Functions in Detail

SAP Advanced Planner & Optimizer

Demand Planning

Contents

SAP AG

Neurottstrae 16

69190 Walldorf

Germany

iii

Copyright 1999 SAP AG. All rights reserved.

No part of this brochure may be reproduced or transmitted in any form or for any purpose without the express permission of SAP AG. The information contained herein may be changed without prior notice. Some software products marketed by SAP AG and its distributors contain proprietary software components of other software vendors. Microsoft, WINDOWS, NT, EXCEL, Word and SQL-Server are registered trademarks of Microsoft Corporation. IBM, DB2, OS/2, DB2/6000, Parallel Sysplex, MVS/ESA, RS/6000, AIX, S/390, AS/400, OS/390, and OS/400 are registered trademarks of IBM Corporation. OSF/Motif is a registered trademark of Open Software Foundation. ORACLE is a registered trademark of ORACLE Corporation, California, USA. INFORMIX-OnLine for SAP is a registered trademark of Informix Software Incorporated. UNIX, X/Open, OSF/1, and Motif are registered trademarks of SCO Santa Cruz Operation. ADABAS is a registered trademark of Software AG. SAP and SAP-Logo, R/2, R/3, RIVA, ABAP, SAP-EDI, SAP Business Workflow, SAP EarlyWatch, SAP ArchiveLink, ALE/WEB, BAPI, SAPPHIRE, Management Cockpit, SEM, are trademarks or registered trademarks of SAP AG in Germany and in several other countries all over the world. Design: SAP Communications Media

Demand Planning

December 1999

iii

Contents

Demand Planning

1 1 2 3 4 4 5 5 6 6 6 9 9 11 11 11 12 13

Demand Planning: State-of-the-Art Forecasting Introduction Overview of SAP Advanced Planner & Optimizer The SAP Business Framework Benefits of SAP Advanced Planner & Optimizer Key Features of SAP Advanced Planner & Optimizer Demand Planning The SAP APO Demand Planning Architecture An Integrated Demand Planning Solution Demand Planning Features and Functions Data Storage and Representation Planning Environment Consensus-Based Forecasting Multiple Views Causal Analysis Multitier Forecasting Statistical Forecasting Toolbox Forecast Accuracy Analysis Conclusion

iii

You can find this and other current literature on our home page in the media centers for each subject at: http://www.sap.com

iii

Demand Planning

Demand Planning: State-of-the-Art Forecasting

Introduction
Problems and Challenges of the Supply Chain
The hallmarks of todays business environment are volatile demand, decreased customer loyalty, shorter product life cycles, and tougher global competition. To survive, organizations need an information infrastructure that allows them to make accurate decisions in real time and to make customer satisfaction a top priority, while still remaining competitive and profitable. The stakes are high. Miscalculations in forecasting that result in excess inventory can prove fatal. Failing to meet promised delivery dates can drive away customers. To handle these challenges, manufacturers are turning to new, advanced planning and scheduling techniques that generate optimized executable plans in response to rapid changes in supply or demand.

Robust Integration Layer

Because SAP has built a robust integration layer between SAP APO and the underlying execution system, SAP APO can gain immediate and seamless access to Online Transaction Processing (OLTP) business data. While the data objects contained within SAP APO are, in most instances, structurally optimized instances of OLTP data, they remain synchronized through a series of real-time triggers and messaging, a task that is seamlessly accomplished through the integration services of the Business Framework. SAP refers to this technique as semantic synchronization. The SAP APO server also integrates with the SAP Business Information Warehouse using this same mechanism providing unprecedented access to vital business decision data.

Advanced Optimization Techniques and Technology

In addition to highly specialized data objects, SAP APO uses a library of advanced optimization algorithms and a high performance, memory resident data processor to perform planning and optimization. You can configure SAP APO to provide task-specific, industry-specific, and company-specific optimization, automated decision, and real-time event notification to the underlying business processes. This brochure provides an introduction to SAP APO and discusses in depth Demand Planing, one of the SAP APO components.

Data-Driven Process
Huge amounts of data drive these planning and scheduling processes. Much of it comes from the organization itself, but other data comes from outside the organization from suppliers, partners, and even customers. Unlike the data models used by existing Enterprise Resource Planning (ERP) systems, supply chain decision support systems require a new breed of memory resident data model that can handle vast amounts of complex data in real time. Until now, if you wanted an end-to-end solution, you had to integrate specialized software with your existing ERP system and built custom interfaces to handle outside data sources. This can work, but only at an enormously high cost.

The SAP Supply Chain Management Solution

SAP has introduced the Supply Chain Management solution to meet the challenges of managing the entire supply chain from end to end. The SAP Advanced Planner and Optimizer (SAP APO) is an important part. With SAP APO, SAP has combined the ERP execution power of the SAP R/3 System with advanced data analysis and supply chain management tools.
Figure 1: SAP APO Overview

Demand Planning

Overview of SAP Advanced Planner & Optimizer

The SAP Advanced Planner and Optimizer builds on the SAP Business Framework to improve information flow by incorporating real-time collaborative decision support, advanced planning, and optimization into the SAP R/3 System. SAP APO uses a powerful memory resident analytical engine and highly specialized, highly configurable data objects that offer major new components:
n Supply Chain Cockpit n Demand Planning n Supply Network Planning and Deployment n Production Planning and Detailed Scheduling n Global Available-to-Promise

factors that affect demand, delivering context-based demand planning, which raises forecasting to a new level of sophistication and accuracy. Like the Supply Chain Cockpit, Demand Planning uses the Alert Monitor to report exceptions, like orders that exceed forecasts or orders that fall short of forecast and therefore may lead to excess inventory if production is not adjusted accordingly. Using Demand Planning, you can:
n Perform collaborative forecasting

You can collect forecast data from multiple sources and store it in a common repository so planners from marketing, sales, logistics, and even third-party vendors and suppliers can work together on a consensus forecast.
n Manage product life cycles

You can manage the life cycles of your products according to such factors as product supersession, substitution, and cannibalization.
n Plan promotions

Supply Chain Cockpit

Graphical Command Center
The Supply Chain Cockpit component is a graphical instrument panel for modeling, navigating, and controlling all the links in the supply chain. It gives you a complete view of the entire length of the supply chain. Its the command center the cockpit from which you manage your supply chain. Using the Supply Chain Engineer, you can build an elaborate graphical representation of even the most complex supply chain. Once you have built a map of the supply chain, you can select any part of it and zoom in to a detailed level. Using a series of event triggers and alarm conditions, the Alert Monitor can automatically identify problems in the supply chain. It can also monitor material, capacity, transportation, and storage constraints, and it can handle such metrics as delivery performance, cost flow, and throughput.

You can model promotional demand based on profitability goals, product availability, and historical patterns. You can even predict how price increases or decreases will affect future demand.
n Forecast new product demand

You can develop accurate forecasts for new products based on models from similar products, demand histories, and other factors. You can monitor the launch of a new product and the end of a products life using point-of-sale data.
n Perform causal analysis

You can identify and predict how such factors as demographic changes, environmental variables, and social or political factors affect demand for your products. You can analyze actual demand using a variety of tools, such as multiple linear regression, and incorporate causal factors such as price.

Demand Planning
Accurate Forecasting
The Demand Planning component is a toolkit of statistical forecasting techniques and demand planning features that helps you create accurate forecasts and plans. Demand Planning is tightly linked to the SAP Business Information Warehouse, so you can use advanced Online Analytical Processing techniques to drill down to detailed levels of data and analyze historical, planning, and business intelligence. Because it integrates such a wide set of data, Demand Planning gives you a sound understanding of all the

Supply Network Planning and Deployment

Model Your Entire Supply Chain
Using the Supply Network Planning and Deployment component you can develop a model of your entire supply network and all of its constraints. Then, using this model, you can synchronize activities and plan the flow of material along the entire length of the supply chain. This allows you to create feasible plans for purchasing, manufacturing, inventory, and transportation and to closely match supply and demand.

Demand Planning

Drawing on data in liveCache, a high-performance memory-resident technology, and using algorithms, user-developed rules, and policies, the Deployment component helps you dynamically rebalance and optimize your distribution network. It also helps you dynamically determine how and when to distribute inventory. Using the Supply Network Planning and Deployment component, you can:
n Model plans at aggregate and detailed levels n Perform what-if analysis n Dynamically match supply and demand using

Global Available-to-Promise
Match Supply and Demand with Global Available-to-Promise
The Global Available-to-Promise (ATP) component uses a rules-based strategy to ensure you can deliver what you promise to your customers. Global ATP performs multilevel component and capacity checks in real time and in simulation mode to ensure that you can match supply and demand. You can also perform these ATP checks against aggregated, memory-resident data for even better performance. Global ATP maintains simultaneous, immediate access to product availability along the supply chain, so you can be confident that you can meet your delivery commitments. Global ATP draws on a number of criteria to arrive at a commitment, including:
n Product substitution

product substitutions
n Use vendor-managed inventory techniques n Determine the optimum distribution of supply to

meet short-term demand

Production Planning and Detailed Scheduling

Generate Production Plans Rapidly
The Production Planning and Detailed Scheduling component is an integrated set of tools that helps you respond rapidly to changing market conditions. Using this component, you can generate production plans and schedules that optimize resources. The component offers optimization based on state-of-the-art methods, such as the theory of constraints and optimization libraries. Using this component, you can:
n Perform forward and backward scheduling on

If a finished product or component is not available, the system automatically selects a substitute using rules-based selection criteria.
n Selection of alternative locations

As with product substitution, Global ATP can source materials from alternative locations. You can also integrate this logic with the product substitution rules.
n Allocation

You can allocate products or components that are in short supply to customers, markets, orders, and so on. The ATP calculation and response take these allocations into consideration.

multiple levels
n Perform detailed capacity planning and material

The SAP Business Framework

Integrate New Technology and Legacy Systems
SAP APO is a separate SAP solution with its own release cycle. It is a Business Component of the Business Framework, SAPs strategic product architecture, which is designed to facilitate the seamless and rapid integration of new business functions and information technology into existing environments. The Business Framework provides an open architecture, allowing its basic elements, the Business Components, to operate through standardized Business Application Programming Interfaces (BAPIs). Each of the SAP APO components, such as Demand Planning, can be implemented as a stand-alone product or as an integrated part of the Business Framework.

planning simultaneously
n Synchronize schedules and make scheduling

changes at multiple levels of the bill of materials (BOM)

n Use what-if scenarios to simulate actual conditions

and consider the effect of various constraints

n Perform interactive scheduling and plan optimization

using a Gantt chart

n Integrate sales and distribution backorder

scheduling into the manufacturing process

Demand Planning

Benefits of SAP Advanced Planner & Optimizer

SAP APO provides a number of benefits, including the following:
n Completeness

SAP APO Demand Planning is integrated with the R/3 System and the SAP Business Information Warehouse. Data can be automatically transferred between SAP BW and SAP APO Demand Planning. SAP APO Demand Planning uses multidimensional data structures that support many different levels and dimensions. Using tree selection and drill-down capabilities, you can navigate through these multidimensional structures. SAP APO Demand Plannings decision-support tools allows you to perform what-if analysis and online simulation to create, store, and compare different planning versions. In Release 2.0 the Notes Management tool enables the documentation of planning activities by multiple users.

SAP APO supports all of the key supply chain planning and optimization functions and processes traditionally found in stand-alone advanced planning and scheduling solutions.
n Performance

The SAP liveCache memory resident computing technology enables forecasting, planning, and optimization functions to be executed in real time.
n Independence

SAP APO performs planning functions and processes outside of the OLTP system, ensuring greater flexibility and high availability of the SAP APO server.
n Openness

Best-of-Breed Solution
SAP APO Demand Planning embodies all of the advantages of a best-of-breed demand planning solution with the following features:
n Multilevel planning n Life cycle management n Promotion planning n Causal analysis n Multitier forecasting n Phase-in and phase-out forecasting n The Statistical Forecasting Toolbox n Forecast accuracy reporting.

SAP built SAP APO to function in heterogeneous environments. It interoperates with SAP R/3, thirdparty, and legacy OLTP systems.
n Integration

SAP APO is seamlessly integrated with the R/3 System so you can integrate all the links in your supply chain. A robust and sophisticated integration layer facilitates the use of SAP APO with additional optimization and forecasting algorithms.

Key Features of SAP Advanced Planner & Optimizer Demand Planning

A Ready-to-Run Solution
SAP APO Demand Planning is a state-of-the-art, readyto-go demand planning solution. It includes a data mart based on the star schema and an easy-to-use front end with powerful planning and analysis tools. You can implement SAP APO Demand Planning quickly, simply, and cost-effectively.

Internet Communications
Using the SAP APO OpenForecast tool, you can collaborate with your business partners using Internetbased interfaces and technology; share event information, such as promotions and store openings; and compare and resolve variances within forecasts. It supports supply chain partnerships, such as VendorManaged-Inventory (VMI), Efficient Consumer Response (ECR), and Collaborative Planning, Forecasting and Replenishment (CPFR).

Demand Planning

The SAP APO Demand Planning Architecture

SAP APO Demand Planning is composed of three layers:
n Graphical user interface n Planning and analysis engine n Data mart

The graphical user interface contains the planning views. The planning and analysis engine contains the Online Analytical Processing processor, the Business Planning Library, and the Statistical Forecasting Toolbox. The data mart contains infocubes, the time series catalog, and notes. The data mart stores different data types in different database structures using a relational database.

Figure 2: SAP APO Demand Planning Architecture

An Integrated Demand Planning Solution

Seamless Integration
You can directly link SAP APO Demand Planning to SAP R/3, the SAP Business Information Warehouse, and any legacy or external systems for a completely integrated system. This integration provides a continuous planning cycle in which actual data, such as orders and shipments, flows from SAP R/3 to SAP APO Demand Planning, and demand data, such as price information, flows from profit-based promotional planning in SAP APO to SAP R/3.

Easy Data Transfer

The APO Demand Planning data mart uses InfoCubes that have the same data structures as those that are used in the SAP Business Information Warehouse. It is therefore extremely easy to transfer data between the two applications. This integration provides many benefits. For example, you can store syndicated point-of-sale data, such as AC Nielsen data, in the SAP Business Information Warehouse. You then analyze this data in SAP APO Demand Planning and determine which of it should be used in causal analysis. You create consensus forecasts in SAP APO Demand Planning and transfer them to the SAP Business Information Warehouse, where users can work with them to create the reports and ad hoc analyses that drive the business. You can use the SAP Business Information Warehouse to pull data from multiple heterogeneous data sources into the data mart.

Ready Access to Data

The SAP Business Information Warehouse is a product that combines state-of-the-art demand planning techniques and state-of-the-art data warehousing features in one consistent model. Its integration with Demand Planning means accurate decision-support information is always at your fingertips.

Integrate the Sales Force and Forecasting

You can integrate the sales force into the forecasting process by using the SAP Mobile Sales solution on a laptop. You can download account data from the data mart into the laptop and then use SAP Mobile Sales on the laptop to enter sales figures for the accounts offline. When youre finished, you can upload the data back into Demand Planning to update the sales figures in the data mart.
Figure 3: SCM Cycle

Demand Planning

Demand Planning Features and Functions

Data Storage and Representation
Multidimensional Data Storage
SAP APO Demand Planning allows you to view data, plan from many different perspectives, and drill down from one level of detail to the next. To make this possible, SAP APO Demand Planning stores data in a multidimensional form in the SAP APO Demand Planning data mart.

Hierarchies and Networks

Develop Aggregation Hierarchies
The hierarchical relationships of characteristics play an important role in planning and reporting. In SAP APO, you model hierarchies and networks as combinations of characteristic values. For example, products are generally grouped into product family hierarchies. Demand Planning allows you to model a hierarchy over the value domain of one or more characteristics. You can use these hierarchies as the basis for aggregation, disaggregation, and drilling down. They provide the ability to plan and view data at varying levels of detail.

InfoCubes
An InfoCube is a multidimensional data structure and the primary container of data used in planning, analysis, and reporting. InfoCubes contain two types of data: key figures and characteristics. Key figures are quantifiable values, such as sales in units. Characteristics determine the organizational levels at which you do aggregation and reporting. Typical characteristics for sales are products and customers.

Time Series Management

Reusable Time Series
Times Series Management is based on catalogs. In each catalog, you can store time series data with related attributes, such as promotional patterns and life cycles. SAP APO Demand Planning allows you to reuse time series, which saves time and ensures consistency. For example, you can reuse a past promotional pattern to estimate the impact of a future promotion of the same type.

Notes Management
Maintain an Audit Trail
Figure 4: Example of an InfoCube

Incorporate Business Rules

The aggregational behavior of key figures must follow business rules. Key figures, such as sales, might be summed by product and time. SAP APO Demand Planning allows you to model these properties. The Online Analytical Processing processor guarantees that all business rules are met and that the computed views present valid results from a business point of view. The InfoCubes allow you to store multiple key figures, such as orders, shipments, POS, and related forecast components, as well as plans and causal factors. InfoCubes also share master data and descriptive text, which are stored in separate tables.

The Notes Management tool maintains all notes entered by planners. The notes discuss planning data, such as promotions and consensus forecasts. You can use the notes to create an audit trail of all demand planning activities, which is especially helpful when multiple sources and people are involved, such as in consensus forecasting.

Planning Environment
Easy Data Access with Planning Books
You access the information in InfoCubes and Time Series Management using planning books and reports. Using slice-and-dice techniques, you can define how information is displayed in a planning book. Demand Planning provides you with a uniform planning

Demand Planning

environment that supports full multilevel planning with top-down, middle-out, and bottom-up reconciliation strategies. Within this planning environment, you can use a wide range of demand planning and forecasting techniques.

Planning Books
Easy-to-Use Planning Books
A planning book is an easy-to-use tree control for selecting data and a frame with a grid and a chart. The tree control in the planning books allows you to explore data easily and to change the perspective of the view by simply clicking the mouse. Tree control provides all data from the various dimensions and hierarchies that matches the given set of selection criteria. For example, you can initially select a general view, jump to a view that is more detailed, and then jump again to a more aggregated view.

Business Planning Library

Library of Powerful Planning Functions
The Business Planning Library is a collection of powerful planning and analysis functions. Using these functions, you can convert the data contained in the InfoCubes into relevant business information. Demand Plannings rich planning and forecasting functions are based on the Statistical Forecasting Toolbox and the Business Planning Library. These functions include:
n Aggregate functions

Jump Between Data Views

You can also jump to a data view for completely different planning issues. For example, a sales manager examining sales forecasts can look at sales for a particular continental region, such as Europe, and then drill down to sales for individual countries within Europe. The manager can stay at this level of detail, but switch views by displaying sales according to product group. The planning books grid displays selected data, such as forecast components and planning versions, using slice-and-dice functions. You can specify which unit of measure and calendar to display and used for planning, and calendars allows you to mix multiple periods of time. For example, you can start with weeks, change to months, and change again to quarters. You can total row and column with a mouse click, and these totals remain fixed as you scroll through the grid. You can make adjustments on the basis of absolutes and percentages. If you change data on one level of aggregation, the system automatically updates the data on other levels of aggregation. The same is true of disaggregation changes. You can also fix the values of certain data elements during the reconciliation process.

Aggregate functions include SUM and AVERAGE. The information content of many types of data is greater when viewed on an aggregate level. For example, product sales are most useful when viewed on a regional or national level.
n Disaggregate functions

Disaggregate functions include disaggregation based on quotas, proportional distribution, and equal distribution. In many cases, planning results, such as forecasts and promotions, are output on an aggregate level and must be disaggregated to the detailed level.
n Comparison functions

Comparison functions include difference, ratio, percent, and so on. These functions are essential for determining the difference between forecasts and actual results so you know quickly how far forecasts deviate from your actual results.
n Financial functions

Financial functions include conversion from units into revenue, currency conversion, and business period conversion.
n Time-series functions

Graphical Data Display

The grid is directly linked to a chart, which displays information from the grid in user-specific graph modes. You can also adjust values directly in the chart. If you edit data either in the grid, the system automatically updates the data in the chart, and vice versa so you never need to worry about inconsistent data.

Time-series functions include the time-phased merge of time series, time series averages, and time-series weighted averages. Time-series functions work with planning techniques, such as composite forecasting.

Demand Planning

Customizable Planning Books

SAP APO Demand Planning Books are highly configurable. You can easily change the properties of the grid and the chart, and you can save different settings as profiles so the system works the way you do, not the other way around. You can run online simulations from the time that data is uploaded into the planning book until the changes are saved. This allows you to see the effects of different judgmental techniques on the forecast.

individual business tasks. Using the macro tool, you can:

n Control how macro steps are processed using

control instructions and conditions

n Build a macro consisting of one or more steps n Control how macro results are calculated using

control instructions and conditions

n Use a wide range of functions and operators n Define offsets so that the result in one period is

Preconfigured Planning Books for Quick Starts

SAP APO Demand Planning comes with preconfigured planning books for promotional planning, causal analysis, statistical forecasting, life cycle management, and so on, so you can get started right away. You can also use these as guides for customized planning books.

determined by a value in the previous period

n Restrict the execution of a macro to a specific

period or periods
n Write macro results to a row, a column, or a cell n Write the results of one macro step to a row,

column, cell, or variable, and use them only in subsequent iterations, macro steps, or macros.
n Analyze forecasting or business data at a glance

Advanced Macros
Quick and Easy Macros
You can use advanced macros to perform complex spreadsheet calculations quickly and easily. The macros are extremely flexible, so you can use them to model your planning environment based on your

using special icons

n Create context-specific and user-specific planning

views
n Trigger an alert in the Alert Monitor to inform you

of particular business situations

Figure 5: Example of Planning Book Design with Advanced Macros

Demand Planning

Integration with Microsoft Excel

Use Familiar Tools
If you are already familiar with Microsoft Excel, you can use Excel to edit data before transferring the results back to the grid. You can also store the content of the grid in Excel on a laptop. This allows you to work in a familiar environment, which reduces errors and training costs.

Synchronize Multiple Plans

Demand Planning can reconcile and combine different plans on the same level and it can synchronize plans created on different levels and different dimensions so you can always arrive at a consensus forecast.

Multilevel Planning
Synchronize Forecasts and Promotions
Multilevel planning capabilities allow you to forecast and plan on any level and in any dimension. Demand Planning provides consistent planning capabilities that synchronize forecasting and promotional data on any level in real time. You can plan from the top down, from the bottom up, and from the middle out. You can base disaggregation of data on user-defined timephased quotas, proportional distribution relative to user-specified data streams, and equal distribution.

Consensus-Based Forecasting
Consensus Planning Drives the Business
SAP APO Demand Planning supports consensus-based sales and operations planning. Its planning and forecasting features allow you to create plans for different business goals (such as a strategic business plan, tactical sales plans, and the operative supply chain plan) and integrate them into one consensus plan that drives the business. The planning books and reports help you gain consensus in meetings because they allow you to see the effects of your decisions.

Life Cycle Management

Life Cycle Management is a function of both the Demand Planning and the Supply Network Planning components. In the demand planning process, the planning strategies for a product depend on the stage in its life cycle. Analysis capabilities help you make decisions. Demand Planning helps you answer questions like:
n Should a new product be introduced? n When should a new product be introduced? n How should a product be promoted during the

Multiple Views
View Data from Multiple Angles
The multidimensional data structure of the InfoCubes allows you to view past sales and forecasts in many different dimensions, using varying units of measure and degrees of granularity, including:
n Product levels for marketing n Sales areas and accounts or channels for sales n Distribution centers and plants for operations n Business units for finance

different stages?
n Should the product be eliminated, and if so, when? n Should a follow-up product be introduced?

Consensus Plan
Create Multiple Versions of Plans
The forecasting techniques in SAP APO Demand Planning allow you to create many planning versions on any level and in any dimension using the same or similar data. For example, you can maintain multiple demand plans, such as a marketing plan, a sales plan, and an operational plan, so you can tailor them to individual situations and needs.

n Should a product be relaunched and when? n Will introducing a new product cannibalize sales

of existing products?

Complete Life Cycle Management

A products life cycle consists of different phases:
n Launch n Growth n Maturity n Discontinuation

Demand Planning

In Demand Planning, you can represent the launch, growth, and discontinuation phases by using phasein and phase-out profiles. A phase-in profile reduces demand history by ever increasing percentages during a specified period or periods, thus mimicking the upward sales curve that you expect the product to display during its launch and growth phases. A phaseout profile reduces the demand forecast of a product by ever decreasing percentages during a specified period or periods, mimicking the downward sales curve that you expect the product to display during its discontinuation phase.

Promotion Planning
Factor Price into Promotional Projections
Promotions of all kinds have a major impact on consumer sales behavior. The impact of promotions must be projected separately from standard forecast components that are based on historical sales data. Promotion-driven components are generated on top of a baseline forecast. Demand Planning extends classical volume-based promotional planning so that you can take prices into account when doing profitability analysis for promotional calendars. You can plan promotions on any level, and you can aggregate or disaggregate them using the Business Planning Library. Reporting capabilities allow you to track promotional activities and related costs and, therefore, assess the efficiency of your companys promotional campaigns.

Create Forecast with Minimal Data

Some products do not have enough historical data to provide the basis for a forecast. With like modeling, you can create a forecast using the historical data on a product with similar demand behavior. You can use like modeling for new products and products with short life cycles. For any product, you can use a phase-in profile, a phase-out profile, a like profile, or any combination of these.

Figure 6: Example of Promotion Planning.

10

Demand Planning

Promotion Patterns
Isolate Promotional Patterns
You can define promotional uplifts in units or percentages with promotion patterns. Using historical sales or estimates, the system can automatically detect a promotion pattern that occurred in the past. You can archive a promotion pattern in a promotion catalog, so you can reuse it if you repeat the promotion at a later date. The promotion catalog has a copy function that you can use to model like products, like regions, and so on. Demand Planning provides several techniques for estimating the effect of a promotion, such as multiple linear regression with or without a trend or seasonality.

Figure 7: Example of Multi-Tier Forecasting

Causal Analysis
Include Causal Factors in Forecasts
To predict consumer behavior, you must first understand all the factors that influence it. With Demand Planning, you can include all significant causal factors in your models and determine how they affect your customers behavior. Causal factors include price, number of displays, number of stores, temperature, and working days. You can even forecast the unknown future of causal factors using Demand Planning forecasting techniques. Breaking down sales to determine the impact of causal variables allows you to model the virtual consumer and simulate sales development according to the mix of causal factors.

Statistical Forecasting Toolbox

Multiple Tools for Multiple Forecasting Needs
Your companys product portfolio probably includes a variety of products that are in different stages of their life cycle and have different demand types. Unfortunately, a single forecasting method that creates accurate statistical forecasts for mature products, slow-moving products, and hot new products does not exist. Approaches that attempt to cover most of such demand types are very complex and tend to be a black box for planners. You have to use different methods to get the right answers. Demand Planning offers a toolbox of all practical, proven forecasting methods. You can choose the best method for a specific type of demand. Composite forecasting extends the idea of picking the best method, allowing you to combine forecasting methods. The Statistical Forecasting Toolbox provides all the tools you need to efficiently create accurate forecasts, including data analysis using time series models and multiple linear regression.

Use Proven Methods

Demand Planning provides the proven forecasting method of multiple linear regression to model the impact of causal factors.

Multitier Forecasting
Multitier forecasting integrates sell-in data, like pointof-sale data, into the process of forecasting sell-through data, like shipments. You can use a causal model based on all significant causal factors to forecast POS. You then use a second causal model to forecast shipments. The shipment model uses past POS data and the POS forecast as the main causal factor while taking the time lag between POS and shipments into account. You can also consider other causal factors, such as forward buys and trade promotions.

Data Analysis
Identify Missing Values
With Demand Planning, you can identify missing values and outliers in the data that drives the forecast and adjust the data or the applied forecasting method. This improves the quality of the statistical forecast.

11

Demand Planning

Intervene as Changes Occur

Structural changes in established patterns, such as level changes, trend changes, amplitude changes, and changes from unstable to stable behavior, must be identified. By automatically detecting structural changes using a tracking signal, Demand Planning allows you to take corrective action. You can either adjust the applied forecasting method, or you can replace it with a more suitable method. You can intervene manually, or the system can do it automatically.

Pick-The-Best Option
With the pick-the-best option, the system either runs all of the available forecasting methods or uses the planner-specified forecasting methods and applies the best one. You can specify the horizon for the fit criteria based on short-term, medium-term, or long-term forecasting goals.

Composite Forecasting
Composite forecasting creates a weighted average of the forecasts of multiple forecasting methods to create one forecast. You can determine the weights as equal (simple average) or using multiple linear regression by taking the residuals of the multiple forecasts into account. In addition, you can time-phase the combination so that short-term and medium-term forecasts are combined with respect to their horizons.

Time Series Models

Time series models are techniques that use past sales to identify level, trend, and seasonal patterns as a basis for creating future projections. The Statistical Forecasting Toolbox provides all of the most important models, such as naive models, moving average, simple linear regression, Browns exponential smoothing, and Holt-Winters so you always have the right tool for the job. For time series models with a linear trend, you can apply a dampening of the trend. You can also integrate life cycle patterns into the time series models using special profiles.

Forecast Accuracy Analysis

Track Forecast Accuracy
Forecast accuracy reporting helps companies assess the accuracy of past forecasts and integrate this knowledge into projections for the future. The primary technique is to store a series of forecasts for a particular period and to compare each deviation of this series to the actual values for the same period. You can then project the deviation into the future. This visualization of expected deviation not only places the forecast in context, it also enables what-if planning. For example, this enables a business to evaluate future scenarios for maximum, planned and minimum production volumes, stock, sales revenue, and cash flow. The real impact of forecast accuracy reporting is realized when you couple it with a simulation at the significant nodes of the supply chain. In Demand Planning, you can also measure the accuracy of the forecast using the following forecast errors:
n Mean absolute deviation (MAD) n Error total (ET) n Mean percentage error (MPE) n Mean absolute percentage error (MAPE) n Mean square error (MSE) n Square root of the mean squared error (RMSE)

Stochastic Models
Accurate Forecast with Sporadic Demand
When sporadic demands exist, the time series models always produce inappropriate stock levels as a result of the high mean absolute deviation from periods without demands. The Crosten method accounts for this inaccuracy by using exponential smoothing to estimate the size of demand during periods in which demands occur as well as the demand frequency. The final forecast values are determined by distributing the size of demand according to the demand frequency.

Multiple Linear Regression

Causal techniques are based on relationships between past sales and other causal factors. Multiple linear regression allows you to estimate these relationships by performing a least squares fit. You can choose from a variety of options to model linear and nonlinear trends, including seasonal patterns, life cycle patterns, dummy variables, and time lags. Using correlation analysis, you can correct variables. The system provides several statistics, including adjusted R-squared and Durbin-Watson.

12

Demand Planning

Real-Time Alerts to Exceptions

In addition, the SAP APO Alert Monitor informs you in real time via e-mail or an exception message in an exception window if an exception occurs in Demand Planning. You can define exception conditions based on thresholds for special statistics, such as MAD and tracking signals, so that the system alerts you if bias, outliers, or trend changes occur or orders exceed the forecast. This allows you to take action quickly when it can do the most good. Forecast accuracy reports show forecast errors at any level and dimension. They compare:
n Actual versus forecast n Actual versus time-lagged forecast n Actual versus different planning versions, such as

Multidimensional data mart The data mart is based on the star schema. The star schema is a structure that supports efficient use of storage space and of CPU cycles, minimizing query response time. You can update long-running mass forecast updates as a batch process so you do not impede online performance.

n Batch forecasting

Conclusion
Demand Planning provides multiple benefits that help you match supply and demand and manage the entire length of your supply chain. Demand Planning is:
n A fast solution

baseline forecast or consensus forecast

n Actual versus budget

Because Demand Planning is a component of the SAP Business Framework, you can implement it rapidly and at a low cost.
n An open solution

You can sort reports by forecast error or restrict them to products with a forecast error greater than a userspecified threshold.

SAP APO Demand Planning does not limit you to using SAP R/3 or R/2. Using open interfaces, you can integrate SAP APO Demand Planning with nonSAP data sources and non-SAP tools to create the best solution for your business.
n A function-rich solution

Performance
Performance is of vital importance in any demand planning solution if users are to benefit fully from available information. Demand Planning architecture includes several features that ensure high performance:
n

Dedicated server SAP APO resides on its own dedicated server. Like the SAP R/3 System, there are a number of configuration options to provide highly scalable performance as the number of users increases. SAP APO takes full advantage of the flexibility of SAP R/3s three-tier client/server technology. All SAP APO server components can reside on the same hardware system, or you can place the database on a dedicated database server. You can also add multiple application servers as the number of users increases.

SAP APO Demand Planning has the tools you need so you dont need to assemble them from multiple, specialized software packages. Demand Planning can improve forecast quality, model consumer sales behavior and market dynamics, integrate processes that drive the forecast, share information, and enable collaborative forecasting with business partners.
n An adaptable solution

If your needs or your IT environment change and they will SAP APO Demand Planning can adapt to these changes.
n A robust, business-driven solution

Based on SAPs proven expertise in SAP R/3 client/ server technology and in real-world business processes, Demand Planning is built to answer the specific demand planning and information requirements of planners and forecasters in all industries, reliably and effectively.

13

Demand Planning

For your note:

14

Demand Planning

15