital assistant (PDA) A small,
portable handheld computer designed primarily
to handle contacts, schedules, e-mail, and
short notes. Some models have more advanced
features to support documents, spreadsheets,
photos, and music. A few have wireless
connections; others have to be synchronized
with desktops to transfer e-mail and update
1021 Glossary
schedules. Replaced by smartphones.
Pebibyte The new IEC defnition for
quadrillion bytes in binary (2
50
) denoted
PiB. It replaces the term petabyte which
is now to be used for decimal values.
Technically 2 raised to the 50th power or
1024*1024*1024*1024*1024 (5 times).
Petabyte One quadrillion bytes of data. One
step above terabyte and one below exabyte. The
new IEC defnition uses pebibyte for binary
data and petabyte for decimal values or 10
15
.
Phased implementation An implementation
method that introduces the new system in
phases or steps. One phase is completed before
the next is undertaken. The pieces could be
software components, different divisions,
different locations, or a similar split.
Phishing Pronounced as fshing. The act
of sending out false messages, typically
pretending to be from a bank, in an attempt to
get users to provide usernames and passwords
to access sensitive systems. Almost any e-mail
message purportedly sent to you by a fnancial
institution should be ignored. Anything that
does not include your name should be deleted
immediately.
Photo-CD A standardized system created
by Kodak to convert photographs to digital
(bitmap) form and store them on optical disks.
Pivot table A tool within Microsoft Excel
used to extract and organize data. It enables
users to examine aggregated data and quickly
see the accompanying detail.
Pixel Picture element, or a single dot on an
image or video screen.
Podcast An audio message distributed via
a Web site designed for storage and playback
on an Apple iPod. But the term today includes
almost any type of audio fle containing
messages.
Point of sale (POS) system A means of
collecting data immediately when items are
sold. Cash registers are actually data terminals
that look up prices and instantly transmit sales
data to a central computer.
Polymorphism In an object design, different
objects can have methods that have the same
name but operate slightly differently. For
example, a checking account object and a
savings account object could each have a
method called pay interest. The checking
account might pay interest monthly, whereas
the savings account pays it quarterly.
Portable document format (PDF) A fle
format often used on the Internet. It can display
documents with detailed precision, including
special fonts and shading. Defned by Adobe,
readers are freely available for many machines.
Special software must be purchased to create
the fles.
Precision (numeric) In computers, numeric
precision represents the number of digits stored
to the right of the decimal point. So, 10.1234
is more precise than 10.12; however, it is not
necessarily more accurate. The original value
might not have been measured beyond two
digits.
Prediction Model parameters can be
estimated from prior data. Sample data is used
to forecast future changes based on the model.
Pretty good privacy (PGP) A dual-key
encryption system based on the Diffe-Hellman
approach similar to RSA. Created by Philip
Zimmermann and commonly used to encrypt
e-mail. Free copies for noncommercial use are
still available from MIT.
Primary key A column or set of columns that
contains data to uniquely identify each row
in a relational database table. For example,
each customer must have a unique identifer,
possibly a phone number or an internally
generated customer number.
Privacy (1) The concept that people should
be able to go about their lives without constant
surveillance, that personal information about
people should not be shared without their
permission. (2) Collecting personal data only
when you have a legitimate use for it, allowing
customers to correct and remove personal
data. Protecting confdential data so that it is
not released to anyone. Giving customers the
option so that you do not sell or lease their
personal data.
Private key In a dual-key encryption system,
the key that is protected by the owner and never
revealed. It is generally a very large number.
Problem boundary The line that identifes
the primary components of the system that are
creating a specifc problem. Subsystems inside
1022 Glossary
the boundary can be modifed to solve the
problem or enhance the system. Subsystems
outside the boundary cannot be altered at this
time.
Procedures Instructions that help people use
the systems. They include items such as user
manuals, documentation, and procedures to
ensure that backups are made regularly.
Process An activity that is part of a data fow
diagram. Systems can be built to process goods
or to process data. Most information system
work focuses on processes that alter data.
Process control The use of computers to
moni tor and control the produc tion machines
and ro bots. Pro duction lines general ly use
many dif fer ent machines, each requir ing
several adjust ments or settings. Computer
control simplifes and speeds the setup.
Process innovation Evaluating the entire
frm to improve individual processes, and to
search for integrated solutions that will reduce
costs, improve quality or boost sales to gain a
competitive advantage. See also reengineering.
Processor The heart of a computer. It carries
out the instructions of the operating system and
the application programs.
Product differentiation The ability to make
your products appear different from those of
your rivals, thus attracting more customers.
Information systems have been used to alter
products and provide new services.
Program logic Writing program code requires
defning the steps or logic that the computer
should follow to complete a task. A program
must also use the correct words, symbols, and
punctuation, known as syntax.
Properties See attributes.
Protect document A method of restricting
changes to Microsoft Offce fles. A limited
version of information rights management that
will allow people to read a document but not
make changes.
Protocols A set of defnitions and standards
that establish the communication links on a
network. Networks are often classifed by their
choice of protocol. Common protocols include
Ethernet, Token Ring, and TCP/IP.
Prototyping An iterative system design tech-
nique that takes advantage of high-level tools
to rapidly create working systems. The main
objec tive of proto typing is to create a work ing
version of the system as quickly as possi ble,
even if some compo nents are not in cluded in
the early ver sions.
Pseudocode A loosely structured method to
describe the logic of a program or outline a
system. It uses basic programming techniques
but ignores issues of syntax and relies on verbal
descriptions.
Public key In a dual-key encryption system,
the key that is given to the public. Each person
wishing to use dual-key encryption must have
a different public key. The key works only in
tandem with the users private key.
Pure Internet plays Dot-com frms that have
no direct tie to traditional business. Firms that
make all their revenue from Internet sales or
other Internet frms. A popular concept in 1999,
but most pure Internet frms failed in 2000 and
2001.
Query by example (QBE) A visual method
of examining data stored in a relational
database. You ask questions and examine the
data by pointing to tables on the screen and
flling in templates.
Query system A method of retrieving data in
a DBMS. It generally uses a formal process to
pose the questions (1) what columns should be
displayed? (2) what conditions are given? (3)
what tables are involved? and (4) how are the
tables connected? See query by example and
SQL.
Radiofrequencyidentifcation(RFID)
Small, passive computer chips that are powered
by radio waves. When triggered by a reader,
the chip returns data stored in its memory by
modulating the radio signals. Readable range is
limited to a few feet or less. If price drops far
enough, they might replace bar codes.
Random access memory (RAM) High-speed
memory chips that hold data for immediate
processing. On most computers, data held in
RAM is lost when the power is removed, so
data must be moved to secondary storage.
Rapid application development (RAD)
The goal of building a system much faster
than with traditional SDLC methods. Using
powerful tools (database management system,
1023 Glossary
high-level languages, graphical toolkits, and
objects), highly trained programmers can build
systems in a matter of weeks or months. Using
workgroups, communication networks, and
CASE tools, small teams can speed up the
development and design steps.
Read Only Memory (ROM) A type of
memory on which data can be stored only one
time. It can be read as often as needed but
cannot be changed. ROM keeps its data when
power is removed, so it is used to hold certain
core programs and system data that is rarely
changed.
Really simple syndication (RSS) A
technique used with blogs to automatically
push new versions down to subscribers. Users
can confgure a Web browser to connect
to a favorite blogger. New versions are
automatically displayed on the browser.
Reduced instruction set computer (RISC)
When designing a RISC processor, the
manufacturer deliberately limits the number
of circuits and instructions on the chip. The
goal is to create a processor that performs a
few simple tasks very fast. More complex
problems are solved in software. Because
RISC processors require fewer circuits, they are
easier to produce.
Redundant array of independent Disks
(RAID) A system consisting of several
smaller drives instead of one large drive. Large
fles are split into pieces stored on several
different physical drives. The data pieces
can be duplicated and stored in more than
one location for backup. RAID systems also
provide faster access to the data, because each
of the drives can be searching through their part
of the fle at the same time.
Reengineering A complete reorganization
of a company. Beginning from scratch, you
identify goals along with the most effcient
means of attaining those goals, and create new
processes that change the company to meet
the new goals. The term reengineering and its
current usage were made popular in 1990 by
management consultants James Champy and
Michael Hammer.
Relational database A database in
which all data is stored in fat tables that
meet the normalization rules. Tables are
logically connected by matching columns of
data. System datasuch as access rights,
descriptions, and data defnitionare also
stored in tables.
Repetitive stress injury (RSI) An injury
that occurs from repeating a stressful
action. For instance, several people have
complained that constant typing damages their
wrists. Ergonomic design, adjusting your
work space, and taking breaks are common
recommendations to avoid repetitive stress.
Replay attack If an attacker captures a set
of network transmissions, the attacker could
replay those same messages by sending them
again to gain access to a server or duplicate a
fnancial transaction.
Replication The intentional process of
duplicating data in a database so that it can be
transported and accessed in multiple locations.
The DBMS has the ability to synchronize data
changes between the master copy and any
replicas.
Report A printed summary or screen
display that is produced on a regular basis by
a database management system. The main
sections of a report are report header, page
header, group/break header, detail, group/break
footer, page footer, and report footer.
Request for proposal (RFP) A list of
specifcations and questions sent to vendors
asking them to propose (sell) a product that
might fll those needs.
Resolution The number of dots or pixels
displayed per inch of horizontal or vertical
space. Input and output devices, as well
as images and video, are measured by their
resolution. Higher values of dots per inch yield
more detailed images.
Reusability The ultimate goal of object-
oriented systems. By defning an object up
front and storing it in a repository the object
can be used in many applications, instead of
requiring developers to recreate it every time a
new system is built.
Reverse engineering The process of taking
older software and rewriting it to modernize
it and make it easier to modify and enhance.
Reverse engineering tools consist of software
that reads the program code from the original
soft ware and converts it to a form that is easier
to modify.
1024 Glossary
RGB Red-green-blue. A color scheme used
for video displays. Colors are specifed by
identifying the desired strength of each primary
color. In most implementations the color value
ranges from 0 to 255 (one byte). See also
CMYK.
Rivals Any group of frms that are
competing for customers and sales. Similar
to competitors, but competition carries an
economic defnition involving many frms.
Even an industry with two frms can experience
rivalry.
Rivest-Shamir-Adelman (RSA) Three
mathematicians who developed and patented
a dual-key encryption system. The term
often refers to the encryption technique. It
is based on the computational diffculty of
factoring very large numbers into their prime
components.
Rocket scientists Mathematically trained
fnancial analysts who build complex
mathematical models of the stock market and
help create and price new securities.
Router A communication device that connects
subnetworks together. Local messages remain
within each subnetwork. Messages between
sub-networks are sent to the proper location
through the router.
Row A horizontal element that contains all
of the data to describe an entity or object in a
relational database or spreadsheet.
Rules A set of conditions that describe a
problem or a potential response. Generally
expressed as If Then conditions. Used by
expert systems to analyze new problems and
suggest alternatives.
Sampler An input device that reads electrical
signals from a microphone and stores the sound
as a collection of numbers. It measures the
frequency and amplitude of the sound waves
thousands of times per second.
Scalability The ability to buy a faster
computer as needed and transfer all software
and data without modifcation. True scalability
enables users to buy a smaller computer today
and upgrade later without incurring huge
conversion costs.
Scope creep The process in any project where
people keep trying to add new features to the
project. An easy way to drive a project out of
control. As the number of features added begins
to exceed the original plan, the costs increase
and the project is delayed.
Script kiddie A lazy attacker who downloads
code from the Internet to attempt to fnd and
exploit known holes. Can be stopped by
ensuring your software contains all current
patches.
Scrolling region On a data entry form, a
subform or section that is designed to collect
multiple rows of data. Much like a spreadsheet,
the user can move back and forth to alter or
examine prior entries.
Search engine optimization A practice of
making a Web site more compatible with
search engines. The goal is to make the page
appear higher in the list of results when users
search for the desired topic. Most good SEO
techniques defne the page precisely so that it
matches exactly how a person perceives the
page. Black hat or bad SEO methods try to trick
the search enginewhich usually results in the
site being blacklisted by the search engine.
Secondary storage Data storage devices that
hold data even if they lose power. Typically
cheaper than RAM, but slower. Disk drives are
common secondary storage devices.
Secure sockets layer (SSL) A system that
provides encryption for Internet transmissions.
Commonly used to establish a secure
connection between client browsers and
e-commerce servers. It is established with dual-
key encryption by installing a digital security
certifcate on the server.
Serifs The small lines, curlicues, and
ornamentation on many typefaces. They
generally make it easier for people to read
words and sentences on printed output. Sans
serif typefaces have more white space between
characters and are often used for signs and
displays that must be read from a longer
distance.
Server farm A collection of dozens or
hundreds of smaller servers. Software allocates
tasks to whichever server is the least busy. This
approach to scalability is fault-tolerant and easy
to expand, but can be diffcult to manage.
Service level agreement (SLA) A formal
written agreement between a user group and
a service provider that specifes guaranteed
1025 Glossary
levels of service and compensation for failure
to meet those levels. SLAs are commonly used
in outsourcing deals to ensure the contracted
party is providing adequate levels of service,
particularly with network providers.
SharePoint Microsofts Web-based tool for
teamwork. It supports fle sharing, version
control, discussion groups, and surveys.
Sign-off In a systems development life-cycle
approach, the approval that managers must give
to forms, reports, and computations at various
stages of the development. This approval
is given when they sign the appropriate
documents.
Simple object access protocol (SOAP) A
standard, easy-to-implement method of
exchanging information and messages among
different computers on the Internet. A protocol
that works with XML to support Web-based
services.
Simulation Models are used to examine what
might happen if we decide to make changes to
the process, to see how the system will react to
external events, or to examine relationships in
more detail.
Single sign-on A comprehensive security
authentication system so that users can log in
(sign on) one time. Once the users identity has
been established, all applications obtain the
credentials from a central server to recognize
the user and determine access rights.
Social engineering A method used by
attackers to obtain usernames and passwords
to obtain illegal access to a system. An attacker
might call a user and pretend to be a system
administrator asking for confrmation of a
password. Relatively easy to stop by never
telling your password to anyone. Systems
administrators will never need your password.
Social legitimacy At one time, mainstream
organizations were identifed by the quality
of their presentation and their image. Large
frms spend millions of dollars on graphic
artists, professional designers, and professional
printing. The decreasing cost of computers
enables even small organizations to create an
image that is hard to distinguish from large
organizations.
Social networking Contacts with friends
and businesspeople. Many Web sites such as
Facebook, Flickr, and YouTube were built to
support social networkingenabling users to
fnd each other and share information online.
Software A collection of computer programs
that are algorithms or logical statements that
control the hardware.
Software as a service (SaaS) Selling software
for usage or monthly fees instead of charging a
fxed upfront price. Easiest to implement with
cloud computing but it can also be done with
in-house software. Upgrades and maintenance
are typically included in the price and
sometimes the maintenance work is handled by
the developing company.
Software maintenance The act of fxing
problems, altering reports, or extending an
existing system to improve it. It refers to
changes in the software, not to hardware tasks
such as cleaning printers.
Software piracy The act of copying software
without paying the copyright owner. With few
exceptions (e.g., backup), copying software
is illegal. Companies and individuals who
are caught have to pay thousands of dollars in
penalties and risk going to jail. It is commonly
accepted that piracy takes money away from
the development of improved software.
Software suites Collections of software
packages that are designed to operate together.
Theoretically, data from each package can be
easily shared with data from the others. So
word processors can incorporate graphics,
and spreadsheets can retrieve data from the
database management system. Suites are often
sold at a substantial discount compared to
buying each package separately.
Solid state drive A disk drive replacement
or data storage mechanism that uses electronic
components to store data instead of a movable
platter. Most SSDs are similar to fash drives
but with greater capacity and faster data
transfer speeds.
Sound One of the basic data types. There are
two methods to describe sound: samples or
MIDI. Digitized (sampled) sound is based on
a specifed sampling and playback rate, and fts
into fre quency and amplitude (volume) ranges.
Spam Unsolicited commercial e-mail, or junk
mail. Unwanted messages sent by commercial
entities or hackers trying to steal your system
1026 Glossary
or your money. It makes up over 50 percent
of e-mail traffc. Most nations have made it
illegal, but it is hard to stop. The name refers
to a Hormel meat product, but its use is often
attributed to a Monty Python sketch.
Speech recognition The ability of a computer
to capture spoken words, convert them into
text, and then take some action based on the
command.
Spyware Software that stealthily installs
itself on your computer, records your activities
or keystrokes. Commonly used by attackers
who collect the data to learn your account
passwords. Extremely dangerous because once
installed it can gain total access to your system.
Software scanning tools can spot most common
spyware programs. Windows Vista has other
tools to prevent programs from installing
themselves without your knowledge.
SQL A structured query language supported
by most major database management systems.
The most common command is of the form:
SELECT column list FROM table list JOIN
how tables are related WHERE condition
ORDER BY columns.
Standard operating procedures A set of
procedures that defne how employees and
managers should deal with certain situations.
Standards An agreement that specifes
certain technical defnitions. Standards can
be established by committees or evolve over
time through market pressures. As technology
changes, new standards are created.
Static HTML Simple HTML pages that
are changed only by humans, so they are
rarely changed. Generally used only for the
prepurchase information stage of e-commerce.
Static integration A means of combining data
from two documents. A copy of the original is
placed into the new document. Because it is
static, changes made to the original document
are not automatically updated. See also
dynamic integration.
Statistical quality control (SQC) The
statistical analysis of measurement data to
improve quality. Several statistical calcula-
tions and graphs are used to determine whether
fuctuations are purely random or represent
major changes that need to be corrected.
Stock options A right to purchase a specifc
stock at a given price. Often granted to workers
and managers in start-up companies. If the
company grows rapidly, its stock price should
increase. The option owner can cash in the
options and receive the difference between the
current price and the option price.
Storage area network (SAN) A method of
storing computer data on devices attached to a
high-speed local connection instead of placing
them into each computer. Separating data from
the computer and centralizing it makes it easier
to upgrade, control, and provide backups.
Fiber optic connections are often used for SAN
drives.
Strategic decisions Decisions that involve
changing the overall structure of the frm. They
are long-term decisions and are unstructured.
They represent an attempt to gain a competitive
advantage over your rivals. They are usually
diffcult and risky decisions. MIS support
for strategic decisions typically consists of
gathering, analyzing, and presenting data on
rivals, customers, and suppliers.
Structured decisions Decisions that can be
de fned by a set of rules or procedures. They
can be highly detailed, but they are defned
without resorting to vague defnitions.
Structured walkthrough A review process
in which the objective is to reveal prob lems,
inaccuracies, ambi guities, and omis sions in
the systems de sign before the program code
is fnal ized. The users are pre sented with a
prototype or mockup of the pro posed system.
Subchapter S corporation A legal variation
of a corporation that can be chosen by the
owners. The IRS and some states impose limits
on the type of company that can elect this
option. It avoids the problem of double taxation
by passing income and losses directly to the
owners personal income tax statements.
Supply chain management (SCM)
Organizing the entire supply process including
vendor selection, parts management, ordering,
tracking, payment, and quality control.
Switch A network device used to connect
machines. Unlike a router, a switch creates a
virtual circuit that is used by a single machine
at a time.
Switching costs The costs incurred in
creating a similar information system when a
1027 Glossary
customer switches to a rival frm. Information
technology creates switching costs because
customers would have to convert data, re-create
reports, and retrain users.
Synchronization A method of sending data
from multiple computers to provide up-to-
date data on both computers. Data changes
are sent to each computer participating in
the synchronization process to ensure each
participant has the same set of data.
Syntax The set of command words,
symbols, and punctuation used by a computer
programming language. When writing
programs, you must type the exact words and
symbols so the computer understands what you
want it to do. See also programming logic.
Synthesizer An electronic device to convert
electrical signals into sound. One basic
technique is FM synthesis, which generates
and combines fxed waves to achieve the
desired sound. A newer method combines short
digitized samples of various instruments with
waveforms to create more realistic sounds.
Sysop System operator. Person in charge of
an electronic bulletin board who organizes fles
and controls access and privileges.
System A collection of interrelated objects
that work toward some goal.
Systems analysis and design A refnement of
the scientifc method that is used to analyze and
build information systems.
Systems analyst A common job in MIS.
The analyst is responsible for designing
new systems. Analysts must understand
the business application and be able to
communicate with users. Analysts must
also understand technical specifcations and
programming details.
Systems development life cycle (SDLC) A
formal method of designing and building
information systems. There are fve basic
phases: (1) feasibility and planning, (2) sys-
tems analy sis, (3) systems design, (4) imple-
men tation, and (5) main tenance and review.
T1, T3 An older communication link provided
by phone companies. Used to carry digitized
analog signals, it is being replaced with ISDN
links. T1 refers to a group of 24 voice-grade
lines and can carry 1.544 megabits per second
(Mbps). A T2 trunk line is equivalent to 96
voice circuits providing 6.312 Mbps. T3
provides 44.736 Mbps, and T4 can carry
139,264 Mbps. Services can be leased at any
of these levels, where greater bandwidth carries
higher costs.
Table A method of storing data in a relational
database. Tables contain data for one entity or
object. The columns represent attributes, and
data for each item is stored in a single row.
Each table must have a primary key.
Tactical decisions Tactical decisions typically
involve time frames of less than a year. They
usually result in making relatively major
changes to operations but staying within the
exist ing structure of the orga nization. MIS
support consists of databases, networks,
integration, decision support systems, and
expert systems.
Telnet A method supported on the Internet
that enables users of one computer to log on to
a different computer. Once logged on to the
new system, the user is treated as any other user
on the system.
Table The basic method of storing data in
a DBMS. Each table represents one object
or entity. Relational databases require that
tables be defned following specifc data
normalization rules.
Tebibyte The new IEC defnition for trillion
in base 2 (2
40
). It replaces terabyte which is to
be used for decimal numbers. Denoted with
TiB.
Telepresence A large, high-resolution video
conferencing system that has the goal of
providing a distance meeting as comfortable as
in person.
Template A method of creating objects such
as forms, reports, and Web sites to ensure that
they follow the same format. A designer creates
a template and all objects follow those design
standards.
Terabyte Approximately 1 trillion bytes of
data. Technically, it is 2 to the 40th power or
1024*1024*1024*1024 (4 times). The step
lower is gigabyte, the step above is petabyte.
Text The simplest of the four basic data types,
it also includes numbers. In its most basic
form, text is made up of individual characters,
which are stored in the computer as numbers.
More sophisticated text is described by its
1028 Glossary
typeface, font size, color, and orientation
(rotation).
Thin client Simpler hardware than a full-
blown personal computer, with minimal
software. It is generally used to display
applications running on the server and to accept
input from the user.
Time division multiplexing (TDM) A
method of sharing a communication medium
with multiple users where each computer is
allowed to send data for a specifed amount
of time, then releases it to the next computer.
Typically handled by requiring devices to send
short packets of data.
Token Ring A communications protocol
that describes when each machine can send
messages. A machine can transmit only when
it receives a special message called a token.
When the message is fnished or a time limit
is reached, the token is passed to the next
machine.
Top-down development An approach to
designing and building systems that begins with
an analysis of the entire company and works
down to increasing detail. A complete top-
down approach is usually impossible because it
takes too long to analyze everything. See also
bottom-up development.
Total cost of ownership (TCO) The cost
of purchasing and running a client computer
(personal computer). A highly subjective
number, it typically includes the hardware cost,
the software license fees, maintenance costs,
and training costs.
Total quality management (TQM) A
management doctrine stating that quali ty must
be built into every process and item. Every
step and each person must be dedicated to
producing quality products and services.
Track changes A method in Microsoft Word
that highlights the changes made by each
person. The original author can then choose
to accept or reject each change. A useful
groupware tool when several people need to
cooperate on writing a document.
Transaction-processing system A system
that records and collects data related to
exchanges between two parties. This data forms
the foundation for all other information system
capabilities. MIS support typically consists
of databases, communication networks, and
security controls.
Transborderdatafow(TBDF)The transfer
of data across national boundaries. Some
countries place restrictions on the transfer of
data, especially data that relates to citizens
(and, of course, data related to national
security). Some people have discussed taxing
the fow of data.
Transmission medium The physical method
of connecting communication devices. The
most common media in use are electrical wires,
fber optics, and radio or micro waves.
Triggered rule In an expert system, if a rule
is used in an application, it is said to have been
triggered or fred.
Trojan Horse A special program that hides
inside another program. Eventually, when the
main program is run, the Trojan Horse program
might delete fles, display a message, or copy
data to an external computer.
True color Humans can distinguish about 16
million colors. Devices that can display that
many colors are said to display true color. It
requires the device to use 3 bytes (24 bits) for
each pixel.
Turing test A test proposed by Alan Turing in
which a machine would be judged intelligent
if the software could use conversation to fool
a human into thinking it was talking with a
person instead of a machine.
Turn-key systems Computer application
systems designed for a specifc task that can
be installed easily. Common examples include
applications for specifc types of businesses,
such as a system for physicians or dentists.
Twisted-pair cable Common dual-line
wire. Often packaged as three or four pairs
of wires. The cable can be run for only a
limited distance, and the signal is subject to
interference.
Two-factor authentication A login process
that requires two types of systems for
identifying users. The most common examples
are: (1) username/password, and (2) a generator
card that creates a random number every
minute that is synchronized to the central
computer.
Typeface A defned way to draw a set of
1029 Glossary
text characters. Several thousand typefaces
have been created to meet different artistic
and communication needs. A common
characterization is serif and sans serif
typefaces.
Ultra-wideband (UWB) A wireless
communication protocol that is radically
different from other wireless systems. Designed
for short range, the system uses very low power
but very high frequency bandwidth. The low
power enables it to be invisible to existing
applications, so it can use large amounts of
bandwidth to send multiple bits of data at the
same timetransmitting in parallel form it
can quickly send large fles. Proposed as a
replacement for physical connection cables,
particularly video connectors.
Unicode An international standard that
defnes character sets for every modern (living)
language and many extinct languages (e.g.,
Latin).
Uninterruptable power supply (UPS) A
large battery and special circuitry that provide
a buffer between the computer and the power
supply. It protects the computer from spikes
and brownouts.
Universal description, discovery, and
integration (UDDI) A public Web-based
directory system designed to enable computers
to fnd and use Web services offered by
other companies. For example, someday
your computer could automatically fnd all
companies that can use current exchange rates
to convert prices.
UNIX A popular operating system created by
Bell Labs. It is designed to operate the same
on hardware from several different vendors.
Unfortunately, there are several varieties of
UNIX, and software that operates on one
version often must be modifed to function on
other machines.
Unstable model A model that cannot be
solved for a single solution. The solution
might continually diverge, or it could oscillate
between several alternatives, generally due to
insuffcient or incorrect feedback mechanisms.
Upload To transfer fles from a local computer
(usually a personal computer) to a distant
computer. See also download.
Usenet See newsgroups.
User resistance People often resist change.
Implementation of a new system highlights
this resistance. Managers and developers must
prepare for this resistance and encourage users
to change. Education and training are common
techniques.
Value chain A description of the many steps
involved in creating a product or service.
Each step adds value to the product or service.
Managers need to evaluate the chain to fnd
opportunities to expand the frm and gain more
sales and profts.
Vector image A stored collection of
mathematical equations, representing lines,
circles, and points. These equations can be
rescaled to ft any output device or to any
desired size. Users deal with the base objects,
not the mathematical defnitions.
Venture capital Money offered by specialized
frms to start up companies. Banks rarely
give money to start-ups, so venture capitalists
fnance risky ventures in the hope of high
profts when the company goes public. Many
strings can be attached to the money
including a loss of control.
Version control Software that tracks changes
made to other documents. Often used in
software development to enable developers
to go back to prior version. It is also available
for common business documents and fles. A
limited version is embedded into Microsoft
Word.
Video One of the basic data types. Video
combines the attributes of images and sound.
An important attribute is the frames per second
defnition. U.S. standard video operates at
30 frames per second, movie flms run at 24
frames per second. Digitizing video requires
capturing and playing back the frames at the
appropriate speed.
Videoconference A meeting tool that
transmits images and sound of at least one
participant. Often, video cameras are available
to everyone involved in the conference. High-
end systems enable the participants to control
the cameras.
View A stored query. If you have a complex
query that you have to run every week, you (or
a da ta base specialist) could create the query
and save it as a view with its own name. It is
then treated much like a simple table.
1030 Glossary
Virtual machine (VM) A software-based
defnition of a computer that is stored and
run on top of a physical computer. A single
physical computer can host several VMs.
Physical processors, RAM, disk space, and
network connections are shared among the
VMs. Generally, the physical elements can be
reallocated to whichever VM currently needs
more capacity.
Virtual mall A collection of Web-based
merchants who join together for marketing
purposes. Generally they share a common Web
host and the same commerce server software.
By sharing costs, they can survive without a
huge amount of sales.
Virtual private network (VPN) Software
installed on a company network and on
each client that automatically encrypts all
communications between the two; useful when
workers travel or need to reach the company
servers from home using the Internet.
Virtual reality (VR) Virtual reality describes
computer displays and techniques that are
designed to provide a realistic image to user
senses, including three-dimensional video,
three-dimensional sound, and sensors that
detect user movement that is translated to on-
screen action.
Virus A malicious program that hides inside
another program. As the main program runs,
the virus copies itself into other programs. At
some point, the virus displays a message, shuts
down the machine, or deletes all of the fles.
Visual Basic A modern variation of the
BASIC programming language created by
Microsoft for application programming in
Windows. A variation resides inside many
of the Microsoft applications, enabling
programmers to manipulate and exchange data
among the database, spreadsheet, and word
processor.
Visual table of contents A graphical design
method that shows how modules of a system
are related. Versions of the tech nique are also
used to display menu trees.
Voice mail A messaging system similar
to telephone answering machines but with
additional features like message store and
forward. You can use your computer to send
messages to coworkers. There are tools that
will read e-mail and fax messages over the
phone, so managers can stay in touch while
they are away from the computer.
Voice over Internet protocol (VoIP)
Connecting telephones to the network
and using the Internet to transfer phone
conversationsinstead of traditional phone
lines.
Voice recognition The ability of a computer
to capture spoken words and convert them into
text.
Web 2.0 A second generation of Web sites
dedicated to providing shared services such
as the social networking sites. The term is not
precisely defned and people disagree on what
to include, but it excludes frst-generation text
and EC sales sites.
Webmaster Specialized IS worker who is
responsible for creating, maintaining, and
revising a companys World Wide Web site.
Webmasters use technical and artistic skills to
create sites that attract browsers.
Whois A utility supported by some operating
systems and the network registrars to provide
information about the ownership of domain
names and Internet access connections.
Unfortunately, many people lie (illegally) when
they fll out the required information.
Wide area network (WAN) A network that
is spread across a larger geographic area. In
most cases, parts of the network are outside
the control of a single frm. Long-distance
connections often use public carriers.
WiFi Short for wireless fdelity. An early
protocol for defning wireless connections,
commonly used in homes and businesses. It has
a relatively short range (perhaps 100 meters).
Transfer speeds initially were less than 11 mbps
(using 802.11b), but have increased to 54 mbps
(802.11a and 802.11g), and are potentially up to
100 or 200 mbps (802.11n). The higher speeds
are achieved by using multiple frequencies at
the same time. Actual speeds are considerably
lower (as low as half the rated maximum).
Wiki A Web site designed to enable multiple
people to create and revise content. The most
famous is Wikipedia where anyone can create
encyclopedic entries. The tools can be used for
business teamwork applications.
Wi-Max A wireless technology designed for
relatively high speeds and medium distances
1031 Glossary
up to several kilometers. It can support point-
to-point transmissions to create high-speed
Internet connections to service providers, or
multipoint systems to handle many users at the
same time. Clearwire and Sprint led the way at
installing the technology in the U.S. and joined
forces in 2007.
Window A portion of the computer screen.
You can move each window or change its size.
Windows enable you to display and use several
applications on the screen at one time.
Wisdom A level above knowledge. Wisdom
represents intelligence, or the ability to analyze,
learn, adapt to changing conditions, and create
knowledge.
WorkfowsoftwareA type of groupware that
is designed to automate forms handling and the
fow of data in a company. Forms and reports
are automatically routed to a list of users on the
network. When each person adds comments or
makes changes, it is routed to the next process.
Workstations Computers attached to
a network, designed for individual use.
Typically, personal computers.
World Wide Web (WWW) A frst attempt to
set up an inter nation al database of information.
Web browsers display graphical pages of
information, including pictures. Hypertext
connections enable you to get related
information by clicking highlighted words.
WYSIWYG What you see is what you get.
With a true WYSIWYG system, documents
will look exactly the same on the screen as
they do when printed. In addition to format,
it means that the printer must have the same
typefaces as the video display. Color printers
use a system to match the colors on the
monitor.
Zebibyte The IEC term for binary counting
with the value 270. It replaces zettabyte, which
is to be used for decimal-based counting.
Zero-day attacks Attacks on computers
that are based on faws found in software
that have not yet been patched. The zero-day
means attackers found the faw before or
simultaneously with security researchers.
Zettabyte In the IEC defnition, it is the
decimal base number 1021. It falls above
zettabyte and below yottabyte. The binary
version is called zebibyte.
zShops Amazon.com offers small companies
a relatively inexpensive e-commerce solution
with little or no fxed costs. Useful for small
frms, the system provides marketing, visibility,
and a payment mechanism. Amazon more
commonly refers to Marketplace shops, but
zShops was the original title. By 2010 it was
renamed to the Webstore.
1032 Index
Index
Symbols
32-bit processor, 70
64-bit processor, 70
* asterisk, 232
% percent, 232
? question mark, 232
_ underscore, 232
A
Accenture, 11, 778
access controls, 264, 313
access point, 173
access to technology, 948
accounting, 365, 747
small business, 725
systems, 370
accounting cycle, 373
accounting journal, 371
acronyms, 92
administrator, 314
Adobe Acrobat, 518
Advanced Encryption Stan-
dard (AES), 316
Advanced Research Projects
Agency, 171
advertising, 460
publisher, 463
advertising revenue
e-commerce, 751
advocacy role, 873
agents, 625
agile development, 800
Akamai, 139, 175
Albertsons, 498
alphabet, 72
Amazon, 305, 434, 438, 501,
958
EC2 disruption, 437
Mechanical Turk, 615
relational database service,
266
Simple Storage Service (S3),
479
WebStore, 471
American Airlines, 712
American Hospital Supply, 33
American Institute of Certi-
fed Public Accountants
(AICPA), 323
American National Standards
Institute (ANSI), 166
amplitude, 81
analysis, 105
AND, 226
Android, 806
location tracking, 952
Angel, Jonathan, 102
Angwin, Julia, 952
anonymity, 931
Anonymous, 322
Antheil, George, 162
Anthony, Robert, 27
antitrust laws, 681
antivirus software, 304
application generator, 261
application layer, 171
applications
Web, 450
application service provider
(ASP), 301, 474
application software, 104
Arab Spring, 925
Arista Networks, 143
Arpanet, 171
artifcial intelligence (AI), 578
aspect ratio, 77
asynchronous, 139
AT&T, 152, 182
auction, 439, 470
audit, 322
audit trail, 375
augmented reality, 109
authentication, 319
Automated Data Processing
(ADP), 377, 726
AVG, 233
B
backhaul, 152
back offce, 3
backup, 143
database, 264
Bahree, Magha, 895
Bahree, Megha, 960
bandwidth, 153, 514
bandwidth hogs, 164
bank loan, 610
banner advertising, 462
Barabba, Vincent, 583
bar-code reader application,
431
bar code scanner, 354
barriers to entry, 662
Baxter Healthcare, 33
Begley, Sharon, 587
Berman, Dennis K., 148
Bersin and Associates, 390
Berzon, Alexandra, 943
best practices
programming, 784
Beth Israel Deaconess Medical
Center, 991
BETWEEN, 229, 230
bicycle, 13
Bigda, Carolyn, 726
bill of materials, 379
bill presentation and payment,
460
binary data, 70
biometrics, 312
bit, 70
bitmap image, 75
BitTorrent, 164
Blackberry, 187
blog, 140, 511
Blogpire, 747
Bluetooth, 153
Blu-Ray, 97
Bode, Karl, 139
Boehret, Katherine, 20
bookkeepers, 3
Borders Group, 682
Boulton, Clint, 806
Brat, Ilan, 372
break footer, 260
break header, 260
breaks, 260
bricks-and-mortar, 429
Bristol-Myers Squibb, 285
broadband, 151
world rankings, 139
broadcast, 150
broadcast stream, 175
1033 Index
Brooks, Fred, 779
brute force attack, 316
Bryan-Low, Cassell, 322
Brynjolfsson, Erik, 674
Burger King, 51
business applications, 66
business operations, 7
business plan, 736, 746
business process management
(BPM), 522
business to business (B2B),
8, 431
business-to-consumer (B2C),
430
business to consumers (B2C),
8
Bussey, John, 28
Bustillo, Miguel, 402, 466
Buy Now button, 469
byte, 70
C
cache memory, 86
calculations, 105
calendar, 108, 141, 512, 531
CAN-SPAM Act, 737, 741
capability maturity model inte-
gration (CMMI), 799
capital costs, 678
Capital One, 643
Cardinal Health, 35
CareerCast.com, 7
Carr, Nicholas, 685
catalog management system,
473
CD-ROM, 97
Census Bureau, 599
Center for Neural Decision
Making, 587
Center for Work-Life Policy,
17
centralization, 144
Web servers, 35
centralization of MIS, 883
certifcate authority, 320
certifcations, 864
Chambers, Nick, 580
Champy, James, 26
change agents, 688, 795
changing industry, 679
character set, 73
chargeback, 866
chart of accounts, 372
check in, 519
check number, 216
check out, 519
ChevronTexaco, 911
Childs, Terry, 300
China, 954
Chiquita Brands, 373
Chrysler-Fiat, 422
Cignet, 937
Cisco, 152
Cisco WebEx, 514
Citigroup, 577, 639
clearXchange, 370
click fraud, 466
click-through rate, 461
client, 145
clip art, 107
Clorox, 880
clothing, 435
cloud computing, 35, 111,
185, 301
private, 99
recentralization, 889
risks, 972
clustering, 604
CMYK, 78
coaxial cable, 149
COBOL, 220
Codd, Edgar, 215
code-division multiple access
(CDMA), 163
cold site, 297
collaboration, 137, 506
co-location, 174, 476
color, 78
column, 213
commerce server, 473
commercial off-the-shelf
(COTS), 105, 780
communication, 506, 507
presentation, 107
problems in system design,
805
voice and mail, 108
writing, 107
Communication
SharePoint, 528
Communications Assistance
for Law Enforcement
Act (CALEA), 959
Comodo, 323
Compaq Computer, 26, 443
competition
follows, 679
competitive, 6
competitive advantage, 32,
655
IS techniques, 662
competitor, 654
components, 779
composite key, 215
compression, 82, 83
computer, 145
computer ethics, 963
computer forensics, 325
concurrency, 217
documents, 510
concurrent changes, 534
condition
AND, 226, 229
BETWEEN, 230
comparison operators, 230
Is Null, 231
LIKE, 231
OR, 225, 229
Cone, Edward, 779
confict management, 894
connection methods, 90
consistent decisions, 610
consultant in a box, 608
consultants, 730
content management system,
473
context diagram, 817
continual improvement, 579
continuous quality improve-
ment, 378
contract programming, 779
contract workers, 21
cookies, 921
Cooney, Michael, 792
Co-operative Group, 894
copyright, 936
correlation, 604
Cortera, 744
COUNT, 233
Covisint, 392
1034 Index
Cowan, Lynn, 725
crime
IT employees, 310
laws, 971
criminals
using technology, 959
cryptography, 315
cube browser, 591
Cullison, Alan, 955
cultural issues, 188
customer, 395
customer relationship manage-
ment (CRM), 395
customer support, 627
custom programming, 777
D
data, 5
centralization, 887
data administrator (DA), 871
database, 4, 105
applications, 257
report, 259
database administrator (DBA),
262, 871
database design, 250
database management system
(DBMS), 105, 212, 589
data capture, 353
data collection, 376
data compression, 83
data conversion, 361
data dictionary, 818
data fow diagram, 791
data fow diagram (DFD), 810,
814
data independence, 219
data integrity, 216, 219, 871
data marts, 589
data mining, 603
data store, 816
data type, 69, 215
numbers and text, 72
pictures, 74
size complications, 82
sound, 81
video, 82
data volume, 218
data warehouse, 589
DB2, 220
Dealer Services Corp., 622
decentralization, 16, 19
decentralization of MIS, 883
decision levels, 27
decision process, 14
decisions, 13, 137, 577
decision support system
(DSS), 30, 594
versus expert system, 623
decision tree, 613
DeCSS, 939
default value, 258
Dell Computer, 26, 66, 123
Delta Air Lines, 654, 700
denial of service (DoS), 311,
330
Department of Defense, 356,
599
cyber attack retaliation, 958
dependent variable, 604
detail section, 260
development, 453
speed, 220
development time, 220
device driver, 98
DHL/Deutsche Post, 568
diagnostic problems, 608
diagnostic situation, 31
diagram
system, 814
Diffe and Hellman, 318
digital camera, 93
digital certifcate, 321
digital content, 94, 937
digital dashboard, 399
digital divide, 948
Digital Millennium Copyright
Act (DMCA), 939
digital products, 448
digital rights management
(DRM), 449, 940
digital signature, 320
digital subscriber line (DSL),
173
direct spread spectrum (DSS),
162
DirecTV, 953
disasters, 297
discussion system, 530
disintermediation, 441
disk drive, 96
distributed services, 819
distribution center (DC), 382
distribution channels, 664
divide and conquer, 812
Dix, John, 880
domain
register, 180
domain name system (DNS),
179
dot-coms, 8
dots per inch (dpi), 75
DoubleClick, 462, 923
double-entry accounting, 375,
725
Dow, 396
Dreazen, Yochi J., 148
drill down, 400
Driptech, 26
Dr. Pepper, 445
dual key encryption, 319
Duffy, Jim, 143
Dun & Bradstreet, 744
DVD, 97
dying industries, 30
dynamic host control protocol
(DHCP), 177
E
ears, 493
eBay, 470
e-business, 9, 430, 625
database, 264
e-commerce, 8
entrepreneurship, 748
security, 326
economics
strategy, 682
Eddy, Nathan, 318
Edifact, 362
e-discovery, 108, 139
education, 944
Efrati, Amir, 370, 870
Eggen, Dan, 939
Eggers, Kelly, 390
Egypt, 925
electrical interference, 150
electronic data interchange
(EDI), 354, 360, 665
proprietary, 361
1035 Index
standards, 362
Electronic Data Systems
(EDS), 778
electronic mail (e-mail), 108
Eli Lilly and Company, 211,
279
e-mail, 137, 509
freedom, 949
most popular communica-
tion method, 546
employee
privacy, 924
employees, 299, 376
evaluation, 325
MIS, 888
employment patterns, 25
encryption, 316, 455
dual key, 319
end-user development, 775,
808, 869
engineering, 673
Enron, 402
enterprise resource planning
(ERP), 29, 383, 385,
720
summarizing, 398
idea, 732
implementation, 742
plan, 736
ergonomics, 92
Ethical Bean Coffee Co., 361
ethics, 963
European Union
privacy, 930
evaluation
system, 797
event-driven, 783
EveryDNS, 958
Excel, 593
Solver, 586
exception report, 372
executive information system
(EIS), 399
expert system (ES), 31, 578,
608
extensible business reporting
language (XBRL), 365
extensible markup language
(XML), 364, 394, 820
external agents, 658
external entity, 815
extraction, transformation, and
loading (ETL), 589
extranet, 140
extreme programming (XP),
800
ExxonMobil, 861, 905
F
Fab@Home, 75
Facebook, 767
face recognition, 591
farming, 24
fault tolerance, 798
feasibility study, 789
Federal Aviation Administra-
tion (FAA), 773, 832
Federal Communications
Commission (FCC),
150
Federal Reserve System, 400
FedEx, 506, 558
fber optic cable, 149
fber optic capacity, 148
fnancial reports, 365
FiOS, 150
Firesheep, 309
frewall, 328
frewire, 91
frst-mover advantage, 686
frst normal form, 252
Five Forces model, 32, 658,
717
Flandez, Raymund, 744
font, 74
food, 434
Ford Motor Company, 351,
415
for each, 239
forecast, 595, 738
forensics, 325
Forrester Research, 21
Fottrell, Quentin, 368
FourSquare, 355
Fowler, Geoffrey A., 445, 728
Fox Broadcasting, 448
franchise, 17
freedom of speech, 960
frequency, 81
frequency division multiplex-
ing (FDM), 161
frequency spectrum, 151
Frykman, Peter, 26
full duplex, 514
G
gambling, 369, 961
poker bots, 611
poker shutdown, 943
Garone, Elizabeth, 926
Gateway and Acer, 126
General Agreement on Tariffs
and Trade (GATT), 21
general ledger, 371
generally accepted accounting
practices (GAAP), 403
General Motors, 581
generic TLDs (gTLD), 180
geographic information sys-
tems (GIS), 598
Giffords, Gabrielle, 922
Gittlen, Sandra, 80
Glader, Paul, 13
GlaxoSmithKline, 282
Glazer, Emily, 361
global positioning system
(GPS), 599
global telecommunications,
186
GNU project, 807
Goldman, David, 152, 182
Google, 20, 100
AdSense, 465
Apps, 510
gmail data loss, 887
government requests, 928
keyword ads, 464
Picasa, 591
self-driving cars, 580
software development, 870
v. Kazakhstan, 955
v. Spain, 924
Google Docs, 113, 546
Gorman, Siobhan, 330, 958
government
responsibilities, 964
technology impact, 950
graphics processing unit
(GPU), 102
1036 Index
graphs, 106
grid computing, 90, 144
Grocer, Stephen, 747
Groove, 541
Grossman, Andrew, 387
GROUP BY, 238, 592
dates, 247
group decision support system
(GDSS), 523
groupware, 137, 509
H
Haislip, Barbara, 962
Hamblen, Matt, 94
Hammer, Michael, 26
Harbert, Tam, 310
hardware, 4, 68, 84, 142
administration, 867
centralization, 884
decentralization, 885
Havenstein, Heather, 380
HBGary Federal, 939
HDTV, 78
health care
government database, 220
Health Insurance Portability
and Accountability Act
(HIPAA), 937
Hecht, Jeff, 170
Helft, Miguel, 928
Hellweg, Eric, 525
help desk, 889
hierarchical command struc-
ture, 15
high-bandwidth digital content
protection (HDCP),
941
high-defnition TV (HDTV),
82
holiday sales forecast, 623
Honda Asimo, 622
Hoover, J. Edgar, 928
Horwitt, Elizabeth, 622
hospital, 33
hosting, 467
hot site, 297
housing, 436
hub, 157
human biases, 581
human resources, 375, 596
hypertext markup language
(HTML), 140
I
icon, 74
ideogram, 74
If-Then statement, 782
implementation, 793
India
do not call, 960
programmers, 22
individuals and technology,
918
Industrial Light & Magic, 90
industry salaries, 726
InfoPath, 537
information, 5
information rights manage-
ment (IRM), 521
information technology (IT), 4
information warfare, 953
Infosys Technologies, 895
initial public offering (IPO),
725, 745
INNER JOIN, 241
innovation, 2, 671
input, 71
devices, 92
input form, 257
input, process, output, 811
instant gratifcation, 434
instant messaging (IM), 108,
140
integration, 382, 389, 506
integrity, 375
intellectual property, 448, 936
intelligence
agents, 625
machine, 623
Internal Revenue Service
(IRS), 840
internationalization, 16, 21,
186, 369, 657
data in multiple countries,
973
ERP, 387
notation, 103
postal codes, 252
small business, 731
Web, 478
Web languages, 23
Web users, 24
International Organization for
Standardization (ISO),
546
International Telecommunica-
tions Union (ITU), 166
Internet, 171
Internet 2, 181
Internet address, 177
Internet advertising bureau
(IAB), 463
Internet Assigned Numbers
Authority (IANA), 181
Internet Corporation for
Assigned Names and
Numbers (ICANN),
181
Internet Engineering Task
Force (IETF), 181
Internet layer, 167, 169
Internet service provider
(ISP), 172
Internet standards, 160
intranet, 140
intrusion detection system
(IDS), 329
intrusion prevention system
(IPS), 330
Intuit, 355
inventory, 372
inventory cost, 435
IP address, 516
iPhone, 806
location tracking, 952
IPv4, 169
IPv6, 169, 177
Izzo, Phil, 30, 623
1037 Index
J
J, 148
JCPenneys, 435
jobs
change from technology, 932
jobs in MIS, 863, 873
Jobs, Steven, 941
John Deere, 149
Johnston, Stuart J., 305
JOIN, 241
joint application design (JAD),
804
Joint Photographic Experts
Group (JPEG), 75
JPMorgan Chase, 259
just-in-time (JIT), 380
K
Kanaracus, Chris, 266, 388
Kane, Daniel, 396
Kane, Yukari Iwatani, 666
Keizer, Gregg, 321, 323
Kerberos, 314
key-stroke loggers, 303
keyword advertising, 461
Kindle, 438
KLM Royal Dutch Airlines,
865
knowledge, 5
knowledge base, 612
knowledge engineer, 612
knowledge management
(KM), 525
Konrath, Joe, 728
Kugler, Logan, 539
Kung, Michelle, 953
L
label, 258
Lahart, Justin, 11, 26
Lai, Eric, 546
Lamar, Hedy, 162
Lands End, 436, 493
language
comprehension, 620
last mile, 172
latency, 155
laws
computer, 965
leader, 687
Leegin Creative Leather Prod-
ucts, 445
Leo Burnett, 9
Levine, Alden Mahler, 807
Levitt and Whisler, 384
liability
from technology and data,
949
Light, Joe, 7
LightSquared, 149
LIKE, 231
limited liability company
(LLC), 743
linear programming, 586
Linux, 99, 110
local area network (LAN), 135
logistics, 388, 674
Long Island Rail Road, 387
long-term-evolution (LTE),
173
loops, 782
Luthans, Fred, 13
M
machine vision, 619
magnetic ink character recog-
nition (MICR), 355
magnetic tape, 97
maintenance
software, 796
Major League Baseball, 346
malware, 305
management by methodology,
16, 17
management information
system (MIS), 4
management levels, 27
management support, 676
management techniques, 300
manager, 11
managerial tasks, 11
managing MIS, 861
manufacturing, 24, 359, 379,
673
map, 598
Marin County, 388
market basket analysis, 605
marketing, 594, 674, 740
Massachusetts General, 937
mass customization, 392, 673
mass production, 391
MasterCard, 402
mathematical formula, 584
McAfee, Andrew, 674
McDonalds Corporation, 2,
17, 44
McKesson, 35
McKinsey and Co, 778
McMillan, Robert, 300, 303
McQuestion, Karen, 728
mean time between failure
(MTBF), 100
Mearian, Lucas, 72, 78, 99
media access control (MAC),
169, 177
meetings, 108, 523
megabyte, 83
menu, 261
mergers, 16, 19
Merrill Lynch, 649, 667
messages, 138
metadata, 590
Microsoft, 321, 778
Microsoft Bing, 20
Microsoft Exchange, 512
Microsoft Live, 546
Microsoft Live Meeting, 514
Microsoft Map Point, 603
Microsoft One Note, 541
Microsoft Project, 537
Microsoft SharedView, 542
microwave, 150
middle management, 933
middle managers, 17
Miller, Claire Cain, 400
Miller, Patrick, 106
millisecond, 85
Mims, Christopher, 615
minimum retail prices, 445
Mintzberg, Henry, 11
MIT Center for Digital Busi-
ness, 674
Mitchell, Robert L., 373, 894
mobile commerce, 476
mobile payments, 360
model, 584
monitoring, 323
1038 Index
Moores Law, 87
motherboard, 86
multidimensional cube, 591
multitasking, 217
multi-touch panels, 544
musical instrument data inter-
change (MIDI), 82
N
nanosecond, 85
Napster, 937
NASDAQ, 749
National Aeronautics and
Space Administration
(NASA), 855
National Basketball Associa-
tion, 343
National Football League,
294, 340
natural language, 620
near feld communication
(NFC), 458
near-feld communication
(NFC), 184
NEC Laboratories, 170
negligence, 300
Nestle, 436
NetFlix, 164
NetSuite, 404
network, 133, 451
management, 869
network address translation
(NAT), 178, 515
network-attached storage, 298
network effect, 748
network interface card (NIC),
147
network service provider
(NSP), 172
neural network, 617
Newman, Barry, 92
New York City, 357
nondisclosure agreement
(NDA), 966
nonlinear relationships, 604
normalization, 250
frst normal form (1NF), 252
second normal form (2NF),
254
third normal form (3NF),
254
North American Free Trade
Area (NAFTA), 21
NULL, 229
numbers, 72
Nvidia, 305
O
Obama, Barack, 893
object hierarchy, 786
object orientation, 70
object oriented
simulation, 588
object-oriented design (OOD),
806, 818
object-oriented programming
(OOP), 785
offshoring, 876
Oil Exploration, 913
ONeill, Shane, 510
one-to-many relationship, 251
online analytical processing
(OLAP), 591
Open Document Format
(ODF), 546
open software, 110
open source development, 807
operating system, 97
operations level, 28
operations research, 586
optical character recognition
(OCR), 93
optimization, 585
opt out lists, 929
OR, 225
Oracle, 220
organizational memory, 526
organizational structure, 15
output, 71
output devices, 95
outsourcing, 779, 876
drawbacks, 879
revenue by main frms, 877
Owens & Minor, 35, 133, 198
P
packet, 159
Pac-West Telecomm, 763
page footer, 260
page header, 260
paired programming, 801
Paletta, Damian, 893
Palin, Sarah, 300
parallel Processor, 88
parameters, 584
Partners Health care System,
996
partnership, 301
password, 302, 311
password generator, 312
Pastore, Richard, 663
patches, 302
patent, 936
pattern recognition, 617
payment fees, 457
payment mechanisms, 456
payment methods, 454
payment risks, 455
PayNet, 744
PayPal, 457
payroll, 386, 388
small business, 726
peer-to-peer collaboration, 540
peer-to-peer network, 146
peer-to-peer Network, 515
people, 4
People Express Airlines, 663
Pepitone, Julianna, 436
Pepitone, Julianne, 435
Personal Computer Intercon-
nect (PCI), 91
Peters, Jeremy W., 686
Petz Enterprises Inc., 716, 760
Pfzer, 287
phased implementation, 796
phishing, 306
physical access, 323
physical disabilities, 934
physical layer, 167
piracy, 963
pitch, 81
PivotTable, 592, 600, 727
pixel, 75
plain old telephone system
(POTS), 176
podcast, 141
pointing device, 93
point of sale (POS), 136, 354
police powers, 959
1039 Index
Porter, Michael, 32, 658
potential competitors, 660
precision (numeric), 71
prediction, 584
Pretty Good Privacy (PGP),
319
price competition, 444
primary key, 215
printer, 95, 142
privacy, 357, 396, 466, 918
consumer, 920
employee, 924
European Union, 930
government, 926
laws, 930, 968
protecting, 929
versus security, 332
private key, 318
procedures, 4
process analysis, 810
process control, 29, 354, 359,
386
processor, 86
product differentiation, 667
production, 670. See manufac-
turing
production chain, 440, 733
production costs, 667
productivity, 2, 6, 105
program logic, 781
programmer, 299
differences, 802
ethics, 963
programming
components, 779
custom, 777
extreme, 800
paired, 801
programming languages, 782
project management, 512
property right, 965
protect document, 518
prototyping, 775
Providian/Washington Mutual,
646
public data on the Internet,
926
public key, 318
public key infrastructure
(PKI), 318
purchase stages, 443
pure Internet plays, 749
Q
QBE, 221
Qualcomm, 162
quality management, 669
quality of service (QOS), 160
query, 221
computation, 233
condition, 225
convert business question,
248
join, 241
multiple tables, 241
single table, 222
sort, 227
subtotal, 238
quick response (QR) code, 357
R
Rachel Sterne, 357
radio frequency identifcation
(RFID), 355
ramdom access memory
(RAM)
DDR4, 72
random access memory
(RAM), 85, 88
Random Hacks of Kindness,
807
rapid application development
(RAD), 805
Rational Rose, 778
really simple syndication
(RSS), 141
recentralization
MIS management, 889
Red Cross, 436
redundant array of indepen-
dent drives (RAID), 96
reengineer, 25
regression, 585
Renesys, 925
repetitive stress injury, 92
replay attacks, 313
replication, 541
report, 259
request for proposal (RFP),
393
research, 672, 734
resolution, 75
responsibility and ethics, 962
retail channel, 442
retail sales, 8
reusability, 785
RGB, 78
Rhoads, Christopher, 925
risks, 366, 788
cloud computing, 972
subcontractors, 975
Rivest, Shamir, and Adleman,
318
Roarke, Shawn P., 611
robotics, 621
Roebuck and Co., 493
Roman, David, 924
Rooney, Ben, 322
Rosenberg, Scott, 779
Roth, Thomas, 305
router, 157
row, 213
row-by-row calculations, 238
(Shell), 908
Royal KPN NV, 478
RSA Security, 330
rules, 613
S
Sabre, 708
salaries, 726
salaries in MIS, 874
Salesforce, 266
sales management, 676
sales tax revenue, 466
Samsung, 72
Sanger, Larry, 520
sans-serif font, 74
Sarbanes-Oxley Act, 402
SATA, 86, 91
scalability, 103, 146
scalable vector graphics
(SVG), 76
scanner, 93
Schatz, Amy, 149
Schechner, Sam, 448
scheduling, 108, 512
1040 Index
trains, 619
Scheier, Robert L., 100
scope, 797
scope creep, 788
script kiddies, 302
scrolling region, 259
search engine, 112
search engine optimization
(SEO), 454
search for extraterrestrial intel-
ligence (SETI), 90
seasonal, 595
secondary storage, 95
second normal form, 254
secure sockets layer (SSL),
327
security, 150, 262, 294, 721
access controls, 264
management, 872
Sega, 332
Segal, David, 435
Segall, Laurie, 355, 887
SELECT, 228
separation of duties, 375
serif, 74
server, 102, 145
security, 327
Web hosting, 467
server farm, 146
service industry, 676
service level agreement
(SLA), 879
service-oriented business, 16,
24
service processes, 527
services, 449
Seward, Zachary M., 357
Shankland, Stephen, 177
share, 113
SharePoint
confguration, 537
sharing
data, 135
documents, 517, 540
hardware, 142
software, 144
sharing data
strategic risk, 680
Sharma, Amol, 955
Sherr, Ian, 332
shielded cable, 148
shortcut keys, 106
Sidel, Robin, 360, 370
SIM card, 186
Simonite, Tom, 75
simple object access protocol
(SOAP), 451, 820
simulation, 587
Singel, Ryan, 164
single sign-on, 314
skills in demand, 792, 875
Skype, 176, 516
small business, 16, 716
social engineering, 302, 318
social group legitimacy, 945
social network, 112, 140, 446,
674
Social Security, 926
software, 4, 68, 104, 144
decentralization, 886
development, 869
patches, 302
support, 868
software development
risks, 788
software maintenance, 796
software piracy., 963
solid state drive (SSD), 78, 96
Solver, 586
Sony, 332, 666
sort, 227
sound, 81
Southwest Airlines, 704
specialists, 720
specialization, 16, 17
specialized problems, 608
speech recognition, 93
speedy decisions, 609
Spitzer, Eliot, 380
spreadsheet, 106, 215
spyware, 305
SQL, 221, 227
SQL Server, 220
Stallman, Richard, 807, 808
standardization, 677
standard operating procedure,
17
standards, 140, 166, 315, 870
database, 263
fle format, 545
Starbucks, 59
State Department lottery
mixup, 793
static HTML, 468
static memory, 96
statistical quality control
(SQC), 669
statistics, 584
Stelter, Brian, 591
Stock, Kyle, 17
stock options, 745
Stoltenberg, Jens, 516
storage area network (SAN),
143
strategic decision, 31
strategy, 31, 654
challenges, 683
sustainability, 685
strategy risks, 678
streaming video, 164
structured decisions, 28
style sheet, 79
subchapter S corporation, 743
subcontractors
risks, 975
subtotals, 238, 591
SUM, 233
Sun, 128
Super Valu, 498
supply chain management
(SCM), 361, 390
survey, 532
Sutter Health, 917, 986
Sutter, John D., 591
switch, 157
switching costs, 34, 666
synchronize, 541
synchronous, 139
syntax, 781
synthesizer, 81, 93
system goal, 798, 813
systems analysis, 789, 791
systems analyst, 787
systems design, 792
systems development life
cycle (SDLC), 774
T
table, 213
tablets, 113
1041 Index
tactics, 29
small business, 727
task list, 531
TaxBrain, 760
taxes, 477
TCP/IP, 167
teamwork, 7, 15, 506
teamwork tools, 539
technology level, 729
Technology Toolbox
Assigning Security Permis-
sions, 324
Browsing Data with a Pivot-
Table, 600
Building Forms in Access,
234
Choosing a Search Engine,
12
Choosing Web Server Tech-
nologies, 459
Collaborating with Share-
Point, 529
Creating a Business Plan,
746
Creating Database Reports,
246
Creating Effective Charts,
101
Creating Forms with Info-
Path, 794
Creating Web Pages, 156
Defning E-Mail Rules, 882
Designing an Executive In-
formation System, 381
Encryption, 317
Finding Government Data,
18
Forecasting a Trend, 607
Locating Customers with a
GIS, 675
Managing Projects, 891
Meeting Space and IM, 536
Paying for Transactions, 452
Privacy, 938
Programming a New Func-
tion in Excel, 803
Selecting an ERP System,
374
Sending E-Mail Legally, 737
Transferring Files on the
Internet, 168
Voice Input, 89
Working in a Global Envi-
ronment, 947
Technology Toolbox:
Analyzing Businesses, 684
telecommuting, 934
tax issues, 962
telepresence, 514
template, 78
temporary workers, 16, 20
test cases, 791
testing, 793
Thibodeau, Patrick, 893
thin-client, 147, 890
third normal form, 254
third-party cookies, 467
threats
shared servers, 974
users, 303
threats to information, 295
Thule, 80
Thurm, Scott, 930
Tibken, Shara, 437
time division multiplexing
(TDM), 161
too much technology, 9
top-level domain (TLD), 180
total cost of ownership (TCO),
867
Toyota, 425
Trachtenberg, Jeffrey A., 682
track changes, 518
tracking people, 924
training, 611
transaction, 136
international, 369
risks, 366
transaction processing, 352
transaction processing system,
29
transactions, 724
transborder data fow (TBDF),
187
transfer price, 867
transmission capacity, 155
transmission medium, 147
transportation, 437
Transportation Security Ad-
ministration, 303
transport layer, 170
trap door, 300
travel search, 626
Treat America Ltd., 372
trends, 3, 14, 67, 134, 212
business, 16
decisions, 579
e-commerce, 430
e-commerce sales, 433
managing MIS, 863
memory prices, 88
MIS hardware decades, 881
paper consumption, 108
processors, 87
security, 295
small business, 718
software development, 775
strategy, 656
teamwork, 508
transactions, 353
Troianovski, Anton, 478
Trojan Horse, 303, 304
Turing test, 620
turn-key system, 720
twisted-pair, 148
two-factor authentication, 311
two-step login, 307
Tyco, 402
typeface, 74
types of data, 69
U
ultra-wideband (UWB), 153
Ultra-wideband wireless
(UWB), 91
Unicode, 73
United Parcel Service, 563
United States Postal Service,
571
Univar: Van Waters & Rogers,
205
universal product codes
(UPC), 354
UNIX, 98
updates, 307
USA Today, 686
USB, 91
USB drives, 303
U.S. economy, 25
user
1042 Index
responsibility, 962
user datagram protocol (UDP),
170
U.S. Patent and Trademark
Offce (USPTO), 525
U.S. Postal Service (USPS),
668
V
value chain, 670
vector image, 75
vendors v. consumers, 942
venture capitalist, 8
venture capital (VC), 745
Verizon Wireless, 152
version control, 518
SharePoint, 533
video, 82
video capture, 94
videoconference, 514
view, 247
Vijayan, Jaikumar, 220, 259,
922, 925, 937
virtual machine (VM), 104
virtual private network (VPN),
309
virus, 304
Visa International, 218, 402
vision
machine, 619
visualization
data, 594
Visual Studio, 540
voice mail, 108, 510
voice over IP (VoIP), 176
volume, 81
data, 218
transaction, 353
voting
electronically, 951
Vranica, Suzanne, 9
W
Wakabayashi, Daisuke, 332
Wales, Jimmy, 520
Wal-Mart, 17, 356, 386, 429,
487
warfare
information, 953
Washington Post, 749
Watson system, 621
Web 2.0, 429, 446
Web ads, 460
Web browser, 99
webcam, 176
Web commerce server, 472
Web hosting summary, 474
Web log. See blog
Web site, 79, 140, 511
WebStore, 471
Weiss, Todd R., 749
Weka, 606
Wendys, 54
what if, 587
WHERE, 230
whois, 180
WiFi, 153
Wiki, 141, 520
Wikileaks, 331
WikiLeaks, 368, 958
Wikipedia, 141, 520
Wilder, Clinton, 663
Williams-Sonoma, 930
WiMax, 152
Windows, 99
Wines, Michael, 954
wireless, 152, 183
high speed, 177
wisdom, 5
Woo, Stu, 434
word processor, 107
workfow, 522, 535
workstation, 144
Worldcom/MCI, 402
world-state, 955
World Trade Organization
(WTO), 21
worms, 305
Worthen, Ben, 865
WPP PLC, 9
W.W. Grainger, 201
X
X12 defnition, 362
XML. See extensible markup
language (XML)
Y
Yahoo, 956
Yum! Brands, 56
Z
Zaltman, Gerald, 583
zero-day attacks, 308
Zillow.com, 436
ZIP Code, 930
Zoltek Companies, 13
zoo
design example, 812

MISPost 600

Uploaded by

Copyright:

Available Formats

MISPost 600

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MISPost 600

Uploaded by

Copyright:

Available Formats

Management

Service contracts for hardware.

Consultants for network and troubleshooting.

Assign someone to do backup.

Security is usually forgotten.

Upgrades and decision software are rare.

You might also like