The Predictive Casino

The Predictive Casino ANDREW PEARSON Copyright © 2017 Andrew Pearson All rights reserved ISBN-13: 978-1543264203 ISBN-10: 1543264204 CONTENTS ACKNOWLEDGMENT ......................................................................... i PREFACE ......................................................................................... iii INTRODUCTION ................................................................................ 1 CHAPTER ONE ................................................................................ 35 The Technology ............................................................................... 35 Overview ................................................................................................ 35 Artificial Intelligence (AI) and Machine Learning (ML)........................... 38 Augmented and Virtual Reality .............................................................. 40 Cryptocurrencies .................................................................................... 46 Facial Recognition .................................................................................. 47 Geofencing Applications ........................................................................ 51 Internet of Things (IoT) .......................................................................... 56 Mobile Marketing .................................................................................. 60 Digital Interactive Marketing: The Five Paradigms ............................ 65 Mobile Payments ................................................................................... 68 OTT ......................................................................................................... 72 Podcasts ................................................................................................. 73 Proximity Marketing .............................................................................. 74 Real-time Technology ............................................................................ 76 Search .................................................................................................... 78 Stream Processing and Stream Analytics ............................................... 79 Comparison of Stream Processing Services ....................................... 81 Videocasting/Livestreaming .................................................................. 84 Wearables .............................................................................................. 85 Conclusion.............................................................................................. 87 CHAPTER TWO ............................................................................... 95 Customer Experience ...................................................................... 95 Introduction ........................................................................................... 95 Customer Relationship Management (CRM) ......................................... 98 Customer Loyalty ................................................................................. 111 CHAPTER THREE ........................................................................... 117 Customer Analytics ....................................................................... 117 Overview .............................................................................................. 117 Sentiment Analysis............................................................................... 129 Clickstream Analysis............................................................................. 132 Location Analytics ................................................................................ 136 Analytics ............................................................................................... 138 Analytical Models............................................................................. 142 Decision Trees .............................................................................. 142 k-Means Cluster ........................................................................... 143 K-Nearest Neighbors .................................................................... 145 ANDREW PEARSON Logistic Regression ....................................................................... 147 A/B Testing ................................................................................... 148 Time Series Model ........................................................................ 150 Neural networks ........................................................................... 151 Discriminant Analysis.................................................................... 152 Survival or Duration Analysis ........................................................ 153 Edge Analytics............................................................................... 154 Casino Analytical Models...................................................................... 155 Customer Segmentation ................................................................... 155 Customer Acquisition Model ............................................................ 156 Recency-Frequency-Monetary (RFM) Models .................................. 157 Propensity to Respond Model .......................................................... 158 Customer Conversion Model ............................................................ 159 Identify When a Patron is Likely to Return ....................................... 160 Identify Patrons Who Come Together .............................................. 161 Patron Worth Model ........................................................................ 162 Customer Churn Model .................................................................... 163 Optimizing Offers ............................................................................. 164 Chronological View of a Casino Analytics Implementation .............. 165 Conclusion ............................................................................................ 166 CHAPTER FOUR ........................................................................... 171 Social Media .................................................................................. 171 Introduction.......................................................................................... 171 Mobile and Social Media in China .................................................... 176 The Four Steps of Social Media ............................................................ 184 Six Types of Social Media ..................................................................... 187 Collaborative Projects .......................................................................... 187 Blogs ..................................................................................................... 189 Microblogs .................................................................................... 192 Content Communities .......................................................................... 193 Social Networks .................................................................................... 194 Customer Understanding ..................................................................... 198 CHAPTER FIVE.............................................................................. 207 Social Business .............................................................................. 207 Overview .............................................................................................. 207 The Uses of Social Media ...................................................................... 212 Add Interactivity to a Website .......................................................... 214 Brand and Anti-Brand Management ................................................ 215 Brand Loyalty Enhancement............................................................. 220 Build Fanbases: Be “Liked” ............................................................... 221 Crisis Management ........................................................................... 223 Develop a Virtual Social World Presence ......................................... 229 Discover a Customer’s Psychological Profile .................................... 232 iv THE PREDICTIVE CASINO Discover Important Brand Trends .................................................... 235 Engage Customers and Potential Customers ................................... 236 Harvest Customer Feedback ............................................................ 238 Reputation Management ................................................................. 241 Social Shopping ................................................................................ 244 Social Media Analytics ......................................................................... 247 Social Media Monitoring ...................................................................... 251 Social Media Monitoring Tools ............................................................ 255 Conclusion............................................................................................ 260 Tips ....................................................................................................... 267 CHAPTER SIX ................................................................................ 273 Casino Operations ......................................................................... 273 Overview .............................................................................................. 273 Smart Operations ................................................................................. 274 Inventory Optimization .................................................................... 277 Waste Management ........................................................................ 279 Data Center ...................................................................................... 281 Smart Parking ................................................................................... 282 Smart Energy .................................................................................... 283 CHAPTER SEVEN ........................................................................... 287 The Predictive Casino .................................................................... 287 The Customer Journey ......................................................................... 287 Mobile Advertising Framework.................................................... 288 System Architecture..................................................................... 289 Amazon Example ......................................................................... 291 Table Games Revenue Management ............................................... 295 Social Media ..................................................................................... 300 Social Media Influencers .................................................................. 304 The Integrated Resort .......................................................................... 305 Casino .............................................................................................. 305 Bars, Coffee Shops, Nightclubs, and Restaurants ............................ 308 Retailers ........................................................................................... 310 The Future............................................................................................ 314 ABOUT THE AUTHOR .................................................................... 323 INDEX ........................................................................................... 325 v ACKNOWLEDGMENT Shinley Uy – thanks for keeping the faith i ANDREW PEARSON ii THE PREDICTIVE CASINO PREFACE The idea for this book came about when I signed on to become a partner to the analytics and IoT provider Hitachi Data Systems (HDS). They wanted to organize an event in Macau for high-level execs within the casino industry and I not only opened my Rolodex (I know that dates me, but why not, it shows experience!), but I also decided a book that delved into the details of the technology we were showcasing would be a great companion piece. Meeting the deadline—twelve weeks, however, I knew, would be tricky, but we live in a modern world, where technology allows us to write, read, and research on the go, so I set about to write this book within that time frame. I didn’t quite make it, but was close. Parts of this book have been written on the Macau-Taipa bus, the Macau-Hong Kong ferry, 30,000 feet above both the Pacific and the Dark Continent of Africa, on the Guangzhou-Macau high-speed rail line, within a taxi stuck driving through the smog-choked streets of Manila, as well as on Hong Kong’s MTR, to say nothing of all the hotel and motel rooms I’ve been scribing in; planes, trains, automobiles and ferries, too. Needless to say, it’s a worldly and wordy piece of writing that I wanted to get out as quickly as possible because this technology is, literally, changing by the day. As I had already written several books on the subjects of casino analytics, casino marketing, mobile technology and social media—which I believe is an integral component of today’s casino marketing plans—the task wasn’t monumental. Difficult, yes. Impossible, no. So here it is… iii INTRODUCTION In December 2014, Amazon was awarded a patent for what it calls a “method 1 and system for anticipatory package shipping,” and, according to Lance 1 Ulanoff’s article Amazon Knows What You Want Before You Buy It , “The patent summary describes a method for shipping a package of one or more items ‘to the destination geographical area without completely specifying the delivery 1 address at time of shipment,’ with the final destination defined en route.” According to the patent, this forecasting model uses data from a customer’s “prior Amazon activity, including time on site, duration of views, links clicked 1 and hovered over, shopping cart activity and wish lists.” When possible, the algorithm also sprinkles in real-world information gleaned from customer telephone inquiries and responses to marketing materials, among other factors. Together, this can provide ‘decision support for speculative shipping of 1 items,’ per the patent,” explains Ulanoff. All of this may sound a little fanciful, but their thesis that predicting customers’ orders could unquestionably increase sales as well as potentially reduce 1 shipping, inventory and supply chain costs holds true. For these reasons alone, Amazon is probably onto something here. Obviously not the sexiest part of the casino business, supply chain and logistics is, however, an imperative and expensive part of it. Ulanoff quotes H. Donald Ratliff, Ph.D., executive director of the Supply Chain and Logistics Institute, who argues that optimization is “the biggest opportunity for most companies to significantly reduce their cost and improve their 1 performance.” “For most…operations, there is an opportunity to reduce cost by 10% to 40% by making better decisions,” Ulanoff argues. This amounts to substantial savings for casino companies, who spend millions of dollars in supplies and products every year. Amazon utilizes predictive analytics, including Machine Learning (ML) to make very sophisticated supply chain, logistics and even customer delivery decision and these concepts and theories work just as well for a casino operator, but the Predictive Casino wants to infuse this anticipatory concept throughout its entire operation, not just in logistics; from patron and customer interactions, to optimization in hotel room and table games pricing, to marketing, data governance, security, call center, sales, advertising, and labor management. The anticipatory concept can make the casino smart, predictive, optimized, and probably more profitable than it would otherwise be. Human beings are, after all, creatures of habit and, if a casino company can 1 ANDREW PEARSON understand these habits on both a micro and macro level, it can not only predict what its customers are going to want and to do, but maybe even, shape these actions. Marketing has always been about influencing people’s actions, what could be different here is the Predictive Casino’s ability to understand how one customer’s actions will affect the company’s entire operation. With this insight extrapolated over a million customers over 365 days of the year, the casino operator can take the most appropriate—and optimized—action to reap the highest profit. 2 According to IBM’s own research , 2.5 quintillion bytes of data are created each day. That is 10 to the power of 18 and that number is growing exponentially each year; 90% of the world’s data was created over the past two years and data creation is certainly not going to slow down any time soon. This data— which has been dubbed “Big Data”–comes from everywhere; our daily financial transactions; our personal online shopping history; our social media uploads; our mobile downloads, even sensor data coming off machines and people, in some cases. The social nature of sharing personal content with family, friends and associates may be the driver behind this growth and it is a growth that several 34 studies suggest will soon outpace revenue generated by commercial media, such as music downloads, video clips, and games. This is the kind of growth that a casino operator ignores at its own peril, but when a casino delves into this Big Data world, they need to ensure that what they’re opening up is a treasure chest of information and not a Pandora’s box of pain. 5 According to Gartner Inc.'s Hype Cycle for Emerging Technologies , 2016, there are three distinct technology trends—transparently immersive experiences, the perceptual smart machine age, and the platform revolution—that are poised to be of the highest priority for organizations facing rapidly accelerating digital business innovation. “The Hype Cycle for Emerging Technologies is unique among most Hype Cycles because it distills insights from more than 2,000 technologies into a succinct set of must-know emerging technologies and trends that will have the single greatest impact on an organization's strategic planning,” explains Mike J. 5 Walker, research director at Gartner. “This Hype Cycle specifically focuses on the set of technologies that is showing promise in delivering a high degree of 5 competitive advantage over the next five to 10 years,” he adds. In its publication, Gartner explains that transparently immersive experiences are technologies like 4D Printing, Brain-Computer Interface, Human Augmentation, Volumetric Displays, Affective Computing, Connected Home, Nanotube Electronics, Augmented Reality, Virtual Reality and Gesture Control 5 Devices. These technologies “will continue to become more human-centric to the point where it will introduce transparency between people, businesses and 2 THE PREDICTIVE CASINO things. This relationship will become much more entwined as the evolution of technology becomes more adaptive, contextual and fluid within the workplace, 5 at home, and interacting with businesses and other people,” Gartner adds. Gartner expects the perceptual smart machine age and the smart machine technologies that utilize them will be the next decade’s most disruptive class of 5 technologies. Because of radical and powerful new computational power, near-endless amounts of data captured from a wide variety of sources, and unprecedented advances in analytical processes such as deep neural networks, organizations with smart machine technologies will be able to harness their 5 data to solve problems that weren’t even thought about a few years ago. Gartner recommends enterprises, including IRs and smaller gaming companies, that are seeking leverage in this area should consider the following technologies: “Smart Dust, Machine Learning, Virtual Personal Assistants, Cognitive Expert Advisors, Smart Data Discovery, Smart Workspace, Conversational User Interfaces, Smart Robots, Commercial UAVs (Drones), Autonomous Vehicles, Natural-Language Question Answering, Personal Analytics, Enterprise Taxonomy and Ontology Management, Data Broker PaaS 5 (dbrPaaS), and Context Brokering.” Many, but not all, of these technologies will be discussed in throughout this book. Gartner argues that the platform revolution—Neuromorphic Hardware, Quantum Computing, Blockchain, IoT Platform, Software-Defined Security and Software-Defined Anything (SDx)—includes emerging technologies that are 5 revolutionizing the concepts of how platforms are defined and used. “The shift from technical infrastructure to ecosystem-enabling platforms is laying the foundations for entirely new business models that are forming the bridge 5 between humans and technology,” Gartner adds. “Within these dynamic ecosystems, organizations must proactively understand and redefine their strategy to create platform-based business models, and to exploit internal and external algorithms in order to generate value,” advises Gartner. When it comes to analytics and Big Data, Caesars was the first casino company to collect and analyze it for Customer Intelligence (CI) purposes and, since the inception of it Total Rewards programme, the company has grown from “being 6 able to trace the journey of 58% of the money spent in their casinos to 85%.” Caesars also credits the widespread adoption of Big Data analytics as the driving force behind its rise from an “also ran” chain to one of the largest 7 casino groups in the country. The Predictive Casino is a casino or an integrated resort that takes into account all kinds of data that can be created throughout a casino property by its employees, vendors, patrons, and customers (we’ll consider these the people who haven’t signed up for a player card yet and aren’t, therefore, as trackable as patrons who are in the casino’s database). Throughout this book, I will look 3 ANDREW PEARSON at the integrated casino resort as a property that contains everything from the gaming floor, to a mall containing retail, food, and entertainment outlets, as well as a property that might house movie theaters, pools, amusement parks and/or rides, as well as arenas for concerts and sporting events. The Predictive Casino utilizes all of the data associated with all of these onproperty businesses to make better business decisions for the company as a whole. The Predictive Casino is viewed holistically and the proverbial butterfly’s wing that flaps somewhere inside the casino can set off a chain of events that can either help or hurt the company’s bottom line months down the line, but captured and analyzed it will be so that surprises and negative impacts can be mitigated. Descriptive analytics, diagnostic analytics, predictive analytics, prescriptive analytics and the newest field of analytics—edge analytics—are exploited throughout the Predictive Casino to try to reach as real-time an IT environment as possible. This data I will be focusing on throughout this book will be culled from the following sources: • • • • • • • • • • • • • • • • • • Gaming data from Bally’s, IGT, or other in-house gaming systems Customer Relationship Management (CRM) software Transaction data from a Point-of-Sales (POS) system A hotel reservation systems Clickstreams from the casino’s website Call center systems Surveillance and security systems, including facial recognition datasets RFID chips for both casino chips and/or IoT tracking devices Geo-location data from in-house Wi-Fi systems Social media data from WeChat, Facebook, Weibo, Twitter, Jeipang, and other mobile and social media apps Table games revenue management systems Angel Eye card shoe data Patron management systems Social media listening hubs Google analytics and web tracking information HR and ERP systems, including virtual roster Transportation data Weather patterns All of this information can be fed into a data lake or an Enterprise Data Warehouse (EDW), where it can be utilized by a multitude of casino departments, including security, call center/customer service, pit bosses, hosts, dealers, HR agents, marketing, including social media marketing, hotel management, slots floor operations, patron management, retail and F&B, all the way up to the top executive branches, including individuals in the C-level 4 THE PREDICTIVE CASINO suite. In recent years, businesses in general and casino companies in particular have come to the realization that data warehouses, while perfectly able to handle the BI and analytics needs of yesterday, don’t always work in today’s complex IT environments, which contain structured, unstructured, and semi-structured data. Normal relational databases worked fine when business users were restricted to proprietary databases and the scope of work was restricted to canned reports and modest dashboards that included limited drill down functionality. Today, however, with the inclusion of so much unstructured data coming from mobile, social, web logs, etc., and semi-structured data originating from a multitude of sources, limitations abound. Standard data warehouses require built-in, understandable schemas, but unstructured data, by definition, doesn’t have a definable schema that is accessible and understandable in every case. Data lakes have been a response to these limitations. James Dixon, “Chief Geek” at Pentaho, is credited with coining the phrase “Data Lake” and Dixon posted that each specialized data mart in a data warehouse could be likened to a bottle of water. The data was ready for use in a small, identifiable container. In contrast, a data “lake” was a massive, intermingled repository of all data in its raw form. A data lake is a hub or a repository of all the data that a casino has access to, where the data is ingested and stored in as close to the raw form as possible, without enforcing any restrictive schema. This provides an unlimited window into the data for anyone to run ad-hoc queries and perform cross-source navigation and analysis on the fly. Successful data lake implementations respond to queries in real-time and provide users an easy and uniform access interface to the disparate sources of data. Data Lakes retain all data, support all data types and all users, as well as adapt easily to changes, while providing faster insights. Today’s IT environment is nothing like the IT environment of even three years ago. Real-time data management capabilities have brought a whole new level of data available to customer intelligence, customer interaction, patron management and social media systems. One of the biggest challenges for IT departments today is scalability. With a Hadoop back-ended data lake, businesses can dynamically scale up or down, according to their storage needs. Over the past few years, the cost of storage has plummeted and virtual servers can be spun up very quickly, as well as quite inexpensively (relative to the outright purchase of hardware). With this instant access to data, a whole new world of real-time interactions should flourish and I will detail how an IR can set up a real-time stream processing environment in chapter one and chapter seven. 5 ANDREW PEARSON The concept of “Edge Analytics”—i.e., the processing of analytics at the point or very close to the point of data collection—exponentially increases the ability to use predictive analytics where it can be utilized best—at the point of interaction between the business and the customer. In short, edge analytics brings analytics to the data rather than vice-versa, which, understandably, can reduce cost and increase its usage as the data is analyzed close to where it can make the most difference. This also reduces latency, which could be the difference between useful and useless analytics. Today, the analytics space is more crowded than it has ever been. Standard ETL-solution providers are adding analytics to their multitude of offerings. Many of these new players in the Master Data Management (MDM) field have BI platforms that combine integration, preparation, analytics and visualization capabilities with governance and security features. Such standard analytics processes as column dependencies, clustering, decision trees, and a recommendation engine are all included in many of these new software packages. Instead of forcing clients to frustratingly purchase module on top of module on top of module, new software companies are creating packages that contain many pre-built analytical functions. Open source products like R, Python, and the WEKA collection can easily be added to many of these software solutions as well, thereby reducing the need for expensive analytics layers. The fact that many of these analytical packages are open source is a further advantage because, since they are free to download and use, they have a robust user base and consultants are sometimes easier to find than analysts with highly developed SAS or SAP Predictive Analytics Library skills, for example. Before going any further, I believe one of the first questions that needs to be answered is, “What exactly is analytics?” The standard answer is that there are four types of analytics and they are: • • • • Descriptive analytics—What happened? Diagnostic analytics—Why did it happen? Predictive analytics—What will happen? Prescriptive analytics—How can we make it happen again? For a casino company, descriptive analytics could include pattern discovery methods such as customer segmentation, i.e., culling through a patron database to understand a patron’s preferred game of choice. Simple cluster segmentation models could divide customers into their preferred choice of games. This information could be given to the marketing department to create lists of baccarat players for a baccarat tournament, for example. 6 THE PREDICTIVE CASINO Market basket analysis, which utilizes association rules, would also be considered a descriptive analytics procedure. Casinos should use market basket analysis to bundle and offer promotions as well as gain insight into its patrons’ gaming habits. Detailed patron shopping and purchasing behavior could also be used to develop future products, whether they be gaming or retail related. I will go into full detail on this topic in chapter three. As Bernard Marr argues in his article Will ‘Analytics on The Edge’ Be The Future 8 Of Big Data? , “Rather than designing centralized systems where all the data is sent back to your data warehouse in a raw state, where it has to be cleaned and analyzed before being of any value, why not do everything at the ‘edge’ of the system?” Marr uses the example of a massive scale CCTV security system 8 that is capturing real-time video feeds from tens of thousands of cameras. “It’s likely that 99.9% of the footage captured by the cameras will be of no use for the job it’s supposed to be doing—e.g. detecting intruders. Hours and hours of still footage is likely to be captured for every second of useful video. So what’s the point of all of that data being streamed in real-time across your network, 8 generating expense as well as possible compliance burdens?” The solution to this problem, Marr argues is for the images themselves to be analyzed within 8 the cameras at the moment the video is captured. Anything deemed out-ofthe-ordinary will trigger alerts, while everything considered to be unimportant will either be discarded or marked as low priority, thereby freeing up 8 centralized resources to work on data of actual value. For a casino company, the CCTV security systems could even be set up to alert a host or a VIP manager when a particular high roller steps aboard the casino bus at the Macau-China border, for example. The CCTV system could also be used to capture problem gamblers or cheats on the casino floor. Edge video analytics might also help in places like Singapore, where the integrated resorts are required to check the passports of each incoming guest because of Singapore law. With cameras that are able to compare the faces of entering customers against the casino’s patron database, these patron records would be onscreen for the security personnel to quickly approve or reject. Rather than waiting for a record to be pulled up, the security personnel would simply confirm or deny the patron in less than half the time it would normally take to pull up a particular patron record. This would cut down on the long lines at the front of the casino that are really costly choke points. A highly positive ROI could be created here; the casino makes money when gamblers are gambling, not when they are standing in line, waiting to get in. As an added benefit this will also be seen as a customer service improvement by the patrons as well. Using edge analytics and real-time stream processing systems, retailers within an IR could “analyze point-of-sales data as it is captured, and enable cross selling or up-selling on-the-fly, while reducing bandwidth overheads of sending 7 ANDREW PEARSON 8 all sales data to a centralized analytics server in real time.” As today’s integrated resorts are also, in many cases, huge retail malls, retail edge analytics could, potentially, become part of a package the IR makes available to its retail clients. Edge analytics, of course, goes hand-in-hand with the Internet of Things–“the network of physical objects that contain embedded technology to communicate and sense or interact with their internal states or the external 9 environment.” 10 In his seminal 2009 article for the RFID Journal, That 'Internet o Things' Thing , Kevin Ashton made the following assessment: Today computers—and, therefore, the Internet—are almost wholly dependent on human beings for information. Nearly all of the roughly 50 petabytes (a petabyte is 1,024 terabytes) of data available on the Internet were first captured and created by human beings—by typing, pressing a record button, taking a digital picture, or scanning a bar code. Conventional diagrams of the Internet include servers and routers and so on, but leave out the most numerous and important routers of all—people. The problem is, people have limited time, attention and accuracy—all of which means they are not very good at capturing data about things in the real world. And that's a big deal. We're physical, and so is our environment. Our economy, society and survival aren't based on ideas or information—they're based on things. You can't eat bits, burn them to stay warm or put them in your gas tank. Ideas and information are important, but things matter much more. Yet today's information technology is so dependent on data originated by people that our computers know more about ideas than things. If we had computers that knew everything there was to know about things—using data they gathered without any help from us—we would be able to track and count everything, and greatly reduce waste, loss and cost. We would know when things needed replacing, repairing or recalling, and whether they were fresh or past their best. The Internet of Things has the potential to change the world, just as the Internet did. Maybe even more so. One of the key points of that quote for the casino industry is “greatly reduce 10 waste, loss and cost” and, in a business that can go through 1.5 tons of food per day just to feed its thousands of employees, anything that can reduce waste should be explored. Sensors that can help keep food fresh or provide alerts when stocks need to be filled are inexpensive enough that strong ROI justifications can be made. I delve into more detail on this subject in chapter 8 THE PREDICTIVE CASINO six. Ashton is correct when he states that computers are far superior to human beings when it comes to data gathering and cheap sensors, such as RFID chips in poker chips, could go a long way to provide highly accurate patron comps 10 totals, which are probably overly generous to patrons right now. Human beings are masters at rigging the system in their favor and casino companies would probably be troubled to learn the real cost and ROI on the free comps they hand out. The term “Big Data” has become a way-too-common and enormously prevalent term and it is being thrown around a lot in the world of IT these days as it has become a kind of catch-all for analytics, IoT, social media, etc., etc. Although not a comprehensive list, Big Data analytics techniques can include association, classification, cluster analysis, crowdsourcing, data fusion, data mining, machine learning (ML), modeling, network analysis, optimization, predictive, regression, rule learning, special analysis, text analytics, time series analysis, amongst many, many others. Which techniques should a casino operator use, well that all depends on what type of data is being analyzed, the available technology, the skills of the business users, and the business problems tried to. be solved. In chapter three, I break down how these analytical processes would work in the concept of the customer journey, and I will specifically explain in what circumstances decision trees, time series, discriminant analysis, K-means clustering, and K-Nearest Neighbor processes, amongst others, would be utilized. None of these techniques, however, will amount to anything if the underlying data environment isn’t robust and cleansed properly; junk in, junk out, as most analysts will tell you. Enormous attention must be paid to ensure the data is prepped and cleansed, otherwise nothing of value will be achieved, no matter how fast and/or robust your analytics software turns out to be. IoT technology costs are coming down, broadband’s price has dropped, while its availability has increased and there is a proliferation of devices with Wi-Fi capabilities and censors built into them. Smart phone penetration is also exploding. All of these individual technological advances were good for the IoT 11 environment, together, however, they have created a perfect storm for it. With less than 0.1% of all the devices that could be connected to the Internet 11 currently connected , there is tremendous growth potential here and those who embrace it now should have the first mover advantage that could prove enormously valuable in terms of ROI over the long term. How can a casino operator utilize IoT technology? Well, today’s integrated resorts are massive structures and sensors have become so small and cheap that they can be put almost anywhere. IoT sensors can be used for smart parking, smart lighting, or as part of a mini smart grid. They can also be used for 9 ANDREW PEARSON silo stock calculation—measuring the emptiness level and weight of goods, as well as waste management, and perimeter access control tools. Liquid presence detection in places like data centers can help ensure the integrity of the IT backbone. For the casino’s retailers, IoT can help with supply chain control, NFC payment systems, inventory shrinkage, as well as smart product management. For the casino’s logistics department, IoT aids quality of shipment conditions, item location, storage incompatibility detection, and fleet tracking. IoT sensors can even be installed to ensure a building’s structural health, as well as part of a swimming pool remote management system. Combining IoT data together with other structure and unstructured data isn’t easy, though. Previous attempts at broad-based data integration has forced users to build data sets around common predetermined schema, or a unifying data model, but this becomes impossible when unstructured and semistructured data are included in the mix. This is where a data lake can come in. Unlike the monolithic view of a single enterprise-wide data model, the data lake relaxes standardization and defers modeling, resulting in a nearly unlimited potential for operational insight and data discovery. As data volumes, data variety, and metadata richness grows, so, too, do the benefits. Today, data is coming from everywhere, from business mainframes, to corporate databases, from log files, cloud services, APIs, RSS feed, as well as from social media live feeds; most of this information does contain meaning, if one knows what and where to look for it. A data lake makes it easier to read and understand that data, at least that’s the theory that is being tested out by several forward-thinking companies right now. Using and understanding all of this data is going to be the challenge. 12 In Tableau’s Top 8 Trends for 2016 Big Data, one of the leading visualization software vendors writes: “We noted the increasing adoption of NoSQL technologies, which are commonly associated with unstructured data, in last year’s version of Trends in Big Data. Going forward, the shift to NoSQL databases becoming a leading piece of the Enterprise IT Landscape becomes clear as the benefits of schema-less database concepts become more pronounced. Nothing shows the picture more starkly than looking at Gartner’s Magic Quadrant for Operational Database Management Systems which in the past was dominated by Oracle, IBM, Microsoft and SAP. In contrast, in the most recent Magic Quadrant, we see the NoSQL companies, including MongoDB, DataStax, Redis Labs, MarkLogic and 10 THE PREDICTIVE CASINO Amazon Web Services (with DynamoDB), outnumbering the traditional database vendors in Gartner’s Leaders quadrant of the report.” 12 Tableau also sees the following eight trends affecting IT departments : 1. 2. 3. 4. 5. 6. 7. The NOSQL Takeover—the benefits of schema-less database concepts are becoming harder and harder to ignore and the proof is in Gartner’s latest Magic quadrant for Operational Database Management Systems, which is dominated by vendors like MongoDB, DataStax, Redis Labs, MarkLogic and Amazon Web Services (with DynamoDB), rather than entrenched players like Oracle, SAP, IBM and Microsoft. Apache Spark lights up big data—according to its originator, Matei Zaharia, Apache Spark is becoming the largest big data open source project on the market because its processing speeds are dramatically faster than Hadoop’s,. The Hadoop project matures—“In a recent survey of 2,200 Hadoop customers, only 3% of respondents anticipate they will be doing less with Hadoop in the next 12 months. 76% of those who already use Hadoop plan on doing more within the next 3 months and, finally, almost half of the companies that haven’t deployed Hadoop say they 12 will within the next 12 months.” Big Data grows up: Hadoop adds to enterprise standards—the “Apache Sentry project provides a system for enforcing fine-grained, role based authorization to data and metadata stored on a Hadoop 12 cluster.” Apache is taking the demands of the enterprise-grade RDBMS very seriously. Big data gets fast: Options expand to add speed to Hadoop—As Hadoop gains more traction in the enterprise market, the need for fast data exploration capabilities is growing. To meet this demand, Tableau sees further “adoption of technologies such as Cloudera Impala, AtScale, Actian Vector and Jethro Data that enable the business user’s old friend, the OLAP cube, for Hadoop—further blurring the lines behind the ‘traditional’ business intelligence concepts and the world 12 of ‘Big Data.’” The number of options to discover all forms of data grows—Because of data discovery tools like Tableau and Qlik, self-service data preparation tools are increasing in popularity. Now that business users have gotten a taste of data discovery and like it, they want to be able to reduce the time and complexity of data preparation—the most important part of the analytics process—and there’s been considerable innovation in this space from companies like Alteryx, Trifacta, Paxata and Lavastorm. Massively Parallel Processing (MPP) data warehouse growth heats up in the cloud—Perhaps the death knell of the data warehouse has been 11 ANDREW PEARSON 8. overhyped and now the trend seems to be towards a hybrid mix of standard DWs and on-demand cloud data warehouses. “Analysts cite 90% of companies who have adopted Hadoop will also keep their data warehouses and with these new cloud offerings, those customers can dynamically scale up or down the amount of storage and compute resources in the data warehouse relative to the larger amounts of 12 information stored in their Hadoop data lake.” Buzzwords converge—Internet of Things, Cloud, and Big Data come together. Although it is in its infancy, IoT technology will soon become one of the “killer apps” for the cloud, resulting in a data explosion unlike we’ve ever seen before. For this reason, Tableau sees “leading cloud and data companies such as Google, Amazon Web Services and Microsoft bringing Internet of Things services to life where the data 12 can move seamlessly to their cloud based analytics engines.” With a normal data warehouse, a casino needs to decide on the structure (schema) of the data when creating the warehouse—before anything is even populated with data (schema-on-write). With a Hadoop-based data lake, however, a casino just has to store the data and structure it later, at a time when it is needed for each query or use case (a schema-on-read framework). Table 1 reveals the main differences between the two systems. DATA WAREHOUSE Structured, processes Schema-on-write Expensive for large data volumes Less agile, fixed configuration Mature Business professionals vs. DATA LAKE DATA Structured / semi-structured / unstructured, raw PROCESSING Schema-on-read STORAGE Designed for low-cost storage AGILITY Highly agile, configure and reconfigure as needed SECURITY USERS Maturing Data scientists, et al Table 1: Differences between a Data Warehouse and a Data Lake With a data lake, the data is ingested and stored in as close to the raw form as possible, without enforcing any restrictive schema on top of it. No OLAP cubes are used to manipulate the data. This provides an unlimited view of the data for anyone within the organization who has been given access to it. The user will be able to run ad-hoc queries and perform cross-source navigation and analysis on the fly. Successful data lake implementations respond to queries in real-time and provide users an easy and uniform access interface. Data Lakes retain all data, support all data types and all users, as well as adapt easily to 12 THE PREDICTIVE CASINO changes, while providing faster insights. Using a data lake solution, a casino can: • • • • • • • • • • • • • • Gain visibility on a patron’s true identity at sign-up by culling through government and social media records during the actual sign up process. Gain a comprehensive view of patron behavior by automatically cleansing and consolidating disparate gaming and nongaming activity into a single version of the truth. Understand a patron’s activity and behaviors—using segmentation to analyze data gathered at multiple touch points and then segmenting patrons according to their subtle gaming differences. Identify the greatest drivers of patron value and project how those drivers will affect profitability and revenue projections months into the future, with enough lead time to take corrective actions, if necessary. Create highly effective personalized promotions that are tailored to appeal to a casino’s patrons by automating and personalizing marketing campaigns on a recurring basis. Maximize patron satisfaction and profitability by optimizing valuable resources to meet patron needs at every property touch point. Get valuable customer insights into the hands of patron-facing employees, decision makers, and others who can exploit it, with highend reporting and data visualization capabilities. Geo-locate a patron from the moment he or she arrives on property and include that individual’s specific gaming play information into models that calculate table games revenue management in close to real time. Understand real-time casino floor traffic flows and heat maps so that casino personnel can be utilized more efficiently. Increase employee and labor management efficiency. Capture fraudulent and Anti-Money Laundering (AML) activity in realtime, thereby reducing costs and creating solid evidence to catch and lock-up criminals. Utilize IoT data to create a table games revenue management model that takes into account patron activity before they even step onto the gaming floor. Add social media as a customer service and marketing channel. Associate online ad marketing with patron conversions, even after a patron signs up for a patron card. For this book, I will consider CRM as a two-part process that allows a casino operator to track and organize its current and prospective customers, as well as to manage the endpoints of customer relationships through its marketing 13 ANDREW PEARSON promotions. When done right, CRM systems enable data to be converted into information that provides insight into customer behavior. From these insights, some form of behavioral influencing can occur. The process of segmenting a market is deceptively simple; seven basic steps describe the entire process, including segmentation, targeting and positioning. In practice, however, the task can be very laborious since it involves poring over loads of data, and requires a great deal of skill in analysis, interpretation and some judgment. Today, personalized web pages can be rendered during the web page load and elements of the page can take into account past purchase history, clickstream information, as well as a whole host of other things. In chapter two, I will delve into the how and why of these systems in detail. Data coming from mobile and social media sources like WeChat, Weibo, Facebook, YouTube, Twitter, YouKu, etc., tend to be highly unstructured, while data coming from CSVs, XML and JSON feeds are considered semistructured. NoSQL databases are also considered semi-structured, while text within documents, logs, survey results, and e-mails also fall into the unstructured category. Structured data coming in from the plethora of casino source system, undoubtedly, can feed into a data lake, where it can be merged with unstructured data and then utilized in ways that are almost impossible for a normal relational DW to handle. Highly structured patron data could be combined with unstructured data coming in from social media to reveal deep customer insights. If a patron tweets from the train heading towards Zhuhai, why shouldn’t the casino marketing department be alerted? Setting up JSON feeds for Twitter user accounts is a very simple process and many other social media companies offer APIs that allow access to customer accounts. These are two-way systems as well, and the casino’s marketing department could include social media as a channel to connect with customers and potential customers. These and other social media marketing campaign ideas will be discussed further in chapters four and five. How does a casino get a player’s WeChat, RenRen, Facebook, Twitter, Weibo, YouTube, or even Twitch account? Easy, just make the patron an offer they can’t refuse and, in most cases, that offer probably wouldn’t be too much more than a coupon for free play or a free meal. Once a patron steps aboard a casino bus at the Macau-Chinese border and taps into the casino’s free Wi-Fi service, he or she is immediately trackable and a whole chain of analytical events could be kicked off. Here, the real customer journey begins. With quick and easy accessibility to a casino’s data, customer conversion rates can be improved, revenue can be increased, and customer churn can be predicted and, hopefully, reduced as much as possible. Customer acquisition 14 THE PREDICTIVE CASINO costs can also be lowered. By utilizing the complex web of customer data coming in from several different channels—mobile, social media, customer loyalty programs, transaction data, e-commerce weblogs, sensors, amongst others—a casino can also work more productively. By understanding customer patterns and patron behavior across the whole spectrum of the business—from patron pickup, to patron navigation throughout the property, to room and restaurant choices, to gaming floor activity, all the way to final check-out—a casino can use these customer behavioral patterns to map out its inventory, food, and human capital needs as well. As Kai Wähner explains in his article Real-Time Stream Processing as Game 13 Changer in a Big Data World with Hadoop and Data Warehouse , “Stream processing is required when data has to be processed fast and/or continuously, i.e. reactions have to be computed and initiated in real time.” Wähner 13 continues : “’Streaming processing’ is the ideal platform to process data streams or sensor data (usually a high ratio of event throughput versus numbers of queries), whereas “complex event processing” (CEP) utilizes event-by-event processing and aggregation (e.g. on potentially out-of-order events from a variety of sources—often with large numbers of rules or business logic). CEP engines are optimized to process discreet ‘business events’ for example, to compare out-of-order or out-of-stream events, applying decisions and reactions to event patterns, and so on. For this reason multiple types of event processing have evolved, described as queries, rules and procedural approaches (to event pattern detection).” Stream processing acts on real-time streaming data feeds, using “continuous 13 queries” (i.e., SQL-type queries that operate over time and buffer windows). With its ability to continuously calculate mathematical or statistical analytics on the fly within the stream, streaming analytics is an essential part of stream processing. “Stream processing solutions are designed to handle high volume in real time with a scalable, highly available and fault tolerant architecture,” adds 13 Wähner. “In contrast to the traditional database model where data is first stored and indexed and then subsequently processed by queries, stream processing takes the inbound data while it is in flight, as it streams through the server,” explains 13 Wähner. Stream processing can also connect to an external data source, thereby adding a whole new dimension to analytical processes. For IRs and casino companies, real-time streaming can help in the following ways: • Customer Service: 15 ANDREW PEARSON Geo-locating a patron when he or she signs onto the casino Wi-Fi, whether that is on a casino bus or while wandering throughout the integrated resort. o Video analytics with facial recognition technology can spot and/or confirm a patron’s true identity. o Social media customer service can cut down on normal customer service expenses as well as connect with customers on the channels that they prefer, including social media. E-Commerce o Clickstream analysis could allow personalized offers to potentially returning guests when they are browsing the casino’s website to make a reservation. Hosts: o Hosts can be alerted when a VIP, or even a premium mass player, steps onto the property. o Executives could be made aware of over-comping. Hotel: o Hotel room revenue management; offer the right price to the right person at the right time, on the right channel, with the right upsell offerings. o Offer high-profit upgrades to not only those who are most likely to use them, but also to pay for them. Human Capital Management: o Employee schedules can be adjusted in real time according to labor managements needs, as well as its predictive and anticipatory needs. o Casinos can take the guesswork out of hiring employees by building templates that show what a model employee should possess in terms of skills. Patron Management: o The ecommerce department can get more accurate attribution analysis—“the process of identifying a set of user actions (‘events’) that contribute in some manner to a desired outcome, and then assigning a value to each of these 14 events” —so that it understands which advertising is associated with which user, making it more quantifiable and, therefore, more actionable. o Customer Relationship Management (CRM) systems can add social media as a channel feeding targeted messages to only those patrons who are most likely to respond to a promotion. o The amount of promotions available and channels through which to market through increases considerably as campaign o • • • • • 16 THE PREDICTIVE CASINO • • lift can be assessed in terms of hours rather than in days or weeks. o Customer acquisition is accelerated because business users throughout the property can quickly derive answers to the following questions: § Which combinations of campaigns accelerate conversion? § What behavior signals churn? § Do web search key words influence deal size? § Which product features do users struggle with? § Which product features drive product adoption and renewal? § What drives customers to use costly sales channels? o Customer interaction data can quickly be turned into business opportunities. o Powerful recommendation engines can ingest data from a multitude of sources and then be made available to frontline staff, who can react in near real time. Pit Bosses / Floor Managers: o Facial recognition technology allows for immediate knowledge of patrons entering the casino floor, which can improve customer service for both VIPs and premium mass players. o Facial recognition can spot self-excluded patrons and alert security about the problem gambler. o Table game minimums can be raised or lowered according to demand forecasts that factor in everything from the change in weather, to traffic problems, and/or border crossing delays. o Table Games Revenue Optimization (TGRO) systems can be fed with additional data (like reservation check-in, mobile, location and/or social data) so that activity beyond the table game floor can be added to predictive models. o Slot floor optimization. o Casino floor maps that reveal floor traffic can assist in the opening and closing of tables. Retail and F & B: o Outliers in a data set can uncover potentially fraudulent activity on POS systems. o Retailers can better target merchandise, sales, and promotions and help redesign store layouts and product placement to improve the customer experience. 17 ANDREW PEARSON • Security: o Uncover AML activity. o Spot blacklisted players trying to gamble. o Reduce exposure to high risk patrons. o Uncover gaming scams. o Spot the use of unauthorized video cameras on the casino floor. o Spot dealing patterns that might reveal the playing cards haven’t been shuffled properly, or at all. o Motion-detecting roulette wheels can catch roulette cheats trying to add chips after the roulette wheel has been spun. In chapter one, I detail the different stream processing engines, each ones pros and cons, as well as explain the required components for a stream processing system. Since this is a highly complex system, there are few market-ready 13 products available, and a lot of custom coding is required to implement them. However, products like Apache Storm, Apache Spark, IBM InfoSphere Streams, Hitachi’s Pentaho platform, TIBCO StreamBase, ITRS’s Valo, and Apache Samza are all interesting platforms to study. In chapter seven, I detail how a stream processing engine would actually work for a digital marketing platform. In his article How Real-time Marketing Technology Can Transform Your 15 Business , Dan Woods makes an amusing comparison of the differing environments that marketers face today as compared to what their 1980s counterparts might have faced: “Technology has changed marketing and market research into something less like golf and more like a multi-player firstperson-shooter game. Crouched behind a hut, the stealthy marketers, dressed in business-casual camouflage, assess their weapons for sending outbound messages. Email campaigns, events, blogging, tweeting, PR, ebooks, white papers, apps, banner ads, Google Ad Words, social media outreach, search engine optimization. The brave marketers rise up and blast away, using weapons not to kill consumers but to attract them to their sites, to their offers, to their communities. If the 15 weapons work, you get incoming traffic.” Real-time stream processing is an integral part of this rapidly changing marketing environment and if casinos and IRs don’t join the real-time marketing world, they will be left behind, I have no doubt. Successful mobile advertising requires three things—reach, purity and analytics; reach can be fostered by accessing accounts through multiple platforms like blogs, geofencing applications, OTT services, mobile apps, QR codes, push and pull services, RSS feeds, search, social media sites, and video- 18 THE PREDICTIVE CASINO 16 casting, amongst others. “Purity” refers to the message and its cleanliness; if the data is unstructured and untrustworthy it is, basically, useless and data 16 governance is paramount for real-time advertising to work properly. The third ingredient, analytics, “involves matching users’ interests–implicit and explicit, context, preferences, network and handset conditions—to ads and promotions 16 in real time.” Knowing what might interest a consumer is only half the battle to making the sale and this is where customer analytics comes in. Customer analytics has evolved from simply reporting customer behavior to segmenting a customer based on his or her profitability, to predicting that profitability, to improving those predictions (because of the inclusion of new data), to actually manipulating customer behavior with target-specific promotional offers and marketing campaigns. These are the channels that real-time thrives in and this is where a casino can gain a powerful competitive advantage when using them. Composing the marketing message, however, is probably the easiest part of the process. In its Delivering New Levels of Personalization In Consumer 17 Engagement , Forrester Research found that survey participants believed that personalization had the potential to increase traffic, raise customer conversion rates, and increase average order value. Surveyed marketers felt that personalization capabilities could improve a variety of business metrics, including customer retention (75%), lifetime customer value (75%), and 17 customer conversion rates (71%). Today, “Personalization” is becoming the optimum word in a radically different business environment and even though this personalization comes at a price— privacy—it is a price most consumers seem more than willing to pay if a recognized value is received in return. For the casino operator, “personalization” requires an investment in software analytics, but casino operators should recognize that this price must be paid because highly sophisticated consumers will soon need an exceptional casino experience to keep them from going over to a competitor. These survey participants see email, call centers, corporate websites, mobile websites and physical locations (such as stadiums, sporting venues and hospitality sites) as today’s key customer interaction channels, but their future marketing efforts would be “focused on mobile websites, applications, and 17 social media channels.” Casino operators should keep these channels in mind as they devise their customer experience (CX) campaigns. Understanding customer-specified preferences is imperative for personalization; “80% of marketing executives currently use them in some or all interaction channels. In addition, 68% of marketers personalize current customer interactions based on past customer interaction history. Other commonly used personalization methods used by nearly 60% of firms in some 19 ANDREW PEARSON or all of their interaction channels are based on the time of day or day of the 17 week of customer interactions.” Forrester Research states that the difficulties 17 of personalization include : 1. 2. 3. 4. 5. Continuously optimizing campaigns in response to a customer’s most recent interactions. Optimizing content or offers for each person by matching identities to available products, promotions, messages, etc. Creating a single repository containing structured and unstructured data about a consumer. Delivering content or offers to a customer’s chosen channel in real time for purposes of conversion. Analyzing all available data in real time to create a comprehensive, contextually sensitive consumer profile. The executives pooled by Forrester Research expected there to be a “huge rise in personalization using consumer’s emotional state, social media sentiment, 17 and context” as well. “Only 29% of respondents claim today to use inferences about the consumer’s emotional state in some or all channels. But 53% expect 17 to do this in two to three years’ time.” Forester’s report goes on to add, “Only 52% of marketers currently use sentiments that consumers express in social media to personalize interactions today, but fully 79% expect to do this in two to three years. In addition, only 54% capitalize on the consumer’s current 17 contextual behavior, but 77% expect to do so in two to three years’ time.” Today, mobile apps, mobile commerce, mobile chat, and mobile gaming have revolutionized the way people do business, seek entertainment, and gamble. Mobile commerce has now evolved into what has become known as “omnicommerce”, a seamless approach to selling that puts the shopper’s experience front and center, giving that shopper access to what he or she wants through these multiple channels. Mobile marketing via Bluetooth, OTT, SMS, MMS, CSC and/or QR codes has become some of the most effective marketing available, while social media has turned the normal channels of marketing on its head. By accessing the Web through a wireless connection, mobile users can surf the Internet as seamlessly as if they were using a PC at home. With little more than the touch of a button, photos and videos can be uploaded seconds after they are taken, then shared with the most intimate of friends or the most distant of peoples. Live streaming channels allow cheap video streams that can be viewed almost anywhere in the world. The mobile platform is so robust and it holds so much promise that if a marketing executive had been asked to dream up the perfect device to connect to, market to, and sell its company's products and/or services to its customers and potential customers, he or she could hardly have come up with something 20 THE PREDICTIVE CASINO more superior to it. One of mobile’s best features is its ability to cross-pollinate the marketing message through several mediums, which include social media— and I will expound upon this throughout the book. 18 In its paper 5 Marketing Prediction For the Next 5 Years , the B2C marketing cloud company Emarsys argues that, “Smart marketers need real-time insights into mobile marketing performance in order to understand how end users are (or aren’t) engaging with their mobile marketing programs or applications.” 18 Emarsys argues that : “We will move from a world focused on designing for mobile as a secondary approach, to designing for mobile first. E-commerce organizations will finally fully alter the online shopping experience from responsive to completely mobile experiences. This mobile-only approach will be different, as it won’t just be a smaller design but will also include more responsive websites and shopping experiences. The mobile-only experience will lead to fully tailored shopping experiences primarily designed for engagement on a mobile device.” Emarsys goes on to add that: “Within the next five years, consumers will be able to swipe right, up, and down to make their selections, all via their mobile devices. And when the consumer is ready to complete the transaction? Easy. It just takes one click; the purchase is complete, and the items arrive at the 18 consumer’s house.” For an IR, this does pose the question as to why this ecommerce concept isn’t utilized more within an IR? Offering guests the option to buy on site and have the products delivered to his or her home so they don’t have to lug all of the purchased items home should be a viable and promoted option as it considerably cuts down on supply chain needs. Much more than a wireless transmitter optimized for voice input and output, a mobile phone, a tablet, or a phablet is an always-on, anytime, anywhere marketing and sales tool that follows a mobile user throughout his or her digital day. It is also an entertainment, CRM, and social networking tool, which makes it, potentially, the most powerful device in the history of marketing and customer relations. The mobile device is, literally, a marketing tool that can— and usually is—personalized by its owner, and it is within reach of that owner almost every hour of every single day—once again a marketer's dream. Push technology even puts the power of communication into the hands of the marketer, allowing casino operators to both initiate contact with an opted-in customer and then send him or her a wide array of products and content. Most large Macau- or U.S.-based casinos now have mobile apps in which they can connect to their patrons. As long as a customer is opted into a CRM system, a casino operator can foster a two-way dialogue with that customer and this dialogue can grow more sophisticated over time as more is learned about the 21 ANDREW PEARSON customer’s wants, desires, habits, and needs. Push technology has moved from clumsy blanket SMS blasts (although this channel still seems to be very popular in Macau and China) to the sophisticated use of mobile apps that allow customers to interact with their personal patron card information, whether that is a points balance or the potential redemption of points within the casino property. I didn’t want this book to only focus on developments in the United States, as I believe some of the most interesting things happening right now in the casino industry occurs in Asia. I had thought this before I moved to Macau in 2011 and my suspicions were confirmed after I made a few trips into China during the ensuing years. Corporations of all kinds now use WeChat to connect with their customers or potential customers in highly unique and sometimes very lucrative ways. WeChat has also introduced an in-app payment system that allows users to make one-click payments from their bank accounts. A scanning feature lets users get pricing information from bar codes as well. In Hong Kong, the WeChat green and white logo is almost as ubiquitous as the blue and white Facebook logo and I would argue that, as a marketing vehicle, it is just as effective. Small mom and pop stores are using WeChat to market their wares, filling up the “Moments” thread with their latest offerings, whether they are clothes, food, shoes, handbags, etc. Since WeChat accepts payments, oftentimes buyers can purchase directly from their mobile phones. One of the most important elements of social media is its inter-connectedness. An upload to YouTube can go viral through Twitter, Facebook, LinkedIn, WeChat, WhatsApp, Youku, as well as a whole host of other social media and mobile media platforms. Within seconds, something uploaded onto a social media website in the US can end up on a mobile application in China or Japan or Korea, or almost anywhere else in the world that has mobile or Wi-Fi access. We are truly living in an interconnected world and this interconnectedness is creating a whole host of ways to market a product, a service, or even a casino. Of course, Macau casino companies cannot market their gambling offerings in China, but there is no restriction to market other IR activities on the mainland. In their 5 Marketing Predictions For the Next 5 Years, Emarsys concludes that, “In an effort to remain competitive and innovative in today’s digital and always-connected world, marketers should continually be piloting and testing mobile strategies with a small subset of their users or target audience. If a brand slows mobile innovation, or pauses testing and optimization for mobile devices, the brand is risking the loyalty of current users as well as jeopardizing 18 new user acquisition.” Social media will also be explored in depth throughout this book. It is quite 22 THE PREDICTIVE CASINO ironic that, in one sense, engaging in social media can be one of the most antisocial behaviors one can do; sitting alone in a room, typing away on a computer was once the realm of solitary computer geeks, but it has now become an activity that most people engage in almost every single day. Perhaps this is because human beings are, first and foremost, social beings and we crave a connectedness that social media offers, even if it is virtual. It should be of no surprise that one of the greatest inventions of the twentieth century–the Internet–would became the watering hole of the twenty-first century; a place where human beings can quickly gather to socialize and connect with friends, family members and acquaintances in a way that was almost unimaginable only fifteen years ago. Smart casino marketers can tap into this interconnectedness to gets its marketing message out far and wide. About a decade ago, “most consumers logged on to the Internet to access email, search the Web, and do some online shopping. Company Web sites functioned as vehicles for corporate communication, product promotion, customer service, and, in some cases, e-commerce. Relatively few people were members of online communities; social networking sites were for college 19 students” and “Liking” something had no social relevance at all. How times have changed. “Today, more than 1.5 billion people around the globe have an account on a social networking site, and almost one in five online hours is spent on social 19 networks—increasingly via mobile devices.” In little more than a decade, social technology has become a cultural, social, political and economic 19 phenomenon. Most importantly, “hundreds of millions of people have adopted new behaviors using social media—conducting social activities on the Internet, creating and joining virtual communities, organizing political 19 activities” , even, as with the case of Egypt’s “Twitter Revolution”, toppling corrupt governments. The secret to social media’s growth is right there in its name—“Social”—as in the fundamental human behavior of seeking “identity and ‘connectedness’ through affiliations with other individuals and groups that share their 19 characteristics, interests, or beliefs.” For Chui: “Social media taps into well known, basic sociological patterns and behaviors, sharing information with members of the family or community, telling stories, comparing experiences and social status with others, embracing stories by people with whom we desire to build relations, forming groups, and defining 19 relationships to others.” Social technologies allow individuals to interact with large groups of people at almost any location in the world, at any time of the day, at marginal, if not no 19 cost at all. With advantages like these, it is not surprising that social media 23 ANDREW PEARSON has become so widespread that almost one in four people worldwide uses it. It is actually surprising that the figure is so low, although with mobile technology rolling out in some of the most remote locations on earth, that figure is sure to climb rapidly over the next few years. Businesses are quickly recognizing the power of social media. “Thousands of companies have found that social technologies can generate rich new forms of 20 consumer insights–at lower cost and faster than conventional methods.” In addition to this, businesses can watch what “consumers do and say to one another on social platforms, which provide unfiltered feedback and behavioral 19 data (i.e., do people who “like” this movie also “like” this brand of vodka?).” This can be a treasure trove of company competitive analysis and casino operators would profit from spending more on their social media listening efforts rather than holding expensive focus groups. Social technologies also “have enormous potential to raise the productivity of knowledge workers,” a very significant development in a world where knowledge workers are becoming highly sought-after assets. “Social technologies promise to extend the capabilities of such high-skill workers (who are increasingly in short supply) by streamlining communication and collaboration, lowering barriers between functional silos, and even redrawing the boundaries of the enterprise to bring in additional knowledge and expertise 20 in ‘extended networked enterprises.’” In a place like Macau, where the labor market is tight, raising the productivity of skilled workers would be a welcome benefit. In this book, I will use Chui et al.’s definition of “social technologies” as the “products and services that enable social interactions in the digital realm, and 20 thus allow people to connect and interact virtually.” These include: “A message to be communicated (a tweet or a blog), adding content to what is already online, or adding information about content (‘liking’ a piece of content). Content creation also includes performing an action that an individual knows will be automatically shared (e.g., listening to a piece of music when you know your music choice will be displayed to others). Social technologies allow anyone within a group to access and consume content or information. They include technologies that also have been described as ‘social media,’ ‘Web 2.0’ and 20 ‘Collaboration tools’.” In their book Marketing Communications: Integrating Offline and Online with 21 Social Media , P.R. Smith and Ze Zook show just how powerful social media marketing can be. Smith and Zook looked at the target audiences for three different types of marketing platforms–broadcast network, telephone and 21 email network, and social media. 24 THE PREDICTIVE CASINO According to Smith and Zook, “Broadcast network is based on a ‘one to many’ 21 model (e.g., old TV advertising). It is the Sarnoff network (after David Sarnoff, 21 the broadcasting legend). A hypothetical Sarnoff network with 20 viewers has 21 a score of 20. The network score is simply the number of nodes (i.e., audience 21 members)” and this equates to a paltry sum of twenty individuals. The telephone and email network is based on the Metcalf model (named after Bob Metcalf, one of the inventors of the Internet) and this is a “many to each 21 other” model. This model allows everyone in the group to connect with 21 everyone else. Because any member of the group can contact anyone else in 21 the group, the total number of potential contacts is 20 squared, or 400. Obviously, this is a much more powerful communication model than the Sarnoff model as the network score is the node number to the power of 2, 21 which is 400. A good number, but it still pales in comparison to the social network model. Named after David Reed (who noticed that people in social situations usually belong to more than just one network), the social network model is a “many 21 belong to numerous networks” model. “The possible value of a Reed network is two to the power of the number of nodes on the network,” explain Smith 21 and Zook. If you take the same group of 20 people in a social situation, a 21 “Reed network generates a score of 2 to the power of the node” , which generates a network score of over one million people; obviously, this is a number exponentially higher than the number of people reached by the Sarnoff and Metcalf models. This is the power of social media and it cannot be underestimated. When coupled with mobile, that number can be even greater and, just as importantly, the reach can be lightning fast. In China, users spend more than 40 percent of their time online on social media websites, a figure that is expected to continue its rapid rise over the next few 19 years. “This appetite for all things social has spawned a dizzying array of companies, many with tools that are more advanced than those in the West: for example, Chinese users were able to embed multimedia content in social media more than 18 months before Twitter users could do so in the United 19 States.” Companies like WeChat are revolutionizing social networks, adding malls as part of their platforms, while Taobao has teamed up with Weibo to allow instant commentary and blogging on purchased items. yy.com has inverted the concept of reality TV, by taking a singing competition and broadcasting it over the Internet, while allowing viewers to directly remunerate the contestants. There is an old adage in social media marketing that says, “Content is king” and, with social media, that adage has never been more true. Those destined to succeed in the social media sphere won’t be the ones with the most content; they will be the ones with the best and most searchable content. And that content will drive eyeballs unlike any other form of marketing available. To 25 ANDREW PEARSON succeed in this new environment, casinos should think of themselves first and 22 foremost as creators and syndicators of content. But “If content is king, then ‘conversion is queen,’” argues John Munsel, CEO of Bizzuka, and he has an important point; social media is about converting customers, driving eyeballs, and building channels to reach them. For casinos, the content they can create is content around their casino floor activities, i.e., baccarat, poker, or blackjack competitions, or the introduction of new slot machines. Since these integrated resorts are more than just casinos, content can be around events held or upcoming musical acts or shows. If the casino contains a sports book, there are countless sporting events to market. IRs are filled with entertainment options and marketing events like movies, concerts, boxing and/or UFC matches, eSports events. These can be marketed through normal PR channels, as well as through inexpensive social media ones. Chapter three delves into the world of social media, including a breakdown of 23 Kaplan and Haelein’s six types of social media. In their influential article Users 23 of the world, unite! The challenge and opportunities of Social Media , Kaplan and Haenlein show how all social media websites can be broken down into one of six different types; collaborative projects; blogs and micro-blogs; content communities; Social networking sites; virtual game worlds; and virtual social 23 worlds. Anyone devising a casino social media marketing plan should find this chapter particularly helpful in understanding how to use each separate platform, both singularly, and combined together. Throughout the book, I will discuss mobile and social media in China. I am looking at that country individually not only because it is the biggest social media market in the world, but also because its censorship rules make it highly unique and a potential minefield for Macau casinos to navigate. Facebook might be the biggest social network in the world, but its penetration in China is minimal and it will probably remain so for a long time to come, not only because China sensors Facebook, but also because Facebook’s Chinese competitors are actually creating some very technologically savvy products. WeChat, in particular, has proven to be highly successful, and it is growing rapidly, both in China and throughout the rest of the world, but companies like QQ, Weibo, Hexun, Youku, Jiepang, Qieke, Ushi and Ku6 are all experiencing exponential growth. With a base of 1.3 billion people, it isn’t too hard for services that catch on in China to rapidly get to tens of millions of users within a year or even sooner. The Predictive Casino is all about understanding as many aspects of the integrated resort’s business as possible; knowing what is expected on the casino floor all the way down to the baccarat or blackjack table dealer needs can feed information into an employee app that would be the perfect channel to connect with dealers who might not be required for a shift. Those that are 26 THE PREDICTIVE CASINO needed for work could have seats automatically reserved on the employee bus that takes them to the casino, an act that could save the employee considerable time he or she normally spends waiting in line for the bus. By tracking patron entrances, the casino could quantify its labor needs almost in real-time and alert the appropriate departments, as necessary. Some of today’s IRs have over ten thousand employees and for the casino, which pays about $10/day to feed its staff, any labor and HR savings would go straight to the bottom line. In chapter four, I explain how a casino can use social media to succeed in today's cutthroat casino business environment, answering such questions as: • • • • How can a casino measure the benefits of social media? How should a casino organize its social media presence or presences? How should a casino spread social media usage throughout its organization? How has social media changed the relationship between a customer and a casino? I will also break down the ways in which a casino should use social media, including: • • • • • • • • • • • • • Adding interactivity to a Website Brand and Anti-brand management Brand loyalty enhancement Building fanbases Crisis management Discover a patron’s psychological profile Discover important brand trends Driving traffic to a Website Engaging customers and potential customers Harvesting customer feedback Marketing to consumers Reputation management Social shopping One company that is certainly doing analytics and customer service right is Disney. As Cliff Kuang explains in his article Disney’s $1 Billion Bet On a Magical 24 Wristband , Disney has created wristbands they call “MagicBands” that look like simple, stylish rubber wristbands. These contain an RFID chip and a radio transmitter that connects the wearer to a vast and powerful system of sensors 24 within the Disneyworld park. If visitors sign up in advance for the so-called “Magical Express”, the MagicBand replaces all of the details and hassles of paper once the visitor touches-down in 27 ANDREW PEARSON 24 Orlando. Express users can board a park-bound shuttle, and check into the 24 hotel wirelessly. Visitors don’t have to mind their luggage, because each piece 24 gets tagged at their home airport. Upon arrival at the park, there are no tickets to hand over, visitors just have to tap their MagicBand at the gate and 24 swipe for the rides they’ve already reserved. For Disney, this technology also helps them cut down on the need for hotel check-in staff as all of that work is handled without human interaction. With the MagicBand, “there’s no need to rent a car or waste time at the baggage carousel. You don’t need to carry cash, because the MagicBand is linked to your credit card. You don’t need to wait in long lines. You don’t even have to go to the trouble of taking out your wallet when your kid grabs a stuffed Olaf, looks up at you, and promises to be good if you’ll just let them 24 have this one thing, please.” Disney has, obviously, thought of everything. For Disney, the MagicBands are nothing less than thousands of sensors that communicate directly with the park’s IT databases that basically turn the park into a giant computer—streaming real-time data about where guests are, what 24 they’re doing, and what they want. Most importantly, the system is designed 24 to anticipate the guest’s desires, i.e., predicting their behavior. Chapter six delves into the operations side of the casino business. There, IoT and edge analytics takes center stage. Fraud can wreak havoc on a casino’s bottom line and every casino must be extremely vigilant to ensure that they don’t get hit by thieves and scam artists. In January 2017, a baccarat player 25 took the Parisian Macau for over HK $100M. As of today, no foul play is suspected, but a confederate dealer could potentially come up with some new way to cheat. Phil Ivey has also won millions from casinos in both London and Atlantic City, although the casinos challenged him in court, and both casinos actually won their cases, but still these cases are proving to be a public relations nightmare for Genting and an ounce of prevention could be worth it when millions of dollars are at stake. In chapter seven, I also attempt to create a holistic view of how the casino of the future—the Predictive Casino—would operate. I have added some realworld examples of how a casino IT department would build a data lake or a real-time streaming system that could surface information from facial recognition cameras, to geo-locating devices, to on-floor patron card swipes. This data would quickly become actionable once it is put into the hands of the front-line staff. My hope is this book can be a blueprint for a casino to step in to the Big Data/predictive analytics/Hadoop/Data Lake world, so that it can not only understand its customers on a truly intimate level, but also shape the experiences of those customers so that a healthy ROI can be created. Technology shouldn’t be embraced simply for technology’s sake and the macro 28 THE PREDICTIVE CASINO perspective should always be kept in mind as well; the IR business is first and foremost a business of tourism and casino companies should understand there are a multitude of factors that affect the business environment. As Ralf Buckley 26 argues in his article Tourism Megatrends : “Tourism is affected by social, political, economic, technological and environmental changes at all scales. Population growth, redistribution of wealth, geopolitical changes and conflicts, rising fuel costs, climate change and its consequences, new technologies and work patterns, and all forms of social fashion influence who wants to travel where, for how long, to do what, and at what prices.” In his article, Buckley examines six large-scale exogenous trends for the global 26 tourism sector that he sees for the next 30 years : 1. 2. 3. 4. 5. 6. The social, economic and environmental consequences of gradual warming and of extreme weather events associated with climate change; The effects of higher fuel costs and social concerns on mass long-haul travel; The role of new technologies, including social media, in marketing, managing, experiencing and monitoring tourism; Economic growth and social change in the highly populous and newly wealthy BRICS nations, especially India and China; The consequences of armed conflict and geopolitical negotiation for tourism, and the use of tourism as a tool for geopolitical interests; The increasing linkages, and also conflicts, between tourism and conservation in many countries. Improved understanding of these megatrends, and the interactions between them. A Predictive Casino should be able to factor in all of these variables and then figure out how this will affect a casino property and all of its staff, including the dealers, the surveillance officers, gaming supervisors, gaming managers, locksmiths, cage cashiers, slot technicians, valet dispatchers, cooks and sous chefs, sales consultants, hosts, bartenders, accountants, etc., etc., etc. This is not a simple task, but, with the right data being captured in the right place and funneled through the right analytical models and then delivered directly into the hands of the right staff members at the right time, it is more than just possible, it is imperative. Since tourism is such a travel-intensive business, the price of oil can wreak havoc on travel plans. As the casinos in Macau very well know, an edict from Beijing can cut the flow of billions of dollars into the SAR. In a business that requires millions of people to pass through its doors every single year, both social and macro-economic factors will play a big part in its success or failure. As Dwyer et al. state in their article Gambling with our 29 ANDREW PEARSON tourism future: the role of research in destination and enterprise strategies to 27 avoid strategic drift , children are growing up faster than ever before, but ironically, more adults want to be teenagers. The aspirational age of a 12 year old is 17. Products for children will need to have cool teen attributes. Adults are behaving more like teenagers in dress sense, eating habits, interests and 28 pastime. Tapping into the mindset of target segments, to determine how they think—not just how they behave will be important for marketing success, 27 conclude Dwyer et al. The experience economy and activity-oriented travel will play a big part in the future of travel and tourism. An increasing number of people will be money rich, but time poor, which will increase the demand for both customized products and customized marketing, 27 Dwyer et al. argue. In search for a deeper meaning in their lives, consumers 27 will seek out newer/richer/deeper experiences. “People will be more experimental, but won’t give a second chance to a product or service that fails 27 to satisfy.” In terms of values and lifestyles, ethical consumption will continue 27 increasing, while the distinction between work and leisure will blur. “Society and men will become more feminized with women having increasing influence 27 on all the key decisions,” Dwyer et al. conclude. Tomorrow’s tourist will be a technologically savvy traveler. Education could 27 prove to be a strong determinative of success in the future of the IR. “This includes innovative businesses that are well attuned to their customers needs, and staffed with highly educated workers valued as ‘human capital’ and organizations with external knowledge focusing on organizational culture that 27 enshrines life-long learning,” argue Dwyer et al. The Predictive Casino takes all of this into account as it shows casino companies how to build a social infrastructure that maximizes opportunities for individuals and businesses to be innovative, knowledge-seeking that gives access to the right knowledge to 27 its staff and executives. I started this introduction with an allusion of Big Data being either a treasure chest or a Pandora’s Box and it can certainly be either one, but going down the Big Data road requires a commitment that is all encompassing and difficult to implement. State of the art technology is required and that always means the potential for severe bumps in the road does exist. However, it is a road that must be traveled as today’s consumer have become highly sophisticated and, unless he or she is made to constantly feel a level of personalization marketing efforts are being reached, he or she will find another company that will provide this level of satisfactory service to him or her. Throughout the book, I will try to avoid what has become known as “wishcasting”, a useful term that the field of meteorology has recently given us. 29 As Rob Tracinski explains in his article How Not to Predict the Future , “It started with the observation that weathermen disproportionately predict sunny weather on the 4th of July and snow on Christmas Day. Their forecasts 30 THE PREDICTIVE CASINO are influenced not just by the evidence, but by what they (or their audience) want to hear.” The writer of any book that delves into current and future technology will, obviously, be susceptible to wishcasting, but I will try to temper my enthusiasm and add a dash of skepticism to all I write, even if I wish it were so. I will also try to avoid “Zeerust”—“The particular kind of datedness which 30 afflicts things that were originally designed to look futuristic.” Taken from The Meaning of Liff by Douglas Adams, TV Tropes explains it this way: “datedness behind zeerusty designs lies in the attempt of the past designers to get an advantage over the technology of their time, only to find out that more mundane designs are actually far more efficient if advanced engineering and 30 craftsmanship are used on them.” Throughout this book, I will offer my honest assessment of the technology I discuss, trying to be as agnostic and objective as possible. Personally, I prefer not to go down rabbit holes of technology that, while proving quite colorful, exciting and interesting, really lead to nowhere, so I will try to point out paths that I think advisable to both take and not to take, always keeping a firm eye on the casino company’s financial bottom line. 1 Ulanoff, Lance. Amazon Knows What You Want Before You Buy It. January 27, 2014. Predictive Analytics Times. http://www.predictiveanalyticsworld.com/patimes/amazonknows-what-you-want-before-you-buy-it/3185/ (accessed February 16, 2017). 2 http://www-01.ibm.com/software/data/bigdata/ (accessed: December 6, 2016) 3 Anderson, C. (2004). Wired. The Long Tail, pp. 171-177. 4 Berman, S. J. (2007). Executive Brief: Navigating the media divide: Innovating and enabling 3 Anderson, newC.business (2004). Wired. models.The IBM Long Institute Tail, pp. for171-177. Business Value. 4 Berman, S. J. (2007). Executive Brief: Navigating the media divide: Innovating and enabling new business models. IBM Institute for Business Value. 5 Gartner. Gartner’s 2016 Hype Cycle for Emerging Technologies Identifies Three Key Trends That Organizations Must Track to Gain Competitive Advantage. August 16, 2016. http://www.gartner.com/newsroom/id/3412017 6 Britt, P. (2013) Big Data Means Big Benefits for Entertainment: Caesars Exec, http://loyalty360.org/resources/article/big-data-means-big-benefits-forentertainment-caesers-exec, accessed 5 January 2016. 7 Marr, Bernard. May 2, 2016. Big Data in Practice. John Wiley & Sons. 8 Marr, Bernard. August 23, 2016. Will ‘Analytics On The Edge’ Be the Future of Big Data? Online: http://www.forbes.com/sites/bernardmarr/2016/08/23/will-analytics-onthe-edge-be-the-future-of-big-data/#124af7ea2b09 9 Gartner. (2013, December 12). Gartner Says the Internet of Things Installed Base Will Grow to 26 Billion Units By 2020. Retrieved from Gartner.com: http://www.gartner.com/newsroom/id/2636073 10 Ashton, K. (2009, June 22). That 'Internet of Things' Thing. Retrieved from RFID Journal: http://www.rfidjournal.com/articles/view?4986 31 ANDREW PEARSON 11 Morgan, J. (2014, May 13). A Simple Explanation of 'the Internet of Things'. Retrieved from Forbes.com: http://www.forbes.com/sites/jacobmorgan/2014/05/13/simpleexplanation-internet-things-that-anyone-can-understand/ 12 Top 8 Trends in Big Data for 2016. http://www.tableau.com/about/blog/2015/12/top-8-trends-big-data-2016-47846 13 Wähner, Kai (2014, September 10). Real-Time Stream Processing as Game Changer in a Big Data World with Hadoop and Data Warehouse. InfoQ. https://www.infoq.com/articles/stream-processing-hadoop/ 14 Interactive Advertising Bureau. (2012). Attribution Primer. Retrieved from iab.net: http://www.iab.net/media/file/AttributionPrimer.pdf 15 Woods, D. (2011, May 6). How Real-time Marketing Technology Can Transform Your Business. Retrieved from Forbes.com: http://www.forbes.com/sites/ciocentral/2011/05/06/how-real-time-marketingtechnology-can-transform-your-business/ 16 Sharma, C. H. (2008). Mobile Advertising: Supercharge Your Brand in the Exploding Wireless Market. John Wiley & Sons, Inc. 17 Forrester Research. (2013, November). Delivering New Levels Of Personalization In Consumer Engagement. Retrieved from sap.com: https://www.sap.com/bin/sapcom/he_il/downloadasset.2013-11-nov-21-22.deliveringnew-levels-of-personalization-in-consumer-engagement-pdf.html 18 Emarsys. 5 E-Commerce Marketing Predictions for the Next 5 Years. https://www.emarsys.com/en/resources/whitepapers/5-e-commerce-marketingpredictions-for-the-next-5-years/ 19 Chiu, C. I. (2012, April). Understanding social media in China. Retrieved from www.mckinsey.com: http://www.mckinsey.com/insights/marketing_sales/understanding_social_media_in_c hina 20 Chui, M. M. et al. (2012). The social economy: Unlocking value and productivity through social technologies. McKinsey Global Institute. 21 Smith, P. Z. (2011). Marketing Communications: Integrating Offline and Online with Social Media; Third Edition. Kogan Page. 22 Black, L. M. (2012, November 11). 7 social media marketing tips for artists and galleries. Retrieved from Mashable: http://mashable.com/2012/11/10/social-mediamarketing-tips-artists-galleries/ 23 Kaplan, A. H. (2010). Users of the world unite! The challenges and opportunities of social media. Business Horizons, Vol. 53, Issue 1. 24 Kuang, Cliff. March 3, 2015. Disney’s $1 Billion Bet On A Magical Wristband. Wired. https://www.wired.com/2015/03/disney-magicband (accessed 19 January 2017). 25 http://hk.apple.nextmedia.com/news/first/20170127/19910578 26 Buckley, Ralf, Gretzel, Ulrike, Scott, Daniel, Weaver, David & Becken, Susanne. 14 November 2013. Tourism Megatrends. Tourism Recreation Research. Volume 40, 2015 – Issue 1. 27 Dwyer, Larry, Mistilis, Nina, Edwards, Deborah, Roman, Carolina. Gambling with our tourism future: the role of research in destination and enterprise strategies to avoid strategic drift. http://scholarworks.umass.edu/cgi/viewcontent.cgi?article=1387&context=ttra 32 THE PREDICTIVE CASINO 28 Profound and Duerbeck, K. (2005) Natural Ingredients For Cosmetics. Centre for Promotion of Imports from Developing Countries (CBI) 29 Tracinski, Rob. October 13, 2016. How Not to Predict the Future. Real Clear Future. http://www.realclearfuture.com/articles/2016/10/13/how_not_to_predict_the_future_ 111945.html 30 http://tvtropes.org/pmwiki/pmwiki.php/Main/Zeerust 33 34 THE PREDICTIVE CASINO CHAPTER ONE The Technology “A little bit of the right information, just a little bit beforehand—whether it is a couple of seconds, minutes or hours—is more valuable than all of the information in the world six months later...this is the two-second advantage.” ~ Vivek Ranadive Founder and CEO of TIBCO Overview We live in an instant gratification culture and the companies that are likely to survive in this culture will be the ones who can not only keep up with the demands of their tough customers, but also predict what these customers will be wanting down the line. Today, IRs need every advantage they can get so that they provide better service than their competitors. A few industry examples from Information Week’s In-memory Databases, IBM, Microsoft, Oracle, and SAP Are Fighting to Become Your In-memory Technology 31 Providers. Do You Really Need the Speed? might shed some light on why speed can be such an important differentiator when it comes to real-time 31 marketing and customer interactions. These include : 1. 2. 3. 4. Online gaming company Bwin.party uses in-memory capabilities to handle 150,000 bets per second. This compares to their normal system rate of 12,000 bets per second. For retail services company Edgenet, “in-memory technology has brought near-real-time insight into product availability for customers of AutoZone, Home Depot, and Lowe’s. That translates into fewer 31 wasted trips and higher customer satisfaction” ConAgra, an $18 billion-a-year consumer packaged goods company, “must quickly respond to the fluctuating costs of 4,000 raw materials that go into more than 20,000 products, from Swiss Miss cocoa to 31 Chef Boyardee pasta” and an in-memory system assists them in material forecasting, planning, and pricing. ConAgra also taps its in-memory solution to make company promotions more relevant by using faster analysis, which allows ConAgra and its retailer customers to command higher prices in an industry notorious for razor-thin profit margins. 35 ANDREW PEARSON 5. 6. Maple Leaf Foods, a $5 billion-a-year Canadian supplier of meats, baked goods, and packaged foods, finds that profit-and-loss reports which “used to take 15 to 18 minutes on conventional databases now 31 take 15 to 18 seconds on their in-memory platform.” Temenos, a banking software provider that uses IBM’s in-memorybased BLU Acceleration for DB2 system, reports that queries that used to take 30 seconds now take one-third of a second thanks to BLU’s columnar compression and in-memory analysis. Bwin.party’s advantage is a hugely important competitive advantage; not only are more bets being taken, but more customers are being made happy. In-play betting has become a huge revenue generator for sports book and speeding up the time it take to close bets means more money flows into the company’s coffers and, in some cases, better odds can be offered to bettors. For Temenos, in particular, that difference in speed means that mobile customers will be able to quickly retrieve all of their banking transactions on their mobile devices, rather than just their last five, which could mean the difference between handling customer issues on a mobile device rather than in 31 a company store. “Online or mobile interaction costs the bank 10 to 20 cents to support versus $5 or more for a branch visit,” therefore the cost savings are 31 substantial. Not all of these examples might seem easily relatable to the casino industry, but they could be as the potential to market to a patron when he or she is primed to accept the advertising is advantageous for both parties involved. Casino marketers don’t waste time advertising to consumers when they aren’t primed to accept the advertisements, and do market to consumers when and where they might want to use said advertisements. Throughout this chapter, I will break down the different technology that I believe can be utilized by the Predictive Casino to speed up the operation process, everything from old standards, like Customer Relationship Management (CRM), facial recognition, and mobile marketing technology, to some of the most interesting and innovative technologies around, like IoT, and stream processing, and up to cutting edge tech like Artificial Intelligence (AI) and Machine Learning (ML), as well as several other technologies in-between. Although the focus of this book is the casino industry, some examples I provide are for industries other than gaming. Many of the business use cases I discuss won’t be in the casino industry because there are no use cases, but explanations of how they are being used in the retail, property management, and advertising industry will still be relevant for a casino executive, I believe. This chapter doesn’t contain an authoritarian list of all the technology available—that would go stale within a few days of publishing—but I plan to continue this book as an ongoing series and will to add new technology to it as 36 THE PREDICTIVE CASINO necessary. However, there is no way a book of this size could contain all of the technology available to an IR and I want to focus on the ones I consider the most important. In his article Patron Analytics in the Casino and Gaming Industry: How the 32 House Always Wins , Scott Sutton lays out the backstory of how analytics is currently used in the casino and gaming industry: “In the 1980’s and 1990’s, casino patron loyalty programs, originally called ‘slot clubs’, started popping up in many of the larger casinos. These slot clubs encouraged customers to sign up for player cards and, in return for loyalty to the casino, patrons would receive rewards such as complimentary rooms, access to special events, and other offers. This was revolutionary, as it allowed casinos to track gaming behavior down to the individual level, leading to more accurate information about patrons’ gaming behavior and interests. The information could then be used to better segment customers, predict future behavior, and improve marketing outcomes. As casino analytics advanced, casino resorts started incorporating the relevant data from hotel, dining, retail, entertainment, and other outlets to get a more complete view of patron’s behaviors. A recent development is that many of the major gaming loyalty programs, especially those in competitive markets such as Las Vegas, are now also rewarding non-gaming spending in order to encourage customers to keep non-gaming spending at their respective properties, in addition to providing additional data about nongaming behavior.” This book is an attempt to chart the next course of analytics for casinos and IRs by utilizing IoT, geo-location capabilities, ML, facial recognition, AR, as well as many of the other technologies discussed here. Take for example a casino that has noticed there might be an interesting way to increase baccarat turnover by filling its baccarat stadium seating areas with as close to an even number of Banker gamblers as Player gamblers (baccarat is a game whereby the house allows the player to choose one of three positions, Banker, Player, or Tie, and then pays off 1-to-1 when Player wins, 1-to-0.95 when Banker wins and 8-to-1 when tie wins). Theoretically, if the house has a stadium seating area in which half of the people play the Banker position the whole time and the other half play the Player position the whole time, the house is in a no-lose position. It’ll pass Banker’s money to the Player when the latter wins, but, when Banker wins, the house gets to keep five cents of every dollar (more in no commission baccarat). The trick here is to try to get the stadium seating area as close to a 50/50 37 ANDREW PEARSON equitable split of Bankers and Players as possible. This is where the multitude of technology I discuss in this book will be of great assistance. This example will be referenced later in the book and I believe it is one of the rare ways a casino could actually legally raise the house edge. It is an interesting example to explore because it utilizes many of the technologies I mention in the book, including ML, geo-location, facial recognition, mobile marketing, etc., etc., and it could, theoretically, manipulate a customer all the way down to putting him or her into a seat that would prove advantageous to the house. For starters, lists of gamblers who have particular gambling propensities could be created for baccarat tournaments so that as close to an equitable draw of people who have a propensity to bet Banker play against those who have a propensity to play Player. One technology I haven’t chosen to discuss is cloud technology. Although it is complicated in its own right, the players are few and the concepts rather simple, so I didn’t feel the need to expound upon them here. Perhaps I am being remiss, but there are plenty of vendors who will gladly take a phone call to explain the intricacies of their systems, and whether you’re crunching numbers in the cloud or on premise, there is really no difference in the crunching of the numbers, while the reasons the numbers need to be crunched in the first place is what I want to focus on. Artificial Intelligence (AI) and Machine Learning (ML) According to Wikipedia, Machine Learning (ML) is the subfield of computer science that “explores the construction and study of algorithms that can learn from data. Such algorithms operate by building a model based on inputs and using that to make predictions or decisions, rather than following only explicitly 33 programmed instructions.” ML “evolved from the study of pattern recognition and computational learning theory in artificial intelligence” and it “explores the study and construction of algorithms that can learn from and make predictions on data—such algorithms overcome following strictly static program instructions by making data driven 33 predictions or decisions, through building a model from sample inputs.” 33 As per Wikipedia, ML can be broken down into the following three categories : 1. 2. Supervised learning: The computer is presented with example inputs and their desired outputs, given by a “teacher”, and the goal is to learn a general rule that maps inputs to outputs. Unsupervised learning: No labels are given to the learning algorithm, leaving it on its own to find structure in its input. Unsupervised learning can be a goal in itself (discovering hidden patterns in data) or a means towards an end (feature learning). 38 THE PREDICTIVE CASINO 3. Reinforcement learning: A computer program interacts with a dynamic environment in which it must perform a certain goal (such as driving a vehicle), without a teacher explicitly telling it whether it has come close to its goal. Another example is learning to play a game by playing against an opponent. There are so many use cases for ML and deep learning that it is impossible to create an exhaustive list here, but it is particularly useful for marketing personalization, customer recommendation, spam filtering, network security, optical character recognition (OCR), voice recognition, computer vision, fraud detection, predictive asset maintenance, optimization, language translations, sentiment analysis, and online search, amongst many others use cases. Machine-learning can be used to spot credit card or transaction fraud; ML can build predictive models of credit card transactions based on their likelihood of being fraudulent and the system can compare real-time transactions against these models. When the system spots potential fraud it can alert either the bank or the retail outlet where the transaction occurred. Although ML and data mining often employ the same methods and overlap 33 significantly, they do differ significantly. As Wikipedia explains : “While machine learning focuses on prediction, based on known properties learned from the training data, data mining focuses on the discovery of (previously) unknown properties in the data (this is the analysis step of Knowledge Discovery in Databases). Data mining uses many machine learning methods, but with different goals; on the other hand, machine learning also employs data mining methods as “unsupervised learning” or as a preprocessing step to improve learner accuracy. Much of the confusion between these two research communities (which do often have separate conferences and separate journals, ECML PKDD being a major exception) comes from the basic assumptions they work with: in machine learning, performance is usually evaluated with respect to the ability to reproduce known knowledge, while in Knowledge Discovery and Data Mining (KDD) the key task is the discovery of previously unknown knowledge. Evaluated with respect to known knowledge, an uninformed (unsupervised) method will easily be outperformed by other supervised methods, while in a typical KDD task, supervised methods cannot be used due to the unavailability of training data.” WEKA, a comprehensive collection of machine-learning algorithms for data mining tasks written in Java and released under the GPL, contains tools for data 39 ANDREW PEARSON pre-processing, classification, regression, clustering, association rules, and visualization. It has a very minimal learning curve compared to products like SAS’s Enterprise Miner. However, unlike SAS, it can become quite inefficient with larger datasets. Python and R are the most popular open sources solutions used for ML and they both have a large user base community. Scikit-learn combined with Pandas, Numpy, Seaborn and Matplotlib make implementing ML algorithms in Python very versatile and these provide more customization and utilization than R does. R does have a large and active user base. Its libraries contain a wide variety of statistical and graphical techniques as well. These include linear and nonlinear modeling, classical statistical tests, time-series analysis, classification, clustering, amongst others. Due to its S heritage, R has strong object-oriented programming capabilities. Other ML software includes Matlab, Scikit, Accord, Apache’s Mahout, Spark’s MLLib, H2O on Hadoop, ConvNteJS, SPSS, SAP’s Predictive Analytics library, even SQL Server is powerful enough to build some of these models. ML can help an IR discover customer segments that they may not realize were there; which customers want to be near the pool, and which ones need three morning papers before they can even get dressed in the morning. Armed with this kind of information, hotels can understand what matters the most to its guests at the individual level, enabling them to anticipate their guest’s needs before even the guest is aware of them. Even more, hotels can understand key characteristics of their most profitable customers and recognize the next important ones when they happen to login onto the hotel’s online reservation system. The use of deep neural networks and image classifiers can analyze and parse images, which can enable hotel marketers to monitor the images that provide the highest booking conversion rate through each channel. ML can also be used to compute dynamic clusters of guests to create fluid segmentation in real-time. As consumer buying habits or booking patterns evolve, fluid segmentation ensures the IR continues to reach the right guests, at the right time and price, through the right channels. Augmented and Virtual Reality Not just the stuff of science fiction anymore, Augmented Reality (AR) is now a part of our everyday life. In his article CrowdOptic and L’Oreal to make history 34 by demonstrating how augmented reality can be a shared experience , Tarun Wadhwa states that augmented reality works by “displaying layers of computer-generated information on top of a view of the physical world.” It is 40 THE PREDICTIVE CASINO “a technology that alters the perception of reality by distorting it, allowing 34 escape from it, and enhancing it—all at the same time.” 35 According to Webopedia.com, Augmented Reality or AR is : “A type of virtual reality that aims to duplicate the world's environment in a computer. An augmented reality system generates a composite view for the user that is the combination of the real scene viewed by the user and a virtual scene generated by the computer that augments the scene with additional information. The virtual scene generated by the computer is designed to enhance the user's sensory perception of the virtual world they are seeing or interacting with. The goal of Augmented Reality is to create a system in which the user cannot tell the difference between the real world and the virtual augmentation of it. Today Augmented Reality is used in entertainment, military training, engineering design, robotics, manufacturing and other industries.” 36 According to Gartner’s Top 10 Strategic Technology Trends 2017 , Augmented reality (AR) and virtual reality (VR) will “transform the way individuals interact with each other and with software systems creating an immersive environment. For example, VR can be used for training scenarios and remote experiences.” AR enables a blending of the real and virtual worlds, which “means businesses 36 can overlay graphics onto real-world objects.” Immersive experiences with AR and VR are reaching tipping points in terms of price and capability but will 36 not replace other interface models.” In the future, AR and VR are expected to expand beyond visual immersion and they might include all of the human 36 senses , although this is a very complicated thing to pull off as smell-o-vision tried many decades ago. According to its press release Gartner Says Augmented Reality Will Become an 37 Important Workplace Tool , “Augmented reality is the real-time use of information in the form of text, graphics, audio and other virtual enhancements integrated with real-world objects.” Tuong Huy Nguyen, principal research analyst at Gartner, states that “AR leverages and optimizes the use of other technologies such as mobility, location, 3D content management and imaging and recognition. It is especially useful in the mobile environment because it enhances the user's senses via digital instruments to 37 allow faster responses or decision-making.” Gartner believes “AR technology has matured to a point where organizations can use it as an internal tool to complement and enhance business processes, 37 workflows and employee training.” Gartner also believes that “AR facilitates business innovation by enabling real-time decision-making through virtual 41 ANDREW PEARSON prototyping and visualization of content.” 37 According to Deloitte, Wearable AR devices can “allow users to access standardized sets of instructions for a particular task in real time, triggered by 38 environmental factors and overlaid on the user’s field of vision.” Research has shown that overlaying 3D instructions over a real-life process can reduce the error rate for an assembly task by 82 percent, with a particularly strong impact 38 on cumulative errors due to previous assembly mistakes. “AR allows for improved senses and memory through the capture and enhancement of the user’s perspective. By recording video/audio, capturing images and removing elements that obscure the senses, AR technology allows users’ eyes to act as cameras, and can enhance the senses in ways not available naturally, such as night vision or the ability to zoom in on far-away objects,” 38 notes Deloitte. AR uses location-based data for navigation, overlaying digital maps and 37 directions on real-world environments. Through the lens of an AR device, a 37 user can receive visual guidance based on GPS technology. AR services generally fall into one of two categories—“location-based or computer vision. Location-based offerings use a device's motion sensors to provide information based on a user's location. Computer-vision-based services use facial, object 38 and motion tracking algorithms to identify images and objects.” Mr. Nguyen claims AR’s benefits include the “potential to improve productivity, provide hands-on experience, simplify current processes, increase available information, provide real-time access to data, offer new ways to visualize 38 problems and solutions, and enhance collaboration.” Augmented reality has many potential applications in the gaming and hospitality industry as well and the following ideas might seem a little like science fiction, but they are certainly within the realm of technical possibilities, and today there is no question that they would take the concept of personalization to a whole new level. These ideas might be a little ahead of their time, but they are perfect for the gaming industry as it might be one of the rare industries that can implement such a system because it has the financial muscle to develop AR applications, the need for in-memory computing platforms, as well as the databases that contain all of the necessary patron information that is required to make these complicated and holistic systems work properly. In his article Augmented Reality and Hospitality…the Next Generation of 39 Hotels? , Matt S-J lays out a very interesting scenario for AR in a hospitality environment, whether that is for a hotel, a standalone casino or an integrated resort. If a casino property provided its front desk staff with a pair of AR glasses that connected to its EDW that provided real-time patron information, the staff 42 THE PREDICTIVE CASINO would be empowered to greet and interact with a patron on a truly personal level. The clerks could know all of the customer’s past history and, perhaps even if these were well-known VIPs, the recent news headlines associated with them. This type of engagement would bring the concept of customer service to a whole new level, a level that would be unlike anything these patrons had ever 39 witnessed before, even if they were high-level celebrities. A guest who had stayed at the property in the past would immediately be identified and all of his or her preferences and necessary patron information 39 could appear on the AR glasses’ virtual screen. “The guest could be checked in before they even reached the door. The extent goes further as restaurants could identify guests allergies or preferences, orders would be recognized by dish then linked to the table and guest images shown to see who has ordered 39 what, so the food would always be served to the correct person.” Birthday or anniversary greetings could be offered up without having to research a patron’s profile or staff who interact with VIP guests could be 39 informed of sensitive topics to be avoided. Many of these things can be achieved through excellent staff, but they all require research, time and a good 39 long memory, which not everyone possesses. For the casino patron, AR could enhance his or her on-property experience considerably. By simply downloading the casino property’s AR app onto his mobile phone, the patron could be checked in virtually and then be given personalized directions to his room, where hotel staff members could greet him. A free bottle of champagne or Chateau Lafite wine could be awaiting him 39 in his suite. The casino’s general manager could even appear in a video to 39 offer a personalized greeting on the television. Continuing with the AR journey, a patron could go to one of the integrated resort’s restaurants and, when seeing an appetizing meal being brought out from the kitchen, he could whip out his mobile phone, snap a picture of the meal, quickly scan it on the app, and then discover that it is a dish of beef 39 wellington, and then, potentially, place an order for it. If interested, the patron could even pay for the dish on his mobile device, possibly utilizing patron points should he chose to cash those in. After dinner, if the patron is interested in going to one of the hotel bars, a quick scan of the line of people waiting to get into the bar would reveal the wait time needed. If the AR system connected with the hotel’s patron system (which revealed that he was a high-roller whose card allowed him to skip the line), the patron could be notified that he could jump the queue. If the patron didn’t have such a vaunted status and didn’t want to wait, he could be shown the 39 name and location of the hotel’s other bars, that might not be so crowded. The AR app could also help with hotel maintenance. As a user scans his or her 39 hotel room, the app could take notice of any minor maintenance issues. 43 ANDREW PEARSON These issues would not be highlighted for the user, but would be relayed to the 39 appropriate hotel maintenance departments so that they could be fixed. This, of course, does raise privacy issues, but they are probably nothing a good corporate lawyer couldn’t overcome. Continuing on the patron’s AR journey: if the patron liked to play golf, a quick 39 scan of the golf course with the AR app would reveal the average par shots. If 39 she chose to play, the app could then keep track of her score. Nearby structures could also be explained so that she could actually discover local 39 areas of interest. Discounts on services could also be pushed out to her and, if they were coupled with a dynamic pricing system, these discounts could actually help sell what might otherwise be empty seats in one of the integrated 39 resort’s venues. Pit bosses could also use AR glasses in a way that could help them adjust table game minimums. The patron’s information (such as his past history, including his average bet, daily theo, how long he normally plays for, whether he tends to buy back in or not) could be projected on AR glasses. This information could help the pit boss to raise or lower his table minimums with much more confidence than he currently can; solid analytics would replace gut instinct. The fact that all of this information was instantly available would make it very actionable. For a sports betting website, augmented reality could be used to offer live odds on players during a soccer match, a basketball game, or on a horse being paraded before a race. A punter could point his phone at a player on a soccer pitch or on a basketball court and see live odds of that player being the next scorer or being the Man of the Match. Bets could be done in one easy click and odds would be updated live throughout the games, or even, potentially, during a horse or dog race. Augmented Reality apps can be used in a whole host of industries, including archaeology, architecture, art, construction, education, entertainment, gaming, industrial design, medicine, the military, museums, navigation, sports, tourism and transportation. I will detail some of these applications later in the book, 34 but for now, I’d like to mention the CrowdOptic L’Oreal App that really did something unique in the AR field. After downloading an app for Toronto’s seventh annual Luminato Festival, attendees could point their phones at different places around David Pecaut 34 Square to see a “virtual gallery” that wasn’t visible to the human eye. As the festival-goers pointed their phones at different places around the square, they 34 could see and interact with artwork on their screens. As people explored the virtual art pieces, a heat-map was created that revealed where they were and 34 what they were looking at. What remained was “an entirely new type of digital art: a giant, crowd sourced, enormous, virtual mural of sorts that each 44 THE PREDICTIVE CASINO person had individually contributed to, just by participating. In other words, it 34 was the world’s first “human heat-map logo.” A casino could use similar technology to get an accurate heat map of the casino floor or the shopping mall or any place throughout the casino, all in real-time. The company behind the app, CrowdOptic, created a technology that utilized the GPS, accelerometer, and compass data from multiple phones to see what 34 people were focusing on. This technology allowed for things like being able “to see the vital stats of an athlete mid-game when you point your phone at 34 them.” Sports betting companies could also overlay the latest odds that this particular player might score the next goal or touchdown. “Another usage is seeing the social media hashtag and discussion of an event as it is occurring; if this can really work the way it is intended, it will have huge ramifications for 34 citizen journalists and emergency management.” So where is AR going? In his article Augmented reality: expanding the user 40 experience , John Moore claims that “app creators have begun to engage more of a mobile device’s sensors–accelerometers and gyroscopes, for example. Augmented reality apps that use detailed animations are also in the works. The objective: inject augmented reality technology in a wider range of apps to boost the user experience.” Pokémon Go was the first location-based augmented reality game that hit it big. Despite mixed reviews, the mobile app quickly became a global phenomenon and it was one of the most used and profitable mobile apps of 41 2016, having been downloaded more than 500 million times worldwide. It certainly revealed the enormous potential of AR and it proved, without a doubt, that the barriers to AR technology were limited and easily scaled by humans seeking out little dueling pocket monsters. Gartner believes “AR technology has matured to a point where organizations can use it as an internal tool to complement and enhance business processes, 37 workflows and employee training.” Gartner also believes that “AR facilitates business innovation by enabling real-time decision-making through virtual 37 prototyping and visualization of content.” AR has a great future and many online gaming companies are already developing software to tap into it. In 2016, Infinity Augmented Reality teamed 42 up with Google to reinvent online casino games. “By using augmented eyewear and a 360-degree camera that can capture the whole set up, it plans to take players right in the middle of the action. It will be a never before experience for people interested in online card games and the like,” Bidisha Gupta explains in her article Augmented Reality Is The future Of Real Money Gaming Industry.42 Since IRs already have so much of the infrastructure in place, there is no reason why they don’t jump on the AR bandwagon, especially as online AR gaming could become a great place to find new patrons for a 45 ANDREW PEARSON 43 bricks and mortar casino. AR glasses like the Icis look like normal eyewear and at a cost of US $590, they are not prohibitively expense for most casino operators. Cryptocurrencies On October 31, 2008, Satoshi Nakamoto published a paper entitled Bitcoin: 44 Peer-to-Peer Electronic Cash System and with it he (whomever “he” is as the name turned out to be an alias) ushered in the world of cryptocurrencies. The most important fact about bitcoin and cryptocurrencies is that the transactions are made with no middle-men, meaning no banks. Transactions can be done 44 anonymously and they cost nothing. According to Nakamoto : “A purely peer-to-peer version of electronic cash would allow online payments to be sent directly from one party to another without going through a financial institution. Digital signatures provide part of the solution, but the main benefits are lost if a trusted third party is still required to prevent double-spending. We propose a solution to the doublespending problem using a peer-to-peer network. The network timestamps transactions by hashing them into an ongoing chain of hash-based proof-of-work, forming a record that cannot be changed without redoing the proof-of-work. The longest chain not only serves as proof of the sequence of events witnessed, but proof that it came from the largest pool of CPU power. As long as a majority of CPU power is controlled by nodes that are not cooperating to attack the network, they'll generate the longest chain and outpace attackers. The network itself requires minimal structure. Messages are broadcast on a best effort basis, and nodes can leave and rejoin the network at will, accepting the longest proof-of-work chain as proof of what happened while they were gone.” 45 As the CNN Money article What is Bitcoin explains, Cryptocurrencies can be used to buy merchandise anonymously. “International payments are easy and cheap because bitcoins are not tied to any country or subject to regulation. Small businesses may like them because there are no credit card fees. Some 45 people just buy bitcoins as an investment, hoping that they’ll go up in value.” Bitcoins can be purchased from several bitcoin exchanges and people send 45 bitcoins to each other exactly the same way cash is sent digitally. The Chinese, in particular, have flocked to bitcoins, so much so that the Chinese government every now and then steps in and tries to stop its citizens from trading them. 46 THE PREDICTIVE CASINO As per CNN Money, “Bitcoins are stored in a ‘digital wallet,’ which exists either in the cloud or on a user’s computer. The wallet is a kind of virtual bank account that allows users to send or receive bitcoins, pay for goods or save their money. Unlike bank accounts, bitcoin wallets are not insured by the 45 FDIC.” Even though “each bitcoin transaction is recorded in a public log, names of buyers and sellers are never revealed—only their wallet IDs. While that keeps bitcoin users’ transactions private, it also lets them buy or sell anything without easily tracing it back to them. That’s why it has become the currency of choice 45 for people online buying drugs or other illicit activities.” Governments are now weighing in on cryptocurrencies and trying to figure out how to oversee and tax them. In his article Controversial tech platform for 46 bitcoin goes mainstream , John Holden notes that “Blockchain technology, the digital ledger underpinning cryptocurrencies such as bitcoin and ethereum, is moving one step closer to legitimisation with the launch of the Blockchain Association of Ireland, a non-profit advocacy group for the technology in Ireland.” As Holden notes, “more and more people in areas such as business, finance and healthcare, are beginning to realise the potential offered by the blockchain— the secure, digital ledger where all cryptocurrency-based transactions are recorded and stored—as a potentially new and reliable data storage facility for any sensitive information—be it private health records, land registry or financial information. Now many of its once vocal critics have had a change of 46 heart.” Companies that take bitcoin as payment run the gamut of services, from Expedia, Virgin Galactic, Tigerdirect, 1-800-Flowers.com, Subway, Dell, Zynga, Bloomberg.com, Etsy, Whole Foods, to even such retail stalwarts as Kmart and Sears. Besides Bitcoin, the cryptocurrencies field is getting crowded, with companies like Litecoin, ethereum, Peercoin, Primecoin, Namecoin, Ripple, Quark, and 47 Mastercoin and Nxt offering similar financial services. These companies have market caps in the tens of millions, even up to over a billion, and some are even attracting millions in venture capital funding, including from Google Ventures (Ripple), so they are probably here to stay, in one form or another. Facial Recognition Facial recognition technology is the capability to identify or verify a person from a digital image or a video frame from a video source by comparing the actual facial features of someone on camera against a database of facial images, or faceprints, as they are also known. As patrons enter a casino property, “security cameras feed video to computers that pick out every face in the crowd and rapidly take many measurements of 47 ANDREW PEARSON each one’s features, using algorithms to encode the data in strings of 48 48 numbers.” These are called faceprints or templates. The faceprints are compared against a database, and when there’s a match, the system alerts casino hosts, bit bosses, the VIP department, salespeople, or security guards, if anyone has been caught cheating, stealing or shoplifting in the past. Casino and/or retail personnel can receive alerts through a mobile app or an SMS when a member of a VIP loyalty program enters the store. A screen can display the shopper’s or casino patron’s name, or a photo just taken from the video feed. Shopping preferences, and other details, like a customer’s average daily Theo or ADT can also be displayed. Currently, facial recognition technology can be more useful for security 48 departments than customer service. At the 2014 Golden Globe Awards, facial 48 recognition technology was used to scan for known celebrity stalkers. The technology has also been used to bar known criminals from soccer matches in 48 Europe and Latin America. “Police forces and national security agencies in the U.S., the United Kingdom, Singapore, South Korea, and elsewhere are experimenting with facial recognition to combat violent crime and tighten 48 border security.” Facial recognition technology is becoming second nature to consumers, who are used to tagging themselves in photos on Facebook, Snapchat, Picasa, and/or WeChat. In 2015, Google launched a photo app that helped users organize their pictures by automatically identifying family members and 48 friends. Google, however, suffered a public relations and social media disaster 48 when its system labeled a photo of two black people as gorillas. The search 48 giant quickly apologized profusely and promised to fix its algorithms , but this does show that the technology isn’t foolproof. Currently, MasterCard is “experimenting with a system that lets users validate purchases by snapping a selfie. Like fingerprint scanners and other biometric technologies, facial recognition has the potential to offer alternatives to 48 passwords and PINs.” This technology is moving so fast, privacy advocates are having trouble keeping up. In this regard, today’s facial recognition is reminiscent of the World Wide 48 Web of the mid-1990s. Back then, few people would have anticipated that every detail about what we read, watch, and buy online would become a commodity traded and used by big business and sometimes, more sinisterly, 48 hacked and used by nefarious individuals to perpetrate crimes. Facial recognition technology “has the potential to move Web-style tracking 48 into the real world, and can erode that sense of control.” Experts such as Alvaro Bedoya, the executive director of Georgetown Law’s Center on Privacy & Technology, and the former chief counsel to the Senate’s subcommittee on 48 privacy, technology, and the law finds this attack on privacy alarming. 48 THE PREDICTIVE CASINO “People would be outraged if they knew how facial recognition” is being 48 developed and promoted, Bedoya states. “Not only because they weren’t told about it, but because there’s nothing they can do about it. When you’re online, everyone has the idea that they’re being tracked. And they also know that there are steps they can take to counter that, like clearing their cookies or installing an ad blocker. But with facial recognition, the tracker is your face. 48 There’s no way to easily block the technology,” Bedoya warns. Right now, facial recognition is largely unregulated and few consumers seem to even be aware of its use. “Companies aren’t barred from using the technology to track individuals the moment we set foot outside. No laws prevent marketers from using faceprints to target consumers with ads. And no regulations require faceprint data to be encrypted to prevent hackers from 48 selling it to stalkers or other criminals,” Bedoya warns. This is true for both the United States, Asia, and Europe. Users might be happy to tag their face and the faces of their friends and acquaintances on a Facebook wall, but they might shudder if every mall worker was jacked into a system that used security-cam footage to access their family’s 48 shopping habits. This could, however, be the future of retail, according to Kelly Gates, associate professor in communication and science studies at the University of California, 49 San Diego. In her article Our Biometric Future: Facial Recognition Technology 49 and the Culture of Surveillance , Gates argues that “Regardless of whether you want to be recognized, you can be sure that you have no right of refusal in public, nor in the myriad private spaces that you enter on a daily basis that are owned by someone other than yourself.” Gates concluded that by entering a retail establishment filled with facial recognition technology, you are tacitly 49 giving your consent to the retailer to use it, even if you are unaware of its use. Facial recognition technology in the offline world is now becoming more and more prevalent, particularly in the hospitality industry. “On Disney’s four cruise ships, photographers roam the decks and dining rooms taking pictures of passengers. The images are sorted using facial recognition software so that photos of people registered to the same set of staterooms are grouped together. Passengers can later swipe their Disney ID at an onboard kiosk to 48 easily call up every shot taken of their families throughout the trip.” There’s no reason why IRs don’t do something similar in their attraction areas. Starting in 2010, the 1,200-room Hilton Americas-Houston in Texas used a facial recognition system that was mainly designed as a security tool to identify 48 VIP guests so the hotel staff could greet them by name. The hotel won’t confirm if the system is still active, but similar technology is being rolled out at 48 hotels and casinos worldwide. Apparently, some surprising uses of facial recognition technology are also 49 ANDREW PEARSON popping up. In 2015, a company called Churchix revealed it had installed facial recognition systems in dozens of churches around the world to track service 48 attendance. Once made public, the technology received a wave of bad publicity, which is hardly a surprise as if there’s one place where people should expect privacy, it is within the confining walls of a church. Hearing that it isn’t God watching over them, but rather, literally, their priest, one would expect church goers to feel as though their privacy had ben violated and the churches that are using the technology seem to agree as none of them seem willing to go 48 on record to discuss it. Surveillance in the pews may seem like a particularly egregious violation of privacy, but evidence suggest facial recognition does tend to make people 48 uncomfortable wherever it appears. “In a recent study of 1,085 U.S. consumers by research firm First Insight, 75 percent of respondents said they would not shop in a store that used the technology for marketing purposes. Notably, the number dropped to 55 percent if it was used to offer good 48 discounts.” However, consumers may warm to facial recognition technology once it becomes more widespread, especially if retailers and/or casinos offer enough incentives to make it worthwhile. In some cases, full facial recognition isn’t needed, some marketers just want to determine the age, sex, and race of shoppers. In Germany, the Astra beer brand recently created an automated billboard 48 directed solely at women, even to the point of shooing men away. The billboard approximated the women’s age, then played one of 80 pre-recorded 48 ads to match. For a casino, this could help if they want to direct specific advertising towards women, or to men, or to a certain age group. Retailers can also utilize “facial recognition systems to see how long people of a particular race or gender remain in the shop, and adjust displays and the store 48 layout to try to enhance sales.” Using related technology, some high-end retailers in the U.S. have experimented with “memory mirrors” that perform tricks such as storing images of what shoppers tried on so that they can be revisited, or emailed directly to friends for feedback. In 2014, Facebook announced a project it calls DeepFace, “a system said to be 97.35 percent accurate in comparing two photos and deciding whether they depicted the same person—even in varied lighting conditions and from different camera angles. In fact, the company’s algorithms are now almost as adept as a human being at recognizing people based just on their silhouette 48 and stance.” “Entities like Facebook hold vast collections of facial images,” says Gates, the 50 THE PREDICTIVE CASINO 48 UC, San Diego professor. “People have voluntarily uploaded millions of images, but for their own personal photo-sharing activities, not for Facebook to 48 develop its facial recognition algorithms on a mass scale.” Potentially Facebook, Instagram, WeChat, Pinterest, Snapchat, Google, or a number of other social media companies could use their vast databases of 48 faceprints to power real-world facial recognition. “Hypothetically, a tech giant wouldn’t need to share the faceprints themselves. It could simply ingest video feeds from a store and let salespeople know when any well-heeled consumer 48 walked through the door.” It could also, potentially, do this for a casino operator as well, to prevent money laundering, Know Your Customer (KYC), or AML activities. Geofencing Applications Today, most smart phones have geofencing capabilities, which tap into GPS or RFID technology to define geographical boundaries. Basically, geofencing programs allow an administrator to set up triggers—usually SMS push notifications or email alerts—so when a device crosses a “geofence” and enters or exits a set area, a user is notified. Applications such as Facebook, Foursquare and China’s WeChat and Jiepang use geofencing to locate users, as well as help them find their friends and/or check into places. 50 As TechTarget explains, geofencing has many uses, including : • • • • • • Mobile Device Management—When a host’s tablet PC leaves the casino property an administrator receives a notification so the device can be disabled. Fleet management—When a truck driver breaks from his route, the dispatcher receives an alert. Human resource management—An employee smart card will send an alert to security if an employee attempts to enter an unauthorized area. Compliance management—Network logs record geofence crossings to document the proper use of devices and their compliance with established rules. Marketing—A retail business can trigger a text message to an opted-in customer when the customer enters a defined geographical area. Asset management—An RFID tag on a pallet can send an alert if the pallet is removed from the warehouse without authorization. With geofencing applications, “users can also offer peer reviews of locations, which add a layer of user-generated content. In exchange for loyalty, more and more businesses—from local retailers to larger organizations like Bravo TV, Starbucks and The History Channel—are offering coupons, discounts, free 51 goods and marketing materials.” 51 ANDREW PEARSON As users continue to enter personal details as well as update and check-in to their locations, geofencing applications like Foursquare can “collect a historical view of consumer habits and preferences and, over time, possibly recommend a much larger variety of targeted marketing materials in real time—as a 51 consumer walks into a store to look for a specific item or service.” In their paper On the Potential Use of Mobile Positioning Technologies in Indoor 52 Environments , Giaglis et al. claim there are six different types of service uses for mobile positioning technology (see Table 2). SERVICES EXAMPLES ACCURACY NEEDS APPLICATION ENVIRONMENT Emergency calls Medium to High Indoor/Outdoor Automotive Assistance Medium Outdoor Traffic Management High Outdoor Indoor Routing Medium Outdoor Group Management Lot to Medium Indoor Travel Services Medium to High Outdoor Mobile Yellow Pages Medium Outdoor Infotainment Services Medium to High Outdoor Banners, Advertisements Medium to High Outdoor People Tracking High Indoor/Outdoor Vehicle Tracking Low Outdoor Personnel Tracking Medium Outdoor Product Tracking High Indoor Location-sensitive billing Low to Medium Indoor/Outdoor EMERGENCY SERVICES NAVIGATION SERVICES INFORMATION SERVICES MARKETING SERVICES Alerts, TRACKING SERVICES BILLING SERVICES Table 2: Taxonomy of mobile location services Source: Durlacher Research Geofencing applications (aka Location Based Services (LBS)) like Jiepang and Foursquare are useful services for hotel and casino marketers as well. Macau casinos, specifically, should be exploiting this medium because of its high 53 concentration of mobile subscribers. In his article LBS Opportunities for Casino 53 Marketers in Macau , Chris Weiners offers the following ideas for casino operators to get their LBS promotions rolling: 1. 2. 3. Pick your LBS service and claim your location. Offer tips to customers via LBS. Reward loyalty creatively. Start by offering your most loyal customers rewards, special access, and other promotions. Those that become your “Mayor”—or any other significant title—should be rewarded for their loyalty. This is a great way to identify potential social influencers 52 THE PREDICTIVE CASINO 4. 5. 6. and utilize them to further promote your venue. Reward new customers: First time check-ins should receive special promotions or incentives as it is important to give people a reason to continuously check in to your establishment. Understand who your loyal customers are online, and work with them. Develop a plan to utilize these ‘influencers’ and tap into their social networks. “Casinos do it offline all of the time; develop a similar approach for high-valued customers online through social connections. Encourage your followers to promote their checked-in status to their friends via social networks and micro blogs like Sina and 53 Twitter.” Promote your services both on- and off-line. In May of 2013, Lighthouse Signal Systems launched its indoor positioning 54 system as an open service for Android app developers. Developers can use the technology to create Android apps that will help users find their way 54 through the vast indoor terrain of Las Vegas’ hotels and casinos. Although global positioning systems have made outdoor navigation as simple as following directions on a mobile device, indoor navigation isn’t so simple, it is actually one of the last major hurdles that smartphones have yet to truly 54 conquer. However, Cambridge, Mass.-based Lighthouse Signal Systems has launched a service that covers 20 million square feet of entertainment and 54 retail space at leading casinos and hotels on the Las Vegas Strip. Lighthouse is “making its service freely available to Android app developers, resort operators, retailers, and others seeking to enhance the visitor experience in Las Vegas. Indoor navigation is the Holy Grail for the mobile industry, and Lighthouse says it is the first to provide GPS-like indoor 54 positioning on a wide scale in a major U.S. metro.” “We are excited to support app developer partners as they create new mobile experiences with indoor positioning in Las Vegas, where large resort interiors have traditionally presented a vexing challenge for visitors,” said Lighthouse co54 founder Parviz Parvizi. The standard line is that casinos create circular floors that differ little from whichever direction you enter or exit them so that patrons will get lost on them and, therefore spend more money at the casino’s tables and slots, but times are changing. Casino operators now recognize the importance of getting their gamblers in front of their preferred gaming table or slot machine as quickly as possible. A line stretching out the door at the entrance of a casino in Singapore (because every guest’s passport must be checked to ensure a Singapore local isn’t attempting to slip in without paying the local’s entrance fee) means minutes of lost gaming time per person, which can add up to thousands of dollars of lost 53 ANDREW PEARSON revenue per day. Giving a gambler directions to his favorite slot machine bank or preferred baccarat table could mean, at minimum, decreasing a player’s frustration at not being able to find what he or she is looking for or, at best, increasing gaming floor revenue by increasing the gaming handle. “Providing location-based services does not really reduce how much time people spend at the resorts but instead has the potential to enhance the 54 overall experience,” said Parviz Parvizi. “From a resort owner perspective, the time that a visitor spends wandering around being lost is a wasted opportunity 54 that could be better and more profitably spent on gaming or entertainment.” The truly massive nature of today’s integrated resorts also means helping people navigate through them as quickly as possible will cut down on property wear and tear. Moving twenty thousand people quickly and efficiently through the halls of an integrated resort will help reduce wear and tear on things like carpets, elevators, escalators, toilets, etc., thereby reducing operational costs for the property quite substantially. Lighthouse’s platform “includes indoor geofencing: a hosting platform for 54 location-based offers and user analytics.” The apps include user opt-in agreements and developers cannot use the service to track mobile phone users 54 without user consent. The technology uses “a combination of WiFi fingerprinting and sensor data. As long as there are WiFi networks in the area, Lighthouse can provide positioning 54 info.” Google, Cisco, Ekahau, Euclid, Shopkick, PointInside, Aisle411, 54 Sensionlab, Indoor.rs, Yfind, and CSR are all developing similar systems. Mobile marketing in general and OTT, MMS and SMS marketing in particular can help casinos create a one-to-one, two-way interactive experience with its patrons. These channels are not just about sending out a simple message, but rather they are about starting a customer relationship that can be analyzed so that the casino has a 360-degree understanding of its patron. It is an understanding that includes his or her wants, desires and needs. Signing onto an OTT service like WeChat in Macau should give a user access to all kinds of information, such as his player card point balance, coupons to onsite restaurants or bars, signups to gaming tournaments, as well as free play gambling coupons. By giving customers instant access to the information they need when they need it most, a casino can enhance their patron’s on-property experience. Whether its patrons are on-property to play baccarat, poker, blackjack, slots, bingo, or if they want to gamble in the sports book, these instant messaging services can provide a patron with instant information that can not only enhance their experience but, potentially, shape it. 54 THE PREDICTIVE CASINO For indoor location-based services such as “libraries, museums, exhibitions, and hypermarkets, location awareness can become a crucial determinant of a superior quality of service, deriving from the ability to locate the co-ordinates of a person or an object with reasonable accuracy at all times, and thus provide 52 spatially-aware services.” For example, location-based services can help customers who are shopping in a large mall; by registering a shopping list before she arrives at the mall, a shopper can, upon entering the mall, receive a 52 suggested route that will guide her to the products she wants. This helps not only the shopper but also the retailer because the retailer will know the exact positions of its customers at all times, which can help with workforce allocation 52 and shelf replenishment. Customers won’t linger in a store as long as they normally would either because they will know exactly where every product they want to purchase is located, meaning the store would probably be less crowded than it otherwise would be. With some casinos today being massive IRs, indoor location-based services can help patrons navigate through an enormous property to the specific slot machine they want to play on. With location-aware technology, casinos can easily create personal guided tours for visitors as well. Perhaps one of the best uses of location-based services is in the Meetings, Incentive, Conferencing and Exhibition (MICE) space. The massive size of some IR exhibition halls can make finding a particular booth or floor section a daunting proposition. Indoor mobile communication technology with location awareness technology can help conference-goers navigate a vast conference 52 floor. Also, before arriving at a conference, a mobile user would be able to register his personal preferences and, once he enters the exhibition hall, a route map would be sent to his or her mobile phone. Vendor appointments could even be set up so that they are located near each other so that the conference-goer wouldn’t have to run around frantically trying to make 52 meetings that are spread out all over the convention floor. Besides geo-fencing applications, social media channels like Facebook, Instagram, Twitter, WeChat, as well as many others can reveal a patron’s location. Instagram tracks a user’s photos even if he or she doesn’t geo-tag them. As Cadie Thompson warns in her article Social media apps are tracking 55 your location in shocking detail , “While the picture sharing app does give users the option to name the location of where they are uploading an image, it also geotags an uploaded pic regardless if the user has selected the ‘Add to 55 Photo Map’ function.” Foursquare's check-in app Swarm also broadcasts 55 users’ location even if they have not selected a specific location for check-in. Many live-streaming apps like Periscope, youtube and several Chinese ones will also show the location of the user and this is information that can be utilized by an IR’s marketing department if it can exploit the information quickly enough. Although YouTube doesn’t have a filter for location, websites like 55 ANDREW PEARSON geosearchtool.com allows users to search by location. Advanced filters allow searching by keyword and within a certain designated area. Besides the normal geo-location apps, IRs should also look into the smaller ones such as Bizzy, Glympse, Neer (neerlife.com), and social gaming app Scvngr. Internet of Things (IoT) As previously mentioned the Internet of Things is “the network of physical objects that contain embedded technology to communicate and sense or 9 interact with their internal states or the external environment.” Technology costs are down, broadband’s price has dropped, while its availability has increased, there is a proliferation of devices with Wi-Fi capabilities and censors built into them, and smart phone penetration is exploding; all of these individual technological advances were good for the IoT, together they have 56 created a perfect storm for IoT. According to its Gartner Says the Internet of Things Installed Base Will Grow to 57 26 Billion Units By 2020 , Gartner claims that, “The Internet of Things (IoT), which excludes PCs, tablets and smartphones, will grow to 26 billion units installed in 2020 [sic] representing an almost 30-fold increase from 0.9 billion in 2009.” Gartner believes IoT product and service suppliers will generate incremental revenue exceeding $300 billion, mostly in services, in 2020. It will result in $1.9 trillion in global economic value-add through sales into diverse 57 end markets.” Simply put, IoT is the concept of basically connecting any device with an on and off switch to the Internet, including cell phones, coffee makers, washing machines, headphones, lamps, wearable devices and almost 11 anything else imaginable. Today, it is almost unimaginable how all encompassing the Internet of Things will be in our daily lives in the not-too-distant future. From such life-changing technology as Google’s driverless cars, which could help optimize traffic, thereby reducing traffic congestion, as well as making people more productive, to sensors that can help regulate room temperature thereby saving energy, IoT is definitely here to stay. “The growth in IoT will far exceed that of other connected devices. By 2020, the number of smartphones tablets and PCs in use will reach about 7.3 billion 9 units,” said Peter Middleton, research director at Gartner. “In contrast, the IoT will have expanded at a much faster rate, resulting in a population of about 26 9 billion units at that time,” Middleton adds. A casino can utilize IoT applications in the following ways: • Internet of things: o Smart parking 56 THE PREDICTIVE CASINO • • • • • • • • o Smartphone detection o Traffic congestion o Smart lighting o Waste management Smart Metering: o Smart grid o Tank level o Water flow o Silos stock calculation o Water leakages Security & Emergencies: o Perimeter access control o Liquid presence o Explosive and hazardous gases alerts Retail: o Supply chain control o NFC payment o Intelligent shopping applications o Smart product management Inventory optimization Logistics: o Quality of shipment conditions o Item location o Storage incompatibility detection o Fleet tracking Industrial Control: o Smart Warehouse o M2M applications o Indoor air quality o Temperature monitoring o Indoor location tracking o Vehicle auto-diagnosis Video analytics: o Object detection o Slip fall analysis o People counting Swimming pool remote measurement IoT isn’t a standalone technology and when combined with wearable technology that is equipped with AR, personalized interactions with the physical world can be created. IoT is faced with the typical problems of new technologies, a lack of standards as the big and small players jockey for position, although there is a movement in place to create a vendorindependent protocol that will allow devices to connect with each other under 57 ANDREW PEARSON the guise of a common service layer. Currently, IoT’s growing pains are being tackled and security issues are being addressed. The addition of edge analytics, which can reduce network and connectivity costs, is also circumventing the need for cloud integration. Improving computer processing power and memory in semiconductors and modules increases by the month. This should allow IoT devices to add an ML component, which should help IoT devices realize the potential of ambient intelligence, allowing them to grow smarter over time. However, heavy duty number crunching power is still needed when performing intensive predictive and prescriptive analytics. In his article The Data of Things: How Edge Analytics and IoT Go Hand In 58 Hand , Gadi Lenz explains that, although IoT data has similar characteristics to Big Data, it is much more complicated. IoT data is: • • • • “Messy, noisy, and sometimes intermittent because sensors are often deployed in the field. IoT data is ultimately collected by sensors sitting somewhere—for example, a sensor could be deployed on a telephone pole or street light. Sensors often cut in and out. Often highly unstructured, and sourced from a variety of sensors (fixed and mobile). Dynamic—‘data in motion’ as opposed to the traditional ‘data at rest’. Sometimes indirect—we cannot measure a certain relevant quantity directly, for example, using a video camera with video analytics to 58 count people in a certain area.” The idea of collecting all of this sensor information and bringing it into one centralized computing station is not viable over the long term, particularly as 58 the volume of IoT devices increases exponentially. “Bringing such a large amount of data into a relatively small number of data centers where it is then 58 analyzed in the cloud, simply [sic] not scale.” The cost, too, would be 58 prohibitive. “With so many devices producing so much data, a correspondingly large array of analytics, compute, storage and networking power and infrastructure is essential. Though analytics will be necessary to the growth and business value of IoT, the traditional approach to analytics won’t be the right fit,” Lenz 58 argues. Edge analytics, which is detailed throughout this book, addresses these problems. A casino operator can “harness the smartness of the myriad of smart devices and their low cost computational power to allow them to run valuable 58 analytics on the device itself.” As Lenz explains, “Multiple devices are usually connected to a local gateway where potentially more compute power is available (like Cisco’s IOx), enabling more complex multi-device analytics close 58 THE PREDICTIVE CASINO to the edge.” 58 Distributed IoT analytics would work in three ways, “simple” analytics would be done on the smart device itself, more complex multi-device analytics on the IoT gateways, and finally the high computational computing—the Big Data 58 analytics, if you will—would connect to and run on the cloud. “This distribution of analytics offloads the network and the data centers by creating a model that scales. Distributing the analytics to the edge is the only way to 58 progress,” advises Lenz. As the DHL Trend Research and Cisco Consulting Services paper Internet of 59 Things in Logistics explains: “With the advent of IoT, Internet connections now extend to physical objects that are not computers in the classic sense and, in fact, serve a multiplicity of other purposes. A shoe, for example, is designed to cushion the foot while walking or running. A street light illuminates a road or sidewalk. A forklift is used to move pallets or other heavy items. None of these have traditionally been connected to the Internet—they did not send, receive, process or store information. Nonetheless, there is information latent in all of these items and their use. When we connect the unconnected—when we light up “dark assets”—vast amounts of information emerge, along with potential new insights and business value.” A connected shoe can reveal the number of footfalls in a given period of time, 59 or the force with which the foot strikes the ground. A connected street light can understand traffic patterns, and “provide information to drivers or city 59 officials for route planning and to optimize the flow of traffic.” A connected forklift can be fitted with predictive asset maintenance alerts that can warn a 59 warehouse manager of an impending mechanical problem. In the Predictive Casino, cameras immediately pick up a customer once he or she enters or even when he or she boards a casino bus at the China-Macau border gate. In this case, it makes sense for the analytics to be done inside the camera itself, rather than having the data sent back to a centralized server as 58 that can be both inefficient and it risks bottlenecking. Lenz adds, “Edge analytics is all about processing and analyzing subsets of all the data collected 58 and then only transmitting the results.” So, the systems is essentially discarding some of the raw data and potentially missing some insights, but it should be a calculated loss as analyzing everything is just not productive in 58 most cases. “Some organizations may never be willing to lose any data, but the vast majority can accept that not everything can be analyzed. This is where we will have to learn by experience as organizations begin to get involved in this new 59 ANDREW PEARSON 58 field of IoT analytics and review the results,” adds Lenz. This is a exploding field and the rules are, literally, been written right now, as the sensors are being rolled out. The term “machine” is used here to mean the computer that ingests the video stream and runs the facial recognition algorithm. This, in principle, could be run in the camera itself, or very near to it. As smarter sensors are deployed throughout the casino, “it makes sense to do the analytics in the devices themselves rather than sending the data back for central analysis in a system, 58 which can be inefficient and risks bottlenecking.” However, some trade-off must be considered with edge analytics. Lenz notes that, “Edge analytics is all about processing and analyzing subsets of all the 58 data collected and then only transmitting the results.” Some of the raw data 58 is discarded and potentially some insights are lost. “The question is, Can we live with this ’loss’ and if so how should we choose which pieces we are willing 58 to ‘discard’ and which need to be kept and analyzed?” It’s also important to learn the lessons of past distributed systems. “For example, when many devices are analyzing and acting on the edge, it may be important to have somewhere a single ‘up-to-date view,’ which in turn, may impose various constraints. The fact that many of the edge devices are also 58 mobile complicates the situation even more.” Although incredibly powerful devices in their own right, mobile phones and tablets will never reach the capacity and compute technology of EDWs. Mobile Marketing If an advertising executive had set about to create the perfect marketing and advertising tool, she could hardly have created something superior to the mobile phone. Not only is the mobile phone within reach of its owner almost every single hour of every single day but, because it can connect to a marketer in a highly personalized way with the simple touch of a button, it has the potential to become not only more effective than television or radio advertising but, just as importantly, more analyzable. 16 As the authors of Mobile Advertising point out that, “With respect to targeting, no other medium can provide the accurate and rich user profile, psychographic, social engagement and demographic data available from mobile. No other medium has the viral capability that mobile possesses–within seconds following a simple click, a unit of advertisement can spread like wildfire.” No other media comes even remotely close to the data measurement capacity that mobile offers, which begins with exposure to the advertisement, followed by the persuasive effect of the advertisement and, finally, to the actual 16 purchase of a product. Just about every link in the marketer’s chain is 60 THE PREDICTIVE CASINO touched by mobile. In 1996, the Internet advertising landscape changed forever when Procter & Gamble convinced Yahoo! that it would only pay for ads on a cost-per-click 16 basis, rather than for banner ads. Procter & Gamble realized the importance of gaining truthful user metrics for Internet advertising and this move ushered in the world of Internet analytics; eyeballs were no longer the goal, click-thrus that showed actual product interest became paramount. According to the American Marketing Association, “advertising is the placement of announcements and persuasive messages in time or space purchased in any of the mass media by business firms, nonprofit organizations, government agencies and individuals who seek to inform and/or persuade members of a particular target market or audience about their products, 60 services, organizations, or ideas.” The time is right for mobile marketing. As Sharma et al., state in their book 16 Mobile Advertising : “the heavy lifting of measurements and metrics; of banner ad standards; of search keyword auctions; of advertising cost models and the new, digital ad networks that support them have been built. The groundwork for digital advertising in mobile is largely in place.” However, because there are so many players involved, the mobile advertising value chain 16 is incredibly complex. As the authors point out in Mobile Advertising, “the mobile value chain comprises advertisers, agencies, solution providers and enablers, content publishers, operators and consumers. Phone manufacturers or original equipment manufacturers (OEMs) are enablers in this value chain rather than 16 active participants.” The bottleneck in the chain arises because, even though there are only a limited number of mobile operators, the number of vendors in 16 the value chain is exceedingly high. Although this was written almost a decade ago, the complexity of the environment still remains and is something that must be kept in mind when developing mobile marketing campaigns. In their article The Typological Classification of the Participants’ Subjectivity to 61 Plan the Policy and Strategy for the Smart Mobile Market , Kim et al. argue that the core technologies of cloud computing can greatly enhance mobile marketing efforts. Without cloud computing, it would be impossible to successfully produce targeting context-aware ads, real-time LBS ads, interactive-rich media ads, mobile semantic webs or in-app ads, advanced banner ads or incentive-based coupon ads, AR or QR codes, social network ads, 61 and n-screen ads. It would be especially difficult integrating and converging 61 multifunctional mash-up ads involving a mix of the aforementioned. “Smart mobile advertising products continuously derive combined services where two or more advertising techniques integrate and interlock due to innovative 61 hardware or software technologies.” 61 ANDREW PEARSON Mobile advertising has the potential to give casinos the best bang for their marketing buck, but a mobile marketing campaign should not simply be viewed as an extension of a company’s Internet marketing brought to the mobile phone. In Mobile Advertising, the authors state that the three basic types of 16 mobile advertising are : • • • Broad-based brand advertising: broad-based campaigns that take advantage of user filtering and targeting. These can include subsidized premium content, sponsorships, video pre-rolls or intromercials, postroll video, on-demand mobile media and contextual or behavioral advertising. Interactive, direct response campaigns: these are opt-in campaigns in which the mobile user usually exchanges some personal information for some type of content. TXT short codes, mobile subscription portals, and user registration campaigns are all examples of this type of campaign. Highly targeted search advertising: mobile’s ability to inform advertiser of the user’s basic age, sex, and address information is far better than any other form of advertising. These campaigns include content targeted search advertising and paid placement or paid inclusion search. Although there were hints that a marketing revolution was underway at the st beginning of the 21 Century, few people would have predicted the radical changes that have transformed the industry today. In their article Interactivitys 62 Unanticipated Consequences for Marketers and Marketing , Deighton and Kornfeld argue that: “Mass communication technology empowered marketers with marketer-to-consumer tools such as radio, television and database-driven direct marketing. The digital innovations of the last decade made it effortless, indeed second nature, for audiences to talk back and talk to each other. They gave us peer-to-peer tools like Napster, eBay, Tivo, MySpace, YouTube, Facebook, Craigslist and blogs, and information search tools like Google and Wikipedia. Mobile platforms have given us ubiquitous connectivity, context-aware search, and the ability to tag and annotate physical spaces with digital information that can be retrieved by others. In sum, new traffic lanes were being built, not for the convenience of marketers, but for consumers.” Successful marketing is about reaching a consumer with an interesting offer when he or she is primed to accept it. Knowing what might interest the consumer is half the battle to making the sale and this is where customer analytics comes in. 62 THE PREDICTIVE CASINO Customer analytics have evolved from simply reporting customer behavior to segmenting customers based on their profitability, to predicting that profitability, to improving those predictions (because of the inclusion of new data), to actually manipulating customer behavior with target-specific promotional offers and marketing campaigns. Going back to the baccarat player example, simple data mining could be used to understand a baccarat player’s propensity to prefer the “Banker” or “Player” side of the bet and then he or she could be rewarded to play at a particular stadium seating bank according to how the real-time performance of the game is currently playing, i.e., is it more filled with more banker or player bettors. Data must be gathered from disparate sources and seamlessly integrated into a 63 data warehouse that can then cleanse it and make it ready for consumption. Trends that surface from the data mining process can help in monetization, as 16 well as in future advertising and service planning. As the authors’ state in 16 Mobile Advertising : “The analytical system must have the capability to digest all the user data, summarize it, and update the master user profile. This functionality is essential to provide the rich user segmentation that is at the heart of recommendations, campaign and offer management, and advertisements. The segmentation engine can cluster users into affinities and different groups based on geographic, demographic or socioeconomic, psychographic, and behavioral characteristics.” Of course, with all of this data collection comes justified privacy concerns and the most important aspect of mobile marketing is ensuring the consumer has 16 control of the advertising. Without this, it is doubtful mobile marketing will 16 reach its true potential. If mobile advertisers do allow users to configure and control the ads depending on where they are, what mood they are in, who they are with, and what their current needs and desires happen to be, mobile marketing could prove to be one of the most successful forms of advertising 16 available to casino marketers. The potential to market to an individual when she is primed to accept the advertising is advantageous for both parties involved. Marketers don’t waste time advertising to consumers when they aren’t primed to accept the advertisements, but do market to consumers when and where they might want to use the advertisements. One of the most interesting companies working in the digital marketing space right now is the customer experience management company Sitecore, whose platform, Sitecore 8 was released in late summer of 2014. In her article Sitecore 64 8 Aims to Add Context to Customer Connections , Ginger Conlon explains that Sitecore 8 helps marketers enhance the customer experience by delivering 63 ANDREW PEARSON consistent, integrated experiences across multiple channels, while responding to customer cues in real time. “The more you know about someone and the context they're in, the better experience you can provide for them,” Sitecore CEO Michael Seifert said during 64 his SItecore conference keynote. “Seifert emphasized that experiences are unique to each individual, so it's essential that marketers understand 64 customers' context and preferences at an individual level.” “The mass 64 experience won't last much longer,” he warned. Sitecore 8’s new features—enhanced data collection, automated testing and optimization, improved customization, and real-time reporting—are designed to make it easier to gather data from such disparate sources as CRM, ERP, and 64 customer service systems to create a holistic view of a customer. The “automated testing and optimization tools are designed to recommend the ideal content to present, the best segments to target, and the optimal paths to 64 conversion.” “Real-time reporting shows customer decision points, and 64 highlights what's working and what's not.” Seifert argues that technology is failing the marketer because each technology gives only a keyhole view of the customers rather than a holistic view, which means marketers aren’t given the 64 full picture. According to Seifert, “experience marketing tools are the next generation of marketing technology, designed to provide a single view of the customer in real 64 time to enable marketers to deliver personalized experiences.” By using these tools, marketers will get the full picture of their customers, rather than only 64 disparate fragments. “Only when you see the full experience can you predict 64 optimally,” Seifert added. 64 Seifert argues that customer experience can make or break a brand. “Brands must provide the right experience at the right time to get the attention and 64 loyalty of today's modern consumer,” he said. “Experience is the magic. It's what ultimately matters, the lasting impression. As you get that single view of the customer you'll know things about your customers that no computer alone can tell you. Imagine the kind of personalization you can do; the experiences 64 you can create,” Conlon marvels. Although I am describing Sitecore here, all of the other major players, including IBM, SAP, SAS, HDS, Microsoft, Oracle, etc., are jumping aboard the customer context bandwagon as it is an integral part of any customer experience (CX) solution. You’re going to find they generally all do the same thing, but with slight variations, some using proprietary codes, others more common programming language code, some using open source code as well. 64 THE PREDICTIVE CASINO Digital Interactive Marketing: The Five Paradigms Deighton and Kornfeld believe that in this new media environment, there are five emerging marketing paradigms that are responses to the decrease of 62 marketing’s power relative to the consumer. Digital interactive marketing has 62 little use for words such as “viewer” and “listener”. Even the label “consumer” is of limited value because today’s interactions with a person will include encounters that have nothing to do with consuming or being part of a “target market.” Deighton and Kornfeld see this new digital interactive 62 marketing breaking down into five different paradigms , as per Table 3. Interactive marketing paradigm Thought tracing How people use interactive technology How firms interpose themselves to pursue marketing goals Resulting digital media markets Firms infer states of People search the web mind from search terms for information and and Web page content browse for and serve relevant entertainment. advertising. A market in search terms develops. Activity tracing People integrate always-on computing into everyday life. Firms exploit information on proximity and pertinence to intrude. A market in access and identity develops. Property exchanges People participate in anonymous exchanges of goods and services. Firms compete with these exchanges, rather than participating with them. A market in service and reputation and reliability develops. Social exchanges People build identities within virtual communities. Firms sponsor or co-opt communities. A market in community develops, competing on functionality and status. Cultural exchanges People observe and participate in cultural production and exchange. Firms offer cultural products or sponsor their production. Firms compete in buzz markets. Table 3: Digital Interactive Marketing: Five Paradigms 62 Source: Journal of Interactive Marketing , 23 pg. 4-10 Today, when a user searches for information or entertainment on sites such as Google, she leaves a trail (also known as a “clickstream”) that reveals what is 62 on her mind. This information, which Deighton and Kornfeld refer to as “thought tracing”, may be “available to marketers in exactly the sense that it is available to marketers through Google, as a clue to our thoughts, goals and 62 feelings.” 65 ANDREW PEARSON Mobile and social media alter the marketing landscape because the ubiquitous nature of computing makes it an “always on” proposition; both the thought 62 and the activity are being traced. “The argument is that when a person is always connected to the Internet, the person is always in the market, always available to be communicated with, and always an audience” contend Deighton 62 and Kornfeld. Of course, most people don’t like to be marketed to continuously throughout the day so technology that allows people to filter out messages that don’t 62 interest them needs to be developed. However, customized marketing messages will be allowed to get through. Just as television demands its audience to sit through commercials in order to enjoy free programming, Deighton and Kornfeld contend that, “we will enjoy ubiquitous computer connectivity for the price of voluntary exposure to context-specific persuasion 62 efforts.” If businesses want to succeed in this new marketing environment they must become an ally to the marketed individual, someone who is actually sought out 62 as a person with cultural capital. “Property exchanges”, “social exchanges” and “cultural exchanges” are all paradigms that are “built on peer-to-peer interactivity motivated by the desire to exchange, to share information, or to 62 express one’s self” state Deighton and Kornfeld. Arguably, Internet property exchanges were introduced on a mass scale by Napster, which was the first company to allow users to share and exchange 62 files in an anonymous way. Unsurprisingly, Napster ran into trouble with copyright holders and quickly left the content exchange business, but sites such as eBay, Flicker and YouTube allow users to share and even sell their property over the Internet. This is a trend that is not going to go away any time soon. While the property exchange deals in things, the social exchange deals in 62 identities and reputations. In general, social networking sites let a person present a face to the world, “including information about whereabouts and action and a ‘wall’ on which friends can post short, often time-sensitive notes, allows people to exchange digital gifts, provides a marketplace for buying and 62 selling, and allows posting of photographs and video clips.” These sites allow for contextually relevant advertising because friends can share information amongst each other and some of this information can include a marketer’s message. Since this messaging is coming from a trusted source, the message is considered much more trustworthy and enticing and, therefore, much more likely to be acted upon. For example, “a recent Nielsen analysis of 79 campaigns on Facebook over six months showed that, on average, social ads—those that are served to users who have friends that are fans of or have interacted with the advertised brand and prominently call the relationship out—generate a 55 percent greater lift in ad recall than non-social 66 THE PREDICTIVE CASINO ads.” 65 One of the key criteria of mobile marketing is that a consumer must opt-in to the service. Mobile marketing is primarily a “pull” media model, meaning a consumer must sign up for the service rather than the traditional “push” media model, which gives the consumer no choice in whether they want to be advertised to or not. Casino mobile marketers must spend money to get users to sign up, but, if they do, the potential market for mobile marketing is huge. It is also a market that is rapidly evolving and its advantages include: • • • • • Ubiquity: mobile devices and their users are everywhere. Effective: over 90% of received text messages are read by the recipient. Powerful two-way dialogue: an instantaneous link between the business and its customer is generated. Economical: compared to other marketing channels, mobile marketing is incredibly cheap per marketed individual. Spam-free: in the U.S. (but not in many other parts of the world) it is illegal to send a text message to someone who hasn’t opted-in to a marketing campaign. 66 In her book, The Mobile Marketing Handbook , Kim Dushinski lists eight types of advertising campaigns that a mobile marketer can engage in: 1. 2. 3. 4. 5. Voice: this includes text-to-call messages in which users are sent a link that, when clicked upon, initiates a phone call to the company sending out the message. These days, Apple’s SIRI, Microsoft’s Tellme and Google’s Now are adding a whole new dimension to voice. Text messaging: this used to be the “now” marketing tool of mobile, and it is still one of the most important tools available. Text messaging includes both SMS and Common Short Codes (CSC), which are abbreviated phone numbers. Text messages are sent to mobile users, the content of which are limited only by SMS character limitations and the marketer’s overall imagination. Mobile web: most smart phones have the ability to connect to the web and many of them have graphic capabilities that rival computer screens. Mobile search: as previously discussed, a mobile user can search company listings through his or her mobile phone, just as he or she can find this information on the Internet. Mobile advertising: placing banner ads and text ads on mobile websites can build brand awareness. 67 ANDREW PEARSON 6. 7. 8. Mobile publicity: presenting a company’s executives as experts in his or her field can be useful to members of the media who need instant information for fast approaching deadlines. Social networking: done right, this can help marketers tap into wordof-mouth campaigns, which will, hopefully, have their marketing messages lighting up social media websites. Proximity marketing: Bluetooth and geofencing campaigns that invite users to accept a multimedia message can deliver unique and locationspecific marketing messages. To these eight, I would add another two—OTT and mobile apps marketing— and I will break each of these campaigns down throughout the rest of this book. Mobile Payments According to research firm Gartner’s Gartner Says Worldwide Mobile 67 Application Store Revenue Forecast to Surpass $15 Billion in 2011 , “The global market for mobile payments is forecast to grow about threefold by 2017 to some $721 billion worth of transactions, with more than 450 million users.” Personally, I think this is under-estimating the market, as this is becoming a very crowded space, but even those figures are quite impressive. Upstarts like WePay and Stripe are competing head-on with the likes of PayPal, Square, Google, and Apple, amongst others. As per Wikipedia.com, “Mobile payments, also referred to as mobile money, mobile money transfer, and mobile wallet generally refer to payment services operated under financial regulation and performed from or via a mobile 68 device.” Instead of paying with cash, a check, or a credit card, a consumer can use his or her mobile phone to pay for a wide range of services and digital or hard goods. Transaction costs, usually in the 2 – 3% ranges, can add up so many companies are clambering into the space. Sending money via a mobile device can be considerably easier–and cheaper– than handling cash. “Square Cash, the payments company’s mobile money app and competitor to Venmo, has today rolled out a new update that lets users send cash by way of Bluetooth Low Energy (BLE). Called ‘Nearby Payments,’ this new feature works on devices running iOS 8, the company says, and offers an alternative to its previous means of sending money by way of text message, 69 mobile number or email.” 70 In his article Three Trends That Will Make a Difference in Mobile Payments , Alberto Jimenez makes a compelling case that the current mobile payments 70 ecosystem has been set up properly for success. Jimenez states that : “When debating the future of mobile payments, there’s a fairly common argument that ‘it is not about payments, it’s about 68 THE PREDICTIVE CASINO commerce.’ This is the belief that mobile payments adoption isn’t just about the actual movement of funds, but more so focused on the broader mobile shopping experience. For the most part, this argument is accurate. However, it misses the point of payments as the enabling platform for monetizing new retail industry engagement services—commerce—in an increasingly mobile world. These two ideas don’t conflict with one another; but secure, frictionless payments, as a standalone capability is simply necessary to succeed in the emerging retail industry reality.” Besides the current mobile payments players, large supermarket chains are getting into the act. Stores such as British supermarket Tesco and France's Auchan hope their “digital wallets” will also give them more comprehensive 71 data about customers' shopping habits than ever before. In the US, retailers like Wal-Mart, Target and Best Buy have announced plans to set up a joint 71 digital wallet service—the Merchant Customer Exchange, or MCX. These stores hope to target advertise to their customers more effectively. This is CRM right down to the micro level, hitting the customer, literally, in their pocketbooks and there’s no reason why IRs and casino operators don’t explore this payment channel as well. These supermarket chains are joining a crowded market that includes banks, 71 credit card companies and tech giants like Google, Apple and Facebook. Each 71 company hopes their app will be the one that becomes the industry standard , but the competition is extremely fierce. Even PayPal, which is well established in the e-commerce space, is also experimenting with mobile payment 71 technology. “Retailers hope to attract customers to their own services by giving discounts and rewards to those using them, while also linking payments automatically to loyalty schemes and offering features like saved shopping 71 lists.” Britain's Centre for Economics and Business Research estimates that “UK retailers could have saved 463 million pounds ($770.32 million) in transaction costs in 2013 by shifting to mobile payments from cash, credit and charge 71 cards.” Besides these monetary advantages, mobile payments can reduce queue lengths in stores by speeding users through tills and they severely 71 reduce the cost of handling cash and card payments. “Handling cash—which accounts for over half retail transactions by volume in Britain—is costly for retailers as it needs to be counted and guarded, costs equivalent to about 2.5 percent of takings, compared with about 2 percent for processing cheques and 71 1 percent for debit and credit cards.” It is no different for retail stores and casinos throughout the world, the cost of physically handling money will always far exceed digital transaction costs. Adding blockchain technology can also reduce fraud and add layers of security that is far superior to most other 69 ANDREW PEARSON systems. Google Wallet “is one of the most hyped mobile payment systems on the market. The app allows users to store loyalty cards from various retailers, send 72 money to family and friends, and buy items instantly on mobile websites.” Although these are nice features, Google Wallet's most desirable attribut is that it allows users to store credit cards in their phone, creating a virtual 72 “wallet”, if you will. “Hundreds of thousands of merchants are supported, but Google Wallet has a long way to go before every consumer can use it at every 72 location.” Besides receiving 24/7 fraud monitoring, as well as purchase protection that covers 100 percent of all eligible unauthorized transactions, Google wallet users can easily disable their phone and/or credit and debit cards 72 within their Google Wallet accounts. Google is not the only company offering fraud protection, Apple is as well. WePay, a Palo Alto startup that provides payment services for marketplaces, crowdfunding platforms and other sites that enable online transaction between third-party buyers and sellers, is taking on another payments juggernaut, 73 Stripe, with a service that offers a no-fraud guarantee. This could be a game 73 changer as online merchants lose up to 1% of all transactions to fraud. “Since WePay takes on fraud management & covers loses, our partners immediately 73 save money,” CEO and co-founder Bill Clerico contends. “These savings come from both what they would have lost to fraud, as well as reduced spending on 73 systems to prevent fraud,” Clerico argues. PayPal is another company that has a massive presence in e-commerce, and it allows mobile users to link their bank account or debit/credit cards to their 72 PayPal accounts. From there, users can shop with a single tap at participating 72 merchants. One of the best features of PayPal’s system is, as David Kennedy, founder of TrustedSec (an information security firm), recently explained on Fox News is that it offers a "one-time credit card number where basically you use a 72 credit card once and then it changes the number again.” Personally, I can’t think of a better fraud prevention system. Another entrant into the mobile payments market is Loop and it has an entirely 72 unique take on the market. “Instead of using NFC (Near Field Communication) or some other familiar technology, Loop uses an add-on device (a Fob or a smartphone case) that can interface with 90 percent of retailers' existing Point 72 of Sale terminals.” The Fob—a small adaptor that plugs into a mobile Phone— retails for US $34 and it works with iPhone and Android phones. Apple is also working in this space. As Bill Hardekopf explains in his article Apple 74 Primed to Become Your Mobile Wallet , “Apple has been slowly laying the foundation to become the leader of e-commerce and mobile payments, and all the pieces are just about in place, including: 70 THE PREDICTIVE CASINO • • • • First, get everybody's credit card numbers in a database via a cool product that everybody wants and uses: iTunes. Done. Second, create a device that people carry with them at all times that will hold these credit card numbers: iPhone. Done. Third, create a very secure system so if a thief stole an iPhone, he would not be able to use the phone or the credit card numbers in it, making it much more secure than a regular leather wallet with credit cards. Use something like fingerprints that a thief couldn't reproduce or fake: TouchID on the iPhone 5S. Done. Fourth, put an NFC (near field communication) chip in the phone so the consumer simply verifies his ID with his fingerprint and then waves the iPhone at the register. The NFC chip then sends the encoded card 74 number. Almost done.” When The Wall Street Journal asked him about mobile payments, Tim Cook, the CEO of Apple, responded: "I think it's a really interesting area. We have almost 800 million iTunes accounts and the majority of those have credit cards behind them. We already have people using Touch ID to buy things across our store, so it's an area of interest to us. And it's an area where nobody has figured it out yet. I realize that there are some companies playing in it, but you still have a wallet in your back pocket and I do, too, which probably 74 means it hasn't been figured out just yet." Cook’s obvious implication is that that wallet may not be in your back pocket 74 for too much longer. Facebook is also working on a friend-to-friend payment system. “Facebook Messenger is all set to allow friends to send each other money. All Facebook has to do is turn on the feature, according to screenshots and video taken using iOS app exploration developer tool CyCript by Stanford computer science 75 student Andrew Aude.” “Messenger’s payment option lets users send money in a message similar to how they can send a photo. Users can add a debit card in Messenger, or use one they already have on file with Facebook. An in-app 75 pin code also exists for added security around payments.” With the likes of Facebook, WeChat, Apple, Google, Square, PayPal, and Twitter (amongst many, many others) working in this space, it is sure to be a robust and quickly evolving space and casino patrons will expect these payment options when they wander through a casino property. The Predictive Casino will allow patrons and customers to purchase via these mobile channels, possibly even letting purchases at a retail, restaurant, or bar locations with Bitcoins. 71 ANDREW PEARSON OTT According to Techopedia, an over-the-top (OTT) application “is any app or service that provides a product over the Internet and bypasses traditional distribution. Services that come over the top are most typically related to media and communication and are generally, if not always, lower in cost than 76 the traditional method of delivery.” Over-the-top content (OTT) refers to the delivery of video, audio and other media over the Internet without a multiple system operator being involved in the control or distribution of the content. The provider may be aware of the contents of the Internet Protocol packets but is not responsible for, nor able to control, the viewing abilities, copyrights, and/or other redistribution of the content. This is in contrast to purchase or rental of video or audio content from an Internet service provider (ISP), such as pay television video on demand or an IPTV video service. According to wikipedia, “Over-the-top messaging refers to a similar idea, where a third party provides instant messaging services as an alternative to text messaging services provided by a mobile network 77 operator.” WeChat is considered an OTT service and one recent example from that service will show just how powerful and disruptive a platform it has become—during the 2014 Chinese New Year, WeChat launched a feature that let users based in China send money to friends and family via a virtual red pocket, which is a 78 traditional gift of money shared during this important Chinese holiday. The feature went viral and on 4 February 2014, WeChat reported that during the first two days of the Spring Festival, over five million users participated in the 78 offering, exchanging over 20 million envelopes. After connecting one’s bank card to the WeChat app, users were able to send cash-filled virtual red envelopes in one of two ways, either directly to an individual or, if he or she wanted to spice things up, senders could put up a set sum of cash that would then be distributed randomly amongst a specified 78 group of friends. The brilliance behind the Red Pocket plan was that WeChat got user to bind 78 their bank cards to WeChat. “While the app originated as a messaging app and social network, it’s quickly evolving into a catch-all solution for ecommerce, gaming, and even consumer banking. WeChat might have hundreds of millions of users texting friends every day, but once those users clip their bank account to the app, they’ve opened their pocketbook to Tencent,” 78 explains Horwitz. OTT services like WeChat have become powerful CRM services in their own right and I will go into further detail about WeChat, in particular, later in this book. 72 THE PREDICTIVE CASINO Podcasts In 2004, Podcasting was developed by software engineer Dave Winer and former MTV video jockey Adam Curry, who wrote a program called iPodder that enabled them to automatically download Internet radio broadcasts to 79 their iPods. Several developers improved upon Curry’s idea, and podcasting 79 was officially born. According to Wikipedia, “a podcast is a type of digital media consisting of an episodic series of audio radio, video, PDF, or ePub files subscribed to and downloaded through web syndication or streamed online to a computer or mobile device. The word is a neologism derived from ‘broadcast’ and ‘pod’ from the success of the iPod, as audio podcasts are often listened to on 80 portable media players.” Since its inception, podcasting has been free of government regulation and it is hard to see how this will change. “Podcasters don't need to buy a license to broadcast their programming, as radio stations do, and they don't need to conform to the Federal Communication Commission's (FCC) broadcast decency regulations. That means anything goes—from four-letter words to sexually 79 explicit content,” Watson explains. Podcasts are considered copyrightable 79 works and they are easily copyrighted under a Creative Commons licenses. Just because podcasts are free of government regulation doesn’t mean they will be free of every government’s oversight, though, so caution should be taken in certain parts of the world. 81 In his article Will the iPod Kill the Radio Star? Profiling Podcasting as Radio , Richard Berry, claims that “podcasting is both a converged medium bringing together audio, the web and portable media player, and a disruptive technology that has caused some in the radio business to reconsider some of the established practices and preconceptions about audiences, consumption, 81 production and distribution.” Because it is free to both listen to and create content for, which departs from the traditional model of 'gate-kept' media and production tools, podcasting can be considered a horizontal media form, i.e., producers are consumers and consumers become producers and engage in 81 conversations with each other. Podcasts are a great way for an IR to keep control of the company’s message, as well as to reach out to an audience that is engaged and probably very interested in the subject matter the casino wants to disseminate, such as information about upcoming gaming tournaments, interviews of stars in soonto-be released movies, or touring musicians, or eSports athletes playing in upcoming tournaments. Podcast subject matter can be just about anything that is going on at the IR. 82 As per his article Start Podcasting on Your WordPress Blog in 7 Easy Steps , Keith Lock recommends the following ways to set up and run a blog: 73 ANDREW PEARSON • • • • • • • Install recording software–use programs like Audacity, Garage Band or Reaper to create audio files in the MP3 file format. Choose a host for your MP3 files–depending on how popular your blog becomes, you might need to choose a hosting service for your podcast from companies like Libsyn, Amazon, and Blybrry. Add a podcast category to your blog—after you have at least one podcast recorded and uploaded, add a category called “Podcasts” in the “Categories” section under “Posts.” Get a player plugin–adding an audio player plugin like Blubrry PowerPress and Compact WP Audio Player will allow listeners to rewind or fast forward your podcast. Create your first podcast post and include download links wherever your audio is hosted. Set up your RSS feed–“While WordPress has a built-in Feeds feature, it provides only the basics and lacks some of the advanced functionality. For this reason, most bloggers prefer setting up their RSS through a 82 Google service called FeedBurner.” Add your podcast to the iTunes store–if you don’t already have an iTunes account, create one. Proximity Marketing Proximity marketing “is the localized wireless distribution of advertising content associated with a particular place. Transmissions can be received by individuals in that location who wish to receive them and have the necessary 83 equipment to do so,” explains Wikipedia.org. There are four main systems used for proximity marketing; Bluetooth-based systems, NFC-based systems; GSM-based systems (via SMS); and iBeacon-based systems. Considered the “killer-app” for mobile commerce, the commercial viability for proximity marketing or “location-aware advertising” (LAA) is enormous. In location-aware advertising, a cellular subscriber receives an advertising message based on his or her location, so a shopper wandering through a mall could set his or her mobile phone to accept all available mobile offers or just offers from a specific store. In their article Foundations of SMS Commerce Success: Lessons from SMS 84 Messaging and Co-opetition , Xu et al. argue that LAA allows advertisers to deliver highly customized promotions, coupons and offers to an individual, specifically taking into account their geographical location, as well as the time of day of the offer LAA also allows advertisers to reach their customers when they are primed to make a purchase. iBeacon is the trademark for an indoor proximity marketing system that Apple calls, “A new class of low-powered, low-cost transmitters that can notify nearby iOS 7 devices of their presence. The technology enables an iOS device or 74 THE PREDICTIVE CASINO other hardware to send push notifications to iOS devices in close proximity. Devices running the Android operating system can receive iBeacon 85 advertisements but cannot emit iBeacon advertisements.” According to Wikipedia, the iBeacon system uses “Bluetooth low energy Proximity sensing to transmit a universally unique identifier picked up by a compatible app or operating system that can be turned into a physical location or trigger an action on the device such as a check-in on social media or a push 85 notification.” In her article Your iPhone is Now a Homing Beacon (But It’s Ridiculously Easy to 86 Turn it Off) , Kashmir Hill warns that this technology opens the door to more aggressive monitoring, tracking and communication from people with apps on their phone, which will vary from convenient to invasive. In those lengthy terms of service and privacy (that few people read), app makers can slip in 86 tracking permission warns Hill. “Hypothetically, a retailer with its app on your phone could tell iBeacon to turn the app on when you’re in or near the store, send information about your being there to a database and then pop up some 86 advertising.” “At this point, every party that wants to communicate with you needs its app on your phone. Inevitably, some monster advertising network will develop a one-stop-iBeacon-shop app that will allow it to act as the conduit for lots of 86 different people to ping your phone,” Hill claims. Currently, as Shane Paul Neil explains in his article Is iBeacon Marketing Finally 87 Taking Off? , “McDonald’s has seen an increase in sales from a test run using the iBeacons, and Virgin Atlantic is among the first to use them as thermostats to supply cold passengers with blankets. iBeacons also have the potential to enhance B2B marketing with its ability to target users’ smartphones at trade shows or other events.” However, as Shane Paul Neil warns, the delay in implementing beacon technology probably has to do with one of the following four possible 87 reasons : 1. 2. 3. 4. Installing, managing, and maintaining beacons can be a struggle Beacon signals are often obstructed by physical objects Beacon marketing requires user opt-in Consumers aren’t sold on the benefits of beacons 88 In his Washington Post article How iBeacons could change the world forever , Matt McFarland sees a world where iBeacon technology can do the following: 1. 2. Send a coupon to a consumer because they have entered a particular area. React when a user walks into his or her home, turning on lights or televisions. 75 ANDREW PEARSON 3. 4. 5. 6. Provide tours of museums. Automatically send concert or sporting events tickets to a phone that approaches an arena’s turnstiles. Win something for visiting a car dealership or a retail outlet. Be warned when someone’s car of bike is no longer in his or her garage. These and many other examples can be created for proximity marketing and even though each upcoming year is claimed to be the “Year of iBeacons technology”, betting against Apple is usually not a smart thing. Real-time Technology 89 As The Cluetrain Manifesto points out, “Real-time marketing is the execution of a thoughtful and strategic plan specifically designed to engage customers on their terms via digital social technologies.” Adding to that description, Wikipedia notes that real-time marketing is: “Marketing performed ‘on-the-fly’ to determine an appropriate or optimal approach to a particular customer at a particular time and place. It is a form of market research inbound marketing that seeks the most appropriate offer for a given customer sales opportunity, reversing the traditional outbound marketing (or interruption marketing) which aims to acquire appropriate customers for a given 'pre-defined' 90 offer.” Real-time marketing can be inexpensive compared to the cost of traditional paid media. “Expensive research, focus groups, and awareness campaigns can be replaced with online surveys, blog comments, and tweets by anyone or any business,” add Macy and Thompson in their book The Power of Real-Time 91 Social Media Marketing. In his article How Real-time Marketing Technology Can Transform Your 92 Business , Dan Woods’ amusing comparison of the differing environments that marketers face today as compared to what their 1980s counterpart faced is highly instructive as today’s marketing executives don’t have time for a market research study in this sort of figurative first-person-shooter game. “The data arrives too late and isn’t connected to the modern weapons of marketing. The world is now bursting with data from social media, web traffic, mobile devices, 92 and tripwires of all kinds.” Today, companies have massive amounts of evidence about consumer behavior coming at them constantly, from all angles. The challenge is to make sense of the data in time to matter, to understand how consumer attitudes and behaviors are changing and how they are being changed by marketing and advertising efforts; to grab the treasure and avoid the pitfalls of unleashing 76 THE PREDICTIVE CASINO Pandora box of furies. The challenge in understanding the modern consumer is making sense of all of 92 the customer data, coming in from vast unstructured sources. Some of this information explains the broad fluctuations in mass opinion, while other 92 evidence clarifies what consumers might be doing on a company Website. 92 Others still explain what consumers have done, en masse or as individuals. Still other data can be collected after a customer trip in the form of surveys, whether they are mobile or physical surveys. For a real-time platform to work, data must be gathered from multiple and disparate sources, which can include Enterprise Resource Planning (ERP), Customer Relationship Management (CRM), and Social CRM (SCRM) platforms, geofencing applications (like Jiepang and Foursquare), Over-The-Top services (like WhatsApp and WeChat), mobile apps, augmented reality apps, and other mobile and social media systems. This data must be collected and then seamlessly integrated into a data warehouse that can cleanse it and make it 63 ready for consumption. As the authors’ state in Mobile Advertising: “The analytical system must have the capability to digest all the user data, summarize it, and update the master user profile. This functionality is essential to provide the rich user segmentation that is at the heart of recommendations, campaign and offer management, and advertisements. The segmentation engine can cluster users into affinities and different groups based on geographic, demographic or socio16 economic, psychographic, and behavioral characteristics." Perhaps the future of real-time marketing was on display during the 2014 World Cup. “On eight different occasions during the 2014 World Cup, Nike and Google cranked out online display and mobile ads in 15 different countries across the globe. The campaigns ran in real time—meaning they went live 93 during the games, and concluded once those games were over.” For example, “on June 23, during a match between Brazil and Cameroon, Nike pumped out an ad featuring Brazil’s star, Neymar da Silva Santos Júnior, who 93 scored two goals that day.” “Within seconds, an ad featuring the star was featured throughout the Google Display Network, pushing it out to thousands 93 of sites and mobile apps across the web, the search giant says.” “Besides being super timely, the Neymar Jr. ad featured some unique 3D technology that utilized the gyroscope found in most smartphones. Mobile users could rotate their phones and see images of the Nike star in the ad at 93 different angles.” Gimmicky, yes, but, probably, effective as fans could interact with these 3D ads as well as add personal touches. Once viewed, users 93 could share the ads via Twitter, Facebook and/or Google+. The eight different World Cup real-time campaigns generated two million fan interactions across 77 ANDREW PEARSON 93 200 different countries. In chapter two, I will break down the required software to create such unique, compelling and sophisticated personalization marketing, while in chapter seven I lay out how a system like this would work. Search A web search engine is a software system designed to search for information on the web and the search results are generally presented in a line of results 94 often referred to as search engine results page (SERPs). “The information may be a mix of web pages, images, and other types of files. Some search engines also mine data available in databases or open directories. Unlike web directories, which are maintained only by human editors, search engines also 94 maintain real-time information by running an algorithm on a web crawler.” In the US, Google is, by far, the biggest search engine around. Google Search (or Google Web Search) is a web search engine owned by Google Inc. and it is the most-used Internet search engine in the world today, handling more than 95 three billion searches every day. In China, Baidu is the search engine of 96 choice, with a market share of 56.5%. Why is search so influential? Because users flock to search engines to organize the vast amounts of information most buyers need to make purchase decisions. “The main purpose of Google Search is to hunt for text in publicly accessible documents offered by web servers, as opposed to other data, such as with 95 Google Image Search.” “The order of search on Google's search-results pages 95 is based, in part, on a priority rank called a ‘PageRank.’” As Shamra et al. state in Mobile Marketing, “Search is one of the best ways to find content and the 16 absolute best way for a marketer to determine consumer intent.” Google Search “provides at least 22 special features beyond the original wordsearch capability, and language translation of displayed pages. In June 2011, Google introduced “Google Voice Search” and “Search by Image” features for allowing the users to search words by speaking and by giving images. In May 2012, Google introduced a new Knowledge Graph semantic search feature to 95 customers in the U.S.” Outside the U.S., Google’s main competitors are “Baidu and Soso.com in China; Naver.com and Daum Communications in South Korea; Yandex in Russia; 95 Seznam.cz in Czech Republic; Yahoo! in Japan, Taiwan [sic].” In the U.S., however, Google dominates all. Bit players like Bing compete with Google on standard search, but today Apple and Amazon are making inroads on Google’s dominance, with Facebook set to be a challenger in the not-toodistant future. With those latter two, search is organically included within their platforms, i.e., when someone searches for an item to buy on Amazon, it gets included in the overall search rankings, ergo, an ecommerce site has become a search engine. 78 THE PREDICTIVE CASINO “When Google was a Stanford research project, it was nicknamed BackRub 95 because the technology checks backlinks to determine a site's importance.” As I explain in the section on Collaborative Projects in the social media chapter, backlinks—and the quality of them—are very important for search engine optimization (SEO). The higher the quality of backlinks, the higher a Website’s ranking. In the early days of the battle for Internet search supremacy, “previous keyword-based methods of ranking search results, used by many search engines that were once more popular than Google, would rank pages by how often the search terms occurred in the page, or how strongly associated the 97 search terms were within each resulting page.” Google’s PageRank algorithm instead “analyzes human-generated links assuming that web pages linked from many important pages are themselves likely to be important. The algorithm computes a recursive score for pages, based on the weighted sum of the 97 PageRanks of the pages linking to them.” As a result, PageRank is thought to 97 correlate well with human concepts of importance. With the introduction of its Knowledge Graph, Google is attempting to give 95 users answers instead of just links. “If you want to compare the nutritional value of olive oil to butter, for example, Google Search will now give you a comparison chart with lots of details. The same holds true for other things, including dog breed and celestial objects. Google says it plans to expand this 95 feature to more things over time.” 95 Knowledge Graph also allows users to filter results. “Say you ask Google: ‘Tell me about Impressionist artists.’ Now, you’ll see who these artists are, and a new bar on top of the results will allow you to dive in to learn more about them and to switch to learn more about abstract art, for example” Lardinois 95 explains. Questions like, “Tell me about the best casino in Las Vegas” could be reverse-engineered to spit out results that a particular casino wants the search engine to relay, but will require considerable thought and engagement from multiple casino property departments. Stream Processing and Stream Analytics In his article Real-Time Stream Processing as Game Changer in a Big Data 13 World with Hadoop and Data Warehouse , Kal Wähner states that: “Stream processing is designed to analyze and act on realtime streaming data, using ‘continuous queries” (i.e. SQL-type queries that operate over time and buffer windows). Essential to stream processing is Streaming Analytics, or the ability to continuously calculate mathematical or statistical analytics on the fly within the stream. Stream processing solutions are designed to handle high volume in real time with a scalable, 79 ANDREW PEARSON highly available and fault tolerant architecture. This enables analysis of data in motion.” As a batch processing framework, Hadoop can’t handle the needs of real time analytics. As the first open source distributed computing environment, Hadoop has garnered a lot of attention recently, but it is not necessarily the best 98 platform for real-time analytics of dynamic information. One recent development in the stream processing methods is the invention of the ‘live data mart’, which “provides end-user, ad-hoc continuous query access 13 to this streaming data that’s aggregated in memory.” “Business user-oriented analytics tools access the data mart for a continuously live view of streaming 13 data” and a “live analytics front ends slices, dices, and aggregates data dynamically in response to business users’ actions, and all in real time,” adds 13 Wähner. For a casino company, streaming data could be coming in from facial recognition software, fraud or anti-money laundering solutions, slot and table games systems, patron card and campaign management databases, redemption systems, social media feeds, as well as employee/labor data sets. Stream processing excels when data has to be processed fast and/or continuously. Many different frameworks and products are available on the market already, however the number of mature solutions with good tools and commercial support today is small. Apache Storm is a good, open source framework, but it suffers from its open source nature and custom coding is required because of limited developer tools. The typical market solution vs. open source questions must be answered; do I want a pre-built product that will require limited—and sometimes not so limited implementation costs—or do I want to start with a solid solution and be required to customize everything? As Wähner explains, “a stream processing solution has to solve different 13 challenges, including : • • • • • Processing massive amounts of streaming events (filter, aggregate, rule, automate, predict, act, monitor, alert). Real-time responsiveness to changing market conditions. Performance and scalability as data volumes increase in size and complexity. Rapid integration with existing infrastructure and data sources: Input (e.g. market data, user inputs, files, history data from a DWH) and output (e.g. trades, email alerts, dashboards, automated reactions). Fast time-to-market for application development and deployment due to quickly changing landscape and requirements. 80 THE PREDICTIVE CASINO • • • • • • Developer productivity throughout all stages of the application development lifecycle by offering good tool support and agile development. Analytics: Live data discovery and monitoring, continuous query processing, automated alerts and reactions. Community (component/connector exchange, education/discussion, training/certification). End-user ad-hoc continuous query access. Alerting. Push-based visualization.” For a casino or a sports book, events could include the following: • • • • • • Real-time gaming systems information. Coupon and/or comp redemptions. Real-time streaming game odds. Table games revenue management. Social media data feeds, including: o Twitter o Facebook o Weibo o WeChat o Live streaming apps such as YouKu, PandaTV, Periscope, etc. Predictive asset maintenance data streams. Comparison of Stream Processing Services From a technical perspective, Wähner explains that the following components 13 are required for a stream processing system : • • • • “Server: An ultra-low-latency application server optimized for processing real-time streaming event data at high throughputs and low latency (usually in-memory). IDE: A development environment, which ideally offers visual development, debugging and testing of stream processing processes using streaming operators for filtering, aggregation, correlation, time windows, transformation, etc. Extendibility, e.g. integration of libraries or building custom operators and connectors, is also important. Connectors: Pre-built data connectivity to communicate with data sources such as databases (e.g. MySQL, Oracle, IBM DB2), DWH (e.g. HP Vertica), market data (e.g. Bloomberg, FIX, Reuters), statistics (e.g. R, MATLAB, TERR) or technology (e.g. JMS, Hadoop, Java, .NET). 81 ANDREW PEARSON • • • Streaming Analytics: A user interface, which allows monitoring, management and real-time analytics for live streaming data. Automated alerts and human reactions should also be possible. Live Data Mart and/or Operational Business Intelligence: Aggregates streaming data for ad-hoc, end-user, query access, alerting, dynamic aggregation, and user management. Live stream visualization, graphing, charting, slice and dice are also important.” Since this is a highly complex set up and system, there are few market-ready 13 products available and a lot of custom coding is required to implement them. However, the following products are a good place to start: • • • • • Apache Storm Apache Spark IBM’s InfoSphere Hitachi’s Streaming Data Platform TIBCO’s StreamBase. Apache Storm is “an open source framework that provides massively scalable event collection. Storm was created by Twitter and is composed of other open source components, especially ZooKeeper for cluster management, ZeroMQ for 13 multicast messaging, and Kafka for queued messaging.” Developed at Twitter, Spark is a stream processing framework and focuses on continuous computation that can process hundreds of millions of tweets 13 generated every day and now is an open source big data analysis system.” Spark is a scalable data analysis platform based on In-Memory Computing and has performance advantage to Hadoop’s cluster storage method. Spark is written in Scala and offers a single data processing environment. Spark supports iteration tasks of distributed data sets. Spark is a “general framework for large-scale data processing that supports lots of different programming languages and concepts such as MapReduce, in-memory processing, stream processing, graph processing or machine learning. This can also be used on top 13 of Hadoop.” Although Storm and Spark were not created to run on Hadoop specifically, they can be integrated into Cloudera, Hortonworks, MapR, and can 13 be used to implement stream processing on top of Hadoop. Amazon’s Kinesis is a managed cloud service that was designed for real-time processing of streaming data. Because it is Amazon owned, it integrates seamlessly with other AWS cloud services such as S3, Redshift or DynamoDB. 13 DataTorrent is a live streaming platform that runs natively on Hadoop. The big players such as SAP and Oracle are also jumping onto the stream processing bandwagon, and open source solutions include Apache Samza—a distributed stream processing framework processor, which was developed by 82 THE PREDICTIVE CASINO LinkedIn—and Esper, a major framework for Java and .NET. 13 InfoSphere Streams is “IBM’s flagship product for stream processing. It offers a highly scalable event server, integration capabilities, and other typical features required for implementing stream processing use cases. The IDE is based on 13 Eclipse and offers visual development and configuration.” Hitachi’s Streaming Data Platform is a real-time streaming software solution that uses CQL, a popular and widely used language similar to SQL for processing and analysis. As per Hitachi Data Systems (HDS), “CQL (Continuous Query Language) is an extension of traditional SQL. CQL executes in memory, designed for high throughput and low latency environments. It has a ‘windowing’ concept that allows the system to treat each stream, packet and flow individually and allows for ‘stateful’ analysis unlike open source technologies where this capability has to be custom coded. Hitachi CQL 99 provides the capability to centrally develop and globally deploy applications.” TIBCO StreamBase is “a high-performance system for rapidly building applications that analyze and act on real-time streaming data. The goal of StreamBase is to offer a product that supports developers in rapidly building 13 real-time systems and deploying them easily.” Stream processing solutions can get very complicated very quickly and Wähner 13 warns that , “Besides evaluating the core features of stream processing products, you also have to check integration with other products. Can a product work together with messaging, Enterprise Service Bus (ESB), Master Data Management (MDM), in-memory stores, etc. in a loosely coupled, but highly integrated way? If not, there will be a lot of integration time and high costs.” Figure 1 shows a potential casino DW system, containing a streams engine and stream processing unit that includes investigation, visualization and analytics systems within it. 83 ANDREW PEARSON Figure 1: Potential Streams Processing engine Videocasting/Livestreaming Videocasting (aka live streaming) is the process of producing digital voice and video files and publishing them for distribution over the Internet. According to Wikipedia, streaming content refers to content delivered live over the Internet, and it requires a camera for the media, an encoder to digitize the content, a media publisher, and a content delivery network to distribute and deliver the 100 content. Streaming content can be an audio or a video file downloaded from the Internet that is played live as the rest of the file is being downloaded. Online radio stations, YouTube, Periscope, yy.com, Twitch, and Youku videos are both good examples of streaming content. On 24 June 1993, the band Severe Tire Damage streamed the Internet’s first 101 live concert. Today, 48 percent of all U.S. adults and 67 percent of young 102 adults watch streaming or downloaded videos during a typical week. “With the explosion of smartphones and digital tablets and the steady rise of Internet-connected home devices, consumers are watching more video when 102 and where they want than ever before.” Users are also creating enormous amounts of content on channels like Twitch, yy.com (in China), Periscope, 84 THE PREDICTIVE CASINO Facebook Live, amongst others. “Mobile is the first screen for watching, streaming or downloading video, with 24 percent of all U.S. adults and 42 102 percent of smartphone owners watching this type of video each week.” “While we are seeing the way we view video drastically changing, television is likely to remain the primary device for consumer video; we just are witnessing the transition of the definition of television,” said John Fetto, senior analyst, 102 marketing and research, at Experian Marketing Services. “A third of Americans live in households with Internet-connected TVs, giving them the option to stream or download video to the television either directly or with devices such as Kindle Fire TV, Roku, Apple TV and Google Chromecast,” Fetto 102 added. “While the growing trend in cord-cutting is understandably disturbing to cable and satellite companies and disruptive to the television advertising revenue model overall, the growth in online viewing creates opportunities for 102 marketers,” said Fetto. “That’s because online video viewers can be targeted more easily and serve up advertising that is more relevant, responsive and measureable. Marketers also can be more confident that their online ad 102 actually was seen given that viewers typically are unable to skip ads.” With cable and satellite video-on-demand (VOD), mobile video viewers in the United States and the rest of the world have become what has been dubbed as “time-shifters”—viewers who choose to watch what they want when they want to watch it. With mobile video, they have also become “place-shifters”— viewers who choose where to watch what they want to watch when they want 103 to watch it. I am often amused by the sight of commuters on the Hong Kong subway watching what are probably pirated copies of the latest Korean, British, or US television shows on websites like Youku, Fushion and Sohu as they make their way home at the end of their day. Sights like this confirm to me that mobile video has altered the consumption of entertainment and has forever changed people’s viewing habits. Other streaming services in China include Tudou, Cntv.cn, Ku6.com, Xunlei.com, Tv.sohu.com, Letv.com, Iqivi.com, Mtime.com, video.sina.com, and 56.com. For a casino, live streaming can be used for Q&A conversations, customer support, as well as answering customer questions, special product announcements, interviews with influencers, as well as actually broadcasting a live event, including backstage and behind-the-scenes of concerts and events. Glimpses of after parties can show the viewer unique content as well. Wearables Wearable products include smart watches, activity trackers, smart jewelry, head-mounted optical displays and earbuds. According to 85 ANDREW PEARSON wearabledevices.com 104 : “The terms ‘wearable technology’, ‘wearable devices’, and ‘wearables’ all refer to electronic technologies or computers that are incorporated into items of clothing and accessories which can comfortably be worn on the body. These wearable devices can perform many of the same computing tasks as mobile phones and laptop computers; however, in some cases, wearable technology can outperform these hand-held devices entirely. Wearable technology tends to be more sophisticated than hand-held technology on the market today because it can provide sensory and scanning features not typically seen in mobile and laptop devices, such as biofeedback and tracking of physiological functions.” In general, wearable technology includes some form of communications 104 capability that allows the wearer to access real-time information. “Datainput capabilities are also a feature of such devices, as is local storage. Examples of wearable devices include watches, glasses, contact lenses, etextiles and smart fabrics, headbands, beanies and caps, jewelry such as rings, bracelets, and hearing aid-like devices that are designed to look like 104 earrings.” Wearable technology isn’t just for items that can be put on and taken off with ease, there are also more invasive and permanent versions of the concept as implanted devices such as micro-chips or even smart tattoos can be considered wearables. Ultimately, whether a device is worn on or incorporated into the body, “the purpose of wearable technology is to create constant, convenient, seamless, portable, and mostly hands-free access to electronics and 104 computers.” 105 In its Adoption of IoT for Warehouse Management , Israel Gogle argues that wearable devices and augmented reality are some of the best technologies to help improve the performance of human operators. John Bermudez, VP of Product Management at Infor, explains that, “In our innovation lab, we are looking into options like providing workers with wearable video cameras that can upload information to the warehouse management system or with smart glasses. This augmented reality solution will give the operator a visual confirmation on a small screen in front of his eye that 105 he is picking the right thing.” “Wearable devices will add a new layer of visibility that does not exist now. It will work in route management, showing wearers where to go via the glasses, and pick and pack verification, where bar code scans or RFID readings in real time can be used to ensure correct pick and order management,” added Douglas Bellin, Global Lead for Manufacturing and Energy Industries at Cisco 86 THE PREDICTIVE CASINO Systems. 105 Another advantage of wearables is their ability to collect information that can 105 also be used for such things as employee safety. This has a double effect, not only does it ensure the employees’ wellbeing, but it can also save the company money, minimizing losses due to injured employees and lost productivity, 105 maybe even saving lives. “Our wearables platform serves as a real-time warning system. It analyzes a vast amount of information gathered from wearable sensors embedded in personal protective equipment, such as smart safety helmets and protective vests, and in the workers’ individual smartphones,” said Asaf Adi, Senior 105 Manager of IoT and Wearables at IBM Research. “Information from the sensors and smart protective equipment feeds directly to the worker’s smartphone, which can then immediately process and analyze the personal 105 data,” explains Gogle. By tracking a worker’s pulse rate, his movement, his body temperature, even, potentially, his hydration level, sensors can continuously monitor a worker’s 105 physical condition. The noise level and/or an employees’ location in relation to moving machinery and forklifts can also be monitored to guard against 105 accidents . Alerts can also be sent out in cases where sensors detect a worker 105 that has fallen or fainted in the warehouse. Sensors can even send alerts if a worker seems to be suffering from low concentration or fatigue, adds Adi. Conclusion With many of the world’s most successful tech companies throwing their development and financial muscle behind mobile, casinos should recognize that a revolution is going on right before their eyes and their ears. We live in a real-time, 24-7 world, a world where 140-character Twitter messages foment political revolutions; a world where marketers should fear not the power of the pen, but the destructive force of the critical tweet or the far-reaching viral impact of an inflammatory social media diatribe that can encircle the digital world in seconds, laying waste to a reputation that might have taken decades to develop. Conversely, it is also a world where an advertiser’s message can go viral and reach more eyeballs in less than an hour than a multi-million dollar television commercial campaign can in a month. Throughout the rest of the book, it is important to keep in mind the individual technologies, as well as understand how each element fits on top of the mobile and social media platforms. Content creation should always be the first goal and clever casino companies can use their customers and potential customers to not only create the content, but also market it to both their current and potential customers. Not every casino marketing department will have the time 87 ANDREW PEARSON to publish its own content, but most will find it rather simple to set up a blog and/or a Facebook, Jiepang, Foursquare, WeChat, or Weibo page. The time spent on these endeavors should be time well spent. The mobile ad of the future will be created by a sophisticated analytical-driven mobile advertisement system that juxtaposes relevant advertiser content that 16 corresponds to the mobile user’s personal profile and context variables. If casinos and IRs want to succeed in this new marketing environment, they must become an ally to the marketed individual, someone who is actually 62 sought out as a person with cultural capital. “Property exchanges”, “social exchanges” and “cultural exchanges” are all paradigms that are “built on peerto-peer interactivity motivated by the desire to exchange, to share information, 62 or to express one’s self,” contend Deighton and Kornfeld. Deighton and Kornfeld argue that of all the paradigms, “the most potent of the new media are those that enable cultural exchange, media currently exemplified by the 62 functionality of YouTube and Facebook.” 16 In Mobile Advertising , the authors state that the three basic types of mobile advertising are such things as broad-based brand advertising, interactive, direct-response campaigns, and highly targeted search advertising. Broadbased campaigns are subsidized premium content, sponsorships, video prerolls or intromercials, post-roll video, on-demand mobile media and contextual or behavioral advertising and each can be utilized by a casino operator. Interactive, direct response campaigns are opt-in campaigns in which the mobile users usually exchange some personal information for some type of content. Highly targeted search advertising can inform an advertiser of the user’s basic age, sex and address information and these campaigns include content targeted search advertising and paid placement or paid inclusion search. All of this demographic information can be utilized in a multitude of ways, not just in marketing but also in gaining a deeper understanding of an IR’s patrons, which might help with produce and service development. To keep consumers interested in paying for their products and services, casinos and IRs must continue to innovate and explore new marketing and messaging avenues. Mobile marketing’s ability to do one-on-one advertising, anytimeanywhere to any individual with a mobile device is vastly superior to any other marketing channel currently available. In the future, it is likely that all marketing will become interactive and the consumer will become a participant rather than a “target audience”. As Shar VanBoskirk states in his article US 106 Interactive Marketing Forecast, 2007 To 2012 , “Instead of planning for a set ‘search budget’ or an ‘online video campaign’, marketers will instead organize around ‘persona planning’–that is, they will plan around generating a desired response from a customer type. In response to changing customer behavior, channel optimization will take place on the fly, shifting between channels 88 THE PREDICTIVE CASINO dynamically.” In the next couple of chapters, I will look at how these technologies can shape the customer experience so that true personalization can be delivered to a market of one. Capturing a first time visitor’s IP address can be an important— and necessary—first step in the customer relation and once a user signs up for a patron card all of his or her customer information becomes relevant. Analyzing clickstream data, patron card data, marketing data, as well as social media data can develop three dimensional profiles on each guest and, once these profiles are perfected, the behavioral marketing work can begin to ensure that the casino is bringing in the patrons that will produce the highest ROI. Matching customer needs with the casino’s staffing and operation requirements then become an added perk. 31 Henschen, D. 2014, March 2. In-Memory Databases: Do You Need The Speed? Retrieved from informationweek.com: http://www.informationweek.com/big-data/bigdata-analytics/in-memory-databases-do-you-need-the-speed/d/d-id/1114076 32 Sutton, Scott. 2011. Patron Analytics in the Casino and Gaming Industry: How the House Always Wins. SAS. http://support.sas.com/resources/papers/proceedings11/3792011.pdf 33 https://en.wikipedia.org/wiki/Machine_learning 34 Wadhwa, T. (2013, June 3). CrowdOptic and L’Oreal to make history by demonstrating how augmented reality can be a shared experience. Retrieved from Forbes.com: http://www.forbes.com/sites/tarunwadhwa/2013/06/03/crowdoptic-and-loreal-areabout-to-make-history-by-demonstrating-how-augmented-reality-can-be-a-sharedexperience/ 35 http://www.webopedia.com/TERM/A/Augmented_Reality.html 36 Gartner. October 18, 2016. Gartner’s Top 10 Strategic Technology Trends for 2017. http://www.gartner.com/smarterwithgartner/gartners-top-10-technology-trends-2017/ 37 Gartner. 2014, January 14). Gartner believes augmented reality will become an important workplace tool. Retrieved from Gartner.com: http://www.gartner.com/newsroom/id/2649315 38 Deloitte. (2013). Augmented Government, Transforming Government Services Through Augmented Reality. Retrieved from Deloitte Development LLC: http://www.deloitte.com/assets/DcomUnitedStates/Local%20Assets/Documents/Federal/us_fed_augmented_government_06 0613.pdf 39 S-J, M. (2013, January 22). Augmented reality and hospitality...the next generation of hotels? Retrieved from Chocolate pillow: chocolatepillow.com/augmented-reality-andhospitality-the-next-generation-of-hotels/ 40 Moore, J. (2012). Augmented reality: expanding the user experience. Retrieved from Digital Innovation Gazette: http://www.digitalinnovationgazette.com/uiux/augmented_reality_app_development/ #axzz2z8tFKN00 41 https://en.wikipedia.org/wiki/Pokémon_Go 89 ANDREW PEARSON 42 Gupta, Bisdisha. July 22, 2016. Augmented Reality Is The Future Of Real Money Gaming Industry. Socialblog.com. http://www.sociableblog.com/2016/07/22/augmented-reality-future-of-real-moneygaming/ 43 http://www.laforgeoptical.com 44 Nakamoto, Satoshi. October 31, 2008. Bitcoin: A Peer-to-Peer Electronic Cash System. http://www.cryptovest.co.uk/resources/Bitcoin%20paper%20Original.pdf 45 What is Bitcoin? CNN Money. http://money.cnn.com/infographic/technology/what-isbitcoin/ 46 Holden, John. Controversial tech platform for bitcoin goes mainstream. December 8, 2016. The Irish Times. http://www.irishtimes.com/business/controversial-tech-platform-for-bitcoingoes-mainstream-1.2896437 47 Ross, Allison. 12 cryptocurrency alternative to bitcoin. Bankrate.com. http://www.bankrate.com/finance/investing/cryptocurrency-alternatives-to-bitcoin1.aspx 48 Facial recognition: Who’s Tracking You In Public. (December 30, 2015) Consumer Reports. Online: http://www.consumerreports.org/privacy/facial-recognition-who-istracking-you-in-public1/ 49 Gates, Kelly A. January 23, 2011. Our Biometric Future: Facial Recognition Technology and the Culture of Surveillance. NYU Press. 50 TechTarget. (2011). Geofencing definition. Retrieved from TechTarget: http://whatis.techtarget.com/definition/geofencing 51 Berman, S. J., Battino Bill, Feldman, Karen. 2007. Executive Brief: Navigating the media divide: Innovating and enabling new business models. IBM Institute for Business Value. 52 Giaglis, G. M. (2002). On the Potential Use of Mobile Positioning Technologies in Indoor Environments. 15th Bled Electronic Commerce Conference eReality: Constructing the eEconomy. Bled, Solvenia. 53 Weiners, C. (2012, March 30). LBS Opportunities for Casino Marketers in Macau. Retrieved from clickz.com: http://www.clickz.com/clickz/column/2281870/lbsopportunities-for-casino-marketers-in-macau 54 Takahashi, D. (2013, May 22). Lighthouse’s new Android location service could give you indoor navigation for Las Vegas’ casinos. Retrieved from Venturebeat.com: http://venturebeat.com/2013/05/22/lighthouse-signal-systemss-android-app-will-letyou-find-your-way-inside-the-biggest-las-vegas-casinos/ 55 Thompson, Cadie. May 28, 2015. Social media apps are tracking your location in shocking detail. Business Insider. http://www.businessinsider.com/three-ways-socialmedia-is-tracking-you-2015-5 56 Morgan, J. (2014, May 13). A Simple Explanation of 'the Internet of Things'. Retrieved from Forbes.com: http://www.forbes 57 Gartner. (2013, December 12). Gartner Says the Internet of Things Installed Base Will Grow to 26 Billion Units By 2020. Retrieved from Gartner.com: http://www.gartner.com/newsroom/id/2636073 90 THE PREDICTIVE CASINO 58 Lenz, Gadi. September 22, 2015. Datanami.com The Data of Things: How Edge Analytics and IoT Go Hand In Hand. https://www.datanami.com/2015/09/22/the-dataof-things-how-edge-analytics-and-iot-go-hand-in-hand/ 59 DHT Trend Research, Cisco Consulting Services. 2015. Internet of Things in Logistics. http://www.dpdhl.com/content/dam/dpdhl/presse/pdf/2015/DHLTrendReport_Interne t_of_things.pdf 60 American Marketing Association. https://archive.ama.org/topics/Advertising/Pages/default.aspx?k=contentsource:%22M ain%22%20AND%20(AMATopicTags:%22Advertising%22%20AND%20(%20%22All%22)) 61 Kim, K. L. (2012). The typological classification of the participants' subjectivity to plan the policy and strategy for the smart mobile market. Korean Management Review, 367393. 62 Deighton, J. &. (2009). Interactivity’s Unanticpated Consequences for Marketers and Marketing. Journal of Interactive Marketing, 23, 4 - 10. 63 Sharma, R. S. (2009). The Economics of Delivering. Journal of Media Business Studies, 1-24. 64 Conlon, G. (2014, September 15). Sitecore 8 Aims to Add Context to Customer Connections. Retrieved from dmnews.com: http://www.dmnews.com/sitecore-8-aimsto-add-context-to-customer-connections/article/371420/ 65 Nielsen Company. (2012). Global Trust in Advertising and Brand Messaging. Nielsen Company. 66 Dushinski, K. (2012). The Mobile Marketing Handbook. Information Today, Inc. 67 Gartner. (2011, January 26). Gartner Says Worldwide Mobile Application Store Revenue Forecast to Surpass $15 Billion in 2011. Retrieved from Gartner.com: www.gartner.com/it/page.jsp?id=1529214 68 https://en.wikipedia.org/wiki/Mobile_payment 69 Perez, S. (2014, October 7). Square Cash for iOS Now Lets You Send Money To Nearby Friends Via Bluetooth. Retrieved from Techcrunch: http://techcrunch.com/2014/10/07/square-cash-for-ios-now-lets-you-send-money-tonearby-friends-via-bluetooth/ 70 Jiminez, A. (2014, August 16). Three Trends That Will Make a Difference in Mobile Payments. Retrieved from Techcrunch: http://techcrunch.com/2014/08/16/threetrends-that-will-make-a-difference-in-mobile-payment/ 71 Thomasson, E. (2014, April 7). RTP-Retailers Push Into Crowded Mobile Payment Market. Retrieved from Reuters: http://www.reuters.com/article/2014/04/07/retailmobilepayment-idUSL5N0MW24520140407 72 Bedigian, L. (2013, December 27). Google Wallet Vs. PayPal Vs. Square Wallet Vs. Loop. Retrieved from Benzinga: http://www.benzinga.com/tech/13/12/4180848/google-wallet-vs-paypal-vs-squarewallet-vs-loop 73 Lunden, I. (2014, October 8). WePay Launches WePay Clear, a Stripe Competitor with Fraud Protection Built IN. Retrieved from Techcrunch.com: http://techcrunch.com/2014/10/08/wepay-launches-wepay-clear-a-stripe-competitorwith-fraud-protection-built-in/ 74 Hardekopf, B. (2014, May 5). Apple Primed to Become Your Mobile Wallet. Retrieved from Cnbc.com: http://www.cnbc.com/id/101642069 91 ANDREW PEARSON 75 Constine, J. (2014, October 15). Hacked Screenshots Show Friend-to-Friend Payments Feature Hidden in Facebook Messenger. Retrieved from Techcrunch.com: http://techcrunch.com/2014/10/05/pay-with-facebook-messenger/ 76 https://www.techopedia.com/definition/29145/over-the-top-application-ott 77 https://en.wikipedia.org/wiki/Over-the-top_content 78 Horwitz, J. (2014, February 5). Chinese WeChat users sent out 20 million cash-filled red envelopes to friends and family within two days. Retrieved from Techinasia: http://www.techinasia.com/wechats-money-gifting-scheme-lures-5-million-chineseusers-alibabas-jack-ma-calls-pearl-harbor-attack-company/ 79 Watson, S. (2005, March 26). How Podcasting works. Retrieved from Howstuffworks.com: http://computer.howstuffworks.com/internet/basics/podcasting1.htm 80 https://en.wikipedia.org/wiki/Podcast 81 Berry, R. (2006). WIll the iPod kill the radio star? Profiling podcasing as radio. Convergence: The Interanational Journal of Research into New Media Technologies, Volume 12, 143-162. Retrieved from Sage Journals. 82 Lock, K. (2014, January 4). Start podcasting on your WordPress blog in 7 easy steps. Retrieved from tipsandtricks-hq.com: http://www.tipsandtricks-hq.com/startpodcasting-on-your-wordpress-blog-in-7-easy-steps-6738 83 https://en.wikipedia.org/wiki/Proximity_marketing 84 Xu, H. T. (2003). “Foundations of SMS Commerce Success: Lessons from SMS Messaging and Co-opetition.” Proceedings of 36th Hawaii International Conference on System Sciences (pp. 90-99). Los Angeles: IEEE Computing Society Press. 85 https://en.wikipedia.org/wiki/IBeacon 86 Hill, K. (2013, December 10). Your iPhone Is Now a Homing Beacon (But It's Ridiculously East to Turn Off). Retrieved from forbes.com: http://forbes.com/sites/kashmirhill/2013/12/10/your-iphone-is-now-a-homing-beacon 87 Neil, Shane Paul. June 17, 2016. Is iBeacon Marketing Finally Taking Off? The Huffington Post. http://www.huffingtonpost.com/shane-paul-neil/is-ibeacon-marketingfina_b_10508218.html 88 McFarland, Matt. How iBeacons could change the world forever. January 7, 2016. Washington Post. https://www.washingtonpost.com/news/innovations/wp/2014/01/07/how-ibeaconscould-change-the-world-forever/?utm_term=.182e91de201b 89 www.cluetrain.com 90 https://en.wikipedia.org/wiki/Real-time_marketing 91 Macy, B. a. (2011). The Power of Real-Time Social Media Marketing. New York: McGraw Hill, 2011. 92 Woods, D. (2011, May 6). How Real-time Marketing Technology Can Transform Your Business. Retrieved from Forbes.com: http://www.forbes.com/sites/ciocentral/2011/05/06/how-real-time-marketingtechnology-can-transform-your-business/ 93 Shields, M. (2014, July 14). Inside Google's World Cup Real-Time Marketing Experiment with Nike. Retrieved from Wall Street Journal Blog: http://blogs.wsj.com/cmo/2014/07/14/inside-googles-world-cup-real-time-marketingexperiment-with-nike/ 92 THE PREDICTIVE CASINO 94 https://en.wikipedia.org/wiki/Web_search_engine Lardinois, F. (2013, September 26). Google improves knowledge graph with comparisons and filters, brings cards & cross-platform notificiations to mobile. Retrieved from Techcrunch: http://techcrunch.com/2013/09/26/google-improves-knowledgegraph-with-comparisons-and-filters-brings-cards-to-mobile-search-adds-cross-platformnotifications/ 96 https://en.wikipedia.org/wiki/Baidu 97 https://en.wikipedia.org/wiki/Google_Search 98 Deng, Lei, Gao, Jerry, Vuppalapati, Chandrasekar. March 2015. Building a Big Data Analytics Service Framework for Mobile Advertising and Marketing. Online: https://www.researchgate.net/profile/Jerry_Gao/publication/273635443_Building_a_Bi g_Data_Analytics_Service_Framework_for_Mobile_Advertising_and_Marketing/links/5 508de220cf26ff55f840c31.pdf 99 https://www.hds.com/en-us/pdf/brochure/hitachi-overview-streaming-dataplatform.pdf (accessed 30 December 2016) 100 https://en.wikipedia.org/wiki/Streaming_media 101 http://www.computerhope.com/jargon/s/streamin.htm (Retrieved: 18 April 2014) 102 Experian Marketing Services. (2014, April 21). “Cord-cutters” grew by 44 percent in the past four years, with 7.6 million households using high-speed Internet for streaming or downloading videos instead of traditional cable or satellite television. Retrieved from Experian: http://press.experian.com/United-States/Press-Release/cord-cutters-grew-by44-percent-in-the-past-four-years-with-7-6-million-households.aspx 103 Nielsen Company. (2009). Tuned into the Phone: Mobile Video Use in the U.S. and Abroad. Nielsen Company. 104 http://www.wearabledevices.com/what-is-a-wearable-device/ 105 Gogle, Israel. Internet of Things. May 2016. Adoption of IoT for Warehouse Management. A & S Internation. Asmag.com. http://www.cisco.com/c/dam/en/us/solutions/collateral/industry-solutions/iotwarehouse.pdf 106 VanBoskirk, S. (2007). US Interactive Marketing Forecast, 2007 To 2012 for Interactive Marketing Professionals. Forrester Research. 95 93 94 THE PREDICTIVE CASINO CHAPTER TWO Customer Experience “In God we trust. All others bring data.” – W. Edwards Deming Introduction It could be argued that the most valuable use of predictive analytics in a casino would be by the marketing and sales department, but that’s not the whole story. Being able to accurately predict not only who are a casino’s best leads and prospects, but when and how it is best to engage them is nice, but understanding how their acceptance of these marketing offers will affect the overall casino property is what the Predictive Casino is all about. This ability will not only empower marketers and salespeople in the coming seasons to be radically more productive and profitable than they are today, but also give multiple casino departments visibility on their macro and micro needs. Used properly, predictive analytics can transform the science of sales forecasting from a dart-throwing exercise to a precision instrument. The concept of sales and marketing automation has already produced some of the highest-flying successes in high-tech. Companies like Salesforce.com have been wildly successful in automating the sales process for salespeople and managers. Organizations like Marketo have enriched the marketing discipline with automation and tools for lead generation, lead nurturing, and lead scoring. For a casino property, analytics could be used not only for marketing purposes but also to spot problem gamblers, for risk management, to increase share of wallet of its gamblers, as well as to incentivize people to spread out their gambling spend, as well as a whole host of other things. Today, “Personalization”—the process of utilizing geo-location, mobile app, WiFi, and OTT technology to tailor messages or experiences to an individual interacting with them—is becoming the optimum word in a radically new customer intelligence environment. Even though this personalization comes at a price—privacy—it is a price most consumers seem more than willing to pay if a recognized value is received in return. For an IR, “personalization” requires an investment in software analytics, but casino companies should recognize that this price must be paid because highly sophisticated consumers will soon need 95 ANDREW PEARSON an exceptional casino and/or shopping experience to keep them from visiting another casino (that will, undoubtedly, offer such services). To compete in this highly competitive industry, casino companies are recognizing the importance of personalization when it comes to customer interactions. Most casinos today have customer loyalty programs that are a part of a CRM and/or a SCRM initiative to provide their guests with an intimate experience that will make them want to return to the casino again and again and again. Mobile and social media channels are some of the best ways to reach these customers. Currently, however, there is an big disconnect between what companies think they are delivering in terms of personalization and what consumers are actually experiencing. In his article Study finds marketers are prioritizing 107 personalization… but are further behind then they realize , Andrew Jones argues that, “Although two-thirds of the marketers surveyed rate their personalization efforts as ‘very good’ or ‘excellent,’ just 31 percent of consumers reported that companies are consistently delivering personalized experiences.” “Aside from this disparity, the report finds that personalization strategies today are immature. It shows that 91 percent of the marketers surveyed are prioritizing personalization over the coming year, yet many still rely on basic 107 segmentation strategies,” Jones notes. This isn’t that surprising as many companies are struggling with the ability to not just capture the information necessary for personalization, but also creating DWs that can silo the data properly, then delivering it to highly complex analytical programs that can make sense of all that data. It’s like finding a needle in a haystack for each and 107 every customer in a massive database; a herculean task, no doubt. It is obvious that creating a consolidated customer view is a necessary component of personalization, but, unfortunately, “most marketers today are working with customer data that is decentralized, spread across the organization in multiple databases that are updated in batch processes. To find success, marketers must prioritize consolidating data into a single database,” 107 states Jones and this is where a data lake can come in. Another important step to bringing personalization efforts up to a user’s expectation level will be using behavioral data. “In order to create these types of customer experiences, marketers must strategically collect and utilize customer data, including real-time signals of intent, which are typically not 107 captured today,” argues Jones. Figure 2 lists out the identity-related data sources that can be used for personalization and it is a considerable amount of data that must be culled through, siloed, and understood. 96 THE PREDICTIVE CASINO Iden[ty-related data sources used for personaliza[on Email address 57 Name 45 Locaƒon 41 40 Demographics Cookies 34 IP address 33 Social ID 30 Job related 25 Device ID 22 Social profile 22 Postal address 20 Owned Account Informaƒon 18 Locaƒon-related data 18 Phone number 17 Specific to your business 15 Social influence 15 Lifestyle Details 15 Family Details 13 Psychographics 8 None of the above 5 Other (please specify) 3 0 10 20 30 40 50 60 Figure 2: Identity-related data sources used for personalization 108 Source: VB Insights With customer attitudes towards personalized content being shaped by recommendation engines like Amazon, Pandora, and Netflix, consumers are becoming more used to receiving what they want, when they want it, and on 107 what ever channel they want it on. Casinos and IRs must keep this in mind when developing personalization programs. The consumer has become highly sophisticated and he or she expects the level of sophistication received on platforms like Amazon and Pandora to filter over to all their other company communication; don’t waste his or her time with non-matching offers or he or she will go down the street to a competitor’s property. 97 ANDREW PEARSON Customer Relationship Management (CRM) CRM is a strategy used to learn more about a customer’s needs and behaviors in order to develop a stronger relationship with him or her, thereby creating a value exchange on both sides. As Lovelock and Wirtz state in Services Marketing, People, Technology, 109 Strategy , “from a customer perspective, well-implemented CRM systems can offer a unified customer interface that delivers customization and personalization.” Lovelock and Wirtz argue that at each transaction point, such relevant patron data as a customer's personal preferences, as well as his or her overall past history transactions are available to the clerk serving the customer, 109 giving them valuable information about how to interact with that person. This is not an easy thing to do, however, especially when unstructured data like social media feeds are added to the equation. However, in this day and age, it is a necessity as consumers expect personalized service of this level from the companies with whom they interact. According to Lovelock and Wirtz, most CRM solutions contain the following 109 stages : • • • • • Data collection: the system captures customer contact details, such as demographics, purchasing history, service preferences, etc. Data Analysis: data captured is analyzed and categorized into a unique set of criteria. This information is then used to tier the customer base and tailor service delivery accordingly. Sales force automation: sales leads, cross-sell, and up-sell opportunities can be effectively identified and processed, and the entire cycle from lead generation to close of sales and after-sales service can be tracked and facilitated through the CRM system. Marketing automation: the mining of customer data can help a company achieve one-on-one marketing to each one of its customers. Loyalty and retention programs can reduce costs, which can result in an increase of marketing expenditure ROI. By analyzing campaign responses, CRM systems can easily assess a marketing campaign's quantifiable success rate. Call center automation: with customer information available right at their fingertips, call center staff can improve customer service levels because they will be able to immediately identify a customer's tier level, as well as compare and contrast him or her against similar customers so that only promotions that are likely to be accepted are offered. Most casinos will have plenty of data collection, data analysis, sales force automation, marketing automation, and call center automation software to help them in their CRM endeavors, but it is not easy getting all of these 98 THE PREDICTIVE CASINO complicated systems and processes working together to provide a level of personalized service that wows a customer. Beyond simple CRM (which, I guess, is never really that simple), Social CRM (SCRM) adds a whole new level of sophistication to the mix. SCRM is the use of “social media services, techniques and technology to enable organizations to 110 engage with customers.” In his article Time to Put a Stake in the Ground on 111 Social CRM , Paul Greenberg argues that: “Social CRM is a philosophy and a business strategy, supported by a technology platform, business rules, workflow, processes and social characteristics, designed to engage and react accordingly in a collaborative conversation in order to promote mutually beneficial value in a trusted and transparent business environment. It’s the company’s response to the customer’s ownership of the conversation.” Another term for “Social CRM” is “Social Media Monitoring,” the process by which companies monitor sites like Facebook, Twitter, LinkedIn, Weibo, and others for relevant brand and anti-brand comments and mentions. Social media monitoring tools allow for continuous customer engagement and, in chapter four, I go into detail about how a casino operator can use these types of social media monitoring tools to build strong, two-way customer relationships. When it comes to social media and implementing it into a casino, the one constant questions should be is, “How does this affect my ROI?” For many businesses, there is the sense that social media is an ethereal, unquantifiable thing, but this shouldn’t be the case. As Figure 3 shows, social media listening can be used in a multitude of ways, like anticipating customer problems, understanding and identifying sentiment, measuring a company’s share of voice, as well as keeping track of a company’s brand. All of these are important in their own right and, together, they can give a casino company deep detail into marketing campaign performances and attribution analysis, which should help with planning and implementing marketing campaigns. 99 ANDREW PEARSON Figure 3: Social Media Listening objectives A good example of how a company can test whether a social media solution would work for it is to consider the experience of a telecommunication company that proactively adopted social media recently as mentioned in R.E. 112 Divol’s article Demystifying social media. “The company had launched Twitter-based customer service capabilities, several promotional campaigns built around social contests, a fan page with discounts and tech tips, and an 112 active response program to engage with people speaking with the brand.” In social-media terms, the investment was not insignificant, and the company’s senior executives wanted quantifiable ROI, not anecdotal evidence that the 112 strategy was paying off. “As a starting point, to ensure that the company was doing a quality job designing and executing its social presence, it benchmarked its efforts against approaches used by other companies known to be successful 112 in social media.” According to Divol, “the telecommunication company 112 advanced the following hypotheses : • • • If all of these social-media activities improve general service perceptions about the brand, that improvement should be reflected in a higher volume of positive online posts. If social sharing is effective, added clicks and traffic should result in higher search placements. If both of these assumption hold true, social-media activity should help drive sales—ideally, at a rate even higher than the company achieves with its average gross rating point (GRP) of advertising 112 expenditures.” The company tested its options. “At various times, it spent less money on conventional advertising, especially as social-media activity ramped up, and it modeled the rising positive sentiment and higher search positions just as it 112 would using traditional metrics.” The results were quite conclusive: “social- 100 THE PREDICTIVE CASINO media activity not only boosted sales but also had higher ROIs than traditional marketing did. Thus, while the company took a risk by shifting emphasis toward social-media efforts before it had data confirming that this was the correct 112 course, the bet paid off.” Just as importantly, the company had now created an analytic baseline that gave the company confidence to continue exploring a 112 growing role for social media. It is very easy to quantify search rankings and it is pretty obvious that if a casino company ranks higher in Google search, it should garner more business, whether that business is for hotel rooms, F & B and/or retail sales. Starbucks has also developed a metric it believes quantifies the value of its social media marketing in terms of media spend—the “company’s 6.5 million Facebook fans are worth the equivalent of a US $23.4 million annual spend, according to calculations by social media specialists Virtue, reported in 113 Adweek.” Virtue claims that, on average, “a fan base of 1 million translates to at least $3.6 million in equivalent media over a year, or $3.6 per fan. Virtue arrived at its $3.6 million figure by working off a $5 CPM, meaning a brand’s 1 113 million fans generate about $300,000 in media value each month.” That’s quite a significant amount of money and, if a coffee company can find success in social media, casino companies, many of which have become full-on integrated resorts, with amusement parks, theaters, clubs, restaurants and bars attached to them, should be able to get a similar ROI. CRM is an integral part of what businesses hope will be a value exchange on both sides of the customer-company equation, one that will, hopefully, create loyal customers who become apostles for the business. Lovelock and Wirtz created the “Wheel of Loyalty” as an organizing structure to help businesses 109 build customer loyalty and it is highly relevant to the gaming industry. The first of its three sequential steps include building a foundation for loyalty, including “targeting the right portfolio of customer segments, attracting the right customers, tiering the service, and delivering high levels of 109 satisfaction.” The second step—creating loyalty bonds that either deepen the relationship through cross-selling and bundling or adding value to the customer through loyalty rewards and higher level bonds—can be achieved by the casino gaining 109 a fuller understanding of the patron. It is important to understand as much about the patron as possible, his wants, desires and needs, all the way down to his preferred choice of game, his preferred type of food and drink, his preferred hotel room, and any other preference he might want to share with the casino company. The third factor—identify and reduce the factors that result in “churn”—is also 109 extremely important to a casino’s bottom line. Engagement is paramount here and mobile apps and social media are great channels to keep customers interested. Patrons who are baccarat or poker players can be made aware of 101 ANDREW PEARSON upcoming tournaments through these channels and reminder calendars can be set up to ping customers as the tournament dates approach. For the rest of this chapter, I will list out the many different technologies that are relevant to a Predictive Casino and, in the ensuing chapters, I will explain how they all come together to produce an unrivaled patron experience. Casino companies should also feel compelled to reward their customers through Facebook, Twitter, WeChat, and Weibo, or any number of social network, blogging, and/or micro-blogging services. The beauty of using these channels is the ability of the customer to share these awards or stories of these 21 awards with friends and family; the Reed Network will do its magic from there. It wouldn’t be that hard to get patrons to share their social media accounts, either, as a casino can ask patrons for their social media accounts at sign up. Social media is now often a preferred contact channel and it does make connecting with users in a real-time way exceptionally easy. Customer satisfaction is the foundation of true customer loyalty, while 109 customer dissatisfaction is one of the main reasons why customers leave. This may sound obvious, but its importance cannot be stressed enough. According to Jones and Sasser, “the satisfaction-loyalty relationship can be divided into three main zones: Defection, indifference, and affection. The zone of defection occurs at low satisfaction levels. Customers will switch unless 114 switching costs are high or there are no viable or convenient alternatives.” This, obviously, isn’t the case with casinos, where switching often constitutes little more than walking across the street to a competing casino's gaming floor or browsing to another sports betting website (that is also probably willing to give a switching user a sign-up bonus for his or her patronage). With the vast echo chamber of social media against them, losing only one disgruntled patron could be the least of the casino’s problems. Jones and Sasser warn that, “Extremely dissatisfied customers can turn into ‘terrorists,’ providing an abundance of negative feedback about the service 114 provider.” Through social media channels, negative feedback can reverberate around the world within seconds. Today, more than ever, casino companies must spot dissatisfied customers and approach them before they do irreparable harm to the company’s image and reputation and social media is one of the best channels in which to engage them. Like the proverbial canary in the coal mine, the Predictive Casino will have systems in place that can warn the business about these customers before they become figurative terrorists. Casino companies need to empower their patrons to post on Facebook or WeChat or Weibo or Twitter or comment about their IR experiences and, hopefully, turn them into apostles. In Jones and Sasser’s zone of affection, satisfaction levels are high and “customers may have such high attitudinal 114 loyalty that they don’t look for alternative service.” It is within this group 102 THE PREDICTIVE CASINO that “Apostles”—members who praise the firm in public—reside and this is the 115 group that is responsible for improved future business performance. The Predictive Casino will not only be able to spot these apostles, but also understand them on such a unique and personal level that their loyalty and patronage will almost be guaranteed. As Darrell Rigby explains in Bain & Company’s Management Tools 2015 An 116 Executive’s Guide , CRM “is a process companies use to understand their customer groups and respond quickly—and at times, instantly—to shifting customer desires. CRM technology allows firms to collect and manage large amounts of customer data and then carry out strategies based on that 116 information.” Casino operators can utilize CRM to: • • • • • • • • • • Create databases of customers segmented into buckets that allow more effective marketing. Generate more accurate sales leads. Gather market research on customers. Rapidly coordinate information between the sales and marketing staff and front-facing hosts and reps, increasing the customer experience. Enable pit bosses to see and understand the financial impact of different product configurations before they set prices. Accurately gauge the return on individual promotional programs and the effect of integrated marketing activities, and redirect spending accordingly. Accumulate data on customer preferences and problems for product and service designers. Increase sales by systematically identifying, managing, and automating sales leads. Improve customer retention by uncovering the reason(s) for customer churn. Design proactive customer service programs. Today, CRM is evolving into what has been dubbed “Customer Centric Relationship Management” (CCRM), a style of CRM that focuses on customer preferences above all else. CCRM attempts to understand the client in a deep, behavioral way, and it engages customers in individual, interactive relationships through tailored marketing and one-to-one customer service. This personalization can help a casino retain customers, build brand loyalty, provide customers not only with the information that they really want, but also with the rewards that they actually might use. Today’s technology allows casino companies to not only surface the information that they need to know about their customers, but it can also provide front-facing employees with offers that these clients will actually like and, therefore, probably use. 103 ANDREW PEARSON As previously mentioned, CRM is a strategy used to learn more about a customer’s needs and behaviors in order to develop a stronger relationship with him or her, thereby creating a value exchange on both sides of the equation. In their comprehensive article on the subject, Application of Data Mining Techniques in Customer Relationship Management: a Literature Review and 117 Classification , Ngai et al. state that CRM “comprises a set of processes and enabling systems supporting a business strategy to build long term, profitable relationships with specific customers.” Although widely recognized as an important element of most businesses’ 117 platforms, there is no universally accepted definition of CRM. Swift defined CRM as an “enterprise approach to understanding and influencing meaningful communications in order to improve customer acquisition, customer retention, 118 customer loyalty, and customer profitability.” Kincaid states that CRM is “the strategic use of information, processes, technology and people to manage the customer’s relationship with your company (Marketing, Sales, Services, and 119 Support) across the whole customer lifecycle.” For Parvatiyar and Sheth, CRM is: “a comprehensive strategy and process of acquiring, retaining, and partnering with selective customers to create superior value for the company and the customer. It involved the integration of marketing, sales, customer service, and the supply chain functions of the organization to achieve greater efficiencies and effectiveness in delivering 120 customer value.” These varying definitions emphasize the importance of “viewing CRM as a comprehensive process of acquiring and retaining customers, with the help of business intelligence, to maximize the customer 117 value to the organization.” For this book, we will consider CRM as a two-part process that allows a company to track and organize its current and prospective customers as well as to manage the endpoints of customer relationships through its marketing promotions. When done right, CRM systems enable data to be converted into information that provides insight into customer behavior. In Bain & Company’s Management Tools 2015 An Executive’s Guide K. Rigby claims, “CRM requires managers to: 1. 2. 116 , Darrell Start by defining strategic ‘pain points’ in the customer relationship cycle. These are problems that have a large impact on customer satisfaction and loyalty, where solutions would lead to superior financial rewards and competitive advantage. Evaluate whether—and what kind of—CRM data can fix those pain points. Calculate the value that such information would bring the company. 104 THE PREDICTIVE CASINO 3. 4. 5. 6. Select the appropriate technology platform, and calculate the cost of implementing it and training employees to use it. Assess whether the benefits of the CRM information outweigh the expense involved. Design incentive programs to ensure that personnel are encouraged to participate in the CRM program. Many companies have discovered that realigning the organization away from product groups and toward a customer-centered structure improves the success of CRM. Measure CRM progress and impact. Aggressively monitor participation of key personnel in the CRM program. In addition, put measurement systems in place to track the improvement in customer profitability with the use of CRM. Once the data is collected, share the information widely with employees to encourage further participation in the 116 program.” Once a casino company starts implementing a CRM program, data segmentation can begin. According to Wikipedia, market segmentation “is the process of dividing a broad consumer or business market, normally consisting of existing and potential customers, into sub-groups of consumers (known as 121 segments) based on some type of shared characteristics.” In dividing or segmenting markets, casino companies can look for shared characteristics, such as similar games played, common spend, similar lifestyles choices, or even similar demographic profiles. Market segmentation tries to identify high yield segments—i.e., those segments that are likely to be the most profitable or that have outsized growth potential—so that these can be selected for special attention (i.e., become target markets). Rigby states that customer segmentation “is the subdivision of a market into discrete customer groups that share similar characteristics. Customer Segmentation can be a powerful means to identify unmet customer needs. Companies that identify underserved segments can then outperform the 116 competition by developing uniquely appealing products and services.” Rigby adds that customer segmentation is most effective when a company can discover its most profitable segments and then tailor offerings to them, 116 thereby providing the customer with a distinct competitive advantage. As Rigby explains, “Customer Segmentation requires managers to: • • • Divide the market into meaningful and measurable segments according to customers’ needs, their past behaviors or their demographic profiles. Determine the profit potential of each segment by analyzing the revenue and cost impacts of serving each segment. Target segments according to their profit potential and the company’s ability to serve them in a proprietary way. 105 ANDREW PEARSON • • Invest resources to tailor product, service, marketing and distribution programs to match the needs of each target segment. Measure performance of each segment and adjust the segmentation approach over time as market conditions change decision making 116 throughout the organization.” For a casino operator, the pain points might be things like customer loyalty and the marketing department should be asking things like, “Why does it cost so much money to retain customers?” “Can we not find cheaper but more meaningful offers that show understanding of the customer?” Also, “How can we drive customer loyalty to such a degree that our customers rave about us on social media?” Beside the above methods, customer segmentation can be used to: • • • • • • Prioritize new product development efforts Develop customized marketing programs Choose specific product features Establish appropriate service options Design an optimal distribution strategy Determine appropriate product pricing. Market segmentation assumes that different market segments require different marketing programs—that is, different offers, prices, promotion, distribution or some combination of marketing variables. Market segmentation is not only designed to identify the most profitable segments, but also to develop profiles of key segments in order to better understand their needs and purchase motivations. Insights from segmentation analysis are subsequently used to support marketing strategy development and planning. Many marketers use the S-T-P approach; Segmentation→ Targeting → Positioning to provide the framework for marketing planning objectives. That is, a market is segmented, one or more segments are selected for targeting, and products or services are positioned in a way that resonates with the selected target market or markets. With real-time technology, segmentation can reach a whole new customer experience level. The process of segmenting the market is deceptively simple. Seven basic steps describe the entire process, including segmentation, targeting and positioning. In practice, however, the task can be very laborious since it involves poring over loads of data, and it requires a great deal of skill in analysis, interpretation and some judgment. Although a great deal of analysis needs to be undertaken, and many decisions need to be made, marketers tend to use the so-called S-T-P process as a broad framework for simplifying the process outlined here: • Segmentation 106 THE PREDICTIVE CASINO Identify market (also known as the universe) to be segmented. o Identify, select and apply base or bases to be used in the segmentation. o Develop segment profiles. Targeting o Evaluate each segment's attractiveness. o Select segment or segments to be targeted. Positioning o Identify optimal positioning for each segment. o Develop the marketing program for each segment. o • • Markets can be broken down into the following segments: • • • • • • • Geographic segment Demographic segment Psychographic segment Behavioral segment Purchase/usage occasion Generational segment Cultural segmentation For the casino industry, customers can be segmented into the following areas: • • • • • • Game preference Day of week Time of day Length of sessions Size of stake Most and least profitable customers Although customer segmentation is a common business practice, it has received the following criticisms: • • • • • That it fails to identify sufficiently meaningful clusters. That it is no better than mass marketing at building brands. That in competitive markets, segments rarely exhibit major differences in the way they use brands. Geographic/demographic segmentation is overly descriptive and lacks sufficient insights into the motivations necessary to drive communications strategy. Difficulties with market dynamics, notably the instability of segments over time and structural change that leads to segment creep and membership migration as individuals move from one segment to another. 107 ANDREW PEARSON Market segmentation has many critics, but, in spite of its limitations, it remains one of the most enduring concepts in marketing and it continues to be widely used in practice. 121 As Wikipedia explains , there are no formulas for evaluating the attractiveness of market segments and a good deal of judgment must be exercised. Nevertheless, a number of considerations can be used to evaluate market segments for attractiveness, including: • • • Segment Size and Growth: o How large is the market? o Is the market segment substantial enough to be profitable? o Segment size can be measured in number of customers, but superior measures are likely to include sales value or volume. o Is the market segment growing or contracting? o What are the indications that growth will be sustained in the long term? Is any observed growth sustainable? o Is the segment stable over time? Segment Structural Attractiveness: o To what extent are competitors targeting this market segment? o Can we carve out a viable position to differentiate from any competitors? o How responsive are members of the market segment to the marketing program? o Is this market segment reachable and accessible? (i.e., with respect to distribution and promotion). Company Objectives and Resources: o Is this market segment aligned with our company's operating philosophy? o Do we have the resources necessary to enter this market segment? o Do we have prior experience with this market segment or similar market segments? o Do we have the skills and/or know-how to enter this market segment successfully? Caesars is one company that has been able to use social media to measure 122 marketing data. In his article At Caesars, Digital Marketing Is No Crap Shoot , 122 Al Urbanski explains that : “While social media networks like Facebook provide metrics that measure activity within its platform, integrating that data to enable visibility across a brand's entire marketing organization is difficult. Caesars, however, unites information from customers coming through social channels across business units, program teams, time zones, and languages. A 108 THE PREDICTIVE CASINO content-building component allows Caesars' marketers to listen in and respond in real time.” No matter where the customer interaction originates, engagement is a key factor in moving those interactions from the top of the sales funnel to an 122 eventual purchase. “It doesn't matter where customers come in or leave or reenter,” says Chris Kahle, Caesar’s Web Analytics Manager, “if they come to your social page and click your button, or if they go into your content or email 122 and click on that, it's all the same app and you've got them.” Caesars IDs a cookie and if the prospects come back around on paid search three days later, 122 Caesars tracks them. “We can track them on every website, even if they came in on a Las Vegas site and then jump markets to Atlantic City,” adds 122 Kahle. 122 Caesars also tracks activity in real time, while responding to customer cues. Unsurprisingly, different types of customers are more responsive to different interactions from Caesars. Aside from dividing customers into categories such as “Frequent Independent Traveler”—or FITs and Total Rewards members, the Caesar’s team uses tracking data to further segment customers by property or market as well as determine how each of their various segments respond to 122 content. Using this data, Caesars evaluates campaigns in regard to KPIs, such as number of nights booked, and adjusts them on the fly to ramp up conversion 122 rates. “When Caesars sponsored free concerts by top artists at several of its properties last year, for instance, it streamed the events live on the Web and used its new analytics suite to fine-tune loyalty program offers on its websites. It resulted in a dramatic spike in Total Rewards program sign-ups during the 122 concerts.” “What's really dramatic about this is that you can determine what is engaging 122 individuals and target them with it,” Adobe's Langie says. “The high-roller segment, for example. They might respond to a very different Web design than the casual visitor and Caesars tailors the page view to who is visiting. Think of the website as a canvas. You can paint a still life of a fruit for one person and 122 something different for another. The canvas is dynamic.” “The speed and the manner with which the chosen website designs and digital marketing tactics are implemented across the Caesars network may well be the most transforming development of the company's new data culture,” Kahle 122 adds. And this was no easy task as the Caesars landscape extends over 60 websites for its various properties and services as well as 40 Facebook pages. “Prior to implementing a data-centric approach to the decision-making process, it could take as long as two weeks to furnish the field with actionable data. 122 They now get it done in a matter of hours,” Kahle adds. In 2013, Caesars’ implemented Adobe’s Digital Marketing Suite, which “includes real-time tracking and segmentation of digital site visitors, analysis of social media’s role 109 ANDREW PEARSON in purchasing, and content testing by segment or individual visitor.” 122 “The people at the individual properties who are managing the content of the websites are not all technically sophisticated, but Adobe system provides them 122 with built-in capabilities,” Kahle says. “Say one of our properties wants to track social. Before, they'd have to spend a lot of time manually adding tracking 122 codes. With Adobe, tracking codes are integrated,” Kahle adds. In this day and age, it is all about one-to-one marketing. “There's a competitive advantage to using customer data to track and customize marketing appeals for targets of one as opposed to solely focusing on the general masses. High rollers frequently drop tens of thousands of dollars at gaming tables, and they are the segment being lured to brand new, luxury casinos in Macau, Singapore, 122 and South Korea.” “Right now we can assign a percentage value to social media if a booking 122 doesn't result right away,” Kahle says. “But with social we're going to be 122 experimenting with a longer funnel, maybe a two-week time frame.” “Values are ascribed to social media for being the site of initial contact with a new customer, for instance, or for numbers of positive reviews by current 122 customers.” Currently, Caesars can’t measure the total value of a reservation booked online and also can't determine how much an online booker spends at the tables 122 during his or her stay. This is important information when it comes to truly understanding a patron. Caesars would also like to know if, for example, “customers left the Caesars' casino in Las Vegas and went to dinner at Gordon Ramsay's restaurant at the Paris Las Vegas, so they could offer them a free 122 dinner at the restaurant to close the deal on a future booking.” “Eventually we're going to set a time frame that will never expire [on the sales 122 funnel],” Kahle says. “But for now we've built a sales allocation model that goes beyond the last click, and that's OK. Most organizations using multiple 122 marketing channels are still stuck on that last click.” Going forward, mobile and social media are going to be important channels for a casino’s CRM, marketing and operations departments for years to come. The mobile phone’s ubiquity, however, could be a double-edged sword. It allows a casino to market directly to its patrons while they are not just on their property, but also anywhere they might be standing. In this changing digital world, if a casino isn’t constantly marketing to its patrons, some other competitor might be and an offer for a rival’s buffet received while a patron is in “decision-mode” could result in that patron leaving one property and eating—and later, potentially, gambling—at another. Companies like Adobe, IBM, Oracle, Microsoft, SAP, Salesforce.com, and SugarCRM all have products that not only include contact management 110 THE PREDICTIVE CASINO systems that integrate emails, documents, jobs and faxes, but also integrate with mobile and social media accounts as well so the market doesn’t lack product, but this will be a case where one side doesn’t fit all. A deep understanding of the casino operator’s current systems and pain points should be explored before any solution is chosen and implemented. Customer Loyalty Loyalty is so important to a casino company because, as repeated studies have shown, customers become more profitable over time. In their study Zero 123 Defections: Quality Comes to Service , Reichheld and Sasser demonstrated that a customer’s profitability increases as his or her loyalty increases. In this study, the authors found that it usually took more than a year to recoup any customer acquisition costs, but then profits increased as customers remained with the service or firm. Reichheld and Sasser believe there are four factors for 123 this growth and, in order of their importance, they are : 1. 2. 3. 4. Profit derived from increased purchases: as a customer ages, he or she will probably become more affluent, therefore will have more money to spend for company products/services. Profit from reduced operating costs: As customers become more experienced, they should make fewer demands on the business, perhaps taking advantage of available self-service options. Profit from referrals to other customers. Profit from price premiums: long-term customers are more likely to pay regular prices for services rather than being tempted into using a businesses’ lower profit products and/or services. Here are a few other facts and figures regarding customers and their loyalty: • • • • On average, loyal customers are worth up to 10 times as much as their 124 first purchase. It is 6-7 times more expensive to acquire a new customer than it is to 124 keep a current one. News of bad customer service reaches more than twice as many ears 124 as praise for a good service experience. For every customer who bothers to complain, 26 other customers 124 remain silent. In their paper, Lovelock and Wirtz introduce the concept of the Wheel of Loyalty as an organizing structure to help businesses build customer loyalty. Its 109 three sequential steps are : • Build a foundation for loyalty, including “targeting the right portfolio of customer segments, attracting the right customers, tiering the 109 service, and delivering high levels of satisfaction." 111 ANDREW PEARSON • • Create loyalty bonds that “either deepen the relationship through cross-selling and bundling or add value to the customer through 109 loyalty rewards and higher level bonds.” Identify and reduce the factors that result in “churn”—the need to replace lost customers with new ones. Customer satisfaction is the foundation of true customer loyalty, while 109 customer dissatisfaction is the key factor that drives customers away. This may sound obvious, but its importance cannot be overstated. The number one thing that creates loyalty in anybody (that includes your customers) is the social construct of Reciprocity—the social norm that's been evaluated and debated since the days of Aristotle. Many scholars believe it to be one of the single most defining aspects of social interaction that keeps society whole. Reciprocity doesn’t have to be a bar of gold, like some casino in Macau like to offer their high rollers, it could simply be an acknowledgement of poor customer service along with the promise to do better the next time. According to Jones and Sasser, “the satisfaction-loyalty relationship can be divided into three main zones: Defection, Indifference, and Affection. The zone of defection occurs at low satisfaction levels. Customers will switch unless 114 switching costs are high or there are no viable or convenient alternatives.” Jones and Sasser warn that, “Extremely dissatisfied customers can turn into ‘terrorists,’ providing an abundance of negative feedback about the service 114 provider.” Through social media channels, negative feedback can reverberate around the world within seconds. Today, more than ever, companies must spot dissatisfied customers and approach them before they do irreparable harm to a company’s image and reputation and, in chapter four, I address the many ways that companies can counter this negative feedback. In the zone of indifference, customers willingly switch if they can find a better alternative, while in the zone of affection, satisfaction levels are high and “customers may have such high attitudinal loyalty that they don’t look for 114 alternative services.” It is within this group that “Apostles”—members who praise the firm in public—reside and this is the group that is responsible for 115 improved future business performance. In the social media world, these people are more likely to be known as “influencers” and I will go into much more detail about these people later in the book. A consumer’s engagement with a brand can be measured along a continuum from no awareness, through early engagement, and, hopefully, if everything 113 goes right, into advocacy. As for the customer-company relationship, “the strength of feeling will develop and vary over time and, as in any healthy relationship, both parties should be aware of feelings so they can react 113 accordingly,” Woodcock advises. 112 THE PREDICTIVE CASINO 125 As was shown in Nielsen’s 2012 Global Trust in Advertising survey, consumers trust their friends and colleagues much more than they trust TV advertising or corporate communications. Today, consumers communicate with each other like never before through a multitude of social and mobile 113 media channels and these channels should be exploited as much as possible. “SCRM is the connection of social data (wherever it is) with existing customer records (customer database) that enable companies to provide new forms of 113 customer insight and relevant context.” With SCRM, marketers can “understand the mood, find new sales leads, respond faster to customer needs and maybe even anticipate needs by listening into their conversations and 113 taking action.” SCRM doesn’t replace CRM systems; it adds value by augmenting traditional 113 systems. As Woodcock notes, “SCRM is a great hunting ground for businesses to find and acquire consumers to full ‘traditional’ CRM programs as well as identify key influencers who can be considered as high value customers. It offers companies an organized approach, using enterprise software that connects business units to the social web giving them the opportunity to 113 respond in near real time, and in a coordinated fashion.” Social media can help amplify the “relationship” in Customer Relationship Management, enabling organizations to connect and engage consumers in a unique way, as well as personalize and monetize customer relationship on a 128 sustained basis, which should increase profitability. “Social media also provides a path to richer customer analysis, using technologies capable of 128 funneling and consolidating customer insights.” Insights derived from this analysis can help companies to “dynamically calibrate, anticipate, and offer products and services that meet perpetually shifting consumer demands in a 128 hyper-competitive marketplace.” Marketers can also “listen into what customers are saying, to better understand their needs, their voices and tie it back to actual customer 113 profiles” , which could contain their Facebook or WeChat pages or Twitter handles. “In addition, marketers will be able to catch leads in ‘mid-air’ by listening for keywords that suggest a customer is getting ready to buy, then 113 sending real-time alerts to sales teams to respond.” Specifically for a casino, it would be advantageous to link a patron’s account with his or her social media accounts so that the casino could get a heads-up on what a patron might be saying about them on social media. A tip off about an upcoming and/or a last minute trip to Vegas, Macau, KL, or Singapore could be captured from social media, then acted upon accordingly. In a lot of cases, ROI is an enormously tricky thing to measure. “Short-term campaign ROI as the main measure for individual campaigns will evolve into correlation analysis between activities, engagement and sales. This will be 113 ANDREW PEARSON 113 unsettling for many traditional marketers.” However, “the explicit use of active and control groups, and experimentation of using different treatments 113 will help marketers understand the impact of specific SM activities.” More direct marketing type disciplines will be required, in a world where there is 113 real-time feedback on attitude and behavior and a plethora of data.” This has become a much more demanding world in terms of capturing and utilizing all of this data, but making the effort to turn this data into actionable intelligence will be noticed by fickle consumers, I have no doubt. 107 Jones, Andrew (2015, December 15). Study finds marketers are prioritizing personalization…but are further behind than they realize, http://venturebeat.com/2015/12/14/study-finds-marketers-are-prioritizingpersonalization-but-are-further-behind-than-they-realize/ (accessed 26 November 2016). 108 http://insight.venturebeat.com/report/marketing-personalization-maximizingrelevance-and-revenue?utm_source=vb&utm_medium=refer&utm_content=editorialpost&utm_campaign=personalization-report (accessed 26 November 2016) 109 Lovelock, C. a. (2010). Services Marketing, People, Technology, Strategy, Seventh Edition. Prentice Hall. 110 https://en.wikipedia.org/wiki/Social_CRM 111 Greenberg, P. (2009, July 6). Time To Put a Stake in the Ground On Social CRM. Retrieved from ZDnet.com: http://www.zdnet.com/blog/crm/time-to-put-a-stake-inthe-ground-on-social-crm/829 112 Divol, R. E. (2012, April). Demystifying social media. Retrieved from Mckinsey.com: http://www.mckinsey.com/insights/marketing_sales/demystifying_social_media 113 Woodcock, N. G. (2011). Social CRM as a business strategy. Database Marketing & Customer Strategy Management, Vol. 18, 1, 50-64. 114 The Customer Satisfaction-Loyalty Relationship from Thomas O. Jones and W. Earl Sasser, Jr., “Why Satisfied Customers Defect” Harvard Business Review, Nov.–Dec. 1995, p. 91. Reprinted by permission of Harvard Business School. 115 Wangenheim, F. v. (2005). Postswitching Negative Word of Mouth. Journal of Service Research, 8, No. 1, 67-78. 116 Rigby, Darrell. 2015. Management Tools 2015. An Executive’s Guide. Bain & Company. http://www.bain.com/publications/articles/management-tools-customerrelationship-management.aspx 117 Ngai, N. X. (2009). Application of data mining techniques in customer relationship management: a literature review and classification. Expert systems with applications, 2592-2602. 118 Swift, R. (2001). Accelerating customer relationships: Using CRM and relationship technologies. Upper Saddle River: Prentice Hall. 119 Kincaid, J. (2003). Customer relationship management: Getting it right. Upper Saddle River, NJ: Prentice Hall. 120 Parvatiyar, A. &. Sheth, JN (2001). Customer relationship management: Emerging practice, process, and discipline. Journal of Economic & Social Research, 3, 1 - 34. 121 https://en.wikipedia.org/wiki/Market_segmentation 114 THE PREDICTIVE CASINO 122 Urbanski, A. (2013, February 1). At Caesars, Digital Marketing Is No Crap Shoot. Retrieved from DM News: http://www.dmnews.com/at-caesars-digital-marketing-is-nocrap-shoot/article/277685/ 123 Reichheld, F. a. (1990). Zero defections: quality comes to services. Harvard Business Review, 105-111. 124 White House Office of Consumer Affairs 125 Nielsen Company. (2012). Global Trust in Advertising and Brand Messaging. Nielsen Company. 115 116 THE PREDICTIVE CASINO CHAPTER THREE Customer Analytics “Anticipate the needs of your customers by understanding the data that increasingly surrounds them.” ~Michael Hinshaw Overview Customer analytics, coupled with insight from social media data from sites like Facebook, Twitter, Pinterest, Weibo, WeChat, amongst others, can enable organizations to make faster strides in predicting retention, attrition, and return rates, with the goal of reducing customer churn, raising customer lift, 126 and/or increasing a whole host of other metrics. Sources such as transactions data, clickstream, as well as service and call center 126 records are also important for customer analytics. These can both improve how a casino organization decides on characteristics for customer segmentation, and also provide clues to emerging characteristics for the 126 definition of new segments. As David Stodder explains in his article Customer 126 Analytics in the Age of Social Media , “Firms can employ predictive modeling to test and learn from campaigns so that they are able to select the most 126 persuasive offers to put in front of the right customers at the right time.” As Webopedia.com explains, customer analytics “exploits behavioral data to identify unique segments in a customer base that the business can act upon. Information obtained through customer analytics is often used to segment markets, in direct marketing to customers, predicate analysis, or even to guide 127 future product and services offered by the business.” In the most basic sense, customer analytics is made possible by combining elements of business intelligence (software such as IBM’s cognos, SAP’s Lumira and Business Object’s suite, and Qlik’s QlikView, amongst a whole host of others) with predictive analytics solutions like SAP’s and SAS’s suite of analytical tools, as well as R, Python, WEKA, etc., etc. 128 In IBM’s Achieving Customer Loyalty with Customer Analytics , IBM argues that customer analytics can uncover “patterns and trends in customer behavior and sentiment hidden among different types of customer data such as transactions, demographics, social media, survey and interactions.” “The results of the analysis are then used to predict future outcomes so businesses 117 ANDREW PEARSON 128 can make smarter decisions and act more effectively.” Results from these models can then be presented back to the business users in easily digestible 128 dashboards and scorecards. “Self-learning predictive models ensure that each new iteration of customer analytics insight and the business decisions it 128 drives become more accurate and effective,” argues IBM. Customer analytics can also help determine which of a casino’s advertising campaign or advertising partner’s pages have the highest landing rates, as well as show conversion rates for all of a casino company’s advertising and marketing budgets. Mobile analytics can also display how many visitors downloaded material from a site, which can help in factoring a company’s advertising and marketing budgets. And, finally, mobile analytics can display which pages have the highest exit rates. With this type of analysis, marketers can rapidly adjust marketing campaigns to exploit the most effective ones and, conversely, trim the non-performing ones. The biggest problem with any analytics procedure is filtering out the noise associated with the data. Without clean data, “the trends, patterns, and other 126 insights hidden in the raw data are lost through aggregation and filtering.” Organizations need an unstructured place “to put all kinds of big data in its pure form, rather than in a more structured data warehousing 126 environment.” This is because what might be considered just “noise” in the raw data from one perspective could be full of important “signals” from a more 126 knowledgeable perspective. “Discovery, including what-if analysis, is an important part of customer analytics because users in marketing and other functions do not always know what they are looking for in the data and must 126 try different types of analysis to produce the insight needed.” As per 126 Stodder, among the frequent targets for analysis are the following : • • • • • Understanding sentiment drivers. Identifying characteristics for better segmentation. Measuring the organization’s share of voice and brand reputation compared with the competition. Determining the effectiveness of marketing touches and messages in buying behavior, i.e., attribution analysis. Using predictive analytics on social media to discover patterns and anticipate customers’ problems with products and/or services. Going back to the baccarat example for a moment, customer analytics can be used to understand the behavioral play of the patron, i.e., is he someone who tends to take the Banker side or the Player side? As the Player side is a slightly more aggressive way to play, perhaps people who make this bet are less conservative than most. A little deep dive into ML might be an interesting exercise here, but the most important thing to understand here is the player’s gambling profile and that should be logged into his or her customer profile. As 118 THE PREDICTIVE CASINO long as the player is using a patron card and he or she is playing at a stadium seating baccarat table, all of his or her bets will be captured and a profile of his or her betting behavior can be created. Once that player’s baccarat preference is known, the casino could, theoretically counter players who had a propensity to play one side of the bet against players who had a propensity to play the opposite side of the bet. If the casino knows the banker/player ratio of the table, i.e., how many players are taking each side of that bet and they know their clientele well enough, there’s no reason why the marketing department can’t send out mobile offers to players that would—if they played at the stadium seating baccarat counter when they were specifically needed to even out the banker/player ratio—help the area reach a more equitable 50/50 banker/payer ratio split. And, most importantly of all, raise the house edge. The marketing department could even create baccarat competitions in which they try to reach as equitable a ratio of players as possible. As the marketing process plays out, the casino’s marketing department should be ready to invite only those players that would keep the ratio as close to 50/50 as possible. 126 TWDI’s research about the general purpose of customer analytics technology and methods (see Figure 4) discovered that “the business functions or operations for which respondents considered customer analytics most important were marketing (81%, with 52% indicating “very important”), sales and sales reporting (79%, with 45% “very important”), and campaign 126 management (74%, with 47% “very important”). Market research (43% “very important”) and customer services and order management (also 43% “very important”) were also high among business functions regarded as critical to 126 developers and consumers of customer analytics. The marketing department, “which in most organizations is empowered with the responsibility for identifying, attracting, satisfying, and keeping customers, 126 is clearly the main stage for customer analytics.” Marketing departments and 126 functions are becoming increasingly qualitative. “Gut feelings” are being 126 replaced by data-driven decision-making. “Data drives the pursuit of efficiency and achievement of measurable results. Marketing functions are key supporters of ‘data science,’ which is the use of scientific methods on data to develop hypotheses and models and apply iterative, test-and-learn strategies 126 to marketing campaigns and related initiatives.” Customer analytics can be a very effective tool for micro-targeting customers 128 with customized marketing offers and promotions. Obviously, when an organization “attempts to cross-sell or up-sell a customer, a product or service 128 they desire, it can enhance satisfaction.” However, unwanted marketing campaigns can do just the opposite, annoying customers, thereby eroding 128 loyalty and, potentially, hurting sales. Even worse, unwanted marketing campaigns can give customers the impression that the organization doesn’t 128 care about their wants, desires, needs and preferences. “Customer analytics 119 ANDREW PEARSON can help determine which marketing interactions are likely to please individual 128 customers and which will not.” Very important Somewhat important Not important Don't know Markeƒng Somewhat unimportant 52 29 7 7 5 Campaign management 47 27 Sales/sales reporƒng 45 34 10 8 8 5 Market research 43 36 9 7 5 Customer services/order management 43 Execuƒve management 40 Adverƒsing 39 26 12 18 5 Fraud/risk management 39 24 16 13 8 Finance 38 29 Product development 32 9 7 33 32 29 Call/contact center 33 Operaƒons management/research 32 Web storefront/online presence 32 10 11 5 32 14 8 6 16 11 6 14 14 12 36 18 15 30 7 8 11 6 14 16 8 New media/social media dept. 31 31 17 13 8 Regulatory complaine/data governance 29 33 15 15 8 Public relaƒons 26 20 14 8 Distribuƒon, fulfillment, or logisƒcs Supply Chain 32 25 27 20 Event Management 14 Procurement 13 21 16 22 35 23 23 26 23 28 18 24 9 11 10 12 0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100% Figure 4: Importance of Customer Analytics Technology Based on one answer per business function from 452 responses. 126 Source: TWDI Research 126 Sales functions can be important beneficiaries of customer analytics as well. Stodder argues that, “Sales reports typically focus on providing visibility into the pipeline. Managers can use data insights to improve sales forecasting of potential revenues based on deeper knowledge of priority opportunities, most 126 valued customer segments, and more.” 120 THE PREDICTIVE CASINO “Customer service and order management can use customer analytics to get a more subtle and substantial view of what actions impact customer experiences 126 and satisfaction.” Contact centers can utilize “customer analytics to help tune performance metrics closer to real time, so that each day’s agents are 126 guided, if not incentivized, to interact with customers in beneficial ways.” Analytics can also “help service and order management functions move away from one-size-fits-all approaches to customers and instead tune and tailor interactions more personally based on knowledge of particular types or 126 segments, such as regions or nationalities.” “Finally, through integrated views of customer data and analytics, service and order management functions are able to work in better synchronicity with the organization’s marketing, 126 sales, and other business functions.” Customer analytics can be used to understand where marketing campaigns are working as well. In the words of business management guru Thomas Davenport, “Organizations are competing on analytics not just because they can—business today is awash 129 in data crunchers—but also because they should.” Although these words were said more than ten years ago, they might be more relevant today than ever before. Davenport adds, “Business processes are among the last remaining points of differentiation. And analytics competitors wring every last 129 drop of value from those processes.” “Customer analytics helps organizations determine what steps will give them competitive advantages, increase profitability, and identify waste in business processes,” Davenport 129 argues. With the steep drop in RAM prices, in-memory solutions are all the rage these days and they allow analytics to reach a whole new level. Today, creativity is becoming the differentiator; today’s overriding philosophy might be “He who analyzes best wins.” With products and services being commoditized at such a rapid rate today, 126 customer loyalty has become more elusive than ever. “Innovation must be constant and must immediately address why an organization is losing customers. Information insights from analytics can help an organization align product and service development with strategic business objectives for 126 customer loyalty.” In addition, these insights can help casino organizations be selective in how they deploy marketing campaigns and customer-touch processes so that they emphasize features in new products and services that are important to customers. When TDWI Research examined the business benefits sought from customer analytics (see Figure 5), respondents cited giving executive management 126 customer and market insight as the most important (71%). The second highest benefit was being able to react more quickly to changing market 126 conditions (62%). Improving customer satisfaction and gaining a complete picture of a customer’s 121 ANDREW PEARSON activity across business channels—two areas that would be considered a part of the “Customer Experience Management” (CEM) process—are critical to identifying what steps an organization must take to build and retain customer 126 loyalty. The remaining items fall mainly into the categories of business intelligence, marketing and brand management and they are extremely important as well. What are the most important business benefits that your organzaton seeks to achieve from implementng customer analytcs technologies and methods? (Please select all that apply) Give execuƒve management customer/market insight 71% React more quickly to changing market condiƒons 62% Improve customer saƒsfacƒon with experiences and engagement 60% Gain complete view of the customer acƒvty across channels 56% Idenƒfy potenƒal compeƒƒve advantages 50% Discover what influences buying behavior 49% Apply insights to product/service development 44% Manage and target markeƒng mix 43% Idenƒfy financial impact of markeƒng acƒons 42% Develop more effecƒve loyalty programs 32% Manage brands effecƒvely in social media 31% Gain accurate a…ribuƒon of conversion to markeƒng touches 26% 0% 10% 20% 30% 40% 50% 60% 70% 80% Figure 5: What Are the Important Business Benefits of Customer Analytics? Based on 2,573 responses from 454 respondents; almost six responses, on average. 126 Source: TDWI Research Studies on loyalty have shown the following: • • It costs five times as much to gain a new customer as it does to retain 130 an existing one. The customer profitability rate tends to increase over the life of a 130 retained customer. 122 THE PREDICTIVE CASINO • • • A 5% reduction in customer churn can increase profits by 25% to 130 125%. The probability of selling to an existing customer is 60% to 70%, while 130 the probability of selling to a new prospect is 5% to 20%. Almost 70% of the identifiable reasons why a customer typically leaves 130 boils down to poor quality of service. Organizations are growing open to customer analytics because they are interested in discovering how a marketing department can be more effective, 126 not just more efficient. “Whereas other types of applications for ecommerce, fulfillment, or marketing automation help organizations determine how to get things done (e.g., getting goods delivered at the right time, executing a marketing campaign), customer analytics helps organizations answer who, what, when, where, and why questions,” argues Scott Groenendal, program director of customer analytics market strategy for IBM 126 Business Analytics. “They can find answers to questions such as: What channel should I communicate through? When is the best time to target this person, and why would they be receptive to this message?” adds 126 Groenendal. Individual creativity, personal experiences, customer behavior and marketing 122 context are critical components of consumer marketing decisions. “The role of customer analytics is not necessarily to replace these, but to help decision makers come to fact-based conclusions through better knowledge of the 126 organization’s customers and markets.” Just as importantly, analytics are 126 needed for scalability. “Just as automation is necessary to run hundreds or thousands of marketing campaigns, customer analytics processes are important for supplying intelligence and guidance to those automated routines. Customer 126 analytics can provide the brains to match the marketing systems’ brawn.” For a casino company, there’s no point in advertising an upcoming poker tournament to a baccarat player and vice-versa, unless, of course, it’s someone to whom both games are compelling. For a sushi restaurant, marketing to a vegan makes no sense at all. With the commoditization of products and services, customer loyalty can be elusive; innovation must be constant and it should help to reveal why an 126 organization might be losing its customer base. “Information insights from analytics can help organizations align product and service development with 126 strategic business objectives for customer loyalty.” These insights can also help an organization be selective about how they deploy their marketing campaigns and customer-touch processes so that they emphasize features in 126 new products and services that are important to each specific customer. 128 In its Achieving Customer Loyalty with Customer Analytics , IBM describes one of its studies that asked some of the world’s leading company CEOs and CMOs 128 what their number one priority was. The CEOs answered that it was to 123 ANDREW PEARSON 128 engage customers, while the CMOs said it was to enhance customer loyalty. The study argued that forward-thinking companies were using customer 128 analytics to : • • • • • • Guide front-line interactions with customers. Create and execute customer retention strategies. Prompt people or systems to proactively address customer satisfaction issues. Guide product planning to fulfill future customer needs. Hire and train employees to act upon customer insights and improve loyalty. Align operations to focus on satisfying customers. 126 The Customer Analytics in the Age of Social Media report concluded that the importance of customer analytics is in the boardroom; “overwhelmingly, respondents cited giving executive management customer and market insight (71%) as the most important business benefit that their organization seeks to 126 achieve from implementing customer analytics.” “This percentage rises to 81% when survey results are filtered to see only the responses from those who indicated ‘strong acceptance’ of data-driven customer analytics over gut 126 feel.” The second highest benefit cited at 62% was “the ability to react more quickly to changing market conditions, which speaks to the need for customer data insights to help decision makers address competitive pressures from rapid 126 product or service commoditization.” Customer analytics can also provide answers to questions like, “When in the life cycle are customers most likely to churn? What types of products or services would prevent them from churning, and when should they be offered 126 complimentary items? When is it too costly to try to keep certain customers? Businesses can realize significant ROI from investing in customer analytics as it 126 can improve the marketing department’s efficiency and effectiveness. However, customer analytics ROI is a difficult thing to fully quantify—especially in the analytics space. Better customer knowledge equates to more optimized marketing spend because a business can focus its resources on those campaigns that have the highest predicted chances of success for particular 126 segments as well as cutting off or avoiding those that have the least. “By using analytics to eliminate mismatches of campaigns targeting the wrong customers or using the wrong messages and offers, marketing functions can 126 reduce wasteful spending and increase gains relative to costs.” Customer segmentation allows organizations to move “away from one-size-fits-all, brandlevel-only marketing and toward the ‘market of one’: that is, personalized, one126 to-one marketing.” Reaching a customization and customer service level that makes a customer 124 THE PREDICTIVE CASINO feel as though he or she is a preferred customer is not easy, scaling that up so that an entire database of customers feel that they are unique and receiving outstanding customer service is even more challenging, but, in this day and age, it is almost a necessity if a company wants to provide good and engaging customer service. In your organizaton, which of the following marketng objectves are most important to achieve for customer analytcs to deliver a return on investment? (Please select all that apply.) Target cross-sell and up-sell opportuni[es 54% Improve customer segmenta[on 49% Predict reten[on, acri[on, and churn rates 47% Determine life[me customer value 42% Increase porbolio penetra[on per customer Op[mize marke[ng across mul[ple channels Impact other business func[ons (sales, service, support) Forecast buying habits and lifestyle preferences Measure types of loyalty for campaign targe[ng 39% 37% 35% 32% 26% Priori[ze marke[ng e-mail messages 23% Implement upli], incremental, or true-li] modeling Increase speed of mul[variate tes[ng analysis 18% 14% 0% 10% 20% 30% 40% 50% 60% Figure 6: Which Are the Most Important Business Objectives When it Comes to Customer Analytics? 126 Source: TWDI Research , Based on 1,625 responses from 432 respondents; almost four responses per respondent, on average. 126 TDWI Research examined the importance of accomplishing various objectives for gaining positive ROI from customer analytics (see Figure 6). “Using customer analytics to target cross-sell and up-sell opportunities was the 126 objective cited by the biggest percentage of respondents (54%).” This objective is about gaining more value from existing customers by understanding their purchasing habits and trying to get them to buy more 126 products more often. “Some organizations (18%) are implementing an advanced technique called ‘uplift modeling’ (also called incremental or true-lift 125 ANDREW PEARSON modeling), which enables marketers to use data mining to measure the impact 126 and influence of marketing actions on customers.” Insights such as these allow marketers to develop new kinds of predictive models to determine the 126 best prospects for up-sell and cross-sell offerings. “As firms scale up to execute large numbers of campaigns across multiple channels, the efficiency gained from predictive modeling can be critical to marketing spending 126 optimization,” Stodder argues. Analytics can improve marketing performance by quantifying a customer’s lifetime value as well as customer worth at the many different stages in the 126 customer’s life cycle. “If organizations can identify their most valuable customers they can determine if they are worthy of retention efforts and 128 resources because of the returns they will provide.” For instance, it may not be worth the time, effort, and expense to retain a low value customer, unless customer analytics reveals that this low-spend customer actually has a lot of 128 social influence. Armed with this information, managers can align their deployment of resources to achieve the highest value, as well as avoid the 126 costs and inefficiencies of marketing to the wrong people at the wrong time. Organizations have long used demographics such as gender, household size, 126 education, occupation, and income to segment customers. Data mining techniques let organizations segment much larger customer populations and, perhaps, more importantly, determine whether to apply new characteristics that refine segmentation to fit the specific attributes of the organization’s 126 products and services. “Customer analytics using data mining tools improves the speed of segmentation analysis over manual and spreadsheet efforts that are often used 126 in less mature organizations.” Speed is a vital ingredient for marketing initiatives that are time sensitive, particularly for those companies that need to provide real-time cross-sell and up-sell offers to customers clicking through 126 Web pages. Today, personalized web pages can be rendered during the web page load and elements of the page can take into account past purchase history, clickstream information, as well as a whole host of other things. “The next most common objectives in the research were predicting retention, attrition, and churn rates (47%) and determining lifetime customer value 126 (42%).” Churn can cost organizations heavily, both from the loss of profits from existing customers as well as in the high price of attracting new ones. “Attrition or churn analysis methods are aimed at discovering which variables 126 have the most influence on customers’ decisions to leave or stay.” With data mining and predictive analytics, organizations can learn which attrition rates are acceptable or expected for particular customer segments and 126 which rates could be highly detrimental to the bottom line. “Predictive customer analytics can play a major role in enabling organizations to discover 126 THE PREDICTIVE CASINO and model which customers are most likely to leave, and from which 126 segments.” With social media added to the mix, as well as clickstreams, and other 126 behavioral data, the volume and variety of data is exploding. “Social networking sites such as Facebook, Twitter, LinkedIn, and MySpace have files 126 containing petabytes of data, often in vast Hadoop clusters.” Weibo and WeChat add another hundreds of millions of users to the mix and with it petabytes of data. “Advertising concerns are recording tens of millions of events daily that organizations want to mine in near real time to identify prospects,” Stodder 126 notes. Businesses of all kinds want to use predictive models and score event and transaction details as fast as they come in so that they can gain insight into 126 individual shopping behavior. Insights that they hope will give them a competitive advantage over their competitors, but this is dangerous and expensive territory to chart, especially if done incorrectly. The “data sources most commonly monitored for customer analytics are customer satisfaction surveys (57%) and customer transactions and online purchases (55%). Just under half (44%) are monitoring Web site logs and clickstream sources. In addition to monitoring customer satisfaction surveys, about half (48%) of organizations surveyed are studying call and contact center 126 interactions.” For a casino operator, additional data sources would be gaming systems, Angel Eye, on-floor redemption voucher systems, PoS and credit card fraud systems, as well as mobile tracking and security systems, amongst many others. Customer satisfaction surveys are usually conducted in person, on a Website, 126 over the phone or through traditional mail and e-mail channels. Because this includes both semi-structured data and unstructured comments, data 126 collection can be difficult. “Standard questions inquire about a customer’s satisfaction with purchases, the services they received, and the company’s brands overall. Other questions address the customer’s likelihood of buying from the company again and whether they would recommend the firm to 126 others.” Text analytics can be used to increase the speed, depth, and consistency of 126 unstructured content analysis far greater than what can be done manually. “More advanced analytics can look for correlations between satisfaction ratings, commented sentiments, and other records, such as first-call-resolution 126 metrics.” In-memory computers can handle these large clusters of data culled both from the significant volumes of customer behavior data, as well as data from the 126 multiple social media channels available. 127 ANDREW PEARSON “To analyze data generated by social media networking services such as Twitter, Facebook, Weibo, and LinkedIn, many organizations are implementing Hadoop and NoSQL technologies, which do not force a schema on the source 126 data prior to storage, as traditional BI and data warehousing systems do.” Because of this, the discovery analytics processes can run against the raw 126 data. “Customer analytics tools need to be able to consume data from sources such as Hadoop clusters and then integrate the insights into overall 126 customer profiles,” advises Stodder. The data sources can be varied for these technologies and methods; “they include transaction data, clickstreams, satisfaction surveys, loyalty card membership data, credit card purchases, voter registration, location data, and 126 a host of [other] demographic data types.” In its Retail Analytics: Game Changer for Customer Loyalty, Cognizant argues that in the retail industry, “predictive models can be used to analyze past performance to assess the likelihood that a customer will exhibit a specific 131 behavior in order to improve marketing effectiveness.” This can help with “predicting customer reactions to a given product and can be leveraged to improve basket size, increase the value of the basket and switch the customer 131 to a better and more profitable offering” Predictive models can also help tailor pricing strategies that take into account both the need for competitive 131 pricing and the bottom line. Predictive analytics and data mining are used to discover which variables out of possibly hundreds are most influential in determining customer loyalty within 126 certain segments. “Advanced analytics generally involves statistical, quantitative, or mathematical analysis and centers on developing, testing, 126 training, scoring, and monitoring predictive models.” Models can be created that will uncover patterns, affinities, anomalies, and other useful insights for marketing campaigns and for determining cross-sell 126 and up-sell opportunities. “The tools and techniques are also used for developing and deploying behavioral scoring models for marketing, deciding whether to adjust customers’ credit limits for purchases, and a variety of highly 126 time-sensitive analytic processes,” Stodder notes. “As more online customer behavior is recorded in Web logs and tracked through cookies and other observation devices, sizeable amounts of information are becoming available to organizations that seek a more accurate 126 view of a customer’s path to purchase,” states Stodder. Attribution analysis is, first and foremost, a big-data problem, given the quantity and variety of data 126 available from today’s multiple platforms. Businesses that are performing attribution analysis will frequently employ Hadoop, MapReduce, with analytic software solutions such as R, SAS’s eMiner, 126 SAP’s InfiniteInsights, Python, and IBM’s SPSS, amongst others. This allows a 128 THE PREDICTIVE CASINO business to run sophisticated algorithms against detailed data to find the correct path of purchase. This analysis can then be integrated with analysis from other data types and sources, including those that might have been 126 generated by any offline customer activity Attribution analysis can reveal such things as what kinds of campaigns most 126 influence customer behavior. “The analysis can help organizations determine where to allocate marketing resources to gain the highest level of success, as well as how to more accurately assign the percentage of credit due to specific 126 marketing and advertising processes,” Stodder concludes. 128 In its Achieving Customer Loyalty with Analytics , IBM argues that customer analytics can provide businesses with the ability to: • • • • • • • Analyze all data types to gain a 360-degree view of each individual customer. Employ advanced algorithms that uncover relevant patterns and causal relationships that impact customer satisfaction and loyalty. Build predictive models that anticipate future outcomes. Learn from every customer interaction and apply lessons to future interactions and strategies. Deploy customer insights to decision-makers and front line systems. Improve sales forecasting and help minimize sales cycles. Measure and report on marketing performance. Sentiment Analysis 126 In the TDWI Customer Analytics in the Age of Social Media Research report about the same percentage (30%) of respondents sought to monitor and measure sentiment drivers. “Sentiment analysis enables organizations to discover positive and negative comments in social media, customer comment and review sites, and similar sources. Sentiment analysis often focuses on monitoring and measuring the ‘buzz’ value, usually through volume and 126 frequency of comments around a topic.” However, it is not just the buzz that is important, many organizations want more analytical depth so that they can understand what the buzz is all about, where it comes from, and who is 126 benefiting or not benefiting from it the most. For more sophisticated sentiment analysis, text analytics tools that use word extraction, natural language processing, pattern matching, and other 126 approaches to examine social media users’ expressions are employed. “Sentiment analysis can give organizations early notice in real time of factors that may be affecting customer churn; the research shows that 14% are 126 interested in monitoring and analyzing social activity in real time.” In 2011, Toyota started testing social media monitoring and sentiment-analysis tools. After a few years of research they discovered that by filtering to find such 129 ANDREW PEARSON words as “Lexus”, “decide”, “buy” and “BMW”, they were able to quickly identify active shoppers who were choosing between theirs and their 132 competitor’s brands. Today, Toyota uses social media data analysis across 132 many areas—sales, service, quality, marketing and product development. For example, if a customer expresses interest in a car, Toyota “can determine engagement by analyzing the frequency of dealership visits via their Foursquare check-ins, understand their dealership experiences, and even understand what features may have sparked their interest in a competitor's 132 product.” This lets Toyota stratify its leads based on their readiness to buy, moving 132 stronger leads to the top of the funnel and weaker ones to the bottom. By analyzing free-form text, Toyota can learn what customers think of specific 132 vehicles. In the quality area, “Toyota can look for information like whether new-car owners are hearing a slight rattle and pass that on to their quality 132 engineers.” They are also working on using sentiment analysis to increase the accuracy of their sales predictions; an important goal everyone, if ever 132 there was one. Casino operators should keep these ideas in mind when developing their own use cases. A “rattle” for the casino wouldn’t be an engine problem, of course (except in a company bus, maybe), but rather a poor customer experience throughout the property. Since each IR offers a considerable choice of entertainment, their social media departments should be seeding social media websites for upcoming shows and events, then following up on the responses. Toyota also wants to deepen its understanding of its customers' other interests, like what a Camry owners' favorite TV show might be, as well as 132 which other brands they might like. This can help with product placement 132 and brand tie-ins down the line. Sentiment analysis is also key to understanding a competitors’ relative 126 strengths and weaknesses in the social sphere. The TDWI research found that “18% of respondents are examining social media data to analyze a 126 competition’s ‘share of voice.’” As Joe Mullich explains in his article Opposition Research: Sentiment Analysis as 133 a Competitive Marketing Tool : “When a leading bank wanted to find out how it stacked up against competitors, it assumed customers would focus on lending terms and interest rates. To the bank’s surprise, the most enthusiastic discourse on blogs and specialized financial forums related to a smartphone app a competing financial institution had just put out. The bank had dismissed apps as a generic marketing gimmick, like the old custom of giving away a toaster for opening an account. After learning how much 130 THE PREDICTIVE CASINO customers valued the app, the bank quickly created its own with the same prized features as its competitor.” “You get the benefits of corporate espionage without doing corporate espionage,” notes Joseph Carrabis, founder of NextStage Evolution, the 133 company that did the analysis for the bank. Sentiment analysis can also provide early insight into a competitor’s new 133 product initiatives. “Very often companies will test market before they 133 release a product,” says Mullich. “And no matter what you get people to sign saying that they won’t share information, they’ll go online and talk about 133 products they’re excited about,” warns Mullich. You can’t change human nature, but sometimes you can make it work for you. In addition, sentiment analysis can alert companies about new competitors 133 who are bubbling to the surface or even coming out of left-field. Ford would obviously consider Chevy a competitor, but it might not think of public 133 transport as being a competitor.” However, Carrabis argues that “a car company which realizes that it can analyze online discussions to get insight into why people are making different transportation choices and change their products or marketing to emphasize growing concerns, like ecological 133 impact.” “We have to think broader and wider than we used to,” Carrabis 133 advises. The lesson here is, don’t just look at your closest competitors as your competition. Gambling’s gambling and sports books and lottery companies are as much of a competitor for the casino gambler’s dollar as is a competing casino. This is why it’s imperative to understand why and how people discuss competitors online. “When car shoppers talk online they don’t talk about 133 ‘quality,’” says Susan Etlinger, an analyst with the Altimeter Group. “They’ll say, ‘I love the leather interior’ or ‘the cup holder fell out.’ It takes meticulous 133 work to roll together all the indicators of quality.” “Etlinger suggests that social-media listening teams work with the groups in the organization that handle keyword search terms and search-engine optimization effort, since they have a solid grasp on how people online actually talk about 133 the industry and products.” Another thing to keep in mind: “At any point in time, the way people feel about a brand can be distorted online, because things like Twitter are so volatile and 133 affected by the news of the day,” warns Etlinger. “But over time, you can get directional trends—why do people love or hate you, how do they feel about 133 your product compared to the competitor’s products.” Casinos in China have the added problem of Internet users who actively write fraudulent blogs and posts about a company’s products or services; separating fact from fiction is not an easy task. 131 ANDREW PEARSON “My belief is that the sweet spot for social media is not conversion, but 126 nurturing,” said Brian Ellefritz, vice president of global social media at SAP. “Whether it’s in your community, through Twitter, or through Facebook pages, you want to build an increasing conviction that your company is the one to do 126 business with,” says Ellefritz. “It’s about establishing a belief system that becomes robust with the support of fans and followers. The question is how 122 you measure that and create value out of that investment,” he added. When it comes to setting strategies for customer and social media analytics, 126 Stodder recommends the following : • • • • Use social media data to support an active, not passive social media strategy. “In competitive, fast-moving markets, organizations cannot just passively listen to and analyze social media data. The analytics should plug into strategies for engaging users and customers on social networks and comment sites. Predictive analytics can help orgnizations anticipate the results of active strategies. Special events such as tweet-ups can build on customer data analysis and create 126 positive exchanges and engagement.” Take a holistic view of the potential contributions of social media data analytics. “Understanding behavior in the social sphere can have a positive impact, not just on marketing and sales functions, but also on services and other processes in the organization. Marketing executives should use social media insights to improve brand awareness and 126 reputation throughout the organization.” Give CMOs and marketing executives the ability to understand the financial impact of certain decisions. Apply analytics to gain a more accurate understanding of marketing attribution. “Last-touch” attribution may be easy to affix, but it is not always reliable. Powerful analytics, along with big data, can help organizations get a better understanding of what truly affects a customer’s purchase decision. Clickstream Analysis When a person surfs a website, he or she leaves behind a digital trail. Clickstream analysis is the process of collecting, aggregating, reporting and analyzing the browsing behavior of a web surfer to better understand the intentions of users and their interests in specific content or products on a website. Clickstream analysis (also called clickstream analytics) is the process of collecting, analyzing and reporting aggregate data about which pages a website visitor visits—and in what order. The path the visitor takes though a website is, basically, the clickstream. There are two levels of clickstream analysis, traffic analytics and e-commerce analytics. Traffic analytics operates at the server level and tracks how many 132 THE PREDICTIVE CASINO pages are served to the user, how long it takes each page to load, how often the user hits the browser's back or stop button and how much data is transmitted before the user moves away from the website. E-commerce-based analysis uses clickstream data to determine the effectiveness of a website as a channel-to-market. It is concerned with what pages the browser lingers on, what he or she puts in or takes out of a shopping cart, what items are purchased, whether or not the buyer belongs to a loyalty program and uses a coupon code, as well as his or her preferred methods of payment. Utilizing clickstream analysis, a casino can help build a Master Marketing Record for each customer in real-time. This allows the casino to test scenarios and options for the website, as well as develop personalized responses for individuals. The system should include a combination of social listening, analytics, content publication and distribution, and tracking, as well as a strong workflow and rules engine that is geared around strong governance. All of these applications are built to ultimately feed a Master Marketing Profile—a centralized customer record that pulls in all data based on digital activity that can be identified by a single customer ID. Figure 7 shows the customer funnel that takes an anonymous web browser to an known patron. Through clickstream analytics, personalization marketing can begin, and associating this activity with a customer once he or she walks through the front Figure 7: Customer funnel door should be a casino’s primary goal. This can be done by enabling new users to log into his or her account via web or mobile applications like a casino’s WeChat app. Of all the available marketing and customer channels, social media represents the biggest issue due to the casino's inability to track the value of social 122 122 connections. In his article, At Caesar’s Digital Marketing Is No Crap Shoot , Al Urbanski explains that at Caesar’s, they “wanted to make better use of the social space, but one of the overwhelming problems had been, ‘How do you measure the effectiveness?' Not a lot of organizations are able to measure it 133 ANDREW PEARSON 122 effectively.” “Caesars marketers didn't want to create a social island that communicated with customers separately and distinctly from all other 122 channels,” Urbanski explains. Each channel is tracked and rated for its ability to turn engagements into booked rooms, and Caesars had been flying blind in the region of social 122 media.” “Top management wants to know, ‘How did this perform? What's the return on ad spend?' and how can we tell the path to purchase from first touch point to last touch point, even if the starting point was in social media,” 122 Kahle, Caesars' Web Analytics Manager, says. “Before, we couldn't understand social's role in the transaction. You could track it to a degree, but you built a social island and there was some guessing involved. You could end 122 up double counting social's contribution,” he warned. “To support these efforts, Caesars invested in Adobe's Digital Marketing Suite, which includes real-time tracking and segmentation of digital site visitors, analysis of social media's role in purchasing, and content testing by segment or 122 individual visitor,” explains Kahle. The problem with driving online conversions among Frequent Independent Travelers (FITs), however, is that—based on their online behaviors—they're not 122 loyal to a particular casino. “They're on Kayak; they're on other casinos' sites. They're looking for a deal,” says Kahle, who adds that Caesars regularly targets 122 these travelers with offers, such as free meals and free gaming play. “Kahle's staff conducted A/B analysis aimed at presenting the company's individual properties with the best option for increasing Total Rewards memberships. Half the people who searched Total Rewards online were sent to the main Caesars Entertainment homepage, while the other half was sent to the homepage of a specific property. While the conversion rate for room reservations was the same for both groups, the latter group signed up for the loyalty program at a significantly higher rate. The practice was adopted across 122 the Caesars Web network and resulted in a 10% increase in sign-ups.” “A similar test was used to maximize business from Total Rewards members, testing its old website interfaces against a new design. The difference was an eye-opener. The conversion rate for the newer interface option was 70% 122 higher,” Kahle explains. “In the past, when planning changes to Web page design or elements, the winning design was often decided by the highest-ranking person in the office,” 122 Kahle says. “With Adobe testing, people's personal opinions aren't the deciding factor. We can look at the numbers, see the results, and clearly 122 identify the best-performing design.” Kahle adds that Caesars deployed 122 these new capabilities without having to increase its IT staff. “Caesars went from a culture of opinion to a culture of data. We essentially 134 THE PREDICTIVE CASINO gave them the…flexibility to test so that the [end-user's] experience is 122 optimized,” says Matt Langie, senior director of product marketing at Adobe. As per Paul Greenberg’s article Is Adobe a Marketing Player Now? 134 , the: “First thing to know about Adobe is that they are a tools company—and this is both good and bad. For the most part and for the purposes of this discussion, the products are a good thing. The second thing to know about Adobe (and the Marketing Cloud) is that what they are currently offering is a significant piece of a digital engagement platform. Keep in mind this is for the marketing side—the first line of engagement—the first place that the customer comes into contact with the brand and either starts interacting or doesn’t.” The core around which the Marketing Cloud is built is the Master Marketing Record which goes to the heart of two Adobe themes, one of which is as old as modern man—the single view of the customer—and the other is the Real Time Enterprise. Just to step into the MCM (Multi-Channel Marketing) arena for a moment, as Greenberg explains, “The Adobe Marketing Cloud essentially consists of a basket of applications and services. At the highest level it is: • • • • • Adobe Analytics—A strong package focused around digital analytics, mobile and web in particular. They also, wisely, have predictive analytics as part of the core offering. Adobe Campaign—For now, this is where the core Neolane integration has occurred. They have done a remarkable job of taking that piece of Neolane’s capabilities and making it seamless in a short period of time. There are still parts of Neolane to be integrated into other areas and here too. Adobe Target—This allows you to test scenarios and options for the web and develop personalized responses for individuals. It's one of the most popular and powerful of Adobe’s tools. It's as close as Adobe gets to what Epiphany and Exact Target does. Adobe Experience Manager—This is Adobe’s digital experience management; its purpose is to manage the assets and create and manage the communities needed to optimize the customer journey across digital channels and media. Adobe Social—This is a combination of social listening, analytics, content publication and distribution, tracking (which they oddly call “campaign”) as well as a strong workflow and rules engine geared around governance and protocols. 135 ANDREW PEARSON • Adobe Media Optimizer—The idea is simple: Who’s your audience, what ads will appeal to them the most and what media mix should we use to maximize that appeal? To do that is complex and difficult, but the tools in Media Optimizer are designed to make it as accurate as 134 possible, while not necessarily as easy as possible.” All of these applications are built to ultimately feed what Adobe calls the Master Marketing Profile—a centralized customer record that pulls in all data based on digital activity that can be identified by a single customer ID. “That means that John Smith's record will have his social profile data, his transactional data, his response to campaigns, his click-throughs, his web browsing, etc. This goes to the heart of their effort—the personalization of the response to individual customers. The Master Marketing Profile is where you 134 find all that data.” To Adobe’s credit, Greenberg feels that the software vendor has done some 134 solid integration of the entire portfolio. The total package, if viewed as an advanced digital marketing cloud, some believe it could very well be the best of its kind on the market. However, Adobe is “competing with other Marketing Clouds—notably, if the name Marketing Cloud is meaningful, salesforce.com and Oracle. If the name Marketing Cloud isn’t—add Marketo, Microsoft, Teradata Applications, SAS, IBM Unica, and Infor to the mix. If niche players count, there are dozens and dozens out there chomping off pieces of the 134 potential revenue stream. This is a hot competition.” Location Analytics Location analytics is a technology that enables firms to capture and analyze location data on customers who are at a physical venue. In his article How 135 Location Analytics Will Transform Retail , Tony Costa argues that, “By leveraging connected mobile devices such as smartphones, existing in-venue Wi-Fi networks, low cost Bluetooth-enabled beacons, and a handful of other technologies, location analytics vendors have made it possible to get location analytics solutions up and running fast at a minimal cost.” “Customer tracking data is typically sent to the location analytics vendor where it is analyzed and accessed via online dashboards that provide actionable data tailored to the needs of specific employees—from the store manager to the executive C135 suite,” adds Costa. Already, the scale of data collected by early adopters is venturing into “Big Data” territory. Location analytics firm RetailNext currently “tracks more than 500 million shoppers per year by collecting data from more than 65,000 sensors installed in thousands of retail stores. A single customer visit alone can result in over 10,000 unique data points, not including the data gathered at the 135 point of sale.” 136 THE PREDICTIVE CASINO RetailNext is not alone; Euclid Analytics—one of the biggest players in this space—“collects six billion customer measurements each day across thousands of locations, and multiple location analytics firms surveyed said they are adding 135 hundreds of new venues each month.” As Costa explains, venue owners are applying insights gathered from location 135 analytics to help in all aspects of their business, including : 1. 2. 3. 4. Design. “After analyzing traffic flows in their stores, a big box retailer realized that less than 10% of customers visiting their shoe department engaged with the self-service wall display where merchandise was stacked. The culprit turned out to be a series of 135 benches placed in front of the wall, limiting customer access.” Simply relocating the benches to enhance accessibility increased sales 135 in the department by double digits. Marketing. A restaurant chain wanted to understand whether or not sponsoring a local music festival had a measurable impact on customer visits. After collecting data on 15,000 visitors passing through the festival entrances and comparing it to customers who visited their restaurants two months before and after the festival, they concluded that the festival resulted in 1,300 net new customer 135 visits. Operations. “A grocery store chain used location analytics to understand customer wait times in various departments and checkout registers. This data not only enabled the company to hold managers accountable for wait times, but it gave additional insight 135 into (and justification for) staffing needs.” Strategy. A regional clothing chain was concerned that opening an outlet store would cannibalize customers from its main stores. “After analyzing the customer base visiting each store, they discovered that less than 2% of their main store customers visited their outlet. The upside: the outlet gave them access to an entirely new customer base 135 with minimal impact to existing store sales.” Just as web analytics is an essential tool on the web, location analytics will become a must-have for designing, managing, and measuring offline 135 experiences. Location analytics is set to have a profound impact on how businesses operate in the very near future. Costa adds that, “Beyond creating more efficient, effective and meaningful services, firms will begin to rethink the 135 notion of customer value.” Costa argues that having the ability “to identify, track, and target customers in physical locations will enable companies to extend preferential status and rewards to customers based on their behaviors, rewarding them on the number and frequency of visits, where they go in 135 venues, and their exclusive loyalty (i.e., not visiting competitor venues).” 137 ANDREW PEARSON Analytics Today, the software analytics space is more crowded than its ever been. Standard ETL-solution providers are adding analytics to their multitude of offerings. Many of these new players in the Master Data Management (MDM) field have BI platforms that combine integration, preparation, analytics and visualization with governance and security features. Such standard analytics processes as column dependencies, clustering, decision trees, and a recommendation engine are all included in many of these software offerings. Instead of forcing clients to purchase modules on top of modules on top of modules, new software companies are creating packages that contain many built-in analytical functions. Open source products like R, Python, and the WEKA collection can easily be added to many of these software solutions, thereby reducing the need for expensive analytical layers. Figure 8: Analytics Value Escalator Source: www.intelligencia.co Before going any further, I believe one of the first questions that needs to be answered is, “What exactly is analytics?” The standard answer is that there are four different types of analytics and they are: 138 THE PREDICTIVE CASINO • • • • Descriptive analytics – What happened? Diagnostic analytics – Why did it happen? Predictive analytics – What will happen? Prescriptive analytics – How can we make it happen again? Figure 8 contains examples of how each of these types of analytics can be utilized by an IR. Diagnostic analytics is a form of advanced analytics that examines data or content to answer the question, “Why did it happen?” It attempts to understand causation and behaviors by utilizing such techniques as drill-down, data discovery, data mining and correlations. Building a decision tree atop a web user’s clickstream behavior pattern could be considered a form of diagnostic analytics as these patterns might reveal why a person clicked his or her way through a website. 136 In his seminal article Predictive Analytics White Paper , Charles Nyce states that, “Predictive analytics is a broad term describing a variety of statistical and analytical techniques used to develop models that predict future events or behaviors. The form of these predictive models varies, depending on the behavior or event that they are predicting. Most predictive models generate a score (a credit score, for example), with a higher score indicating a higher likelihood of the given behavior or event occurring.” Data mining, which is used to identify trends, patterns, and/or relationships 136 within a data set, can then be used to develop a predictive model. Prediction of future events is the key here and these analyses can be used in a multitude of ways, including forecasting behavior that could lead to a competitive advantage over rivals. Gut instinct can sometimes punch you in the gut and predictive analytics can help factor in variables that are inaccessible to the human mind and often the amount of variables in an analytical problem are beyond human mental comprehension. Predictive analytics (or supervised learning) is the use of statistics, machine learning, data mining, and modeling to analyze current and historical facts to make predictions about future events. Said another way, it gives mere mortals the ability to predict the future like Nostradamus. In recent years, data-mining has become one of the most valuable tools for extracting and manipulating data and for establishing patterns in order to produce useful information for decision-making. Whether you love it or hate it, predictive analytics has already helped elect presidents, discover new energy sources, score consumer credit, assess health risks, detect fraud, and target prospective buyers. It is here to stay, and technology advances ranging from faster hardware to software that analyzes increasingly vast quantities of data are making the use of predictive analytics 139 ANDREW PEARSON more creative and efficient than ever before. Predictive analytics is an area of data mining that deals with extracting information from data and using it to predict trends and behavioral patterns. Often the unknown event of interest is in the future, but predictive analytics can be applied to any type of unknown, whether that is in the past, the present, or the future. Predictive analytics uses many techniques from data mining to analyze current data to make predictions about the future, including statistics, modeling, machine learning, and artificial intelligence. For example, logistic regression can be used to turn a market basket analysis into a predictor so that a casino can understand what items are usually purchased together. Of course, the old beer and diapers story market basket wouldn’t fit for a casino, but gleaning data from the casino floor could reveal second favorite games that patrons like to play. This could be useful information when a patron is having a run of bad luck on his or her favorite game. Perhaps a marketing offer for a game he or she sometimes plays would be appreciated rather than an offer on his or her favorite game, as that might not be seen in such a positive light while the patron is in the midst of a losing run. For a casino, predictive analytics can also be used for CRM, collection analysis, cross-sell, customer retention, direct marketing, fraud detection, product prediction, project risk management, amongst many other things. Prescriptive analytics tries to optimize a key metric, such as profit, by not only anticipating what will happen, but also when it will happen and why it happens. Wikipedia states that, “Prescriptive analytics suggests decision options on how to take advantage of a future opportunity or mitigate a future risk and shows the implication of each decision option. Prescriptive analytics can continually take in new data to re-predict and re-prescribe, thus automatically improving 137 prediction accuracy and prescribing better decision options.” Prescriptive analytics can ingest a mixture of structured, unstructured, and semi-structured data, and utilize business rules that can predict what lies ahead, as well as advise how to exploit this predicted future without compromising other priorities. Stream processing can add an entirely new component to prescriptive analytics for this. Predictive analytics utilizes the following techniques: • • • • • • Regression Linear regression Discrete choice models Logistic regression Multinomial logistic regression Probit regression 140 THE PREDICTIVE CASINO • • • • • • • • Time series models Survival or duration analysis Classification and regression tress Multivariate adaptive regression splines Machine learning Neural networks Naïve Bayes K-nearest neighbors The analytics powerhouse SAS is finding its vaunted place atop the analytics pyramid challenged not just by their typical acronymed competitors—SAP, IBM, EMC, HDS, and the like—but also by the simpler visualization toolmakers like Tableau, Qlik, and Alteryx, who are muscling their way into the mix, with offers that include data blending and in-memory technology that allows business users to access complete datasets at the touch of a button. These solutions offer less complex analytical capabilities, but such things as market basket analysis or simple decision tree networks can be created with them and the costs associated with them can be one quarter or one fifth of what the top echelon providers charge. Throughout the rest of this chapter, I will break down many of the different types of analytical models that can be used to strengthen the customer experience for casino companies. In its conference paper How Predictive Analytics is Changing the Retail 138 Industry from the International Conference on Management and Information Systems, the writers argue that predictive models incorporate the following steps: • • • • • • Project Definition: Define the business objectives and desired outcomes for the project and translate them into predictive analytic objectives and tasks. Exploration: Analyze source data to determine the most appropriate data and model building approach, and scope the effort. Data Preparation: Select, extract, and transform data upon which to create models. Model Building: Create, test, and validate models, and evaluate whether they will meet project metrics and goals. Deployment: Apply model results to business decisions or processes. This ranges from sharing insights with business users to embedding models into applications to automating decisions and business processes Model Management: Manage models to improve performance (i.e., accuracy), control access, promote reuse, standardize toolsets, and minimize redundant activities. 141 ANDREW PEARSON Even though this was a paper around the subject of retail, it can be utilized for model building in the casino industry as well. Analytical Models Decision Trees According to Wikipedia, a decision tree is “a decision support tool that uses a tree-like graph or model of decisions and their possible consequences, including chance event outcomes, resource costs, and utility. It is one way to 139 display an algorithm.” Decision tress are used to identify the strategy that is most likely to reach a goal. It is a decision support tool that uses a graph or model of decisions and their possible consequences, including chance event outcomes, resource costs, and utility. Decision trees are sequential partitions of a set of data that maximize the differences of a dependent variable (response or output variable). They offer a concise way of defining groups that are consistent in their attributes, but which vary in terms of the dependent variable. A decision tree consists of three types of nodes: 1. 2. 3. Decision nodes – represented by squares. Chance nodes – represented by circles End nodes – represented by triangles The construction of a decision tree is based on the principle of “divide and conquer”: through a supervised learning algorithm, successive divisions of the multivariable space are carried out in order to maximize the distance between groups in each division (that is, carry out partitions that discriminate). The division process finalizes when all of the entries of a branch have the same value in the output variable, giving rise to the complete model. The further down the input variables are in the tree, the less important they are in the output classification (and the less generalization they allow, due to the decrease in the number of inputs in the descending branches). For the casino and hospitality industry, decision trees can be utilized in operations management and marketing, where they can predict whether a person will respond to an offer or not, or whether they are likely to abuse an 32 offer. According to Deng et al. in their paper Building a Big Data Analytics Service 140 Framework for Mobile Advertising and Marketing , the decision tree algorithm is: “Used to classify the attributes and decide the outcome of the class attribute. In order to construct decision tree both class attribute and item attributes are required. Decision tree is a 142 THE PREDICTIVE CASINO tree like structure where the intermediate nodes represent attributes of the data, leaf nodes represents the outcome of the data and the branches hold the attribute value. Decision trees are widely used in the classification process because no domain knowledge is needed to construct the decision tree.” The main step in the decision tree algorithm is to identify the root node for any 140 given set of data. “Multiple methods exist to decide the root node of the decision tree. Information gain and Gini impurity are the primary methods used to identify the root node. Root node plays important role in deciding which side of the decision tree the data falls into. Like every classification methods, decision trees are also constructed using the training data and tested with the 140 test data.” Advantages Disadvantages • • • • • • Simple and robust Useful to predict the outcomes of future data Little cleansing is enough to remove the missing values data Useful for large data sets Decision trees can handle both categorical and numerical data • • • Possibility of creating complex decision trees for simple data Replication problem makes the decision trees complex. So remove the replicated data before constructing a decision tree Pruning is required to avoid complex decision trees It is hard to find out the correct root node Table 4: Advantages and disadvantages of decision trees 140 Source: Researchgate k-Means Cluster As its name suggests, the k-Means cluster is a clustering algorithm and it is one of the most common analytical models because of its simplicity and ease of use. The fact that it is still going after over fifty years after it was created speaks as much to its ease-of-use as it does to the difficulty of designing a general purpose clustering algorithm. According to Telgarsky and Vattani, “The goal of cluster analysis is to partition a given set of items into clusters such that similar items are assigned to the same cluster whereas dissimilar ones are not. Perhaps the most popular clustering formulation is K-means in which the goal is to maximize the expected similarity 141 between data items and their associated cluster centroids.” Hartigan and Wong explain that the: “aim of the k-means algorithm is to divide M points in N dimensions into k clusters so that the within-cluster sum of squares is minimized. It is not practical to require that the solution has minimal sum of squares against all partitions, except when M, N are small and k = 2. We 143 ANDREW PEARSON seek instead ‘local’ optima, solutions that no movement of a point from one 142 cluster to another will reduce the within-cluster sum of squares.” K Means Clustering identifies and classifies items into groups based on their similarity. K is the number of clusters that needs to be decided upon before the 140 clustering process begins. “The whole solution depends on the K value. So, it is very important to choose a correct K value. The data point is grouped in to a cluster based on the Euclidean distance between the point and the centroid of 140 the cluster,” explains Deng et al. For Deng et al., initial clustering can be done in one of three ways. 1. 2. 3. Dynamically Chosen: In this method, the first K items are chosen and then assigned to K clusters. Randomly Chosen: In this method, the values are randomly selected and then assigned to K clusters. Choosing from Upper and Lower Boundaries: In this method, the values that are very distant from each other are chosen and they are 140 used as initial values for each cluster.” Figure 9: Clustering Algorithm Source: Researchgate 140 According to Deng at al., the K-Means methodology is as follows 144 140 : THE PREDICTIVE CASINO • • • • • • • • • • Step 1: Choose the initial values using one of the above three methods Step 2: For each additional value Step 3: Calculate the Euclidean distance between this point and centroid of the clusters. Step 4: Move the value to the nearest cluster. Step 5: Calculate the new centroid for the cluster. Step 6: Repeat steps 3 to 5. Step 7: Calculate centroid of the cluster. Step 8: For each value Step 9: Calculate the Euclidean distance between this value and the centroid of all the clusters. Step 10: Move the value to the nearest cluster. Advantages Disadvantages • • • • • • Faster computations than hierarchical clustering It produces tighter clusters than other clustering techniques Gives best result when data sets are distinct Easy to understand • • • • Sensitive to noise Numbers of clusters must be decided before starting clustering Choosing correct initial clustering process Choosing correct number of clusters The centroid of the group changes because we calculate centroid every time a new item joins the cluster Large data sets needed to cluster the data correctly Table 5: Advantages and disadvantages of decision trees 140 Source: Researchgate K-Nearest Neighbors First described in the early 1950s, the k-nearest neighbors method is a classification (or regression) algorithm that in order to determine the classification of a point, combines the classification of the K nearest points. It is supervised because you are trying to classify a point based on the known classification of other points, is labor intensive when given large training sets, and it did not gain popularity until the computer revolution in the 1960s 140 brought processing powers that were able to handle large data sets. Today, it 140 is widely used in the area of pattern recognition. As Deng et al. explain 140 : “Nearest-neighbor classifiers are based on learning by analogy, that is, by comparing a given test tuple with training tuples that are similar to it. The training tuples are described by n attributes. Each tuple represents a point in an n- 145 ANDREW PEARSON dimensional space. In this way, all of the training tuples are stored in an n-dimensional pattern space. When given an unknown tuple, a k-nearest-neighbor classifier searches the pattern space for the k training tuples that are closest to the unknown tuple. These k training tuples are the k ‘nearest neighbors’ of the unknown tuple. When the ‘k’ closest points are obtained, the unknown sample is then assigned to the most common class among those k-points. In case of k=1, the unknown sample is assigned to the closest point in the pattern space. The closeness is measured using the distance between the two points. The following table defines some of the approaches to find distances between two points.” the k-means clustering and k-nearest neighbor methodologies seek to accomplish different goals; k-nearest neighbors is a classification algorithm, which is a subset of supervised learning, while k-means is a clustering algorithm, which is a subset of unsupervised learning. K-nearest neighbor techniques can be used to prevent theft in the retail and casino business. Modern surveillance system are intelligent enough to analyze and interpret video data on their own, utilizing k-nearest neighbor for visual pattern recognition to scan and detect hidden packages in the bottom bin of a shopping cart at check-out, for example. As she explains in her article Solving Real-World Problems with Nearest 143 Neighbor Algorithms , Lillian Pierson explains that, “If an object is detected that’s an exact match for an object listed in the database, then the price of the spotted product could even automatically be added to the customer’s bill. While this automated billing practice is not used extensively at this time, the technology has been developed and is available for use.” The K-nearest neighbor algorithm can also be used to detect patterns in credit card usage to root out credit card fraud. “Many new transaction-scrutinizing software applications use kNN algorithms to analyze register data and spot 143 unusual patterns that indicate suspicious activity,” Pierson adds. “If register data indicates that a lot of customer information is being entered manually rather than through automated scanning and swiping, this could indicate that the employee who’s using that register is in fact stealing 143 customer’s personal information,” states Pierson. Another example would be “if register data indicates that a particular good is being returned or exchanged multiple times, this could indicate that employees are misusing the return 143 policy or trying to make money from doing fake returns.” kNN is not just about fraud. It can also be used to increase retail sales. “Average nearest neighbor algorithm classification and point pattern detection can be used in grocery retail to identify key patterns in customer purchasing 146 THE PREDICTIVE CASINO behavior, and subsequently increase sales and customer satisfaction by 143 anticipating customer behavior,” explains Pierson. Advantages Disadvantages • • • • It produces tighter clusters than other clustering techniques Gives best result when data sets are distinct Easy to understand • • KNN neither doesn’t follow any nor have any standard for selecting the value ‘k’, which is one of the key factors in the success of an algorithm As KNN is a Lazy Learner algorithm, it has high storage requirements and requires efficient indexing techniques The efficiency of the KNN algorithm also depends on the choice of the distance metric used. The results of the algorithm differ for each similarity metric Table 6: Advantages and disadvantages of decision trees 140 Source: Researchgate. Logistic Regression According to Wikipedia, logistic regression is a regression model where the dependent variable (DV) is categorical, i.e., a variable that can take on one of a 144 limited, and usually fixed, number of possible values. This compares to a variable that would be continuous. Developed in 1958 by statistician David Cox, “The binary logistic model is used to estimate the probability of a binary response based on one or more predictor (or independent) variables (features). It allows one to say that the presence of a risk factor increases the probability 144 of a given outcome by a specific percentage.” In his article Using Logistic Regression to Predict Customer Retention Karp explains that: 145 , Andrew “Logistic regression is an increasingly popular statistical technique used to model the probability of discrete (i.e., binary or multinomial) outcomes. When properly applied, logistic regression analyses yield very powerful insights in to what attributes (i.e., variables) are more or less likely to predict event outcome in a population of interest. These models also show the extent to which changes in the values of the attributes may increase or decrease the predicted probability of event outcome.” Logistic regression techniques may be used to classify a new observation whose group is unknown, in one of the groups, based on the values of the predictor variables. According to Karp, “Logistic regression models are frequently employed to assess the chance that a customer will: a) re-purchase a product, 147 ANDREW PEARSON b) remain a customer, or c) respond to a direct mail or other marketing 145 stimulus.” Karp adds that “Economists frequently call logistic regression a ‘qualitative choice’ model, and for obvious reasons: a logistic regression model helps us assess probability which ‘qualities’ or ‘outcomes’ will be chosen (selected) by 145 the population under analysis.” As can be expected, Karp argues that, “When proper care is taken to create an appropriate dependent variable, logistic regression is often a superior (both substantively and statistically) alternative to 145 other tools available to model event outcomes.” Karp uses a health care example to make his point that the analyst has several independent variables to use in the modeling process, but this example can be 145 illustrative of how they could be used in the casino industry. Karp explains that “An analyst developing a model predicting re-enrollment in a health insurance plan may have data for each member’s interaction with both the health plans administrative apparatus and health care utilization in the prior 145 ‘plan year.’” The analyst can then construct variables such as the “number of times member called the health plan for information, number of physician office visits, whether or not the member changed primary care physicians during the 145 previous ‘plan year,’ and answers to a customer satisfaction survey.” These can be employed in the modeling process and, once the model has been constructed, the analyst must decide which variable can be employed as the 145 “outcome” or the “dependent” variable. In logistic regression analyses “it is often the analyst’s responsibility to construct the dependent variable based on an agreed-upon definition of what 145 constitutes the ‘event of interest’ which is being modeled.” In the health care re-enrollment example, “a health plan’s management team may define ‘attrition’ or ‘failure to re-enroll’ as situations where a member fails to return the re-enrollment card within 30 days of its due date. Or, in a response modeling scenario, a direct mail firm may define ‘non-response’ to an ”145 advertisement as failure to respond within 45 days of mailout. Logistic regression models can be powerful tool in building models to 145 understand customer retention. “When applied properly, logistic regression models can yield powerful insights into why some customers leave and others stay. These insights can then be employed to modify organizational strategies and/or assess the impact of the implementation of these strategies,” Karp 145 adds. A/B Testing Also known as split testing or bucket testing, A/B testing is a method of 148 THE PREDICTIVE CASINO marketing testing by which a baseline control sample is compared to a variety of single-variable test samples in order to improve response rates. A classic direct mail tactic, this method has recently been adopted within the interactive space to test tactics such as banner ads, emails and landing pages. As Scott Sutton explains in his article Patron Analytics in the Casino and 32 Hospitality Industry: How the House Always Wins , for casino marketers, A/B Testing is the most effective way to identify the best available marketing 32 offer. It can test “two different offers against one another in order to identify 32 the offer that drives the highest response and the most revenue/profit.” As Dan Siroker and Peter Komen explain in their book A/B Testing: The Most 146 Powerful Way to Turn Clicks Into Customers , “The hardest part of A/B testing is determining what to test in the fist place. Having worked with thousands of customers who do A/B testing every day, one of the most common questions we hear is, ‘Where do I begin?’” The mistake many companies make is they jump in head first without any detailed planning. Siroker and Komen propose the following deliberate five146 step process : 1. 2. 3. 4. 5. Define success Identify bottlenecks Construct a hypothesis Prioritize Test A/B testing is particularly good for website marketing, particularly for landing pages. As Siroker and Komen explain, “Defining success in the context of A/B testing involves taking the answer to the question of your site’s ultimate purpose and turning it into something more precise: quantifiable success metrics. Your success metrics are the specific numbers you hope will be 146 improved by your tests.” Whereas an e-commerce could easily define its success metrics in terms of 146 revenue per visitor , a casino website could similarly look at completed hotel reservations, as well as other types of event bookings for A/B testing. Site Type e-Commerce A site that sells things for users to purchase online. Media/Content A site focused on article or Common Conversion & Aggregate Goals • • • • Completed purchase Each step within the checkout funnel Products added to cart Product page views • • • Page views Articles read Bounce rate (when measuring within an A/B 149 ANDREW PEARSON Site Type Common Conversion & Aggregate Goals other content consumption. Lead Generation A site business capture. that acquires through name Donation testing tool, this is often measured by seeing if the user clicked anywhere on the page) • • Form completion Clicks to a form page (links may read “Contact us” for example) • • Form completion Clicks to a form page (links may read “Send a donation” for example Table 7: Typical A/B conversion & aggregate goals Source: A/B Testing: The Most Powerful Way to Turn Clicks Into Customers 146 As Siroker and Komen state, “Part of building out your testing strategy is identifying what constitutes—and does not constitute—a “conversion” for your particular site. In online terms, a conversion is the point at which a visitor takes the desire action on your website. Pinpointing the specific actions you want people to take most on your site and that are most critical to your business will 146 lead you to the tests that have an impact.” Once the site’s quantifiable success metrics are agreed upon, attention can be 146 paid trying to discover where the bottlenecks are. These are the places where users are dropping off, or the places where momentum in moving users 146 through the desired series of actions weakens. Time Series Model A time series is an ordered sequence of values of a variable at uniformly spaced time intervals. A Time Series model can be used to predict or forecast the future behavior of a variable. In his article Time Series Analysis, Muhammad Imdadullah explains that “Time series analysis is the analysis of a series of data-points over time, allowing one to answer question such as what is the causal effect on a variable Y of a change in variable X over time? An important difference between time series and cross 147 section data is that the ordering of cases does matter in time series.” These models account for the fact that data points taken over time may have an internal structure (such as autocorrelation, trend or seasonal variation) that should be accounted for. For the casino and hospitality industry, a Time Series Analysis can be used to forecast sales, project yields and workloads, as well as analyze budgets. Time series can be broken down into two variations: 150 THE PREDICTIVE CASINO • • Continuous Time Series—“A time series is said to be continuous when observation are made continuously in time. The term continuous is used for series of this type even when the measured variable can only 147 take a discrete set of values.” Discrete Time Series—“A time series is said to be discrete when observations are taken at specific time, usually equally spaced. The term discrete is used for series of this type even when the measured 147 variable is continuous variable.” As Sang and Dong explain in their Determining Revenue-Generating Casino Visitors Using a Vector Autoregressive Model: The Case of the G Casino in 148 Korea , time-series data was analyzed to: “investigate the characteristics of casino visitors that affect casinos’ revenue generation. Exchange rates—a traditional measure relevant to tourism—and customer types and nationalities were empirically analyzed with a vector autoregressive model using data acquired from all branches of Korea’s G casino. The results suggest that the casinos’ revenues were affected by the customers’ type and nationality: VIP customers were very important factors in the casinos’ revenue generation; moreover, the revenue impact of Russian visitors was quite strong despite their small numbers.” Time series can be used to compare seasonal estimation and trend estimation in forecasting models on both a state or an national level. Neural networks Artificial Neural Networks (ANN) or just “Neural Networks” are non-linear statistical data modeling tools that are used when the exact nature of a relationship between input and output is unknown. In their article Neural 149 Networks in Data Mining , Singh and Chauhan claim that a neural network is: “A mathematical model or computational model based on biological neural networks, in other words, is an emulation of biological neural system. It consists of an interconnected group of artificial neurons and processes information using a connectionist approach to computation. In most cases an ANN is an adaptive system that changes its structure based on external or internal information that flows through the network during the learning phase.” As Jim Gao explains in his article Machine Learning Applications for Data Center 150 Optimization : 151 ANDREW PEARSON “Neural networks are a class of machine learning algorithms that mimic cognitive behavior via interactions between artificial neurons. They are advantageous for modeling intricate systems because neural networks do not require the user to predefine the feature interactions in the model, which assumes relationships within the data. Instead, the neural network searches for patterns and interactions between features to automatically generate a bestfit model. Common applications for this branch of machine learning include speech recognition, image processing, and autonomous software agents. As with most learning systems, the model accuracy improves over time as new training data is acquired.” Neural networks can be used to find patterns in data. A key feature of neural networks is that they learn the relationship between inputs and output through training. There are three types of training in neural networks; reinforcement learning, supervised and unsupervised training, with supervised being the most common one. Neural Networks are data processing systems whose structure and functioning are inspired by biological neural networks. Their fundamental characteristics include parallel processing, distributed memory and adaptability to their surroundings. For casino and hospitality marketing purposes, neural networks can be used to classify a consumer's spending pattern, analyze a new product, identify a 149 patron's characteristics as well as forecast sales. The advantages of neural networks include high accuracy, high noise tolerance and ease of use as they can be updated with fresh data, which makes them useful for dynamic 149 environments. Discriminant Analysis According to Wikipedia, “Discriminant function analysis is a statistical analysis to predict a categorical dependent variable (called a grouping variable) by one or more continuous or binary independent variables (called predictor variables). The original dichotomous discriminant analysis was developed by Sir Ronald Fisher in 1936. It differs from an ANOVA or MANOVA, which is used to predict one (ANOVA) or multiple (MANOVA) continuous dependent variables 151 by one or more independent categorical variables.” Discriminant or discriminant function analysis is a method used to determine which weightings of quantitative variables or predictors best discriminate between two or more than two groups of cases and do so better than chance. It is a method used in statistics, pattern recognition and machine learning to 152 THE PREDICTIVE CASINO find a linear combination of features that characterizes or separates two or more classes of objects or events. Because of its ability to classify individuals or experimental units into two or more uniquely defined populations, discriminate analysis can be used for market segmentation and the prediction of group membership. The discriminant score can be the basis on which a prediction about group membership is made. For example, the discriminant weights of each predictive variable (age, sex, income, etc.) indicate the relative importance of each variable. For example, if age has a low discriminant weight then it is less important than the other variables. For a casino and hospitality marketing department, use of discriminant analysis can help predict why a patron frequents one casino over another. Discriminant analysis is specifically useful in product research, perception/image research, advertising research and direct marketing. Survival or Duration Analysis As per Wikipedia, “Survival analysis is a branch of statistics for analyzing the expected duration of time until one or more events happen, such as death in biological organisms and failure in mechanical systems. This topic is called reliability theory or reliability analysis in engineering, duration analysis or 152 duration modeling in economics, and event history analysis in sociology.” 152 Survival analysis attempts to answer questions such as : • • • • what is the proportion of a population which will survive past a certain time? Of those that survive, at what rate will they die or fail? Can multiple causes of death or failure be taken into account? How do particular circumstances or characteristics increase or decrease the probability of survival? A branch of statistics that deals with death in biological organisms and failure in mechanical systems. It involves the modeling of time to event data; in this context, death or failure is considered an “event” in the survival analysis literature – traditionally only a single event occurs, after which the organism or mechanism is dead or broken. Survival Analysis is the study of lifetimes and their distributions. It usually involves one or more of the following objectives: 1. 2. 3. 4. To explore the behavior of the distribution of a lifetime. To model the distribution of a lifetime. To test for differences between the distributions of two or more lifetimes. To model the impact of one or more explanatory variables on a lifetime distribution. 153 ANDREW PEARSON By applying survival analysis to revenue management models, casino operators 153 can gain a truer picture of their table games revenue. There are several other data mining techniques that can be used but the ones listed above are the most commonly used ones in the industry and much of what you will need to glean from your data can be discovered by using them. Once the data has been mined, a business intelligence solution can tell you what's going on in your data while a predictive analytics program can actually analyze current and historical trends to make predictions about future events. Edge Analytics The driving concept behind edge analytics is the fact that data loses its value as it ages. As previously mentioned, the concept of “Edge Analytics”—i.e., the processing of analytics at the point or very close to the point where the data is being collected—exponentially increases the ability to use predictive analytics where it can be utilized best. As Patrick McGarry explains in his article Why Edge Computing is Here to 154 Stay , Edge analytics is easier to implement than ever before because in the field micro data centers use a fraction of the space, power and cost of a traditional analytics infrastructure, but they can provide massive performance gains. These systems use “hybrid computing technology, seamlessly integrating diverse computing technologies, whether they are x86, GPU or FPGA technologies, or any combination thereof. They are extremely compact in space and require very little power, yet still provide performance that is several orders of magnitude more than what today’s traditional systems can 154 provide.” “It’s a win/win situation for all involved; insights come faster than ever before, operational expenses are lower, [sic] power and administration 154 needed to run the systems,” McGarry ads. “Emergency repair work and equipment down-time can be reduced when manufacturers build edge-based analytical systems into machinery and vehicles, allowing them to decide for themselves when it is time to reduce power output or send an alert that a part may be due for 8 replacement.” Integrated resorts, which buy equipment on a massive scale can connect these IoT devices into their data warehouse and enact predictive asset maintenance to reduce plant and equipment costs. Although building an edge analytics platform does require a shift in corporate thinking, the ROI benefits should far outweigh the costs. “The cost savings by scaling back central data analytics infrastructures to handle non-time sensitive analysis while installing cost-efficient platforms purpose-built for edge analytics 154 can have a real impact on an organization’s budget,” McGarry notes. The value of near-instant analysis and insight cannot be underestimated in a business so dependent on customer excellence like the casino business. 154 THE PREDICTIVE CASINO Avoiding latency and eliminating the time and costs associated with transporting the data to and from the edge is a major step toward achieving 154 that goal. IoT sensors can help spot patrons arriving in a casino, or track employees, suppliers, and supplies throughout the casino, as well as help save energy and water usage. Edge analytics can help analyze retail customer behavior, as well as spot upcoming equipment malfunction. Other areas where it can help include compliance analysis and mobile data thinning, i.e., the culling of mobile data noise from social media or direct mobile streams. Personally, I believe edge analytics could be one of the top technologies that will give a casino company or an IR a competitive advantage over its rivals and I will provide more detail on how this technology can be implemented within a casino environment throughout the rest of this book. Casino Analytical Models Customer Segmentation A customer segmentation model provides a view of the casino from a customer perspective: such models have many and varied applications. Customers are segmented according to what they present to the casino. Views include: 1. 2. 3. 4. 5. Game preference: the games offered by a casino are grouped according to business needs and customer turnover by game is analyzed to derive a preferential game for every customer above a threshold turnover. Day of week: customer turnover is analyzed by day of week and clusters are derived in line with how the casino is visited. It is often the case that customers group into single day segments, as well as some longer segments, such as weekend players, midweek players, etc., etc. Time of day: split the 24 hours of the day into meaningful bins and analyze customer turnover by time period from when their session began. This leads to a view of the customer according to when they are most likely to frequent the casino, e.g. matinee players, night owls, etc., etc. Length of session: this view clusters customers according to how long their play session is likely to be. Whilst this isn’t usually as pure a view as the above models, it can give an idea of impulse players, session players, etc., etc. Size of stake: how the customer stakes can be an important view for planning. Depending on how granular the venue information is, individual stakes or the table limit or credits played per hand on a slot machine are used to segment customers according to their stakes. Generally, the data is used to determine the appropriate segments for these 155 ANDREW PEARSON views. However, the casino has the ability to select the intervals that are preferential and relevant to their venue. For example, it may be desired to split time of day into three, eight-hour periods or six, four-hour periods. The results of this analysis presents a detailed view of how the casino is populated at different times and can allow for appropriate strategic decisions to be made. These decisions could be a function of marketing, operations or strategy. The output is also used for the building of acquisition models as discussed below. Other potential for analysis would be a master segmentation model that uses the preference results described. Customers are clustered based on their preferences to gain a global view of the casino that is concise and understandable. Furthermore, such models can help measure the impact of strategic decisions, e.g. the addition or removal of a sports or lottery game can be measured against how particular metrics are affected. Customer Acquisition Model Casino operators are always looking for new customers. With the market getting more and more competitive and saturated by the day, there is always a constant need to know where to attract customers from and what type of customer to target. The results of the segmentation modeling previously described can be used to build a predictive model that identifies likely characteristics of attractive customers. Obviously the casino will have no internal data available on customers they don’t already have on their books so the analysis becomes a data mining exercise using publically available input variables. Casinos can then target these customers with a view to attracting those who have the traits that they see in their already valuable customers. The best external data to use would be population census data, linked to the internal customers by a location identifier (such as postcode or mesh block). It is acknowledged that in some jurisdictions robust and accurate census data may not be available so the model would be relying on whatever information the sports book records on its customers from a demographic and lifestyle point of view. This approach becomes a classical data-mining problem, where a pool of independent variables would be tested for the strength of association with the response variable. Once the relevant predictors are identified and the characteristics and traits are defined, marketing and acquisition campaigns could be targeted at the population towards these kinds of people. This would be something that looks to predict a metric derived from current/past customers. Such a metric could come from a segmentation model that identified the high value customers that are most attractive to the casino. There are several approaches that can be used and once the target has been 156 THE PREDICTIVE CASINO defined, this allows for a parametric equation to be derived. This equation attempts to predict the characteristics that distinguish the desirable customers from the rest. This model can only use publically available information (although other casino information might be acceptable) as that is how a potential customer would be identified. Current information that the company would have on hand would be age, nationality, gender, and address. Where available, third party data should be looked at to further enhance the findings. This could be census data that gives an indication of further customer demographics and this enhances the ability to hone in on customer sweet spots. Recency-Frequency-Monetary (RFM) Models RFM is a method used for analyzing customer value. It is commonly used in database marketing and direct marketing and has received particular attention in the casino and retail industries. RFM stands for: • • • Recency—How recently did the customer purchase? Frequency—How often do they purchase? Monetary Value—How much do they spend? Most businesses will keep scores of data about a customer’s purchases. All that is needed is a table with the customer name, date of purchase and purchase value. One methodology is to assign a scale of 1 to 10, whereby 10 is the maximum value and to stipulate a formula by which the data suits the scale. For example, in a service based business like the casino business, you could have the following: • • • Recency = 10—the number of months that have passed since the customer last purchased Frequency = number of purchases in the last 12 months (maximum of 10) Monetary = value of the highest order from a given customer (benchmarked against $10k) Alternatively, one can create categories for each attribute. For instance, the Recency attribute might be broken into three categories: customers with purchases within the last 90 days; purchases between 91 and 365 days; and purchases longer than 365 days. Such categories may be arrived at by applying business rules, or using a data mining technique to find meaningful breaks. Once each of the attributes has appropriate categories defined, segments are created from the intersection of the values. If there were three categories for each attribute, then the resulting matrix would have twenty-seven possible combinations (one well-known commercial approach uses five bins per 157 ANDREW PEARSON attributes, which yields 125 segments). Segments could also be collapsed into sub-segments, if the gradations appear too small to be useful. The resulting segments can be ordered from most valuable (highest recency, frequency, and value) to least valuable (lowest recency, frequency, and value). Identifying the most valuable RFM segments can capitalize on chance relationships in the data used for this analysis. For this reason, it is highly recommended that another set of data be used to validate the results of the RFM segmentation process. Advocates of this technique point out that it has the virtue of simplicity: no specialized statistical software is required, and the results are readily understood by business people. In the absence of other targeting techniques, it can provide a lift in response rates for promotions. Whichever approach is adopted, profiling will be done on the final results to determine what makes up group membership. Categorical factors such as gender, nationality/locality can be used as well as age (or, indeed, any other demographic feature that is available) to understand the “type” of customer that resides in each group. These factors can be used for each segment and applied against the population metrics to determine how much more or less likely a segment is to exhibit a particular feature or type of behavior when compared to the customer base as a whole. A few words of caution in the gaming field: a major drawback of classical RFM modeling is the high propensity of a casino and/or a sports book to be continually hitting the same segment(s) with the same marketing message. Propensity to Respond Model A Propensity to Response model is the theoretical probability that a sampled person (or unit) will become a respondent in an offer or survey. They are specifically useful in the marketing field. A response likelihood model can have substantial cost savings as it can lead to lower mailing costs by identifying patrons who are very unlikely to respond to a particular offer. After segmenting these people out, the casino can then focus on only those most likely to take up the offer. A casino can identify the 32 likelihood of response from all eligible patrons. After that, it can identify the 32 most valuable patrons that are most likely to respond. This allows the casino to estimate the expected response from the most valuable patrons and eliminate mailing(s) to the patrons that are of lower worth and/or are unlikely 32 to respond. Sutton warns that, “Occasionally, response likelihood models will lead to easy decisions, such as cutting out low worth patrons with a low likelihood of responding. However, more complex situations might arise since response 32 models are never perfect.” It doesn’t matter how good a model is or how accurate the historical data is, there is always a chance that a patron identified 158 THE PREDICTIVE CASINO 32 as unlikely to respond will respond. “Thus, when making a decision about patrons identified as unlikely to respond to an offer, it is also important to balance that likelihood of response with the potential return on response,” 32 advises Sutton. A propensity to respond model would be built using historical information around marketing campaigns and it looks at predicting the likelihood a customer will respond to a marketing communication. The advantage of this model is that it strengthens the marketing strategy even more, beyond purely segmenting the customer base. It can further allow for improved ROI on the marketing budget, by identifying the likely number of respondents to be returned by a campaign. Often a business’ marketing department will have an expected number of 32 respondents or an expected response rate. By identifying those who are most likely to respond, the chances of meeting that expected number or rate of 32 response is greatly improved. Gone are the days of marketing to an entire customer base. This is an unnecessary waste of the marketing budget and also runs the risk of annoying customers by touching them too often or with the 32 wrong offer. Again, a predictive model would be built which identifies those most likely to respond through to those least likely to respond. This would be done using customer metrics and historical campaign/marketing information that identifies those who responded and those who didn’t. Variables that have a significant association with the customer action are extracted and these form part of the prediction algorithm. Every customer is then given a score according to how likely they are to respond to marketing campaigns. This information can be used for strategies such as extracting the top 40% of customers most likely to respond, or a fixed number, such as 100,000 customers. The end result is the marketing function becomes more efficient and effective with better returns for the company’s marketing dollar. Customer Conversion Model This model would be used to score customers based on information contained in a casino’s source systems as it would only be applicable for customers who had pre-booked their room (as opposed to walk-in customers). Historical information would be extracted from the casino’s IT systems around desirable customers. This would include spending patterns and profitability. To identify the relationships that may exist between how these customers come to the casino and his or her desirability metric, information would be extracted from the casino’s source systems. For a sports book, this would include information such as source of betting, channel of betting, lead-time for betting and the incentives offered to attract the customer. Basically, anything 159 ANDREW PEARSON that can be attributed to the initial transaction the customer has with the casino would be used as a potential input. These models might also have to be stratified by itinerary to identify the most relevant relationships. The major advantage of a predictive model with this intention would be that it allows the casino to identify customers that they need to interact with once they step onto the casino floor. This would give the casino hosts the potential to get the required information they need to successfully foster a strong customer relationship. Furthermore, if every potential customer has a score associated with them as to their long-term likelihood of being attractive, the casino can further hone in on its customers by monitoring their behavior once they are on the casino floor. It is imperative that the casino interact with desirable customers before they have left the property. If customers are made to feel like they are valuable and worthwhile, the likelihood of them returning under their own volition significantly increases. Identify When a Patron is Likely to Return Besides knowing which offers a patron is most likely to respond to, it would also be nice to know exactly when a patron was planning to make his or her 32 next bet. Although it might not be possible to know exactly when a patron plans to return, the casino’s marketing department might be able to make an 32 accurate prediction around it. “There are a variety of methods that range in complexity that can be used to assess when a patron will return to bet, including frequency analysis, regression, and survival analysis. Knowing when a patron is likely to bet is beneficial as it helps to identify patrons that haven’t made a trip in the expected amount of time and are at risk of leaving,” advises Sutton. “First, the business needs to have an idea of the average or median time between trips. This might need to be segmented based on geography, worth, or even historical frequency,” recommends Sutton. Patrons who haven’t made a bet within the set amount of time for his or her segment will be flagged and dealt with appropriately, perhaps marketed to more aggressively, perhaps given marketing content referencing “We haven’t seen you in a while”, or “Last 32 chance type of offers.” “Historical data can help to identify segments of patrons that are expected to 32 make bets weekly, monthly, quarterly, annually, bi-annually” , or around specific events such as the World Cup, the Euro championship, or the Champions League tournament, say for a sports book. “Marketing can integrate information from predicted worth, optimal offers, and time to next trip to maximize campaign success in a number of ways,” says Sutton. “The business can save money by adjusting the frequency of offers for 160 THE PREDICTIVE CASINO patrons that are not identified as likely to bet on rarer events. Instead of sending the patron monthly offers, they can send quarterly offers with longer validation windows that allow more time to bet. Conversely, campaigns might be created with the goal of increasing the frequency of bets from higher worth patrons,” recommends Sutton. The casino marketing department’s goal should be to generate customer bets sooner than expected and converting patrons into more frequent bettors or perhaps manipulating behavior so that bettors could spread out their bets and use their typical weekly bet bankroll over the course of several days, rather 32 than blowing it all at once. Although the casino property might prefer the player to spend his or her money all at once, spreading out the gambling losses would mean more money would be spent on other property endeavors. Identify Patrons Who Come Together “Another important consideration in the discussion of patron worth is household worth,” Sutton explains. “This refers to the combined worth of multiple patrons that tend to make their trips together. This can be difficult to identify, as these patrons might stay in one room or separate rooms, or one 32 patron might only come when accompanied by their spouse or their partner.” “Additionally, the other patron might make trips without the first patron. Although identifying household worth can be tricky, it can pay huge dividends by helping to account for revenue that looks like two separate individuals but can be combined into one ‘household,’” advises Sutton. Most patron management and/or CRM systems “contain the functionality to link accounts so that patrons who come together (i.e., married couples) can be 32 easily identified.” Unfortunately because of system or regulatory limitations, 32 patrons might not be allowed to have linked accounts. Nevertheless, data mining can be used to identify groups of patrons who come together without linked accounts. “There are ways to identify patrons with trips over the same period of time that have some commonality about their behavior that can be used to be fairly certain the patrons are together,” Sutton explains. “For some of the more subjective measures (i.e., room, floor, city, time/type of play) it’s a good idea to be more conservative about how many overlapping trips the patrons have,” advises Sutton. “For instance, two patrons with one overlapping trip and rooms next to each other may or may not be in the same ‘household’ group. However, those same patrons with five overlapping trips, each with rooms next to one 32 another, are much more likely to be in the same household grouping.” “In this manner, household grouping can identify a group of four patrons that are of ‘middle of the pack’ worth individually, but come together and stay in the same room every time and thus are worth much more as a group,” says Sutton. Once a “household” has been discovered, the casino’s marketing department can 161 ANDREW PEARSON adjust its marketing efforts and send a better offer based on the group’s combined worth, safe in the knowledge that they’re not really marketing to 32 four unique individuals, but rather to a group of related patrons. Patron Worth Model As Sutton explains, “Most industry experts would agree that determining a patron’s worth is the first and foremost responsibility of patron analytics in the 32 casino industry.” Of course, predicting a patron’s future behavior is not easy and it is affected by a number of variables, “many of which are outside factors that the business might not have insight into, including total income, 32 expendable income, ethnicity, reasons for a trip to the casino, etc., etc.” Even where a patron lives or information gleaned from his or her social media accounts could be very revealing. There is also “plenty of information to be found with in-house data that can be used to build models and metrics to 32 predict a patron’s future worth.” Once patron worth has been determined, “patrons can then be segmented into groups based on other behaviors and 32 effective marketing campaigns can be developed around those behaviors.” The first thing to do is to “determine what worth is, as the definition of worth is critical for deciding how valuable a patron is and how much to reinvest in the 32 patron in the future.” As Sutton explains, “There are two main components of worth—the financial sources of worth (i.e., gambling) and the unit of time to which it refers (daily, weekly, monthly, etc.). Additionally, worth can refer to 32 historical worth, which is already known, or future worth, which is unknown.” “The definition of worth will likely depend on both the various financial sources of revenue that affect the business directly and the exact business problems that are being addressed. Gambling worth can also be broken down into various sources (i.e., what types of betting does the patron like to do) 32 depending on the business issues being addressed,” Sutton explains. In the casino industry, there are “two important measures used to assess a patron’s gaming worth—actual and theoretical loss. Actual loss is how much money the patron actually lost (or won), whereas theoretical loss usually refers to the amount of money a patron is expected to lose based on the amount of money wagered, the time spent playing, and the probability associated with 32 type of games played.” Theoretical loss or Theo loss “tends to be more heavily relied upon for predictive analysis and is a much stronger predictor of future behavior, as actual loss is usually used to measure campaign 32 performance and profitability.” Actual win could be heavily affected by a lucky winning streak, while Theo win won’t. “Once patron worth has been defined, the business can then use data mining 32 and modeling to estimate predicted worth into the future,” states Sutton. “There are a variety of techniques that are used to develop models to predict 162 THE PREDICTIVE CASINO future worth, the most common being regression models. Multiple regression models are the most common because they utilize a variety of predictors and the relationships between those predictors to predict future worth,” adds 32 Sutton. “Regression models can also be built using such categorical variables predictors as gender, ethnicity, age range, or other demographic variables. Developing separate models based on categorical variables, such as separate models predicting worth for slot and table players, might produce models with less 32 error and better predictions.” “Regression models are particularly effective because the model can be used to score historical data to predict an unknown outcome, which is worth in this case, within a certain degree of confidence,” 32 adds Sutton. All casino and sports betting analytics departments should have a solid method for predicting the various types of patron worth based on the sources and time periods they need for making informed marketing decisions. For instance, daily gambling worth would be most useful for building a campaign with a daily free play offer. Customer Churn Model The use of analytics and data management to help detect and avoid the act of attrition is something that can benefit all sports books and casinos. Churn questions that casinos should be asking include: • • How is a casino detecting behavior changes in is patrons? Does the casino have steps in place to identify when the customer experience is going wrong, or when the customer is about to leave? Casino operators can use Master Data Management (MDM) techniques to communicate important customer preference information to staff who sit at interaction points throughout the casino property. MDM is the processes, governance, policies, standards and tools that consistently define and manage the critical data of an organization to provide a single point of reference. One of the benefits of using MDM is that when that single point of reference is a customer profile, the master data can ensure that the treatment of a customer is consistent and that preference information reaches all customer points of contact. To ensure customer retention is front and center, casinos and sports books should be scoring its database on a regular basis in order to understand the likelihood of a customer churning from their venue. This kind of modeling is prevalent in the telecommunications, finance, and utilities industries, and should be utilized in the gaming industry. Whilst a slightly different set up due to those industries mostly having their customers locked into contracts, gaming companies need to stay ahead of the game in retaining their customers. 163 ANDREW PEARSON Anecdotal evidence collected in discussions with gaming venues have indicated a tendency to ignore customers until they have not been seen for up to two years. At this stage, there might be a marketing activity targeted at the customer for up to 12 months. It could be proffered that, by this stage, it is too late to win the customer back; the customer has probably already made up his or her mind and, once a decision like that has been made, it is almost impossible to reverse it, no matter how attractive any competing offer might be. One of the hardest parts for a gaming company to determine—as opposed to commercial entities that have their customers on contract and definitely know they are tied down—is whether the customer has categorically churned. It may be that a change in location, circumstances or something else has caused a customer to disappear from the casino or sports book, with every intention of returning. However, statistical measures could be used to identify customer’s whose behavior has changed and wouldn’t be attributed to chance. Historical internal data can be used to model the difference between a churned customer and one who is still engaged. There would be significant metrics in the data that identify the likelihood of churning. Similar to the acquisition model described above, a parametric equation could be constructed that elicits the association and relationship between the target variable and the predictors. This model would serve as an early warning system for the casino. It would also be a strategic tool useful to predict whether a customer was deemed worth retaining or not. The model should be run on a regular basis across the entire customer database to understand which customers have reached or are reaching a critical value in their churn score. The theory is, these customers would then be targeted with an offer to return to the casino or sports book, in the process avoiding the likelihood of them churning. Alternatively, if the customer is deemed to be of little or no value, there would be no offer forthcoming to entice them to return. Optimizing Offers As Sutton explains, “In addition to predicting the future worth of patrons, it is important to know which marketing campaigns are the most effective for driving response, revenue, and profit. In general, certain offers are better than 32 others, and specifically certain offers will be better for certain patrons.” “While knowing the probable future worth of a patron is critical for determining the reinvestment level for which a patron is eligible, patrons’ behaviors and interests can be used to identify the offer(s) that will be most appealing to each patron as well as the ones generating the most profitable 32 response,” Sutton explains. By analyzing the likelihood that a patron will 164 THE PREDICTIVE CASINO respond to a certain offer or offers, casino and sports book analysts can optimize the offer that each patron is given in order to maximize the amount of 32 revenue and profit driven by the marketing campaigns as a whole. As previously mentioned, A/B testing is one of the best ways to identify which offers work best. A/B testing involves “testing two different offers against one another in order to identify the offer that drives the highest response and the 32 most revenue/profit,” explains Sutton. “More advanced statistical methods can be used to generate likelihood of response scores and classification scores. Some of the more common statistical approaches are logistic regression, 32 decision trees, and discriminant analysis,” Sutton states. “Essentially, these statistical methods use historical data to find the factors that are related as to why a patron responds. Those factors can then be used to assess the likelihood of response based on the similarity of a patron profile to 32 that of responders,” adds Sutton. “These methods have historically been used in direct marketing analysis to 32 identify the best types of offers and the most likely responders,” says Sutton. “In order to build accurate and predictive response models, historical data about response is required. The likelihood of response might be a broad measure of response that refers to the likelihood a patron will respond to any offer, or it might be specific to the likelihood of response to a specific type of 32 offer.” In addition, Sutton adds, “it’s a good idea to select test segments of customers for the purpose of continually testing new offers. Doing so will help to ensure that there is a large amount of response data that can be used to 32 build models and continually improve the efficacy of marketing.” “Effective response models will help identify which patrons are most likely to respond to 32 an offer, and in turn to which offer patrons are most likely to respond.” Chronological View of a Casino Analytics Implementation 1. 2. 3. Data reduction via cluster analysis and segmentation is a logical starting point and initial work should be around identifying patron preference(s). Reducing the customer database into more manageable and meaningful segments has many advantages; the preferences that can be derived are dependent on the availability of meaningful distinguishing factors. Segmentation models use customer metrics that help reduce and profile the customer data base should be constructed as this information can be the underpinning for further analyses, such as acquisition patron worth models. First Trip Scoring Model would require a view of the customer across the entire business and a rich history of engaged customers. The casino would then need to build a modeling data set that is adequate to investigate the relationship between a metric for “valuable” and the 165 ANDREW PEARSON inputs that are extracted, derived, and constructed from the first trip of each customer. 4. A Propensity to Respond Model is heavily dependent on the marketing data and the veracity and richness of it. The casino would need to develop the whole view of the customer first, but, once developed, this is one of the most powerful marketing models available. 5. Customer Conversion Model could be viewed as an extension of a number of the above models, with the idea to derive a data driven metric that scores a customer’s likelihood of returning after his or her first trip. 6. Patron Likelihood to Return Model requires a complete view of the customer along with considerable marketing data. This would help with offers sent, who was sent offers, who responded to the offers, etc., etc. The derived metric on its own would have value, but it could also be a significant input into a two stage model to predict next trip value and worth. 7. A Patron Worth Model would identify the casino’s most valuable patrons. The assumption is that a casino would be looking to predict different metrics, such as worth on the next trip, worth over the next 12 months, lifetime value, etc., etc. 8. A Customer Acquisition Model would then be built by using the results of the segmentation modeling models (or a different metric for desirable customers). A deeper investigation of a casino’s source systems is needed and this could be part of the analysis to help understand what is available, and what might be able to be used from external parties. Different jurisdictions would have different models. 9. Customer Churn Models would require preliminary analysis to extract only engaged customers. The casino would need to derive a statsitcially dirven metric that indicated whether a customer had churned or not. The casino could then build models to detect upcoming patron attrition. 10. An Identifying Patrons at Risk of Abuse Model would likely take into account the factors that predict whether a guest will play on a future trip, but it also makes sense to build a separate model to identify patrons who are likely to use a future offer and not play at all. Conclusion In this chapter I wanted to lay out the many ways in which the Predictive Casino can track and understand its customer base on both a micro and a macro level. Many of the analytical models I mention in this chapter have been around for decades and every IR should be aware that creativity with these models is what will separate them from their competitors. With IT budgets in the millions of dollar per year, every casino can afford software that segments 166 THE PREDICTIVE CASINO its customers, creates marketing campaigns, and predicts churn, but it’s what it does with this information that matters. Customers want to be wowed and that’s not easy. In the ensuing chapters, I will explain how IRs can get a deeper psychological understanding of their patrons, potential patrons, as well as their employees and potential employees. For an industry that sometimes extends credit to its players, it is imperative that they understand these players as much they can. For a company like Genting Singapore, which had bad debt write-offs in Q1 155 2015 of S$76.3m , this can of client information can be very valuable. In its quarterly report, Genting “warned that its VIP markers were becoming increasingly difficult to collect, which led the company to tighten its lending policies.” Getting a fuller picture of the people one is lending money to is imperative in this day and age, and with social media making the world a smaller place, there are some way in which a casino can get a better understanding of whom it is standing on the other side of that felt baccarat table and I will detail these in the next few chapters. 126 Stodder, D. (2012). Customer Analytics in the Age of Social Media. Retrieved from Business Times: http://www.businesstimes.com.sg/archive/monday/sites/businesstimes.com.sg/files/Cu stomer%20Analytics%20in%20the%20Age%20of%20Social%20Media.pdf 127 http://www.webopedia.com/TERM/C/customer_analytics.html 128 IBM. (2013). Achieving Customer Loyalty with Customer Analytics. Retrieved from adma.com: http://www.adma.com.au/assets/Uploads/Downloads/IBM-AchievingCustomer-Loyalty-with-Analytics.pdf 129 Davenport, T. (January 2006). Competing on Analytics. Harvard Business Review. 130 Breski, A. (2013). Customer Retention: The New Acquisition. Madddness Marketing. 131 Cognizant. (2014, January). Retail Analytics: Game Changer for Customer Loyalty. Retrieved from congnizant.com: http://www.cognizant.com/InsightsWhitepapers/Retail-Analytics-Game-Changer-forCustomer-Loyalty.pdf 132 Hicks, Z. (2013, August 28). Toyota Goes All-in With Social Media Monitoring. Retrieved from CIO.com: http://www.cio.com/article/2383143/social-media/toyotagoes-all-in-with-social-media-monitoring.html 133 Mullich, J. (2012, December 10). Opposition Research: Sentiment Analysis as a Competitive Marketing Tool. Retrieved from Wellesley Information Services: http://data-informed.com/opposition-research-sentiment-analysis-as-a-competitivemarketing-tool/ 134 Greenberg, Paul. March 31, 2014. Is Adobe a Marketing Player Now? http://www.zdnet.com/article/is-adobe-a-marketing-player-now/ 135 Costa, T. (2014, March 12). How Location Analytics Will Transform Retail. Retrieved from Harvard Business Review: http://blogs.hbr.org/2014/03/how-location-analyticswill-transform-retail/ 167 ANDREW PEARSON 136 Nyce, Charles. 2007. Predictive Analytics White Paper. https://www.scribd.com/document/200505883/Predictive-Analytics-White-Paper 137 https://en.wikipedia.org/wiki/Prescriptive_analytics 138 How Predictive Analytics is Changing the Retail Industry. International Conference on Management and Information Systems. September 23-24, 2016. http://www.icmis.net/icmis16/ICMIS16CD/pdf/S154.pdf 139 https://en.wikipedia.org/wiki/Decision_tree 140 Deng, L., Gao, J., Vuppalapatie, C. Building a Big Data Analytics Service Framework for Mobile Advertising and Marketing. March 2015. https://www.researchgate.net/profile/Jerry_Gao/publication/273635443_Building_a_Bi g_Data_Analytics_Service_Framework_for_Mobile_Advertising_and_Marketing/links/5 508de220cf26ff55f840c31.pdf 141 Telgarsky, Matus, and Andrea Vattani. "Hartigan's Method: k-means Clustering without Voronoi." AISTATS. 2010. 142 Hartigan, J.A., Wong, M.A. A K-Means Clustering Algorithm. Journal of the Royal Statistical Society. Series C (Applied Statistics), Vol. 28, No. 1 (1979) http://www.cs.otago.ac.nz/cosc430/hartigan_1979_kmeans.pdf 143 Pierson, Lillian. Solving Real-World Problems with Nearest Neighbor Algorithms. www.dummies.com. http://www.dummies.com/programming/big-data/datascience/solving-real-world-problems-with-nearest-neighbor-algorithms/ 144 https://en.wikipedia.org/wiki/Logistic_regression 145 Karp, A. H. (2009). Using Logistic Regression To Predict Customer Retention. New York. Sierra Information Service, Inc. http://www.lexjansen.com/nesug/nesug98/solu/p095.pdf 146 Siroker, Dan, Koomen, Pete. A/B Testing: The Most Powerful Way to Turn Clicks Into Customers. Google Books. 147 Imdadullah, Muhammad, December 27, 2013. Time Series Analysis and Forecasting. http://itfeature.com/time-series-analysis-and-forecasting/time-series-analysisforecasting 148 Sang, Hyuck Kim, Dong Jin Kim. March 23, 2016 Determining Revenue-Generating Casino Visitors Using a Vector Autoregressive Model: The Case of the G Casino in Korea. Journal of Quality Assurance in Hospitality & Tourism. 149 Singh Y, Chauhan AS. Neural Networks in Data Mining. Journal of Theoretical and Applied Information Technology. 2009; 5:37-42 http://jatit.org/volumes/researchpapers/Vol5No1/1Vol5No6.pdf 150 Goa, Jim. Machine Learning Applications for Data Center Optimization. https://static.googleusercontent.com/media/www.google.com/en//about/datacenters/ efficiency/internal/assets/machine-learning-applicationsfor-datacenter-optimizationfinalv2.pdf 151 https://en.wikipedia.org/wiki/Discriminant_function_analysis#cite_note-cohen-1 152 https://en.wikipedia.org/wiki/Survival_analysis 153 Peister, Clayton. 2007. Table-games revenue management: Apply- ing survival analysis. Cornell Hotel and Restaurant Administration Quarterly 48 (1): 70-87. 154 Patrick McGarry. Why Edge Computing Is Here to Stay: Five Use Cases. https://www.rtinsights.com/why-edge-computing-is-here-to-stay-five-use-cases/ 168 THE PREDICTIVE CASINO 155 Stradbrooke, Steven. Bad debts eat into Genting Singapore profits; no new Singapore casino licenses. ClavinAyre.com May 16, 2015. http://calvinayre.com/2015/05/16/casino/genting-singapore-profit-tumbles-bad-debtsrise/ 169 170 THE PREDICTIVE CASINO CHAPTER FOUR Social Media “Think like a publishers, not a marketer” ~David Meerman Scott Introduction Although it is one of today’s buzzwords, “Social Media” is a generic term that refers to websites that allow one or more of the following services: social networking, content management, social bookmarking, blogging and microblogging, live video-casting, and access into virtual worlds. Social Media—the technology as we know it today—has its roots in Usenet, a worldwide 23 discussion system that allows users to post public messages to it. Social media refers to online resources that people use to share content. This content can include images, photos, videos, text messages, pins, opinions and 156 ideas, insights, humor, gossip, and news of almost any kind. Drury’s list of 156 social media includes the following : Blogs, vlogs, social networks, message boards, podcasts, public bookmarking and wikis. Popular examples of social media applications include Flickr (online photosharing); Wikipedia (reference); Bebo, Facebook and MySpace (networking); del.icio.us (bookmarking) and World of Warcraft (online gaming). Unlike traditional marketing models that are nothing more than one-way delivery systems from a company to its consumers, social media is about building a relationship with an audience and starting a two-way dialogue 156 between a company and its consumers. In this new environment, marketing becomes a multi-dimensional discipline that is about receiving and exchanging 156 perceptions and ideas. The consumer is seen as a participant rather than as a 157 “target audience.” The old Source-Message-Channel-Receiver model is evolving into “a collaborative and dynamic communication model in which marketers don’t design ‘messages’ for priority audiences but create worlds in which consumers communicate both with the company and with each 158 other.” Drury argues that confusion exists when pundits talk about social media because the emphasis is often placed on the “media” aspect of social media 156 rather than the “social” aspect, where he feels it correctly belongs. By giving 171 ANDREW PEARSON people a platform to share and interact with each other, social media allows 156 “content” to become more democratized than ever before. In their influential article Users of the World, Unite! The Challenges and 23 Opportunities of Social Media, Kaplan and Haenlein explain that a formal definition of social media first requires an understanding of two related concepts that are often referred to when describing it: Web 2.0 and User 23 23 Generated Content. As Kaplan and Haenlein see it : “Web 2.0 is a term that was first used in 2004 to describe a new way in which software developers and end-users started to utilize the World Wide Web; that is, as a platform whereby content and applications are no longer created and published by individuals, but instead are continuously modified by all users in a participatory and collaborative fashion. While applications such as personal web pages, Encyclopedia Britannica Online, and the idea of content publishing belong to the era of Web 1.0, they are replaced by blogs, wikis, and collaborative projects in Web 2.0. Although Web 2.0 does not refer to any specific technical update of the World Wide Web, there is a set of basic functionalities that are necessary for its functioning.” The “basic functionalities” that Kaplan and Haenlein refer to are; Adobe Flash, the popular animation tool, interactivity, and web streaming audio/video program, Really Simple Syndication (RSS), a family of web feed formats used to publish frequently updated works—such as blog entries or news headlines, as well as audio and video—in a standardized format; and Asynchronous Java Scrip (AJAX), a group of web development methods that can retrieve data from web servers asynchronously, allowing the update of one source of web content 23 without interfering with the display and behavior of an entire page. This is important because it means that a web page for a casino and/or a sports book could, while it is loading on a customer’s or patron’s computer or mobile phone, be accessing and returning specific personalized customer content, including appropriate coupons that have been chosen because they are highly likely to be used and, potentially, could cost the casino or sports book the least to redeem. For Kaplan and Haenlein, Web 2.0 represents the ideological and technological foundation, while “User Generated Content (UGC) can be seen as the sum of all the ways in which people make use of social media. The term, which achieved broad popularity in 2005, is usually applied to describe the various forms of 23 media content that are publicly available and created by end-users.” Also known as Consumer-Generated Media (CGM), User-Generated Content (UGC) refers to a wide range of applications, including blogs, news, digital 172 THE PREDICTIVE CASINO video, podcasting, mobile phone photography, video, online encyclopedias and user reviews. According to Juniper Research, User Generated Content can be 159 broken down into the following three categories : 1. 2. 3. Mobile dating and chat room services—destinations for people to meet. Personal content distribution—audio and video files uploaded onto third party sites for other mobile users to consume. Social networking—social structures made of nodes “that are tied by one or more specific types of interdependency, such as values, visions, ideas, financial exchange, friendship, kinship, dislike, conflict or trade.” UGC can be seen as the sum of all the ways in which people make use of social media and, according to the Organisation for Economic Cooperation and Development, UGC needs to fulfill the following three basic requirements in 160 order to be considered as such : 1. 2. 3. It must be published either on a publicly accessible website or on a social networking site accessible to a selected group of individuals. It must show a certain amount of creative effort. It must have been created outside of professional routines and practices. For Kaplan and Haenlein, the first condition can't include content exchanged in e-mails or instant messages; the second precludes mere replications of already existing content (e.g., posting a copy of an existing newspaper article on a personal blog without any modifications or commenting); and the third condition implies that all created content must exclude a commercial market 23 context. Kaplan and Haenlein believe that Social Media isn’t just “a group of Internetbased applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of User Generated 23 Content.” For Kaplan and Haenlein, this general definition should be broken down further because such disparate sites as Facebook, LinkedIn, Wikipedia, Weibo, and yy.com have little in common with each other when their offered 23 services are looked at individually. As new sites are also popping up on a daily basis, a classification system created for social media should be able to include any future applications that 23 are developed as well. To create such a classification system, Kaplan and Haenlein rely on a “set of theories in the field of media research (social presence, media richness) and social process (self-presentation, self23 disclosure), the two key elements of Social Media.” “Regarding the media161 related component of Social Media, social presence theory states that media differ in the degree of ‘social presence’—defined as the acoustic, visual, and physical contact that can be achieved—they allow to emerge between two 173 ANDREW PEARSON communication partners.” 23 Kaplan and Haenlein argue that: “Social presence is influenced by the intimacy (interpersonal vs. mediated) and immediacy (asynchronous vs. synchronous) of the medium, and can be expected to be lower for mediated (e.g., telephone conversations) than interpersonal (e.g., face-to-face discussions) and for asynchronous (e.g., e-mail) than synchronous (e.g., live chat) 23 communications.” The higher the social presence, the more social influence the communication partners will have on each other. 162 Media Richness Theory is a framework to describe a communications medium by its ability to reproduce the information sent over it and it implies that “the goal of any communication is the resolution of ambiguity and the 23 162 reduction of uncertainty.” For Daft & Lengel , Media Richness is a function of: • • • • The medium's capacity for immediate feedback The number of cues and channels available Language variety: and The degree to which intent is focused on the recipient. Regarding the social dimension of social media, the concept of selfpresentation states that when an individual comes in contact with other people, that individual will attempt to guide or control the impression that others form of them and all participants in social interactions are attempting to 163 avoid being embarrassed or embarrassing others. “Usually such a presence is done through self-disclosure; that is, the conscious or unconscious revelation of personal information (e.g., thoughts, feelings, likes, dislikes) that is consistent 23 with the image one would like to give.” This is important for casino operators because someone’s gambling isn’t always something that patrons want to share with his or her friends. A trip to Macau might actually be done in secret (for various reasons), while a trip to Las Vegas might include more shots of an integrated resort’s offerings beyond gambling; the pool, the nightclub, the steak dinner, maybe a winning sports betting ticket? Applied to the context of social media, a classification is made based on the 23 richness of the medium and the degree of social presence it allows. By 23 combining both dimensions (see Table 8) , create the Classification of Social Media by social presence/media richness and self-presentation/self-disclosure table revealing: • Low: Collaborative projects such as Wikipedia and blogs score the lowest, mostly because they are text based and only allow relatively simple exchanges. Blogs usually score higher than collaborative projects because the former aren't focused on specific content domains. 174 THE PREDICTIVE CASINO • • Medium: Content communities and social networking sites, which allow users to share text-based communication, as well as other forms of media. Social network sites score higher than content communities because they allow more self-disclosure. High: Virtual games and social worlds, which attempt to replicate faceto-face interactions in a virtual environment. Virtual social worlds score higher on the self-presentation scale as the latter are ruled by strict guidelines that users have to either follow or, if they don’t, they risk losing their account access altogether. Social Presence/ Media Richness Low Medium High High Blogs Social networking sites (e.g., Facebook) Virtual social worlds (e.g., Second Life) Low Collaborative projects (e.g., Wikipedia) Content communities (e.g., YouTube) Virtual game worlds (e.g., World of Warcraft) Selfpresentation/ Self-disclosure Table 8: Classification of Social Media by social presence/media richness and self-presentation/self-disclosure Source: Users of the world unite! The challenges and opportunities of Social 23 Media. User Generated Content can be very useful in website Search Engine Optimization (SEO). Search engines are constantly looking for updated information on websites and adding such things as blogs and customer forums can be a cheap and effective way to get customers and/or clients to generate new content for you, which should increase search engine rankings. Giving casino visitors the power to spread images or stories of their casino visits can help market a property virally. Casino properties should actively promote the sharing of social media content. Remembering the Reed Network, i.e., one million people can be marketed to when only 20 people are reached via social media, casinos and IRs should empower their guests to share as much of their on-property experiences as possible. Casino operators should look beyond the most well known social platforms as well and try to be creative. For example, in 2014, Snapchat teamed up with Betfair to offer “self destructing” odds to gamblers during two football games 164 in February 2014. “The offer was extended to anyone following the company’s official Shapchat account, betfairofficial, during the Chelsea versus Everton and Crystal Palace versus Manchester United Premier League 175 ANDREW PEARSON 164 fixtures” and enhanced odds were given to bettors. This may seem like a gimmick but, in this day and age, this is the kind of thing that gets noticed. Mobile and Social Media in China I want to bring China into this discussion because it is not only home to more than 1.3 billion people, who are right next door to Macau—many of whom are incredibly tech-savvy—but also because, I believe, China is on the cutting edge of mobile and social media technology. Macau also generates more gambling revenue than any other country, by far. Living in Macau, a Special Administration Region (SAR) of China, has given me a unique vantage point to witness the explosive growth of social media in China. Since it is an autonomous region like Hong Kong, Macau has no Internet censorship, but it is right next to Zhuhai, a bustling Chinese border-town of a million-and-a-half people. Censorship is very much alive there and it has been interesting to see the different types of censorship that occurs there, a city, literally—and figuratively—on the edge of China. 165 According to the Global Web Index , six out of 10 of the most widely used social systems are Chinese, including Qzone (19%), Sina Weibo (18%), Tencent Weibo (16%), RenRen (11%), Kaixin (8%) and 51.com (6%). For every one of 23 Kaplan and Haelein’s American social media types, there is a corresponding Chinese social media type that includes sites that either mimic or supersede the American original; for every Facebook in the US, there is a RenRen in China (and Qzone, and Douban, etc., etc.); for every American microblog site such as Twitter there is a corresponding site like Tencent Weibo (see Table 9). Type Social Media Site Comparable Chinese site Baidu Bookmarks Delicoius QQ bookmarks Sina viv bookmarks Hudong Collaborative projects Soso baike Wikipedia Baidu baike MBAlib Blogger Weibo Instablogs Hexum Livejournal Sina blog Tumblr Blogus Blog 176 THE PREDICTIVE CASINO Wordpress Bolaa Sina Weibo Tencent Weibo Micro-blogging Twitter Netease Weibo Souhu Weibo Youku Content Community YouTube Ku6 Qivi RenRen Kaixin Facebook Qzone Douban Pengyou Social Network Jiepang Foursquare Qieke Ushi LinkedIn Wealink Jingwei Virtual Game Worlds League of Legends League of Legends Virtual Social Worlds Stageit yy.com Table 9: Chinese Comparable Social Media Sites Source: Intelligencia.co According to Statista, China’s mobile Internet users should top 650 million by 166 the end of 2017. Like their brethren in other countries, Chinese mobile subscribers do everything from make phone calls to send text and email messages, to tweeting and blogging, to watching videos and listening to music, to reading mobile books, to playing mobile games, to shopping at online stores, to checking in at physical stores through geofencing applications, as well as accessing social networking services. Basically, whatever other worldwide mobile users are doing, the Chinese are doing it too. And, probably, much, much more of it, as I will explain throughout this book. Live streamers even figured out how to monetize eating a banana, although the Chinese government’s censorship police brought a quick end to that sexually suggestive 167 practice. 177 ANDREW PEARSON Chinese mobile users are very much on the cutting edge of mobile and social media technology and their behaviors can be utilized to market everything within an IR but the casino offerings. A walk through the electronics market of Huaqiangbei’s commercial district of Shenzhen—reputed to be the largest electronics market in China, which, probably, makes it the largest electronics market in the world—reveals not just a bevy of counterfeit technology, but also a window into the future of mobile and social media. China has almost half a billion social media users, who are engaging with each other on mobile. Mobile instant messaging is the most popular activity, followed by mobile search, mobile news, mobile music, mobile literature, mobile social networking sites, mobile microblogs, mobile games, mobile posts and reposts, mobile emails, mobile videos, mobile payment, mobile banking, mobile shopping, travel booking, and mobile daily deals. To understand the popularity and the potential that social media holds for China in the future, I think it illustrative to look back at the country’s long and tumultuous history. As Boye Lafayette de Mente explains in his book The Chinese Mind: Understanding Traditional Chinese Beliefs and Their Influence on 168 Contemporary Culture : “For over three thousand years the vast majority of Chinese did not have the political or social freedom to make decisions on their own. They were culturally conditioned to suppress their own personal needs and ambitions and to think and behave in terms of the collective responsibility—first for their family, then for their community, next for their clan, and ultimately for the nation at large.” Early Chinese society was built on a strong foundation of Confucian philosophy, which was “based on the already old Chinese idea that social stability was far 168 more important than allowing people to make decisions on their own.” Over the ensuing centuries, “the concept and practice of collective behavior rather than personal actions became so deeply embedded in Chinese culture that 168 individualism virtually disappeared.” Despite the rise and fall of one imperial dynasty after another, “the Confucian concept of collectivism continued to be 168 the bedrock of Chinese culture until the late 1970s.” When Mao Zedong rose to power in 1949, he systematically attempted to remold the Chinese into paragons of communism, but the effort was a 168 complete failure. In 1966, desperate to destroy all of the vestiges of traditional Chinese thought and behavior, Mao set in motion a violent “Cultural Revolution,” whose intention was to enforce communism in the country by removing capitalist, traditional and cultural elements from Chinese society, and 168 to impose Maoist orthodoxy within the Party. During the ensuing 10-year 178 THE PREDICTIVE CASINO period, agriculture and industry were mismanaged to the point where the 168 economy became dysfunctional. Mao died in 1976 and, thankfully, that revolution died with him. However, “the damage inflicted upon the Confucian-oriented culture was profound and set the stage for a second but peaceful and entirely different kind of revolution inaugurated in the early 1970s by his successor, his former but disillusioned 168 Communist ally Deng Xiaoping.” Deng ushered in a new era that was 168 euphemistically referred to as “socialistic capitalism.” Although it seems impossible to prove Deng made the statement “To get rich is glorious!” it has been attributed to him and, whether he said those words or not, the statement embodies a philosophy he would have wholeheartedly 168 embraced. Over the next decade, Deng made it possible for ordinary Chinese people “to utilize their long suppressed ambitions and skills, to begin thinking and acting as individuals, and to help themselves for the first time in the long 168 history of the country.” Incredibly, less than two decades after they went into effect, Deng’s policies “freed over one billion people from a kind of cultural and political enslavement that had often treated them more like objects rather 168 than human beings throughout their existence.” There was not one area or one aspect of Chinese life in the large eastern urban areas of the country that was not fundamentally changed by the economic and 168 social revolution initiated by Deng. Today, a visit to Chinese cities like Shenzhen, Guangzhou, Dongguan and Zhuhai is an eye-opening experience. The typical London taxi cabs (painted a dull blue rather than the standard London shiny black) aren’t quite as ubiquitous on the streets of Shenzhen as they are in London, but they certainly stand out as much because of their uniqueness and seemingly out-of-place oddity. Frank Gehry-esque skyscrapers pierce the smoggy skies of cities like Shenzhen, Shanghai, Chengdu and Guangzhou, turning them into modern architectural marvels. Although sometimes I think they need better translators; for example, the V hotel in Shenzhen is filled with giant ‘X’s on the outside, which makes for a confusingly mixed marketing message. Many restrictions that had prevented people from changing jobs and moving 168 away from their birthplaces have been lifted. Virtually everyone in the country, from teenagers on up, tried to figure out how they could get a piece of 168 the action. Literally, for the first time in the history of the country, individual Chinese people were free to look out for number one, and millions of them 168 began doing just that with a vengeance. 168 In fact, the remaking of the Chinese mindset required almost no time at all. The intelligence and skills the Chinese needed to start remaking their country 168 had always been there. It had just been locked down at the point of a gun or a bayonet, in some cases, literally. 179 ANDREW PEARSON The appearance of computers and the Internet in China had an equally pervasive influence on the thinking and behavior of the Chinese—further weaning them away from the traditional culture as well as the communist 168 culture of the Mao era. One of the key elements in the cultural changes brought on by computers was the fact that the computer itself is culturally neutral—that is, unlike human 168 beings it does not come with any culture hardwired into it. It is not pre168 programmed to require any obedience to existing cultural norms. Like Americans, Japanese, Koreans, and other computer users before them, large numbers of Chinese were freed for the first time in the history of the country to think like and act like individuals, without any thought of their social status, 168 gender, or relationships with others, including with the government. One might say that computers and the Internet helped lower the Great Wall of 168 the traditional Chinese mindset. The combination of the “get rich quick” policy initiated by Mao’s successor in the late 1970s and the widespread use of computers in China in the following decade resulted in a second revolution— this time one that was proposed and aided by the government, but was 168 primarily the work of the people themselves—by individual. The appearance of digital video games also had a profound effect on the 168 attitudes and behaviors of the young Chinese mind. Millions of children in families whose income had risen above subsistence levels began spending countless hours playing video games that had both an obvious and subtle 168 influence on their way of thinking and acting. Internet cafes throughout China, Hong Kong, and Macau are filled with gamers playing game titles like League of Legends, Counter Strike, Call of Duty and Halo. Tencent has stepped into the fray and purchased Riot Games, which is the publisher of League of Legends. For a casino, these eSports players and fans can now be expected to fill stadiums in Macau and Henquin Island and casino operators in Macau should be thinking about monetizing these highly motivated potential audiences. It is not a big jump from playing videos games to playing electronic gambling games. There are even companies developing gambling games that are centered around first person shooter games, rather than spinning slot wheels of dragons. The role models of the first video games to hit China were not the selfless, self168 sacrificing, well-mannered heroes of Confucian China. They were the individualistic, independent, self-serving, fashion-oriented, sensual-minded characters embellished by the imagination and creativity of Japanese anime 168 and manga comic masters. Unintentionally, the war that Chairman Mao had set in motion in 1966 against the traditional mindset and behavior of the Chinese took one giant leap 168 forward by the creators of these video games. By the end of 2008, the 180 THE PREDICTIVE CASINO number of Internet users in China had surpassed that of all the other countries in the world combined, and Chinese Internet companies started expanding 168 globally, first into Japan and then into other neighboring countries. Today, WeChat is taking China and the rest of the world by storm. As the Internet caught on in China, it gave voice to millions of people who before it had been mute and isolated, with few means beyond putting up 168 posters on walls—a dangerous activity—to display their discontent. Despite the fact that the content of the Internet was—and still is—controlled to a considerable degree by the Chinese government, the impact it had on the 168 ability of ordinary Chinese people to make their voices heard was seminal. In the eyes of the Chinese government, Facebook has probably committed a cardinal sin—helping foment a revolution. “Given Facebook's high-profile role in mobilizing people and facilitating protests, such as those that helped topple the Egyptian government in 2011, it seems unlikely that the Chinese government would be interested in granting Facebook a license to operate 169 locally.” Unfortunately for Facebook, its time to make inroads and grow in China have probably come and gone. The local market is dominated by competitors, who have diversified business models that have helped them achieve deep mobile 169 and social media engagement. Robust startups are sprouting up in cities across China, from Shenzhen to Shanghai to Beijing and Chengdu, and many points in-between. Those aforementioned cities have populations of over 10 million people, and their tech workforce are all keyed in and turned on to mobile and social. Sina’s Weibo is China’s most popular and most influential social media service. It is a “mashup of Twitter and Facebook that makes money selling marketing services to business customers and offering paid memberships to individual 169 VIPs.” Users post 140-character messages, which in Chinese conveys 169 considerably more information than it does in English. Images and video can be included as well as comments on post threads. “With an educated, urban audience of 368 million, Sina Weibo has become the top platform for breaking 169 news and adding editorial commentary.” When Wal-Mart-backed retailer 360buy.com wanted to publicize a 10% drop in prices on big home appliances, its CEO first announced it on Weibo, a sign of Weibo’s growing importance to 169 advertisers. Tencent is Sina's biggest competitor. Its platform–called QQ–today boasts 784 million accounts and it “helps Tencent attract users for its other services, such 169 as Tencent Weibo, which has some 469 million users.” Tencent also owns WeChat and offers Qzone, the country's largest straight-up Facebook clone, which is “especially popular among teens, who post photos and videos, keep 169 journal entries, and play games on it.” All those games and posts add up: in 181 ANDREW PEARSON 2011, Tencent reported $4.5 billion in revenue, 22% more than Facebook's 169 2011 revenues. In China, Internet users spend more than 40 percent of their time online on social media, a figure that is expected to continue its rapid rise over the next 19 few years. “This appetite for all things social has spawned a dizzying array of companies, many with tools more advanced than those in the West: for example, Chinese users were able to embed multimedia content in social media more than 18 months before Twitter users could do so in the United States.” “Social media began in China in 1994 with online forums and communities and migrated to instant messaging in 1999. User review sites such as Dianping emerged around 2003. Blogging took off in 2004, followed a year later by social-networking sites with chatting capabilities such as RenRen. Sina Weibo launched in 2009, offering microblogging with multimedia. Location-based player Jiepang appeared in 2010, offering services similar to 19 Foursquare’s.” This explosive growth should continue into the foreseeable future, “a trend that’s at least partially attributable to the fact that it’s harder 19 for the government to censor social media than other information channels.” The Chinese government puts the onus of censorship on the Internet providers, but it is very hard for them to keep up with technologically savvy users, who are constantly on the lookout for the latest technology and newest and hippest platforms. In China, the competition for consumers is fierce, especially in the social-media 19 space. “Many companies regularly employ ‘artificial writers’ to seed positive content about themselves online and attack competitors with negative news 19 they hope will go viral.” “In several instances, negative publicity about companies—such as allegations of product contamination—has prompted 19 waves of microblog posts from competitors and disguised users.” Businesses trying to manage social-media crises in China should “carefully identify the source of negative posts and base countermeasures on whether 19 they came from competitors or real consumers.” Companies should also be aware of the impact of artificial writers when mining for social-media consumer 19 insights and analytics. They should compare “the performance of their brands against those of their competitors. Otherwise, they risk drawing the wrong 19 conclusions about consumer behavior and brand preferences.” This is something that casino companies should do as their competitors could potentially be creating negative reviews to affect Macau holiday choices for Mainland visitors. Casino and hotel operators should already be using social media to manage their brand, to enhance brand loyalty, as well as engage both their current customers and their potential customers. Most casinos are using social media, but in a limited 182 THE PREDICTIVE CASINO way. Not only the perfect avenue to reach patrons, the social media world is also the perfect place to harvest customer feedback, provide real-time customer service, build fanbases, and drive traffic to a casino’s website. Social media can have a predictive quality as well, i.e., it can be used to discover patterns that reveal upcoming customer problems with products and services. It can reveal brand sentiment as well show what drives that sentiment. Customer churn could be spotted early enough to head it off, as well. Social media is a great place to get competitive intel, as well. Benchmarking a competitor’s social media footprint and sentiment can help a casino operator understand how its products and services measure up to its competitors and this is unfiltered information, coming directly from the voice of the customer so it is trustable. Casino and hotel operators should not be reluctant to dive into social media because of its unfiltered nature. These forums will exist with or without the casino’s involvement, therefore it is better to stay ahead of the game rather than to be painfully stuck behind it. Regarding social media, engagement is the key when it comes to successful ROI and profitable customer relations. To compete in this highly competitive industry, casino companies are recognizing the importance of personalization when it comes to customer interactions. Most casinos today have customer loyalty programs that are a part of a CRM and/or a SCRM initiative to provide their guests with an intimate experience that will make them want to return again and again and again. Mobile and social media channels are some of the best ways to reach these customers. Blogs and micro-blogging sites are also important mobile and social media channels and casinos should monitor Twitter feeds for both their satisfied and dissatisfied customers. This is where brand and anti-brand management comes into play. The invert of that old Paul Newman gambling chestnut that “Money won is twice as sweet as money earned” is probably “Money gambled away is twice as painful as money spent on necessities,” so casinos must be cognizant of the emotional toll that gambling losses can induce upon players. Monitoring what losing gamblers are saying on social media is paramount to any good CRM strategy. Real time technology gives hotels and casinos the ability to see—and know— what is going on in real time around them, and this allows them to instantly counter negative brand perceptions. Social media marketing makes good economic sense as well. Given the explosive growth of social media sites, “these might become more cost-effective than using traditional advertising and 170 marketing methods.” Social media is also universal, for every Facebook in the US, there is a corresponding RenRen or WeChat in Asia, yet there is no reason why a casino 183 ANDREW PEARSON in Macau isn’t on Facebook today; in reality, most of them already are. Facebook helps Macau’s integrated resorts market their properties and services to customers in Hong Kong, Japan, South Korea, Taiwan, the Philippines, and even such faraway places as Singapore and Indonesia. Conversely, there’s no reason why a US land-based casino shouldn’t market itself on WeChat either. The fact that the Chinese market is such a lucrative market would seem to dictate that Las Vegas casinos should use such channels to court these highly-motivated Asian gamblers, many of whom can easily afford a trip to Las Vegas. To maintain credibility with customers, casino companies shouldn’t remove negative comments or constructive criticism from these social media sites unless the person posting the comment uses foul language or says something 170 offensive to others. Fostering this kind of goodwill can reap many rewards. Another example of great near real-time customer service is from the MGM Mirage, which won plaudits from fans—and perhaps more business—for how it responded to a 170 disgruntled dinner couple. “After a customer posted on Facebook that he was unhappy with his meal at one of the company’s Strip resorts, the property’s concierge contacted the customer, who was still at the hotel, and 170 offered to fix the problem.” “In another instance, a customer who had won show tickets complained online that he couldn’t use the tickets because he had a conflict. MGM Mirage gave 170 the man free tickets for another date.” This kind of social media proactivity will, undoubtedly, go far in customer relations. I highly doubt, on that customer’s next visit to Vegas, he or she won’t think, first and foremost, of MGM when deciding where to stay and gamble. The fact is, this is the kind of service that people love to tell their friends about so, for the small price of a steak dinner and some concert tickets, MGM probably got some invaluable word-of-mouth advertising. Facebook should be a part of every casino’s social and mobile media marketing plan, but simply putting up a Facebook page won’t cut it these days; creativity and uniqueness are needed to get noticed in today’s highly competitive social media landscape. The Four Steps of Social Media In their book Online Marketing Inside Out: Reach New Buyers Using Modern 171 Marketing Techniques , Eley & Tiley state that, when a company is first delving into social media, there are four steps of social media that should be followed—listen, join, participate and create—and these steps must be strictly followed in that order. Listening is the most important step. People online are frequently mentioning 184 THE PREDICTIVE CASINO and commenting on a company and its products, so all one has to do is listen. Even if a casino operator does not choose to participate in the discussion itself, 171 it will discover valuable information about the company by just listening. Instead of doing expensive surveys, focus groups or other experiments, the best information is often found right there in front of you at minimal or maybe 171 even at no cost. A casino can find out what its customers think of its property, rooms, gaming floors, and service, as well as what they might want improved. Problems and frustrations that might not make it onto corporate surveys might be detailed enough on blogs to affect real change. Most importantly, a casino operator might get the inside scoop of what is actually 171 important to its target audience , whether that audience is a gaming, hotel, spa, retail, or MICE customer. 172 In her article 50 ways to drive traffic to your website with social media Amanda Nelson recommends, listening can be used in the following ways: 1. 2. 3. 4. 5. , Monitor for buying indication terms and reply with helpful links Listen for recommendation requests and share helpful links Listen for discussions of your product or category and provide web links Share relevant web content with prospects Discover relevant blogs and ask for backlinks. Once the casino operator understands the community and what it is all about, it is time to join a social network. Many networks require that you have an account on their site to participate in the discussions and the casino operator should sign up for the account as it is always better to have an account even if it is not required to have one because one always want to claim its brand and/or company name to gain credibility. A casino operator should also join communities where it is most likely to find its 171 customers. If you start out by listening, you will know where your customers tend to congregate online. Facebook, LinkedIn, YouTube, Flickr, Delicious, Digg 171 and Twitter are big networks which should be on your radar. Many of these sites can be used to listen to your audience or to start a discussion. Chinese listening social media sites include Sina Weibo, Tencent Weibo, and Netease Weibo, amongst others. Casino operators should set up accounts at all the major social networking sites and link back to their website(s), as well as link content and similar keywords throughout their social channels. Once the discussion has been joined, then it is time to participate in the community. Participating includes replying and posting to online forums and blogs, reviewing products and services and bookmarking sites that are like171 minded. 185 ANDREW PEARSON By participating, casino operators will build their online brand and people will 171 start to respect them as a valuable contributor to the community. When respected, others will help to promote the property and, possibly, the integrated resorts services without even being asked to do so, which, as most marketers will tell you, is some of the best marketing around. Not only is wordof-mouth marketing one of the most trusted forms of marketing, but it can also spread virally. Two words of warning, however; your role models should always be very experienced and remain very active users in the community; and, most 171 importantly of all, remember that it is never okay to spam. In her article 50 Ways to Drive Traffic to Your Website using the following methods to increase participation: 1. 2. 3. 4. 5. 6. 7. 8. 9. 172 , Nelson recommends Ask readers to sign up for an RSS feed Answer all questions and share peer referrals Feature community members on your site Share customer stories Ask influencers to share your web links Interview an influencer for web content Have an influencer guest blog Help an influencer write content about the brand Share products with influencers for feedback and web content. Finally, it is time to create. Once a casino operator has built itself an online brand by listening, joining and participating, it is time to create its own 171 content. It will now have an audience to share its content with and they will help the casino operator spread its content far and wide. It should be noted here that the casino operator has to create value; ads are not generally seen as 171 valuable. Posting “buy my stuff” on twitter will fail to achieve the results you 171 want, and this practice may even get you banned. By making beneficial contributions to the community, people will notice you and want to know more 171 about the company. If you have listened properly, you should have a solid 171 idea of the type of content people would like to see. Then, simply, give it to 171 them. Nelson recommends companies be creative in the following ways : 1. 2. 3. 4. 5. Divide a piece of content into multiple Slideshare presentations that link to your site Start a LinkedIn group Tie content together so an ebook links to a relevant blog post, which, in turn, links to a topical webinar Build a forum or community section on the company website Create referral programs. The four steps of social media fit well within the six types of social media, which I will detail next. Throughout the rest of this chapter, I will explain the differing types of social media and, in chapter five, I will explain how each of 186 THE PREDICTIVE CASINO these social media platforms can be used individually as well as, sometimes, together. Six Types of Social Media According to Kaplan and Haenlein’s article Users of the World, Unite! The 23 Challenges and Opportunities of Social Media , the writers break Social Media down into the following six different types: 1. 2. 3. 4. 5. 6. Collaborative projects Blogs and micro-blogs Content communities Social networking sites Virtual game worlds Virtual social worlds Throughout the rest of this chapter, I will break down each of these types of social media separately, as well as explain how a casino operator can use them on their own or, preferably, combined together. Chapter five goes into much more detail about how these sites should be used in a social media marketing and listening way, but, in this chapter, I will lay out the foundations and provide a list of sites available to the social media marketer and listener. Collaborative Projects Probably the most democratic form of all UGC, collaborative projects enable 23 the joint and simultaneous creation of content by many end-users. Kaplan and Haenlein believe collaborative projects can be split into two different 23 categories : 1. 2. Wikis—these are websites that allow users to add, remove, and change text-based content; and Social bookmarking applications—these enable the group-based collection and rating of Internet links or media content. The main idea behind collaborative projects is that joint efforts can lead to a 23 better outcome than individual action. Examples of collaborative projects include the web-based encyclopedia Wikipedia and social bookmarking sites such as Delicious and Stumbleupon. Social bookmarking is both the method of storing and the managing of Web page bookmarks with individually chosen keywords as well as the sharing of this information with others. At social bookmarking sites, users can tag, save, manage and share Websites with their friends and their connections. Users can add descriptions in the form of metadata and these descriptions can be anything from free text comments, favorable or unfavorable votes, or tags that collectively form a social thread of information. This kind of thread is also known as a folksonomy—“the process by which many users add metadata in 187 ANDREW PEARSON 173 the form of keywords to shared content.” 174 In his article How to Use Social Bookmarking for Business , Lou Dubois explains that, “Social bookmarking, at its most basic form, is a simple way to organize all of the best content from around the web based off your interests, all in one place.” It is a handy way to “sort the relevant from the irrelevant, according to their interests and the value of the information provided. And perhaps most importantly, the bookmarks are transferable between computers 175 and locations.” Founded in 2003, Delicious (then known as del.icio.us) coined the term social 176 bookmarking and pioneered the concept of tagging. The following year, similar sites such as Furl, Simpy, Citeulike and Connotea came online. Stumbleupon also appeared around the same time. Compared to search engines and traditional automated resource location and classification software, social bookmarking systems are advantageous because the tag-based classification is done by a human being, who usually understands the content and context of a resource better than any algorithm-based computer program. Human beings are also adept at finding and bookmarking 176 Web pages that often go unnoticed by web spiders. In addition, a user will probably find a system that ranks a resource based on how many times it has been bookmarked by other users more valuable than a system that simply ranks resources based on the number of external links pointing to it. For the promotion of a casino operator, social bookmarking is important because it helps a website get quality backlinks. When a website is submitted for ranking by a search engine, the search engine considers the quality of the backlinks, i.e., the quality of the sites linking back to it. This means that if you bookmark popular sites, the search engine spiders will automatically follow the links back to your site. SEOMoz’s Linkscape and Majestic SEO’s Link Intelligence are both very good tools to discover current backlinks to a site. Kaplan and Haenlein argue that, “From a corporate perspective, firms must be aware that collaborative projects are trending toward becoming the main source of information for many consumers. As such, although not everything written on Wikipedia may actually be true, it is believed to be true by more and more Internet users.” This can have particularly damaging repercussions during 23 a corporate crisis. I go into specific detail about how social media should be used in crises situations in chapter four, while discussing a few examples such 177 as the Red Cross alcohol Tweet , the United Airlines’ “United Breaks Guitars” 178 179 viral video and the Domino’s YouTube fiasco. Collaborative projects can also be used to increase productivity, for example, the Finnish mobile manufacturer Nokia “uses internal wikis to update employees on project status and to trade ideas, which are used by about 20% 23 of its 68,000 staff members.” Also, the U.S. application software company 188 THE PREDICTIVE CASINO Adobe Systems “maintains a list of bookmarks to company-related websites 23 and conversations on Delicious.” Dubois explains that “From an individual consumption perspective for Internet readers, social bookmarking can make great sense to filter your news and 174 information all into one place.” But it also makes great sense for businesses to utilize these tools as they can increase website traffic and grow brand 174 recognition by curating information and disseminating client testimonials. Throughout the business world, content curators are “considered the gatekeepers to information for businesses and individuals. As a company, curating, or aggregating the best content from around the web, can make you 174 an industry leader.” For companies you already work with, showing that you 174 are on top of industry news gives you a vaunted level of credibility. “Similarly, if you think of it from the perspective of businesses who you don't already do business with, you're going to be seen as a resource for 174 information,” which should give you an immediate leg up on your competition. Another way to utilize these tools is by pulling together all of your company’s best customer testimonials in a social bookmark. Just about every business gets questions about its client list and testimonials from its potential business partners and when asked the question: "What have others said about your work?", wouldn’t it be better to direct potential clients to a site that has all of the company’s testimonials in one place, in a simple format rather than sending 174 them to a Yelp page, argues Dubois. Social bookmarking isn’t as intuitive a process as blogging or social networking on sites like Facebook or Twitter, but it is a very valuable tool in its own right and it should be one part of an IR’s social media marketing plan. Chinese collaborative projects include Baidu bookmarks, QQ Bookmarks, Sina viv, Hudong, Soso baike, Baidu baiki and MBAlib. Blogs In 2005, Merriam-Webster added the word “blog” to its dictionary, calling it, “a web site that contains an online personal journal with reflections, comments, 180 and often hyperlinks provided by the writer.” The Website Webopedia defines a blog as, “a web page that serves as a publicly accessible personal 181 journal for an individual.” The term originated from the word “weblog”, which was coined by Jorn Barger on 17 December 1997 when he used it to describe the list of links on his Robot Wisdom website that “logged” his 182 internet wanderings. In April or May of 1999, Peter Merholz broke the word “weblog” into the two 183 words “we blog” in the sidebar of his blog Peterme.com. The term “blog” 189 ANDREW PEARSON was picked up by Evan Williams at Pyra Labs who used “blog” as a noun and a verb to mean “to edit one's weblog or to post to one's weblog” and created the term “blogger” for Pyra Labs' Blogger product, which led to the term's 184 popularity. Representing the earliest form of Social Media, blogs are the “Equivalent of personal web pages and can come in a multitude of different variations, from personal diaries describing the author’s life to summaries of all relevant 23 information in one specific content area.” In its article It's the Links, Stupid 183 , The Economist claims that a blog is: “A web page to which its owner regularly adds new entries, or ‘posts’, which tend to be (but need not be) short and often contain hyperlinks to other blogs or websites. Besides text and hypertext, posts can also contain pictures (‘photoblogs’) and video (“vlogs”). Each post is stored on its own distinct archive page, the so-called ‘permalink’, where it can always be found.” The Economist explains that blogging is a quintessentially social activity, 183 highlighted by two features : “A ‘blogroll’, along the side of the blog page, which is a list of links to other blogs that the author recommends (not to be confused with the hyperlinks inside the posts). In practice, the blogroll is an attempt by the author to place his blog in a specific genre or group, and a reciprocal effort by a posse of bloggers to raise each other's visibility on the internet (because the number of incoming links pushes a blog higher in searchengine results). The other feature is ‘trackback’, which notifies (‘pings’) a blog about each new incoming link from the outside—a sort of gossip-meter, in short.” According to Dave Winer, the influential software engineer who pioneered several blogging techniques and has, by his own estimate, the longest running 183 blog of all time , weblogs should be: 1. 2. 3. Personalized: Weblogs are designed for individual use (a multi-person weblog is also possible through collaboration, such as the ‘‘team blog’’ offered by www.blogger.com). A Weblog style is personal and informal. Web-based: Weblogs can be updated frequently. They are easy to maintain and accessible via a Web browser. Community-supported: Weblogs can link to other weblogs and Websites, enabling the linkage of ideas, and hence stimulating knowledge generation and sharing between bloggers. 190 THE PREDICTIVE CASINO 4. Automated: Blogging tools help bloggers to present their words without the hassle of writing HTML code or any other programming language; instead, bloggers can just concentrate on the content. 183 Winer argues that blogging should have a raw, unpolished authenticity to it. “Blogging is all about style” and the essence of blogginess is “the unedited 183 voice of a single person,” preferably an amateur. For Winer, editors do not 183 belong in the Blogosphere, even though, today, they very much are in it. Blogs are incredibly popular because they are cheap, easy to set up and they provide maximum exposure with limited effort. As Jeff Jarvis, Director of the Interactive Journalism at City University of New York's Graduate School of Journalism points out, they are the “easiest, cheapest, fastest publishing tool 182 ever invented.” Blogs are everywhere, affecting every sector of society and, because of their ease of use and low barrier to entry, they will continue to be a 182 big part of the national and worldwide conversation. Blogs can take many forms, including a diary, a news service, a collection of links to Internet resources, a series of book reviews, reports of activity on a project, the journal of an expedition, a photographic record of a building project, or any one of a number of other forms. Integrated resorts have so much going on that there should be an endless amount of available things to blog about; concerts, gaming tournaments, eSports competitions, TV shows filming on property, restaurant, bar, and club news, etc., etc., etc. According to Winer, a successful blog should include the following key 183 elements : 1. 5. 6. 7. 8. Great content: as the old adage goes, “Content is king” and that old axiom should be kept very much in mind when it comes to blogging. Competition is fierce so one’s content better be relevant, valuable and captivating. Post frequently: along with having great content, bloggers should constantly post new material. A constant stream of new material will garner more views, which should result in more followers. User friendly navigation: readers prefer navigation that is simple and straightforward so have links that make logical sense. Eye pleasing content: as with any other type of marketing, the prettier something looks, the more likely it is to be viewed, so keep the design element in mind when creating a blog. Connect to other content: linking and back-linking is exceptionally important so feel free to add links to other content that expands upon or references your content. Although, China censors its social media site, Chinese consumers are some of the biggest bloggers and commentators around and Chinese blogging sites like Weibo, Hexum, Sina blog, Blogus and Bolaa are filled with a constant stream of 191 ANDREW PEARSON observations and explanations about any and everything, including the goingson in Macau. IRs should exploit any and every available blog opportunity they have to get their message out, as well as use any and all platforms that help them to connect with their customers and potential customers anywhere they happen to be. Microblogs Although similar to a blogging website, a microblog site differs from a traditional blog in that its content is typically smaller in both actual and aggregate size. “Social networking and microblogging services such as Twitter, Facebook, or Google+ allow people to broadcast short messages, so-called microposts, in continuous streams. These posts usually consist of a text message enriched with contextual metadata, such as the author, date and 185 time, and sometimes also the location of origin.” While individual posts can be no longer than 140 characters, “aggregated posts of multiple users can provide a rich source of time-critical information that can point to events and 185 trends needing attention.” The most used microblog in the English-speaking world is Twitter, which, according to its website “is a real-time short messaging service that works over 186 multiple networks and devices.” A free social networking and micro-blogging service, Twitter allows users to send and receive Tweets—messages that can be up to 140 characters in length. “Connected to each Tweet is a rich details pane 186 that provides additional information, deeper context and embedded media.” Because it is happening in near real-time, “Twitter is a ‘what’s-happening-rightnow’ tool that enables interested parties to follow individual users’ thoughts 187 and commentary on events in their lives.” On its Website, Twitter recommends building a following, increasing a businesses' reputation, and raising a customer's trust by following these best 186 practices : 1. 2. 3. 4. 5. 6. Share: disseminate photos and behind the scenes info about your business. Even better, give a glimpse of developing projects and events. Users come to Twitter to get and share the latest, so give it to them! Listen: regularly monitor the comments about your company, brand, and products. Ask: question your followers to glean valuable insights and show them that you are listening. Respond: reply to compliments and feedback in real time. Reward: Tweet updates about special offers, discounts and timesensitive deals. Demonstrate wider leadership and know-how: Reference articles and links about the bigger picture as it relates to your business. 192 THE PREDICTIVE CASINO 7. 8. Champion your stakeholders: Retweet and publicly reply to great tweets posted by your followers and customers. Establish the right voice: Twitter users tend to prefer a direct, genuine, and, of course, likable tone from your business, but think about your voice as you Tweet. How do you want your business to appear to the Twitter community? Twitter also offers three ways to advertise on its service; promoted tweets; promoted trends; and promoted accounts. Promoted tweets are regular Tweets that are amplified to a broader audience and they are offered on a Cost-per-Engagement (CPE) basis. A business is charged when a user Retweets, 186 replies to, clicks on, or favorites the Promoted Tweet. Retweeted impressions by engaged users are free, and can exponentially amplify the reach 186 and cost-effectiveness of a marketing campaign. Twitter is a very useful tool that connects businesses to customers in real-time. It can help a business quickly share information with people who are interested in their products and/or services, as well as gather real-time market 186 intelligence and customer feedback. Using Twitter, a business can build strong relationships with its customers and partners as well as raise the profile 186 of its brands, direct sales, and engage a primed audience. Twitter can help a business build a following, increase its reputation as well as raise a customer’s trust by sharing, listening, asking questions, responding to replies, rewarding customers with special offers and discounts, demonstrating wider leadership and championing the right stakeholders. “Promoted Trends” give a business the exclusive opportunity to feature a Trend 186 related to its business at the top of the “Twitter Trends” list. When a user clicks on the “Trend”, he or she is taken to the conversation for that trend and a “Promoted Tweets” tag is attached to the tweet at the top of the timeline. Because of its placement, the ad receives substantial exposure, thereby 186 initiating or amplifying a conversation on Twitter and beyond. “Promoted Accounts” can help companies quickly increase their Twitter 186 followers. Part of “Who to follow” (Twitter's account recommendation engine), “Promoted Accounts” will highlight a business account to users who 186 will most likely find it interesting. According to Twitter's Website, “Users find Promoted Accounts a useful part of discovering new businesses, content, and 186 people on Twitter.” Chinese Microblogging sites include Sina Weibo, Tencent Weibo, Netease Weibo and Sohu Weibo are useful to utilize in Macau. Content Communities Content communities exist for a wide range of media types, including text, 23 photos, videos, and PowerPoint presentations. In general, users are not 193 ANDREW PEARSON required to create a personal profile page or, if one is required, only basic 23 information is needed. Kaplan and Haenlein state that, “The main objective of 23 content communities is the sharing of media content between users.” Although businesses run the risk of these platforms being used for the purpose of sharing copyright-protected materials, the advantages of getting one’s content into the social media community seriously outweighs the 23 disadvantages of potential copyright infringement. The popularity of these content communities make them a very attractive contact channel for many businesses. This fact isn’t surprising when one considers that a site such as 23 YouTube has over 2 billion views per day. Social Networks Perhaps the most familiar of all social media sites are the social networks, including platforms such as Facebook, Foursquare, Instagram, LinkedIn, WeChat, and Pinterest, amongst hundreds of others. According to Wikipedia, “a social network is a social structure made up of individuals (or organizations) called ‘nodes’, which are tied (connected) by one or more specific types of interdependency, such as friendship, kinship, common interest, financial exchange, dislike, sexual relationships, or relationships of beliefs, knowledge or 188 prestige.” Boyd and Ellison define social network sites (SNS) as: “web-based services that allow individuals to (1) construct a public or semi-public profile within a bounded system, (2) articulate a list of other users with whom they share a connection, and (3) view and traverse their list of connections and those made by others within the system. The nature and nomenclature of these 189 connections may vary from site to site.” What makes a social network site unique is its “ability to enable users to 189 articulate and make visible their social networks,” which can result in 189 connections between individuals that would otherwise not have been made. This is one of the key aspects of social networking that makes it so primed for marketing, an area I will delve into further in chapter four. Social networks can also be important platforms for businesses, including IRs. 190 In their paper Expanding Opportunities in a Shrinking World , Avimanyu Datta and Len Jessup state that “Social networks promote social entrepreneurship by means of (a) technology and knowledge transfer; (b) locating information; (c) generating entrepreneurial opportunities; (d) building entrepreneurial competency; (e) financing innovation; and (f) building effective networks for commercialization of innovations. WeChat is a Chinese mobile text and voice messaging communication service developed by Tencent. It was first released in January 2011. According to Wikipedia, “WeChat provides multimedia communication with text messaging, 194 THE PREDICTIVE CASINO hold-to-talk voice messaging, broadcast (one-to-many) messaging, photo/video sharing, location sharing, and contact information exchange. WeChat supports social networking via shared streaming content feeds and location-based social plug-ins (‘Shake’, ‘Look Around’, and ‘Drift Bottle’) to chat with and connect 191 with local and international WeChat users.” WeChat is a communication platform made up of what WeChat has dubbed “The 4 Pillars”—Instant Messaging, Location Based Services, Moments and 192 Official Accounts. The “Instant Messaging” section is pretty straightforward, it allows users to message other users via text, which has become many people’s communication 192 channel of choice these days. “Location Based Services” is the section that allows the user to find information that is relevant relating to the area they are 192 in, a channel good for IRs to disseminate information through. Beyond just finding the nearest ATM, the Radar feature launched in March 2014 allows 192 users to find friends around them without revealing their cell phone number. In the “Moments” section, users can post pics, comments and “Like” or share their pictures or videos with the general public or simply share them with a 192 select few. I’ve seen small business owners also use this feature to showcase their products to any and every one in their address book, so it can have alternate uses. The “Official Accounts” section is where brands come in. “WeChat has the ability to integrate into the company’s system so that content from the 192 company can be posted across all its channels.” Official accounts allow companies to send out blanket messages to multiple users, but then it also 192 enables individual and private conversations, too. “This means that WeChat can be used to resolve issues in a private forum unlike other platforms such as 192 Twitter.” In terms of both functionality and user activity, WeChat has gone through the following three distinct phases of growth: 1. 2. 3. Replacement of SMS with a basic—and free—messaging function that 193 was similar to WhatsApp. The addition of a social networking function through its “moments” section, where users can send status updates with pictures and short 193 snippets of text. Similar to Facebook’s News Feed, this allows users to create a diary of personal memes. “It was an important addition because it also introduced to WeChat a more public, visible social media channel where updates can be openly shared and viewed by a 193 wide group of connections.” E-commerce—users can link their consumer bank cards, credit cards, and Tenpay and WePay accounts to their WeChat accounts. “The linking of these payment options allow WeChat to be a totally 195 ANDREW PEARSON enclosed ecosystem where social can be linked seamlessly with sales. A user inside of WeChat never has to leave the app on their mobile 193 phone” to make a purchase. Brands have taken advantage of WeChat’s “public accounts” to create awareness and spread messages virally by teaming up with influencers, 193 celebrities, and key opinion leaders.” IRs could use these avenues to market events, even selling tickets through these channels. “While WeChat's social functions are not as open as Weibo's and the structure makes it harder to create massive followings, features like “look around” (people can add each other based on proximity), “Shake” and “Message Bottle” (for random connections), and “QR Codes” (a path to the user's account from wherever they share the code) have all helped to create more growth in the number of 193 connections.” There are four e-commerce avenues inside of WeChat that brands can also exploit; subscription accounts; online-to-offline sales channels; WeChat shops; and affiliate sales. With Subscription Accounts, brands can create content and present new products and offers to followers. This content can be linked to an e-commerce 193 store built inside or outside the WeChat application. A Subscription Account is simple to build by using the tools provided by WeChat's Fengling.me 193 service. The only catch is that you have to have a company registered in 193 China, which is, admittedly, not an easy hurdle to overcome. For companies without a China registered office, there are other ways to access the market, 193 but they don't allow for the same control over presentation and process. Most casinos in Macau already have WeChat apps so what they could sell would include just about anything they can sell in Macau, including many touristy items. If customers are offered delivery to their home of the normal things they would buy in Macau, they might just take the casino up on the offer as it would alleviate carrying lots of items home. As a mobile application with the GPS features of a smartphone, WeChat allows 193 for some promising location-based opportunities for casinos. The app allows for location-based messaging from retailers or casinos or any number of other industries who have a one-to-one connection with its customers. Brands can take advantage of the location-based capabilities of the app by creating a loyalty card and/or by encouraging users at a specific location (a retail shop) to add (follow) the brand account. The “Loyalty Card” account inside of WeChat is basically a CRM tool which audiences can opt-in for and find locations nearby 193 (of retail shops), receive discounts, promotions, points, and rewards.” Retailers, restaurants, casinos, consumer staples, and a whole host of other industries are using QR codes and other invitations to encourage customers to 193 sign up for WeChat accounts. “They typically do this on-location, taking 196 THE PREDICTIVE CASINO customers from offline to online, and thereby collecting contacts within the CRM accounts of the brand. Moving from online to offline, brands are starting to experiment with creating promotions online that drive users to a retail 193 location.” For a casino, a restaurant, or a café this could mean sending out a “flash” alert to followers about a promotion taking place “in the next hour” for a free trial or a discount. For a fashion company it could allow them to activate a pop-up 193 shop within a short timeframe for a launch or product demo/trial. Once a user is in a restaurant, café, pop-up shop, or on a casino floor, he or she can pay on location, provided the payments function has been set up on their 193 WeChat account. It makes the retail process much more flexible as brands can take pre-payments or set up small sales without cash registers at WeChat 193 shops. Currently, there are “a growing number of shops, malls, group-buy (TuanGou), 193 and flash-sales (MianGou) channels being built into WeChat.” Companies such as Xiaomi, ONLY, and Sephora have created branded stores (as “Service 193 Accounts”) where they sell products directly” to their customers. WeChat only allows access to this channel to brands that have “a plan for building awareness (traffic to their store) and to have a logistics/fulfillment 193 capability,” so this might be an avenue for IRs, but you never know. Most products sold on WeChat are moving through “malls” of one type of 193 another—many of which are controlled by Tencent. “Tencent has done a good job of implementing its most important companies, applications, and investments into WeChat. Grouped together inside the payments section, key WeChat/Tencent owned/invested channels are highlighted, including; “Specials” (linked to its e-commerce mall yixun.com), Weituangou (linked to its 193 group buy site gaopeng.cn) and Dianping for restaurants.” “Tencent also has accounts for other invested companies, including eLong, 193 JD.com, OKBuy, Tongcheng, and Sougou.” There are also “malls” for Dangdang, Amazon, VIP.com, Lefeng, Mougujie, Meilishou, Suning, Guomei, 193 No1Shop, and Qunar, just to name a few. The biggest challenges in selling through these channels is, firstly, gaining enough visibility in a very crowded 193 channel and, secondly, managing the presentation of the brand. These are significant challenges, especially the former. To get significant visibility in these channels, “brands often have to pay hefty fees to the ‘malls’ to get priority listings. Ultimately, the ‘malls’ control which products get sold and so there is a 193 real loss of control for brand owners.” The products OKWei currently offers are not great and the process has the potential to become very complicated, very quickly, but affiliate sales network are resilient; the invisible hand of profit is just too seductive a motivator to bet against, I believe. “Despite these hurdles, ‘affiliate networks’ built 197 ANDREW PEARSON inside/around WeChat hold a lot of potential. It is this type of link between social and e-commerce which makes WeChat very powerful for brands, even, 193 potentially, casino operators. Other Chinese social networks include RenRen, Kaixin, Qzone, Douban, Pengyou and the Foursquares of China, Jiepang and Qieke. Another recent addition to the social networking landscape is Snapchat. It was one of the 194 runaway success stories of 2012. Users send about 50 million pics (called "Snaps") a day on the platform. Bearing truth to that statement that copying is the highest form of flattery, the success of Snapchat prompted Facebook to release a competitor, Poke, although it met with limited success and has since 194 been shut down. In what might be a first for marketing through Snapchat, the New York frozen yogurt chain 16 Handles “is leveraging Snapchat for a promotion that presents 194 users with a coupon that self-destructs within 10 seconds.” Noticing that a lot of its young users were using Snapchat handles to interact on social media, the yogurt chain started an advertising campaign that asked users to send them a picture of themselves and their friends at a 16 Handles location tasting one of 194 their frozen yogurts. In return, users received a coupon for anywhere from 16% to 100% off their purchase, but they only had 10 seconds to let the cashier 194 scan the coupon. In the gambling space, Snapchat teamed up with Betfair to offer “self 164 destructing” odds to gamblers during two football games in February 2014. “The offer was extended to anyone following the company’s official Snapchat account, betfairofficial, during the Chelsea versus Everton and Crystal Palace 164 versus Manchester United Premier League fixtures” and enhanced odds were given to bettors. Customer Understanding In this chapter, I have tried to lay the foundations for an understanding of the social media milieu. Few casino operators will succeed in this new millennium without embracing social media. When first delving into social media, casino companies should follow the four steps of social media—listen, join, participate 171 and create—and these steps must be strictly followed in that order. Listening can be done on blogs, content communities, and social networks. By keeping an eye on any comments made on these blogs, or on the pages of Facebook, Instagram, Pinterest, Twitter, or a while host of other social network pages, social media marketers can get a sense of what the community feels about its business and its products/services. To understand how important this process can be for a company, I’d like to take the example of one bourbon manufacturer who found itself in the midst of a self-created social media disaster. In February, 2013, because it was faced 198 THE PREDICTIVE CASINO with both a high demand for its product and a low supply of bourbon whiskey, Maker’s Mark announced plans to cut the amount of alcohol in its drink from 195 45 to 42 percent. Needless to say the internet wasn't pleased. As Laura Stampler explains in her article Makers Mark Turned Watered Down Whiskey 195 Debacle Into a Social Media Win , “It's the age of social media, so consumers were tweeting and Facebooking their complaints to anyone who would 195 listen.” 195 There were angry tweets as well as Facebook petitions against the company. A normal Valentine's Day post on the company's Facebook page was flooded 195 with negative comments about the shift. Immediately realizing that it had made a huge mistake, the brand decided to embrace the social media platforms where they have been receiving such negativity and quickly put out the message that it had made a huge mistake, it was sorry and that it was 195 reversing its decision about lowering the alcohol content. The link to the 195 company’s Facebook apology soon became a popular Hashtag. “Customers 195 went from feeling abandoned to listened to and respected in record time.” The apology noted that even though the social media reaction was highly 195 negative, the company wanted the conversation to continue. Maker’s Mark even took this conversation into their print advertising, using the tagline line: “You spoke. We listened. Here’s proof”, with an arrow cleverly pointing to the label, which showed that the alcohol content (or proof) was still 45%. By listening, joining, participating and creating, Maker’s Mark built its online brand and it now has an audience to share its content with, an audience which should help them spread their content far and wide, as well as, more importantly, sell a lot more 45 proof whiskey. Social media is all about adding value to communities of customers and prospects by providing interesting content (blogs, podcasts, webinars, etc.). It allows immediate engagement with groups of customers and potential customers. Today, the traditional model of blasting messages to customers and potential customers is fading as trust in corporate America is at an all-time low. In today’s difficult economic climate, peer referrals are becoming more and more important. Consumers are tuning out regular advertising and tapping into social media for advice. Listening and joining these conversations could prove highly lucrative to casino companies. The Predictive Casino utilizes social media in many facets of its business because of its inexpensive and real-time aspects. Social Media is constrained only by the imagination of a casino company’s marketers and it offers enormous potential both creatively and financially to any company willing to enter the arena. In the next few chapters, I will reveal how social media can be used in a casino’s mobile and social media marketing plans, as well as discuss many of the available tools one can use to track and 199 ANDREW PEARSON monetize these platforms. The power of social media influencers will also be discussed and, as the casino industry is an industry where travel and hospitality take center stage, connecting with the social influencers should be a key part of every IR and casino company’s marketing plan. Social media can play an integral part in customer understanding. Continuing the journey from an anonymous customer to one who is known so well that their actions can be manipulated is the addition of a customer’s IP address, user agent/browser finger print, his or her device ID, his or her digital cookies, and, finally, his or her social ID (see figure 10). But how can a casino capture a gambler’s social IDs? It’s actually pretty simple, offer them the ability to sign in with their social media accounts. Many people enjoy the simplicity of signing into websites with their Facebook, Twitter, Pinterest, Snapchat, or other social media accounts and once the casino marketing department can connect the patron account with a player’s social media account, a whole new world of customer interaction is possible. China is an important player in social media. As Chiu et al. state in their paper Understanding Social Media in China, “The sheer number of the more than 300 million social-media users in China creates unique challenges for effective consumer 196 engagement,” but I believe the potential market is too massive to ignore. But it is a tricky market; in China, “People expect responses to each and every post, for example, so companies must develop new models and processes for effectively engaging individuals in a way that communicates brand identity and values, satisfies consumer concerns, and doesn’t lead to a negative 196 viral spiral.” Other problems Figure 10: Customer funnel include the “difficulty of developing and tracking reliable metrics to gauge a social-media strategy’s performance, given the size of the user base, a lack of analytical tools (such as those offered by Facebook and Google in other markets), and limited transparency into leading platforms,” 196 Chiu et al. warn. 200 THE PREDICTIVE CASINO China is, unquestionably, making great strides, but they still have a long ways to go before their social media analytics technology can rival the US’s. However, as Chiu et al. make clear this is not a barrier that should stand in the way of companies as “The similarity between the ingredients of success in China and in other markets makes it easier—and well worth the trouble—to cope with the 196 country’s many peculiarities.” 201 ANDREW PEARSON • • • • • • • • • • • • • • • • • Tips Claim and secure your brand and/or company name. Websites like knowem.com allow you to check for the use of your brand, product, personal name on over 500 popular and emerging social media websites. Join communities where you are most likely to find your customers. Once you have joined the discussion, it is time to participate in the community; post to online forums and blogs, review products and services and bookmark sites that interest you. For the promotion of a business, social bookmarking sites like delicious.com are important because they help a Website get quality backlinks. Create a blog/blogs—they are incredibly popular because they are cheap, easy to set up and they provide maximum exposure with limited effort. When blogging, use lists. For some reason, lists generate more interest than long-winded diatribes. Although similar to a blogging website, microblog sites such as Twitter allow people to broadcast short messages, so-called microposts that can consist of text messages enriched with contextual metadata. Twitter can also be used to build a following, increase a businesses' reputation, and raise customer trust. Use social media listening tools to do highly affordable market research. Free Twitter listening tools alone can provide priceless market research. Use a site like Listorious to find a list of prominent Tweeters in your business or industry. Create interest on Pinterest with their “Wish lists” feature. Network with other people in the community who share similar interests. Use SEOMoz’s Linkscape and Majestic SEO’s Link Intelligence to check backlinks to a site. Use Viralwoot to build up Pintrest follower inexpensively. Stay active by sharing engaging content and adding Google +1s and comments to the pages and content in your community. Encourage reviews from your customers on Google+, Weibo, Facebook, etc., etc. Download the Google+ mobile app so you can monitor and share content on the go. 202 THE PREDICTIVE CASINO 156 Drury, G. (2008). Opinion piece: Social media: Should marketers engage and how can it be done effectively? Journal of Direct, Data and Digital Marketing Practice, Volume 9, pages 274-277. 157 As identified in Claude E. Shannon and Warren Weaver’s The Mathematical Theory of Communication, the Source-Message-Channel-Receiver model is a basic model of communication; Source is the person who encodes the message and transmits it to the receiver; the Message is the intended meaning the source hopes the receiver will understand; the Channel is the medium through which the message is conveyed and it must tap into the receiver’s sensory system; the Receiver is the person at the end of the communication, someone who will decode the message and create their own meaning. 158 Lefebvre, R. C. (2007). The New Technology: The Consumer as Participant Rather Than Target Audience. Social Marketing Quarterly, 31-42. 159 Juniper Research. (2008). Mobile User Generated Content: Dating, Social Networking & Personal Content Deliver. Juniper Research. 160 OECD. (2007). Participative web and user-created content: Web 2.0, wikis, and social networking. Organisation for Economic Co-operation and Development. Paris. 161 Short, J. W. (1976). The Social Psychology of telecommunication. Hoboken, NJ: John Wiley & Sons, Ltd. 162 Daft, R. &. (1986). Organization information requirements, media richness, and structural design. Management Science, 32(5), 554-571. 163 Goffman, E. (1959). The Presentation of Self In Everyday Life. New York: Doubleday. 164 Sparkes, M. (2014, February 24). Betfair to offer ‘self-destructing odds’ via Snapchat. The Telegraph. 165 https://www.globalwebindex.net/ 166 https://www.statista.com/statistics/265146/number-of-mobile-internet-users-inchina/ 167 http://www.bbc.com/news/blogs-news-from-elsewhere-36226141 168 De Mente, B. L. (2009). The Chinese Mind: Understanding traditional Chinese beliefs and their influence on contemporary culture. Tuttle Publishing. 169 Hempel, J. (2012, September 10). Facebook’s China Problem. Retrieved from Fortune Tech: http://tech.fortune.cnn.com/2012/09/10/facebook-china-problem 170 Benson, L. (2009, October 26). Casinos saving face online. Retrieved from Las Vegas Sun: www.lasvegassun.com/news/2009/oct/26/saving-face-online/ 171 Eley, B & Tilley S. Online Marketing Inside Out: Reach New Buyers Using Modern Marketing Techniques. May 28, 2009. Sitepoint 172 Nelson, A. (2013, November 21). 50 ways to drive traffic to your website with social media. Retrieved from Exact Target Cloud Blog: http://www.exacttarget.com/blog/50ways-to-drive-traffic-to-your-website-with-social-media/ 173 Golder, S., & Huberman, B. A. (2006). Usage Patterns of Collaborative Tagging Systems. Journal of Information Science, Volume 32 (2), pages 198-208. 174 Dubois, L. (2010, September 16). How to Use Social Bookmarking for Business. Inc. Retrieved from http://www.inc.com/guides/2010/09/how-to-use-social-bookmarkingfor-business.html 175 DuBois, S. (2014, July 4). Google Glass Hits the Operating Room. Retrieved from tennessean.com: http://www.tennessean.com/story/money/industries/healthcare/2014/07/05/google-glass-hits-operating-room/12228547/ 203 ANDREW PEARSON 176 Mathes, A. (2008). Folksonomies – Cooperative Classification and Communication Through Shared Metadata. Computer Mediated Communication – LIS590CMC. University of Illinois Urbana-Champaign: Graduate School of Library and Information Science. 177 http://redcrosschat.org/2011/02/16/twitter-faux-pas/ (Retrieved: 7 July 2014) 178 178 http://www.davecarrollmusic.com/music/ubg/ (Retrieved: 7 July 2014) 179 https://www.youtube.com/watch?v=xaNuE3DsJHM (Retrieved: 7 July 2014) 180 https://www.merriam-webster.com/dictionary/blog 181 http://www.webopedia.com/TERM/B/blog.html 182 Wortham, J. (2007, December 17). After 10 Years of Blogs, the Future's Brighter Than Ever. Wired Magazine. 183 Economist, The. (2006, April 20). It’s the links, stupid. Retrieved from Economist.com: http://www.economist.com/node/6794172 184 Baker, J. (2008, April 20). Origins of “Blog” and “Blogger. Retrieved from linguistlist.org: http://listserv.linguistlist.org/cgi-bin/wa?A2=ind0804C&L=ADSL&P=R16795&I=-3 185 Lohmann, S. B. (2012). Visual Analysis of Microblog Content Using Time-Varying Cooccurrence Highlighting in Tag Clouds. New York, NY: AVI 2012 Conference. 186 Twitter.com 187 Bifet, A. a. (2010). Sentiment knowledge discovery in twitter streaming data. Retrieved from University of Waikato, Hamilton, New Zealand: http://www.cs.waikato.ac.nz/~ml/publications/2010/Twitter-crc.pdf 188 https://en.wikipedia.org/wiki/Social_network 189 Boyd, D. a. (2007). Social Network Sites: Definition, History, and Scholarship. Journal of Computer-Mediated Communication, Vol. 13. 190 Datta, A. J. (2009). Expanding Opportunities in a Shrinking World: A Conceptual Model explicating the Role of Social Networks and Internet-based Virtual Environments in Social Entrepreneurship. International Journal of Virtual Communities and Social Networking, 1 (4), pp. 33-49. 191 https://en.wikipedia.org/wiki/WeChat 192 Segev, L. (2014, March 20). WeChat is so much more than just Instant Messaging. Retrieved from Thetechieguy.com: http://thetechieguy.com/2014/03/20/wechat-is-somuch-more-than-just-instant-messaging/ 193 Baker, C. (2014, May 26). 4 Ways Brands Can Use WeChat for Sales. Retrieved from clickz.com: http://www.clickz.com/clickz/column/2346596/4-ways-brands-can-usewechat-for-sales 194 Wasserman, T. (. (2013, January 3). Is Snapchat the Next Frontier for Marketers? Retrieved from Mashable.com: http://mashable.com/2013/01/02/snap-chatmarketers.com 195 Stampler, L. (2013, February 19). How Maker’s Mark turned its watered down whiskey debacle into a social media win. Retrieved from Business Insider: http://www.businessinsider.com/makers-mark-turns-whiskey-fail-into-win-2013-2 196 Chiu, C. I. (2012, April). Understanding social media in China. Retrieved from www.mckinsey.com: http://www.mckinsey.com/insights/marketing_sales/understanding_social_media_in_c hina 204 THE PREDICTIVE CASINO 205 206 THE PREDICTIVE CASINO CHAPTER FIVE Social Business th “If the 20 Century was known in marketing circles as the st advertising century, the 21 Century may be the advertising measurement century. Marketers are increasingly focused on the effectiveness of their pitches, trying to figure out the return on investment for ad spending...The ability of newer Digital media to provide more precise data has led traditional media like television, radio, magazines and newspapers to try upgrading the ways they count consumers.” ~Stuart Elliot, NY Times Overview The most important thing to recognize about social media is the fact that the content is user generated. Social networks provide all of the tools their members require to become content producers and social network members submit photos, videos and other forms of multimedia as well as provide customer reviews, content for blogs and vlogs and links to other social 197 networking websites that they find noteworthy. The content comes from the users themselves, not from the publishers, and this is an important 197 distinction. The publisher supplies all of the necessary tools for the content’s distribution, but it must remain at arm’s length from the actual content to ensure that the integrity of the content remains intact. 198 Business.com's Top Tools to Measure Your Social Media Success states that there are five Ws that must be kept in mind when devising a social media strategy. These are: 1. 2. Who within the company will be using this tool? Will one person or several people be using the tools and will they be inside or outside the organization? Will the primary user be tech savvy or will he or she require an intuitive interface? What key performance indicators (KPI) are to be measured with this tool? It is imperative to know how you are going to measure and benchmark your social media efforts as this will dictate what social media monitoring tools are the best to use. If sales revenue is a key 207 ANDREW PEARSON 3. 4. 5. KPI, businesses should invest in a tool that integrates with a CRM system to track impact. Where on the Web will the business be engaging customers, and where does it plan to monitor its social media conversations? If a business is only interested in tracking specific channels such as Facebook or Twitter, tools such as Facebook (obviously), 48ers.com and socialmention.com can help with the former, while Twazzup, TweetEffect and Twittercounter can track the latter. All-encompassing tools that monitor new sites and forums are useful to monitor mentions from across the entire Web. When should the company be alerted of conversations and mentions within the social media sphere? Options here include general reporting dashboards or instant notifications via e-mail alerts or RSS feeds. Why is the company engaging in social media? This is, perhaps, the most important question of all, and a casino operator must decide whether it is turning to social media to manage its online brand reputation, to engage with its customers and/or potential customers, to provide real-time customer service, or simply to drive traffic to its website to influence SEO. A company is only as strong as its weakest customer relationship and I believe that social media can help both businesses reach their customers in highly efficient and, what can be extraordinarily affordable, ways. Businesses should look to social media to help them in the following ways: • • • • • • • • • • • • • Add interactivity to a Website Brand and Anti-Brand management Brand loyalty enhancement Build fanbases Crisis management Develop a virtual social world presence Discover a customer’s psychological profile Discover important brand trends Engage customers and potential customers Harvesting customer feedback Market to consumers Reputation management Social Shopping Studies have shown that 80% of social media users prefer to connect with brands through Facebook and 43% of people prefer Pinterest over associating 199 directly with retailers and/or brands. This fact alone should underscore the importance of social media in a business context. This fact, coupled with the 21 power of a Reed Network —i.e., one million people marketed to when only 20 208 THE PREDICTIVE CASINO people are reached—should show how important marketing through social media should be. This is the true power of social media and it cannot be underestimated. And, when coupled with mobile, that number can be even greater and the reach lightning fast. Throughout the rest of this chapter, I will go into detail about each of the different ways in which businesses should use social media. First, I will lay out the foundations for social media marketing, including a description of Jantsch’s 200 Hierarchy of Social Marketing , as well as reveal the differences between social media users. In his article The Hierarchy of Social Marketing John Jantsch argues that when venturing into social media businesses should be following six distinct steps 201 (see Figure 11). Using Abraham Maslow's Hierarchy of Needs as a blueprint to look at social marketing, Jantsch states implicitly that each step should be 200 fully understood and implemented before any ensuing step is undertaken. Just as Maslow claimed that “Self-Actualization” can only be reached after all the needs below it have been fulfilled, so too does Jantsch believe that, without a mastery of the first five steps, the sixth step–“Micro”–will be useless 200 because engagement with the intended audience will be shallow. Figure 11: Jantsch's Hierarchy of Social Media Jantsch's breaks the hierarchy down into the following steps 1. 2. 3. 200 : Blogging: The pyramid's foundation and the doorway through which all other social marketing should flow. Businesses and individuals should read blogs, comment on blogs, and then blog themselves. RSS: Aggregate and filter content around subjects and use RSS technology as a tool to help repurposing, republishing, and creating content. Social Search: participation is important at this stage as is stimulating and managing one’s reputation. 209 ANDREW PEARSON 4. 5. 6. Social Bookmarking: Tagging content to and participating in social bookmarking communities can both open up more channels for a business as well as generate extra search traffic to a site. Social Networks: Creating profiles on any social networking site will prove frustrating if the steps below it haven't been completed. These networks take time to understand and thrive on ideas and content, therefore a lot of content is needed to build a strong business case. Micro: With their instant tracking, joining, and engagement capabilities, platforms such as Twitter, Thwirl, Plurk and FriendFeed are very important elements of a social media strategy. They are atop the pyramid because, without content created below them, engagement will be superficial, at best. Although this hierarchy is an interesting framework, it should be taken with a grain of salt. Social and mobile media is moving at such a high rate of speed as well as splintering off in a thousand different directions that nothing in this field is set in stone. To create an effective and engaging marketing campaign, a marketer must understand the individual being marketed to as best as he can. As there are literally billions of people using social media, categorizing them into a simple classification system is not easy, but Li and Bernoff circumvent this problem by differentiating social media users into six different categories and they have created the Participation Ladder for Social Media (Table 10) that includes the 202 following stages : TYPE OF PARTICIPATION ACTIVITY Publish a blog Publish your own Web page Creators Upload video you created Upload audio music you created Write articles or stories and post them Post ratings/reviews of products or services Comment on someone else’s blog Critics Contribute to online forums Contribute to/edit articles in a wiki Use RSS feeds Collectors “Vote” for Web sites online Add “tags” to Web pages or photos Joiners Maintain profiles on a social networking site 210 THE PREDICTIVE CASINO Visit social networking sites Read blogs Listen to podcasts Spectators Watch video from other users Read online forums Read customer ratings/reviews Inactives None of the above Table 10: Participation Ladder for Social Media Source: Harvard Business Review 202 As the “Inactives” at the bottom rung do not use social media, they can be ignored, but the other five types should be looked at and marketed to individually. It should also be noted that there are two other superficial types; the “contributors,” who actively participate by starting their own conversations or replying to other threads and the “lurkers,” who only read and follow 203 content created by others. It is important to recognize that the vast majority of social media users fall into the “lurkers” category and these people are highly influenced by the 203 “contributors” , so reaching the contributors is paramount. Casino operators should recognize the unique symbiotic relationship between “contributors” and “lurkers” that is similar to a political commentator and his or her reading 203 public. These commentators do have enormous sway over their audience. Dovetailing Li and Bernoff’s participation ladder is the 1:9:90 rule, which as 204 Paul Bates states in his article Social Media Theory the 1:9:90 rule the 1:9:90 rule is not really a rule, but, rather, it is a useful concept to help understand how Social Media selling works. The theory proposes that for a mature website, just 1% of its website or social media visitors will actually produce original 204 material or user generated content. Meanwhile, 9% of visitors “will be editors or more likely commentators on that material and 90% of visitors will 204 only ever read the material without ever making a comment.” Respectively, these groups are known as: 1. 2. 3. Originators—1% Editors (or Commentators)—9% Lurkers—90% Bates points out that, “UGC is a very desirable thing on a social media site [and] if you can get your audience participating and creating original content, it will allow you to punch considerably above your weight and appear to have a large 204 and active user base.” As previously mentioned, UGC can be a very good thing in terms of SEO and allowing users to add comments and/or create 211 ANDREW PEARSON original content will go a long way to increase SEO rankings. Bates also recommends that social media managers and professionals go out of their way to produce content that will attract comments and readership, but “it is even more profitable if social media managers or business owners can 204 persuade others to create relevant and original content” ; getting such comments help business’s gain credibility, especially if the comments are positive. A Facebook “like” is nice, but it doesn’t take a lot of effort to “like” a Facebook page, whereas a comment is much more valuable because someone had to actually have gone through the process of explaining why they might 204 recommend a product or a service, or both. The “lurker” category is the one Bates believes should not be taken for granted, 204 though. “This is the pay dirt for a social media based sales strategy,” Bates 204 argues, adding that this is where the big numbers lie. By definition, lurkers have no engagement with a business and social media theory suggests that 204 engagement must first be developed before any sale can be made. However, the good news is that, although the ratios remain relatively stable, the 204 individuals within these groups are constantly in flux. “A lurker might indeed get interested and become a commentator before eventually buying and then just disappear or go back to lurking, or your 1% of originators might suddenly 204 stop creating content and become lurkers” argues Bates. What the 1:9:90 theory fails to address is that beyond the active user base is a whole world of 204 potential prospects that you hope one day, at the very least, become lurkers. The Uses of Social Media Kaplan and Haenlein argue that there are several pre-conditions that should be 23 met before a company embraces social media. First, because there are so many social media sites available, it is imperative to choose social media 23 applications based on where a target audience tends of congregate. Once the social media sites have been selected, “the next decision involves 23 whether to make or buy.” In some cases, it is best to join an existing social media platform and take advantage of its popularity and built-in user base, while in other cases the right application might not be available so it is 23 necessary to build one from scratch. Whichever way an IR decides to go, “it is vital that there is an understanding of the basic idea behind Social Media. It's all about participation, sharing, and collaboration, rather than straightforward 23 advertising and selling,” contend Kaplan and Haenlein. Since by its very nature social media is a splintered platform, it is crucial to 23 ensure that all social media activities are aligned with each other. As an 23 example, Kaplan and Haenlein reference Dell and its “Digital Nomads” campaign, which used “a combination of social networking sites (Facebook, LinkedIn), blogs, and content communities (YouTube videos) to show how its 212 THE PREDICTIVE CASINO range of laptop computers enable individuals to become a nomadic 23 workforce.” The main goal was to keep the message consistent across all 23 channels. Social media shouldn't exist in a vacuum and it should be integrated with traditional media, contend Kaplan and Haenlein, who also argue that there is 23 little differentiation between the two in the eyes of the consumer. The ideas for social media content for a casino company are almost endless and casino marketing departments will have no shortage of events, shows, fights, concerts, etc., etc., to promote on a weekly and monthly basis, but these events should all be coordinated through an holistic social media marketing lens. Finally, a company that wants to create a social media presence should ensure that all of its employees have access to all of the company’s social media platforms. Obviously, a fine balance between giving everyone access to social media platforms and complete freedom of expression must be struck so that not everyone is wasting their time producing and uploading irreverent YouTube 23 videos. “One possible approach involves defining groups of employees whose primary objective is the management of corporate social media; all other staff members are treated as occasional participants. Under this scenario, the first group is given administrator rights—rights which allows the opening of new discussion threads and deletion of inappropriate posts—while the second 23 group is not,” Kaplan and Haenlein argue. This is important because it could be rather embarrassing if certain people who shouldn’t have access to a company’s social media accounts do, as music chain HMW discovered when an employee who was about to be fired started live205 tweeting about the mass firing of 190 staff. In her article The Top 10 205 corporate social media disasters , Burn-Calandar The first tweet announced: “’We're tweeting live from HR where we're all being fired! Exciting!!!", followed by: “There are over 60 of us being fired at once! Mass execution, of loyal 205 employees who love the brand.” “Unfortunately for HMV, even its inability to lock the outgoing staff out of the social network was broadcast to the masses. The rogue @hmvtweets tweeter wrote: “Just overheard our Marketing 205 Director (he's staying, folks!) ask ‘How do I shut down Twitter?’” It is also important to inform the social media community that employees need to identify themselves before they post so that “end-consumers don't get the impression that anonymous accounts are used by employees to post fake 23 messages and overly-positive feedback.” Kaplan and Haenlein warn that this 23 could severely damage the credibility of the whole social media campaign. Once these pre-conditions have been met, Kaplan and Haenlein lay out five 23 specific points to be aware of when going social and these include : 213 ANDREW PEARSON 1. 2. 3. 4. 5. Be active: keep your content fresh and constantly engage the customer, taking the lead whenever necessary. Be interesting: After listening to your customers to hear what is important to them, “find out what they would like to hear; what they would like to talk about; what they might find interesting, enjoyable, 23 and valuable.” Be humble: you should engage only when you have learned the basic rules and gained the necessary understanding, but remember to do so with humility. Social media has been around for a while now so don't claim to be an expert if you're only just wading into the waters. Be unprofessional: social media is not about projecting absolute professionalism, it is about being human and that entails making mistakes. Social media consumers much prefer someone who occasionally gets his or her hands dirty than someone with a stuffy, boorish attitude. Be honest: honesty is always the best policy, especially when a little dishonesty can be highly embarrassing, as well as ignite a firestorm of negativity in the blogosphere. As Kaplan and Haenlein warn, “Never expect that other participants may not find out who stands behind some anonymous user account; after all, you're dealing with some of 23 the most technologically sophisticated people on the planet.” Now that I have built what I believe is a strong base of understanding for why a company or an individual would want to delve into social media, I will introduce the very specific ways a casino can use social media. Add Interactivity to a Website Adding social media to a static Website can turn a site into a highly interactive destination that allows interested users and/or customers to actually become a participant in the marketing of the site. Adding “Like” buttons makes it easy for users to share content; sites such as Facebook, Twitter, and YouTube are some of the more well-known and highly understood social networks and are good places to start, but other platforms such as Google+, Pinterest, Tumblr, Vimeo, Instagram and LinkedIn can also be used to spread your message far and wide. On those sites, businesses can add content that keeps customers interested and returning for more. All of these sites usually provide widgets that are easy to include on other Websites. RSS feeds, blogs, and podcasts are also great platforms that can keep customers engaged. Adding videos and presentation to a website is also a great way to drive traffic. One of the most important things to remember is to update content often. This is probably the single most important rule of SEO, give the search engines a reason to keep coming back to your site with strong content. 214 THE PREDICTIVE CASINO Brand and Anti-Brand Management In 2001, Hislop defined branding as “the process of creating a relationship or a connection between a company's product and emotional perception of the customer for the purpose of generation segregation among competition and 206 building loyalty among customers.” According to the Management Study 207 Guide , “Brand management begins with having a thorough knowledge of the term ‘brand’. It includes developing a promise, making that promise and maintaining it. It means defining the brand, positioning the brand, and delivering the brand.” Although brand management can be considered to be “nothing but the art of 207 creating and sustaining the brand,” Jones and Huang note that, “unlike traditional consumer-created brand communities where the brands appear to occupy minimal presence, online brand communities are increasingly hosted by 208 the brands.” 209 Building upon Schau et al.’s research into collective value creation within brand communities, Jones and Huang found that “Social media as an interactive technology both enables and becomes a site of value creation for 208 the individual consumers, the brand community and the brand itself.” Unlike user-generated brand communities, where brands are often unwelcomed, these online brand communities “revealed that brand content contributions on their brand offerings and interactions with consumers’ 208 content contributions were appreciated by the community.” “These consumer-brand interactions, functional and otherwise, fostered emotional bonding between the members and brands—because the brand has become individual and approachable, and the boundaries between consumer and brand 208 interactions have somewhat blurred,” Jones and Huang conclude. Jones and Huang list six different types of avenues that can be exploited by 208 brands, including : 1. 2. 3. 4. Branded social networks: These are customized platforms for interacting with consumers. They often include a fusion of applications, such as discussion forums or wikis. External social networks: While some brands choose to create their own community websites, others just have a presence on an external social networking site, thereby taking advantage of the network’s built-in platform and audience. Innovation hubs: These are unique platforms provided by the brands for users to post their ideas to the company. Content aggregation sites: These are websites where users share media content with other users. They are Websites such as BookCrossing, Flickr, YouTube, and Slideshare and they often include brand-sponsored “channels” through which the brand distributes 215 ANDREW PEARSON 5. 6. media content and allows users to engage with and comment upon the content. Blogs: A normal type of blog, but filled with corporate information. Micro-blogging: Brand-owned micro-blogging websites are normally used for data collection (i.e., a Twitter account that is used to track Tweets and mentions). Brand is one thing and anti-brand is another; when it comes to anti-brand management, businesses should be well aware of the threat this unique problem entails. In his article Negative Double Jeopardy: The Role of Anti-brand 210 Sites On the Internet , S. Umit Kucuk claims that anti-brand websites are today‘s “form of boycott and protest, developed through consumer activism as a result of increasing consumer power.” According to Kucuk and Krishnamurthy because of the advent of the Internet, “Consumers are able to clearly broadcast their messages and organize with other like-minded consumers,” which allows them to use “anti-brand websites as weapons of empowerment to battle the 210 corporate world and its brand power on a day-to-day basis.” It doesn’t take much to set up a website these days and as Kucuk explains, “the 210 corporation has a website and so does the consumer.” Anti-brand sites are “attacking targeted brands and corporations by using their most powerful online branding tool against them: ‘domain names’. Many such anti-brand domain names are easy to remember and catchy in nature (such as Northwest Airlines’ Northworstair.org, Safeway’s Shameway.com, Starbucks’ 210 Starbucked.com, Coca-Cola’s Killercoke.org, etc.).” Kucuk notes that, “Antibrand sites purposefully use the targeted corporation’s brand name in their domain name to insult the corporation’s brand identity and to express their anger and frustration while entertaining and educating consumers and 210 audiences” alike. As the attack on Las Vegas Sands proved, there are bad actors out there and a corporate owner can just as easily be targeted for his political views as anything to do with his company. Although their names can be quite humorous, these websites are no laughing matter and they must be taken very seriously and the courts in the United States have been of little help; many have found in favor of these anti-brand sites, stating that, “usage of a targeted brand’s name in a domain name is not trademark infringement, but is protected under the First Amendment—as long 210 as the site owner does not use the anti-brand site to make a profit.” Like a virus living off its host, these anti-brand sites “benefit by sharing the link popularity, brand awareness and web traffic of the targeted brands’ site in many search engine results and in consumer surfing decisions on the 210 Internet.” Troublingly, anti-brand sites “often show up in the top ten search results when a corporate brand is researched on major search engines,” Kucuk 216 THE PREDICTIVE CASINO 210 adds. Other sites are also taking “advantage of mistyping (called typosquatting) to steal traffic directed to the targeted brands as in the case of 210 Untied.com, a hate site targeting United Airlines (United.com).” “For these oppositional consumer groups, anti-brand sites have turned out to be major message dissemination venues and a powerful communication tool. Today, hate sites exchange information, organize boycotts and coordinate 210 lawsuits, thus revolutionizing consumer movements,” Kucuk adds. The Internet’s tagline could almost be “Seek and you shall find” and, in the case of disgruntled airline passengers or slighted employees, these sites are easy to discover and even easier to add complaints to, so companies should be wary of engagement on them. Another thing to keep in mind is that if one of these antibrand websites discusses company policy and invites other employees to comment, it “can be considered ‘concerted activity’ and is protected by the 210 National Labor Relations Act—NLRA.” Kucuk breaks down anti-brand sites into four different categories—Experts, 210 Symbolic Haters, Complainers, and Opportunists. The developers of an “Experts” site usually have detailed knowledge about a company’s markets and their alternatives as well as expertise about business practices, products and 210 technologies. “Because of their advanced level of expertise, they are capable 210 of sensing and following market changes in real time,” Kucuk notes. These Websites are often sophisticated and some “apply strong expressiveness and communication strategies with unforgettably powerful images to maximize 210 their impact on visitors to the site,” with the ultimate purpose of hurting 210 brand identity. “This, in effect, creates some level of economic pressure by 210 stimulating anti-consumption against the targeted brands,” concludes Kucuk. The brands targeted by “Symbolic Haters” have high brand awareness, but they 210 are not as valuable as those targeted by the experts. This group of anti-brand protesters is predominantly sustained by rumors, supposition and negative word-of-mouth and they focus more on the myths behind the brand’s 210 success. In order to create opposition to a targeted brand, “Complainers” reflect their anger by bringing negative attention to a company with service failure 210 scandals. “Complainers” are “more interested in operational and product210 related problems than business philosophy or system order,” notes Kucuk. Acting like rejected lovers, these complainers “might have initially tried to build communication with the company regarding their concerns, but their insight was not appreciated by the company, and they chose to protest them on hate 210 sites using ‘wake-up call’-type attention grabbers to get their point across.” These types of sites are not as advanced as either experts or symbolic haters 210 sites , and “their expressiveness is limited to depictions of actual service failures (pictures of smashed packages, etc.) or scanned and posted documents 217 ANDREW PEARSON about the unresolved communications with the targeted company rather than 210 advanced interactive website designs.” The message, however, is clear and often the examples given involve personal experiences that the Website’s readers can empathize with, and thus the reader may develop a negative opinion of the targeted company based on another person’s experience. “Opportunists” could be considered the scavengers or hyenas of the anti-brand website world as they “rely on a company’s service failures reflected in the 210 media news as their main source of information.” Kucuk argues that “opportunists are fed by media, not personal expertise nor experience, but they are trying to use flashy news stories to influence potential consumers into 210 viewing their own website in order to increase site traffic.” “Opportunists” are driven not by personal experience, but rather by a desire to trumpet 210 scandalous news so that their websites gain more traction and attention. Although anti-brand sites are usually created to attempt to hurt brands, smart companies can—in a Vito Corleone “Keep your friends close but your enemies closer” kind of way—use these websites to their advantage. Kucuk offers four basic strategies for companies to counter these anti-brand Websites; work with experts; monitor symbolic haters; talk to complainers; and combat 210 opportunists. The anti-brand “Experts” websites might actually be helpful because they can alert a company to problems they might never have known existed, and the expertise shared on the sites can be used in a company’s market value creation 210 process going forward. Unlike customer-satisfaction surveys or pricey 210 consultants, feedback gained from such sites is free to the company. Some very useful information can be gleaned from an ex-employee’s tell-all Website, as former employees often have extensive knowledge of the inner workings of 210 a company , and this is knowledge that is oftentimes overlooked by higherlevel company executives. Of course, before engaging with these experts, a company should first analyze 210 and determine the “hostility level” and “expertise level” of such sites. If the hostility level is extremely high and the expertise level low, the site should be monitored but not necessarily engaged with, until, that is, the hate generated 210 has reached such a harmful level that it needs to be addressed. “If a company reaches a somewhat manageable level of hostility along with a good level of consumer expertise through these sites, it should, however, encourage consumer involvement in the market co-value-creation process,” advises 210 Kucuk. Because symbolic haters might be under the influence of negative word-ofmouth stories, the challenge with these sites is to counter these negative 210 stories with timely positive and credible ones. The insight symbolic haters provide on their sites isn’t usually as useful or informative as that provided by 218 THE PREDICTIVE CASINO 210 experts, however. Kucuk advises that, “Companies should closely monitor what these sites are talking about and be open to any communication form directed to consumers in order to control this symbolic (or sometimes even 210 disingenuous) hate targeted towards their brands.” “In other words, a company cannot defend its perspectives without knowing the truth behind the 210 news broadcasted by such sites:” once again, keeping one’s enemy close. “Complainers” are the consumers who “might have been satisfied with a 210 company’s products and services for a while, but have grown dissatisfied.” They might still be “looking for the spark or enjoyment they once felt when 210 they met with that brand,” but they have currently fallen out of love with it. Since these sites are trying to garner attention by focusing on major service failure scandals, Kucuk recommends that companies “contact site owners to solve such consumer dissatisfaction problems before the aggression begins to 210 impact the company’s brand identity on the Internet.” Because negative word-of-mouth can go viral in seconds and live on social media sites almost indefinitely, it is imperative for companies to challenge negative word-of-mouth stories and transform them into positive word-of210 mouth ones as quickly as possible. Companies should keep an eye on ecomplaint sites, as well as consumer blogs to monitor the type and duration of 210 problems that can flare up. When appropriate, companies should email upset customers, perhaps attaching discount coupons or gift cards to show its remorse for poor service and/or bad quality,y as well as its sincerity towards 210 improving customer service. As their name implies, “opportunists” are mostly looking to exploit an opportunity to be recognized so that they can reach a level of awareness that 210 will gain them public notoriety. “Opportunists” “can be very harmful once they find scandalous events regarding a targeted brand, which brings the site 210 higher visibility and traffic,” warns Kucuk. “Opportunists” are trying to steal web traffic from the targeted company, and will search the news media for anything they can use to gain attention by attacking brands. “In a pre-emptive attempt to prevent the creation of such anti-brand sites, corporations can buy potential negative domain names that can be targeted on 210 the Internet,” advises Kucuk. For every Facebook.com there is probably going to be a Facebooksucks.com so companies should be aware of the existence of these opportunists as well as be ready to combat them to stop any potential 210 brand erosion. The “.sucks” web address is also available so any disgruntled customer could purchase that and quickly become a headache for any casino company. Snapping up possible anti-brand sites when creating regular brand sites can help avert many potentially embarrassing anti-brand headaches in the future. Casino companies are ripe for anti-brand messaging as the 2014 cyber attack 219 ANDREW PEARSON on Las Vegas Sands showed. According to US intelligence, this attack was 211 orchestrated by the Iranian government. Culling through sites like Twitter, Facebook, Weibo, etc., etc., for customer complaints should be a daily, if not hourly practice. Alerts can be set up to peg comments by known patrons, with an appropriate email response automatically prepared. Approval from a host, manager, or executive would only take moments and responses could, literally, be in the hands of patrons moments after they made their critical tweet, Facebook post and/or Weibo comment. Brand Loyalty Enhancement As Reichheld and Sasser showed in their study Zero Defections: Quality Comes 123 to Service , loyalty is important because a customer’s profitability increases as his or her loyalty increases. In their study, the authors found that it usually took more than a year to recoup any customer acquisition costs, but then profits 123 increased as customers remained with the service or firm. The cost of attaining a new customer is also higher than maintaining a recurring 123 customer. Threadless, a T-shirt company based in Chicago is a perfect example of a business that has built a community by embracing its customers and creating strong brand loyalty. Today, Threadless is as much an online community of artists as it is an ecommerce site. Founded in 2000, Threadless “asks consumers to submit shirt designs they've created—it gets as many as 300 submissions a 212 day—and [sic] allows its large fan base to vote on the ones they like best.” Threadless “picks the best of the most popular T-shirt designs, screens them for copyright violations and obscenities, and sells them on its site within three to 212 eight weeks for $18.” Competition winners earn about $2,000 for their 212 creations and Threadless aims to release seven new designs per week. With 2.1 million followers on Twitter and 615,000 fans on Facebook, Threadless has used social media masterfully to promote both its designs and its designers 212 as well as, just as importantly, keep its community engaged. Threadless injects its personality into every engagement, sending news to people when new T-shirts are available, as well as informing customers of such quirky things as what music is playing in the company warehouse and stories 212 about the interesting people stopping by the office. All of this work in social media has been very successful. According to Threadless, “The investment in Twitter has bumped our traffic. Sales from Twitter alone are in the high six212 figures.” Threadless believes “the other key is that we act like humans on our own site and social networking sites. We act like we're interacting with our friends, posting videos of our employees talking about their favorite bands. It's not all 220 THE PREDICTIVE CASINO 212 direct promotion; it's human.” Burkitt concludes that the takeaway is: “Know 212 what interests your consumers and build on it.” Another example of a great use of social media is the coffee company Starbucks; it is a company that has taken the lead in harnessing the power of 213 social media. As Hodge explains, “With 32 million 'Likes' on Facebook and almost 3 million followers on Twitter, the numbers alone are impressive. But it's the depth of engagement Starbucks achieves with its customers that makes 213 it stand out.” A few years ago, Starbucks recognized that more and more of its users were accessing their accounts from a smart phone or a tablet, so they launched a 212 Frappuccino Happy Hour photo competition. “Each day for two weeks 212 Starbucks gave its Twitter followers a different photographic challenge,” Hodge explains. “Users had to tweet a photo of themselves, including the @StarbucksUK username, for the chance to win a £10 Starbucks card,” Hodge 212 added. According to Hodge, the campaign achieved “maximum exposure for minimal cost, recruiting many hundreds of thousands of unofficial 'brand 212 ambassadors', who were generating online buzz about Starbucks.” Build Fanbases: Be “Liked” Whether you call them “fans”, “friends”, “followers” or “subscribers”, building a community of users for your casino or integrated resort will help you grow your customer base, as well as participate in the conversation about your brand. Facebook no longer uses the term “Fan”, and, instead of becoming a “Fan”, Facebook would rather users “Like” or “Find” a brand instead. Not wanting to get into too much of a semantic argument, I believe the concept of a “Fan” makes more sense when discussing the creation of user communities. “Fandom” connotes devotion, enthusiasm, advocacy, and affection and all of these positive emotions can be used to drive customer loyalty and, just as importantly, customer spend. Few would argue that customer satisfaction is the foundation of true customer loyalty, while customer dissatisfaction is one of the key factors that drive 109 customers away. This may sound obvious but its importance cannot be stressed enough. Increasing loyalty is what building fanbases is all about, but an IR should be aware that building large fanbases and gaining thousands of followers shouldn’t be an end-goal in itself, it is important to put in place a system that produces enough content to keep these fans engaged. As per Jones and Sasser’s zone of affection, satisfaction levels are high and “customers may have such high attitudinal loyalty that they don’t look for 114 alternative service.” It is within this group that “Apostles” or “Influencers” reside, and this is the group that is responsible for improved future business 221 ANDREW PEARSON 114 performance. It is within this group of devotees that businesses can find their most vocal and valuable customers. Through social media, these people can be discovered, nurtured, and stimulated to spread the company’s message far and wide. With little marketing expenditure, a company can get its message out to the wider community and, once again, because these recommendations would be coming in through a marketed person’s contacts, followers and friends, they are messages that are much more likely to be acted upon. To increase a company’s fan base, Facebook recommends doing the following: • • • • • • Encourage visitors to “like” your page. Partner with other brands or local organizations: this can promote viral sharing of customers between businesses. Expand the reach of your posts: when you mention a person or an organization in your posts, be sure to connect to them in a post so that your post will automatically appear on their wall. Use social plugins on your website: Installing a “Like” box on your homepage or on your newsletters will drive people to your Facebook page. Encourage physical check-ins at your business: these posts will appear in a customer’s News feed, thereby providing more exposure for your business. Promote with ads and sponsored stories. The best way to encourage engagement and interaction on Facebook is to post 214 multimedia links on your page. You should post images regularly so you have a better chance of standing out, even on status updates as images will catch the eye of your current followers and images are shared much more often than 214 simple text. Running competitions is another great way of increasing fans and fostering engagement, but be aware of Facebook’s rules as they can be 214 quite restrictive. Just as on Facebook, images are also a great way of increasing engagement on 214 Twitter. “With the likes of Instagram and Camera+ allowing Twitter integration, it's easier than ever to tweet images. There's no harm in cross posting from different social media sites so don't be afraid to link to a Facebook or Pinterest post if it's interesting. Better yet, post the image and provide the 214 link to the other site.” “A handy way of gaining a few new followers is to use hashtags when tweeting. While there are always trending topics, you shouldn't latch onto them for the sake of it. Instead, identify hashtags that are relevant to your brand and tweet witty or useful information using it. If you provide value, you could be 214 retweeted, which will result in more exposure and some new followers.” 222 THE PREDICTIVE CASINO Twitter can also be used for competitions, but, once again, check out the rules. Twitter’s rules might not be as strict as Facebook’s, but you should refer to them to understand what is acceptable and what isn’t. Building up a following on LinkedIn is not as easy as it is on Facebook and 214 Twitter as there isn't as much scope for posting content. However, LinkedIn recently bought Slideshare and they have made it easy to integrate any Slideshare presentations on a LinkedIn page. With LinkedIn, it is important to get involved in groups and discussions as that’s a good way to get one’s brand 214 name out there. In this case, you might be promoting yourself more than your actual company page, but your strategy should be about building up your 214 own reputation, which should, subsequently, build up your brand. As it is entirely based around visuals, Pinterest has a dedicated following that is worth the time to engage with, even if your products aren't of a visual 214 nature. The great thing about it is the fact that you can create boards that 214 only vaguely relate to your company. The first step you should take is to pin high quality images. Your followers are going to judge your pins based on their appearance, no surprise there since this is a visual medium so you should use 214 the best quality images available. “Something else worth remembering is that for each Pinterest page, each board has its own number of followers that mightn't even be following your page. If you have a number of boards, check to see just how many people are following and prioritize the boards with the most members,” the digital agency Simply Zesty recommends. “That doesn't mean you should neglect your other boards, but it can be useful to focus on one or two boards as you build up your 214 following,” Simply Zesty concludes. Hashtag marketing is, in itself, a science. Casino operators should use hashtags to market the many events that occur within its halls, from sporting events, concerts, to conferences, to eSports events, and beyond. Crisis Management “Crisis management” refers to the art, technique or practice of averting or dealing with crisis situations that threaten to harm an organization, its stakeholders, or the general public. It is also the attempt to limit the damage of a known or an unforeseen problem. According to Seeger, Sellnow & Ulmer (1998), there are three elements that are common to most definitions of a crisis: (a) a threat to the organization, (b) the element of surprise, and (c) a short decision time. 215 In his paper A Typology of Social Media Crisis , Ashwin Malshe argues that one of the key benefits of using social media marketing is the two-way interaction it gives to both businesses and their consumers. “Consumers like it 215 because they can engage in conversations with the brands they buy,” while 223 ANDREW PEARSON businesses recognize the value of keeping their customers interested and informed. However, this can be a double-edged sword as a customer can as 215 easily complain about his bad experiences as he can trumpet a good one. Malshe argues that 215 : “The increased efficiency of communication and intimacy with consumers come at the cost of higher riskiness of the business. Users can share an article, video, or photograph with their social network at the click of a button, thus spreading a firm’s content “virally,” generating wide visibility instantaneously. In exactly the same way, users can harm firms by sharing bad experiences, rumors, or events that were pure accidents.” Malshe adds that because there is little accountability on the Internet, firms are put in a tight spot. The fact that there are no moderators on social networks monitoring the flow of information over this super highway only exacerbates 215 the problem. Because of this, social media crisis can flare up quickly and 215 spread at the speed of light , turning the very viral nature of the Internet—a nature that normally makes it so appealing to marketers—into a very serious threat against them. “There are several ways in which one can group social media crises,” argues Malshe, adding that “An intuitive criterion for categorization can be the type of 215 social medium.” Whether the crisis erupted on Facebook, Twitter, or YouTube will call for different strategies to be used to address and contain the 215 crisis. For example, since human beings are, first-and-foremost visual creatures, a video on YouTube will probably elicit more negative reaction than 215 a Tweet on Twitter. However, Malshe argues that crises should not be viewed in a vacuum as they can quickly spread from one network to the 215 next. Something initiated on Facebook can easily be shared with others on 215 Google+, Twitter, or the hundreds of other social networks available today. As Malshe notes, since a YouTube video may generate more attention on Facebook than on YouTube, it is not that important to identify which social 215 media site the crisis first blew up on. What is important is countering the 215 crisis on as many social media sites as it is affecting. 216 Using the Comprehensive Typology of Crises Model proposed by Gundel as a starting point, Malshe proposes the typology of social media crises. Gundel categorizes crises into “four types based on their predictability (high vs. low) 216 and controllability (high vs. low).” Grundel “puts an emphasis on predictability because, in general, organizations can design measures to proactively eliminate a crisis or at least prepare the response to a crisis based 216 on its predictability.” However, for crises flaring up across multiple social media networking sites, the consideration of predictability becomes less 216 important. Malshe argues that, “due to its open and viral nature, social 224 THE PREDICTIVE CASINO media makes crises almost unpredictable. Therefore, although organizations can put in place measures to avoid crisis-like situations, the predictability of the 215 time, place, or the nature of crisis [sic] is almost zero.” The global nature of social media also makes it exceptionally difficult to predict these types of crises, with language, geographical and cultural differences exacerbating not only the complexity of the problems, but also the responses necessary to deal with 215 them. Another possible reason for low predictability of social media crises is that companies currently lack an in-depth understanding of the social media 215 world. Social media marketing still isn’t very well understood and often is not 215 considered worthy of large IT investments. Although this is starting to change “this underinvestment leads to two related problems. First, the employees delegated to handle social media have little resources or incentives to systematically analyze crises that have already been taking place. Second, a lack of resources leads to relatively small social media teams, which in some 215 cases handle momentous amounts of data,” warns Malshe. “Controllability is a critical factor for general crisis management as well as specifically for social media,” states Malshe, adding “Crises with low controllability are dangerous and can scar the credibility of the firms [involved] 215 for a long time.” Because content can be easily captured through screengrabs, even small mistakes can have huge effects on a company’s brand identity. “Controversial articles that were yanked from the original websites remain alive on several blogs which display all or part of the original content,” Malshe notes. Even though rogue Tweets can be deleted in seconds, they can still jeopardize multi-million dollar ad campaigns because screenshots of the offending Tweets can encircle the globe as enthusiastic Twitter users share 215 them with friends and followers. Like a virus replicating itself, the damage can be quick and sometimes fatal. Although it seems that all social media crises have low controllability, that is 215 not necessarily the case. By tackling a problem head-on (oftentimes with disarming humor and/or clever irreverence), a few companies have been able 215 to avoid their crisis-like situations turning into full-blown disasters. An example of this is the case of the Red Cross beer Tweeting fiasco: when a Red Cross social media strategist mistakenly posted a Tweet about alcohol on the Red Cross’s official Twitter account. The Red Cross not only removed the offending Tweet quickly but also offered a witty retort on the fiasco, thereby 215 making light of the situation. Such incidences as these are not outliers and, in 215 a few situations, social media crises can be highly controllable. For Malshe: “It is critical to understand the difference between the two constructs: ease of control and the degree of controllability. In some cases, the overall controllability of the crisis can be low. Within that set of cases, a few crises can be managed, to whatever limited extent, with less effort while the 225 ANDREW PEARSON rest may require more effort.” 215 Malshe uses several examples to illustrate his point; including the crisis faced by United Airlines when one of its disgruntled passengers, Dave Carroll, 215 uploaded his parody song “United Breaks Guitars” to YouTube. In the song, Carroll and his band mates narrate an experience about how United broke Carroll’s pricey Taylor guitar and refused to reimburse him for the damages. 217 The song went viral and has now garnered more than 16.2 million views. United was very slow to react to the situation, which exacerbated it even more. The story was picked up by the traditional media and quickly went viral. Overall, for United, this was a situation with very low controllability, but their inaction made the problem considerably worse, turning it into a full-blown PR disaster. Considering the triviality of the amount involved in this crisis (about US $1,500), it was a relatively “low cost solution for United to offer the money to Carroll 215 and control the crisis,” Malshe notes. However, even though United may have reimbursed Carroll, there was no guarantee that he wouldn’t still have 215 uploaded the song to YouTube , but, at the very least, United wouldn’t have looked so tin-eared. In one television interview (more bad press for United), Carroll explained that he was emotionally attached to the guitar and, even after it had been fixed, it 215 didn’t sound as good as it had before it was broken. Therefore, in this situation, there was little United could have done to control the events from getting out of hand and becoming a full-blown crisis. However, “they could have easily averted additional negative publicity by reimbursing Carroll for his 215 troubles,” argues Malshe , thereby mitigating the crisis to a certain extent. After all of the social media attention the video received, United did actually offer to reimburse Carroll, but, even then, Carroll took the high road, refusing the money and asking United to offer it to charity, which, once again, did little 215 to help United’s negative public perception. It was a lose-lose situation for them all around; a deserved, self-inflicted lose-lose situation, admittedly. Malshe proposes that, to some extent, the shock value of the trigger that starts 215 the crisis can explain the ease with which the crisis can be controlled. Simply put, the higher the shock value of the trigger, the more difficult the crisis is to control, although Malshe does admit that, since there has been no large scale study of this issue, it is difficult to say to what extent shock value explains the 215 correlation. Malshe uses two dimensions to categorize social media crises. The first, “Controllability,” is based on Gundel’s definition and is described as something that is “controllable if responses to limit or eliminate the crisis by influencing its 215 causes are known as well as executable.” “Controllability” can either be high 226 THE PREDICTIVE CASINO 215 or low as can the effort required to control it. This effort, Malshe argues, is a function of the shock value of the trigger, which refers to the extent to which a 215 crisis-triggering incidence offends the social media community. The amount to which the masses can be shocked by a single event is completely dependent 215 on the context. The shock value of the trigger “refers to the degree to which an incident that is embedded in the context and time offends members of the social media 215 community.” This shock value can’t be isolated from the context and time, 215 making the predictability of social media crises even more difficult to gauge. Malshe divides crises into four categories and they are based on whether a crisis has high or low levels of controllability and the shock value of the trigger 215 is either high or low. For example : • • • • Soft Crises: Like the Red Cross Tweet previously mentioned, soft crises have high levels of controllability and low levels of shock value. Left unattended, a soft crisis is likely to turn into a full-blown crisis so it is important to recognize the crisis, isolate the classification and immediately issue a response. Firefighting: This type of crises is highly controllable, but it also has a high trigger of shock value. Even though controllability is high, the high shock value of the crisis has the potential to overwhelm a company’s social media department, which could mean the situation unravels very quickly. Wait and watch: These require patience and a thick skin. These crises are caused by less shocking triggers, but they have very low controllability. The crisis can go viral very quickly and the more an organization responds to the crisis, the more disastrous it could become. Disaster: These are highly uncontrollable and shocking crisis situations and they can break organizations. The Red Cross Tweet was a soft crisis because the original Tweet was offensive, but its shock value was relatively low and a clarification on the origin of the 215 Tweet was enough to dissipate the tension in the social media world. It is really not surprising that the Red Cross’s employees think about drinking beer after work, but the Tweet made it sound as if the employees were drinking 215 215 while at work. Once the facts were presented, the crisis was averted. Ashton Kutcher’s November 2011 Tweet stream about the firing of Penn State Football team coach Joe Paterno is one of the most celebrated examples of a 215 firefighting crisis. Kutcher, who had more than eight million Twitter followers at the time, expressed his disappointment in the firing and thought that the 215 whole incident lacked class. However, Kutcher was ignorant of the fact that Paterno had been let go because of his reluctance to fire a coach who had been 227 ANDREW PEARSON 215 accused of raping a child in the Penn State men’s locker room. Understandably, there was a huge backlash against Kutcher’s Tweet and his 215 image was badly bruised. Kutcher “immediately decided to stop engaging 215 directly with his Twitter fans and followers” and has since outsourced his social media responses. This incident is a perfect example of a firefighting crisis. The trigger—the 215 offending Tweet—was highly shocking to the social media community. Had he acted to counter his offensive Tweet, Kutcher could have controlled the crisis before it had spread so wide, Malshe argues. Looking back, it is surprising to see that one person—a busy celebrity at that—insisted on communicating with his followers on a one-to-one basis, thereby so easily exposing himself to a 215 crisis. “The crisis was controllable because Kutcher could have put in place a 215 system to tackle such situations,” Malshe concludes. The United Airlines crisis is a perfect example of a wait and watch crisis. The song was hardly shocking, but it had huge amusement value, which helped it go viral. Once the video was available on YouTube, there was little United could have done to defuse the crisis. In April 2009, the video of two Domino’s Pizza employees tainting food at a 215 Domino’s franchisee hit YouTube and, unsurprisingly, it instantly went viral. 215 Domino’s Pizza immediately found itself in a classic disaster situation. “The enormous shock value of the YouTube video coupled with extremely low controllability of the crisis made it the worst possible situation for Domino’s,” 215 states Malshe. Domino’s responded with a YouTube video of its own: Patrick Doyle, the President of Domino’s USA, acknowledged that the two people in the video were, indeed, employees of a Domino’s franchisee, but he explained that warrants for their arrests had been issued and that the store where they had 215 worked had been completely sanitized since the incident. This crisis became a disaster because the trigger—the YouTube video—had a very high shock 215 value—but the crisis was highly uncontrollable. The employees’ actions were 215 suicidal and completely unpredictable. Domino’s couldn’t really be faulted for not having a plan for such an extreme scenario as the sheer bizarreness of 215 the act of sabotage made it uncontrollable and completely unavoidable. Going forward, Domino’s promised to toughen its employment hiring practices, while also hammering home the message that this was an isolated incident that 215 didn’t represent normal Domino’s employee behavior. Once Domino’s responded, however, it could do little but hope for the situation to blow over, which, after a short while, it did. “In contrast to United Airlines, Domino’s Pizza faced a highly shocking social media crisis and experienced a major damage to their brand name,” states Malshe, however, neither company was in a position to do much about it. 228 THE PREDICTIVE CASINO As with normal crisis management, the response to a social media crisis should involve the three steps of crisis management—prevention, response, and 218 recovery—and Malshe suggests use of the Hale, Dulek, and Hale model to mitigate crises. In this model, Hale, Dulek, and Hale propose that the response can be divided into four sub-processes: observation, interpretation, choice, and 215 dissemination. Malshe breaks these processes down in the following way : “Observation entails collecting all the relevant information at the onset of the crisis. Interpretation involves assessing information within the context of the current crisis to determine both its accuracy and its relevance. Choice involves assessing viability of different alternative actions and choosing the most appropriate among them. Finally, dissemination leads to information exchange with the public.” According to Malshe, the four-step response to a social media crises is • • 215 : Soft crises and firefighting: For soft crisis and firefighting, Malshe suggests implementing Hale, Dulek, and Hale’s spiral crisis response communication model. In this model, the four steps of the crisis communication—observation, interpretation, choice, and 215 dissemination—are repeated one after another in a kind of circle. “Such a model is most appropriate for the crises with high controllability,” Malshe argues, adding that, “By actively engaging in the four steps, crisis can be managed more effectively and prevented 215 from growing into a disaster.” Wait and watch disaster: For the two crises situations where controllability is low, Malshe contends the spiral communication 215 model is redundant. By constantly changing the response based on developments within various social media avenues, an organization is very likely to make a bad situation worse. Therefore, it is more appropriate to select a linear model that follows Hale, Dulek, and Hale’s four steps—observation, interpretation, choice, and dissemination. After the information is disseminated to the public, there may not be a need to actively keep on responding to the crisis as the organization doesn’t have any control over the crisis. Develop a Virtual Social World Presence Although virtual social worlds did hold a lot of promise when they were first introduced several years ago, they have not gained much traction over the last few years. Several high-profile sites, such as Teleplace, have actually shut down, the remnants of it 3D immersive environment is now in the open source software Openqwaq. 219 In her article Why Virtual Worlds Suck for Business—and Some Solutions 229 , ANDREW PEARSON Maria Korolov makes a good case as to why virtual worlds haven’t taken off when she discusses a typical virtual world scenario as such: “You spend several minutes (at least!) loading up the software, finding your microphone, adjusting your sound levels, and logging into a virtual scene. If you’re lucky, you have a choice of avatars, none of whom look anything like you. If you’re unlucky, you show up as a big plastic Gumby-like thing or as an ugly woman in orange tights (known as ‘Ruth’ in OpenSim and Second Life). You already feel stupid, and then your hosts show up, also as cartoon characters. You can’t make eye contact with them, you can’t read their body language—if they’re animated at all, it’s usually jerky and inappropriate to the situation. Occasionally, you’ll get the tour from someone in a funky costume, or a cross-dresser, or a robot. Even when they’re dressed in business clothing, the clothes are often too tight, too sexy or otherwise ‘off’.” Korolov goes on to explain that even the act of walking is difficult and, for some 219 reason, each platform does it differently. This would all be worth it if there was something worthwhile to see while you were in these immersive 219 environments, but all you get “is a screen hanging on a wall.” The tour guide explains that you can put anything on the screen, some video, a webpage, even share a desktop with some of the enterprise platforms but what’s the use of that, Korolov asks, adding sarcastically, “You already own a screen you can put anything on, it’s your own computer screen. And it already does all of those 219 things.” The argument that these worlds are so compelling because they are 219 “immersive” as Second Life likes to point out is specious at best. “The environments are bland. The clothes are bland. All virtual conference rooms look the same, anyway. And there’s nothing to do there, except walk around and look at screens. You are not immersed. And if you’re not immersed, than [sic] the virtual environment is, in fact, nothing more than a chat room with 219 bad 3D graphics.” Korolov does, however, make some compelling arguments for companies to use virtual social worlds as virtual work environments—fewer centralized offices, more telecommuting, lower facilities budgets, lower travel budgets, and shorter commutes that result in a better work-life balance, and, probably 219 higher employee retention. “If a company is able to create a successful, engaging and immersive virtual workplace for its employees and managers, it will also be able to recruit from a wider pool of potential workers,” offers 219 Korolov. As explained in chapter four, websites like YY.com have created powerful 230 THE PREDICTIVE CASINO business models that are resulting in highly profitable virtual world endeavors. YY’s Music platform allows users to choose from thousands of live performances and each performer has his or her own theatre “in which fans’ avatars cluster in seats around the main stage. A live video feed of the 220 performer rises from the middle.” “Users can chat with the performer and buy all sorts of virtual gifts for her; their avatars hurl the favors onto the stage. (The performer gets pretty much the same view, along with some administrative controls.) The performances can seem something like a 220 combination of a pop concert and a peep show.” YY Music fills a need for both parties involved; the singers get an avenue to reach his or her fans and the audience gets to enjoy a concert-type of experience at a much lower price than they would spend attending a real concert. The audience also gets something they’d never get at a real concert, the ability to reach out and connect with the artist. Virtual worlds like YY.com and stageit.com provide virtual venues where musicians can connect with a virtual audience and actually make money. In its 221 “Sample show idea” section, Stageit suggest the following shows that a performer can offer: 1. 2. 3. 4. 5. 6. 7. 8. Interactive backstage show—let fans hang out in an intimate setting, where they can hear some songs played, get questions answered, and get a front row seat to a backstage experience they can’t get anywhere else. Fundraiser—raise money for an upcoming tour, record, or charity that by telling fans to head to Stageit, where they can win prizes, and/or giveaways. Live Q&A—hang out and answer questions for fans. It is an environment that can foster valuable feedback on what fans want to see and hear on an upcoming tour. Musicians will get free, instant data as well as any revenue generated from ticket and/or album sales. CD listening party—preview upcoming album songs for fans at home. Stream directly from a venue. Backstage + Venue stream—Give fans a backstage experience unlike any other by streaming backstage before the show. Take viewers on a tour of the venue, then stream live from the set, even following the artist backstage after the show. Tour rehearsal—Give fans a sneak peak of an upcoming tour by bringing them into a final rehearsal or practice session. Stream your venue sound check—Give fans an inside look at a soundcheck process, even allowing the online audience to make requests. 231 ANDREW PEARSON Discover a Customer’s Psychological Profile Since every “Like”, every purchase made, every video watched, every cellphone movement, and every website visit is logged somewhere on some server and they are all analyzable. As Hannes Grassegger and Mikael Krogerus explain in 222 their Das Magazin article I Just Showed That the Bomb Was There , “Psychologist Michal Kosinski developed a method of analyzing people’s 222 behavior down to the minutest detail by looking at their Facebook activity.” 222 According to Grassegger and Krogerus : “Psychometrics, sometimes also known as psychography, is a scientific attempt to ‘measure’ the personality of a person. The so-called Ocean Method has become the standard approach. Two psychologists were able to demonstrate in the 1980s that the character profile of a person can be measured and expressed in five dimensions, the Big Five: Openness (how open are you to new experiences?), Conscientiousness (how much of a perfectionist are you?), Extroversion (how sociable are you?), Agreeableness (how considerate and cooperative are you?), and Neuroticism (how sensitive/vulnerable are you?). With these five dimensions (O.C.E.A.N.), you can determine fairly precisely what kind of person you are dealing with—her needs and fears as well as how she will generally behave. For a long time, however, the problem was data collection, because to produce such a character profile meant asking subjects to fill out a complicated survey asking quite personal questions. Then came the internet. And Facebook. And Kosinski.” In 2008, With a fellow Cambridge student, Kosinski created a small app for Facebook called MyPersonality that asked users a handful of questions from the Ocean survey and they would receive a rating, or a “Personality Profile” 222 consisting of traits defined by the Ocean method. The researchers, in turn, got the users’ personal data, which soon amounted to millions and millions of 222 people. “It was, literally, the then-largest psychological data set ever 222 produced.” In the ensuing years, Kosinski and his colleagues continued the research; “first surveys are distributed to test subjects—this is the online quiz. From the subjects’ responses, their personal Ocean traits are calculated. Then Kosinski’s team would compile every other possible online data point of a test subject— what they’ve liked, shared, or posted on Facebook; gender, age, and 222 location.” Once the researchers dug into the data, they discovered that amazingly reliable conclusions could be drawn about a person by observing 222 their online behavior. For example, “men who ‘like’ the cosmetics brand MAC are, to a high degree of probability, gay,” which isn’t that surprising, but 232 THE PREDICTIVE CASINO there are other interesting findings, such as one of the best indicators of 222 heterosexuality is liking Wu-Tang Clan. Also, people who follow Lady Gaga are most probably extroverts, while someone who likes philosophy is probably 222 an introvert. In the ensuing years, Kosinski and his team continued, tirelessly refining their models. “In 2012, Kosinski demonstrated that from a mere 68 Facebook likes, a lot about a user could be reliably predicted: skin color (95% certainty), sexual orientation (88% certainty), Democrat or Republican (85%),” explains 222 Grassegger and Krogerus. Level of intellect, religious affiliation, alcohol-, cigarette-, and drug use could all be calculated as well, something that a casino company might find quite interesting as there are strong correlations between 222 alcohol and/or drug abuse and gambling problems. Employee Facebook pages could be scanned to screen out problem candidates as well. As Kosinski continued refining his model, he discovered that with a mere ten “likes” as input, his model could appraise a person’s character better than an 222 average coworker. With seventy, “it could ‘know’ a subject better than a friend; with 150 likes, better than their parents. With 300 likes, Kosinski’s machine could predict a subject’s behavior better than their partner. With even more likes it could exceed what a person thinks they know about 222 themselves,” which is a pretty frightening thought. The day Kosinski published his findings, he received two phone calls, both from 222 Facebook; one a threat to sue, the other a job offer. Since the publication of Kosinski’s article, Facebook has introduced a differentiation between public and private posts so the data isn’t as easily 222 accessible now. In “private” mode, “only one’s own friends can see what one likes. This is still no obstacle for data-collectors: while Kosinski always requests the consent of the Facebook users he tests, many online quizzes these days demand access to private information as a precondition to taking a personality 222 test.” 222 Kosinski and his team are now adding variables beyond Facebook “Likes”. Offline activity is now traceable and “motion sensors can show, for example, how fast we are moving a smartphone around or how far we are traveling 222 (correlates with emotional instability).” Flipping this idea on its head, Kosinski speculated his research could become a 222 search engine for people. By using all of this data, psychological profiles 222 could not only be constructed, but they could also be sought and found. For example, if a company, or a politician, wants to find worried fathers, or angry introverts, or undecided Democrats, for example, these profiles could be 222 uncovered in the data. To Kosinski’s chagrin, one company he had been partnered with—Cambridge 233 ANDREW PEARSON 222 Analytica—was involved with Donald Trump’s 2016 presidential election. Cambridge Analytica bought up extensive personal data on American voters: “What car you drive, what products you purchase in shops, what magazines 222 you read, what clubs you belong to.” Voter and medical records were 222 purchased as well. In America, detailed personal consumer data is available for a price and Cambridge Analytica snapped it up and they crosschecked these data sets with 222 Republican Party voter rolls and online data such as Facebook likes. Ocean personality profiles were built from this data and, from a selection of digital signatures there suddenly emerged real individual people with real fears, 222 needs, and interests—and home addresses. Today, Cambridge Analytics has assembled psychograms for all adult US citizens, 220 million people, and they have used this data to influence electoral outcomes, as was seen with the 2016 222 U.S. Presidential election. “Trump’s conspicuous contradictions and his oft-criticized habit of staking out multiple positions on a single issue result in a gigantic number of resulting messaging options that creates a huge advantage for a firm like Cambridge Analytica: for every voter, a different message,” explains Grassegger and 222 Krogerus. Mathematician Cathy O’Neil notes that Trump is like a machine learning 222 algorithm that adjusts to public reactions. On the day of the third 2016 presidential debate, “Trump’s team blasted out 175,000 distinct variations on 222 his arguments, mostly via Facebook,” an astounding number of unique ads. “The messages varied mostly in their microscopic details, in order to communicate optimally with their recipients: different titles, colors, subtitles, with different images or videos” were utilized, explains Grassegger and 222 Krogerus. Beyond towns, city districts, apartment buildings, and even individual people 222 could be targeted, explains Grassegger and Krogerus. Blanket advertising— the idea that a hundred million people will be sent the same piece of marketing collateral, the same television advert, the same digital advert—is no more, note 222 Grassegger and Krogerus. Micro and personalization targeting has reached the point where companies can advertise to a market of one. Cambridge Analytica separated the entire US population into 32 different 222 personality types, and focused their efforts on only seventeen states. “Just as Kosinski had determined that men who like MAC cosmetics on Facebook are probably gay, Cambridge Analytica found that a predilection for American222 produced cars is the best predictor of a possible Trump voter.” Among other things, this kind of information helped the Trump campaign focus in on what messages to use, and where to use them, perhaps even what channel to use 222 them on. In effect, the candidate himself became an implementation 234 THE PREDICTIVE CASINO instrument of the model. 222 As Grassegger and Krogerus note, the first results seen by Das Magazin were amazing: psychological targeting increased the clickthru rate on Facebook ads by more than sixty percent. And the so-called conversion rate (the term for how likely a person is to act upon a personally-tailored ad, i.e., whether they 222 buy a product or, yes, go vote) increases by a staggering 1400 percent.” Now, what does all of this mean for an IR? How can they use “Likes” to gain a deeper understanding of their patrons? Well, potentially, by analyzing Facebook Likes, an IR could be able to predict how open, conscientious, outgoing and neurotic an individual user and/or patron was. It could be as simple as doing a Facebook graph search of “Pictures liked“ or “Videos liked” and/or “Stories Liked” with the patron’s name. In addition to predicting a user's personality, these tests could estimate a user/patron's age, relationship status, intelligence level, life satisfaction, political and religious beliefs, and education. An IR’s HR department would also find these personality test results interesting as matching a candidate with jobs based on their personality might make more sense than the current scattershot approach HR often takes in hiring. These personality tests could also reveal troubling traits that should not be ignored. Discover Important Brand Trends As Bifet and Frank explain in their paper Sentiment Knowledge Discovery in 223 Twitter Streaming Data , Twitter is a: “potentially valuable source of data that can be used to delve into the thoughts of millions of people as they are uttering them. Twitter makes these utterances immediately available in a data stream, which can be mined for information by using appropriate stream mining techniques. In principle, this could make it possible to infer people’s opinions, both at an individual level as well as in aggregate, regarding potentially any subject or event.” 224 Services offered by companies like Rival iQ can track a list of brands of one’s choosing and monitor their activity on Facebook, Twitter, and Google. Rival IQ could not only provides insight into an IR’s competitor, but also insight into an industry as a whole. For instance, IR’s could learn from the “Day of the Week” chart when content from the casino and hospitality industry is most likely to go viral. 225 Buzz Sumo also has a search tool that tracks the most popular content on any given topic or website and ranks it according to shares on Facebook, Twitter, LinkedIn, and Google. Later in this next chapter, I will discuss the importance of sentiment and influencers. 235 ANDREW PEARSON As Bifet and Frank note, “There are also a number of interesting tasks that have been tackled using Twitter text mining: sentiment analysis, classification of tweets into categories, clustering of tweets and trending topic detection. 223 Considering sentiment analysis.” O’Connor et al. found that surveys of consumer confidence and correlate with sentiment word frequencies in tweets, 226 and they proposed text stream mining as a substitute for traditional polling. Jansen et al. discuss the implications for organizations of using micro-blogging as part of their marketing strategy. Free sentiment analysis services like twittersentiment.appspot.com can be used to analyze a company’s sentiment. Engage Customers and Potential Customers Social media has really upped the ante when it comes to customer engagement because, through these channels, customers can not only connect with the brand they like, but also with other people who like the brand. What is new now “is that customer engagement is not just a brand's connection with the customer. It is also the customers' engagement with one another in the 92 horizontal, viral aspects,” argues Macy and Thompson. It is these horizontal and viral aspects that can be so important—and, potentially, so lucrative. “Tactically, brands must begin writing and publishing content with embedded links to other content, pictures, and videos to meet the expectations of the 92 online audience,” recommends Macy and Thompson. This “encourages 92 engagement and facilitates sharing.” Conversely, Macy and Thompson warn that a “lack of engagement limits brand leadership effectiveness and ultimately 92 defeats the purpose of the medium.” “A smart combination of listening to the online conversation already taking place, learning what people want, and then providing what they are open to receive from the brand constitutes the winning ticket,” advises Macy and 92 Thompson. Whether the engagement is through video, online polls, games, photo sharing, e-mail, blogging, PowerPoint presentations, or podcasting, “engagement strategies present an opportunity for brands to align content creation for social media with a company's priorities and involve cross92 functional interaction and collaboration,” state Macy and Thompson. “Social media engagement can also be used for front-end campaigns and appearances to help guide the conversation and generate buzz,” conclude Macy and 92 Thompson. To engage customers and potential customers, an IR should set up blogs that provide customers a place to comment on their products and/or services. This helps an IR stay connected with its customers. An environment that keeps the dialogue open and honest should be fostered; even criticism should be welcomed and the blog should not be censored, except for comments that are out-of-bounds or indecent, of course. 236 THE PREDICTIVE CASINO IRs should also keep a keen eye on other blogs and websites where their customers might be blogging as this can lead to criticism that is more open and honest—and, potentially, more helpful—than criticism on the company’s own social media websites. The instantaneous nature of the blogosphere also allows companies to counteract any negative press that might be slung their way. This could be especially useful to a casino that is hit with negative publicity or product recalls as the blogosphere allows them to immediately present their side of the story. Today, many companies use blogs to update their employees, their customers, and their shareholders about important company developments, but anyone planning to use a blog for business should be aware that blogs do come with their own built-in risks. Customers who are dissatisfied with a company’s products or services may decide to engage in virtual complaints in the form of protest websites or blogs, which could result in potentially damaging information being released online, where it can easily be read and picked up by 23 any customer a company is trying to court. The mobile phone has added a new element to the world of blogging as well. Moblogging, m-blogging or phone blogging are blogs created or updated from a mobile phone, a tablet or a phablet. With moblogging, text updates can be sent via SMS or email from a mobile phone, while photographs and/or video files can be uploaded using the mobile device’s camera feature. 200 Jantsch’s Hierarchy of Social Media is a good framework to consider here. Once you have set up your blog, it’s time to move up to the RSS step, where aggregating and filtering content takes place. RSS can help an IR repurpose and republish content. Content can be spread far and wide by leveraging it into other social media channels. “For those looking to feed an RSS of a blog straight to Twitter, Facebook, or LinkedIn profiles, Twitter Feed has you covered. Simply enter your feed, connect your social accounts, and send your posts complete 227 with tracking tools for follow-up.” TweetDeck is another service that allows rapid sharing of content; it is a free service owned by Twitter. The website ShareRoot has several tools that can help boost Pinterest 227 engagement, promotion, and it measures engagement as well. Its Pinterest Board Cover Creator lets users create images to use as the cover for their 227 different pin boards, which can help them stand out. Another good example of excellent customer engagement is for the camera manufacturer GoPro, who “launched its channel four years ago and started producing videos as a way to promote its cameras. But the popularity of the campaign took [on] a life of its own, and now GoPro is taking a cue from its customers and producing many of its own videos as well. The change has 228 tripled their amount of views so far this year.” GoPro’s channel is now the 228 most popular one on YouTube. 237 ANDREW PEARSON GoPro’s videos “include the exciting dramatic events like water sports or bungee jumping, but people also used the camera to capture some emotional and personal moments up close, such as a student graduating high school or 228 their baby falling asleep.” “What they were capturing in videos were things we could not imagine to go out and shoot. They were doing things with cameras better than anything we could script,” said Adam Dornbusch, GoPro's senior director of content 228 distribution. “They capture stuff we never even thought of. They're taking us 228 to the next level of what's possible with what they're capturing.” Toward the end of 2013, GoPro realized the inherent potential in this social media marketing opportunity and so launched a couple of submitted videos 228 that received a large number of hits, Dornbusch explained. “We realized there was such an opportunity here. The users were not only using our cameras but were attaching our name (in the titles and descriptions) to incorporate with us,” Dornbusch said. “Sometimes I'll get a call and someone will say, ‘That's a great video you guys did.’ We have nothing to do with that. 228 People just want to be associated with our product,” explained Dornbusch. Suddenly GoPro's YouTube strategy was at a much larger scale than they had 228 even realized, said Dornbusch. “GoPro began multiplying the amount of videos they released going from a few a week—with only one or two submitted 228 from the public a month—to as many as four a day.” Of the 823 videos on its 228 page, 359—or 44%—were made since October. “We get users to submit content and are able to distribute it to as many 228 people as we can over the world,” Dornbusch said. “They are using their 228 [cameras to capture] their greatest passion.” GoPro has also announced a new “bonus program” for users whose videos are 228 licensed by GoPro. They will receive $1,000 if their videos reach a million 228 views, Dornbusch said. This is a very cheap way to incentivize an audience, who hardly needs to be incentivized it seems, but some of these videos do require substantial organization and expenditure. “We want to work with users on a much larger scale,” Dornbusch said. “This is the heart of GoPro. We love it. It's a surprise. We want to nurture it as much as 228 possible.” Casino operators obviously are not high-end, action camera sellers, but they are selling experiences and anything they can do to help patrons share their experiences should be fostered, after all most people love to snap pictures and show off photos of their vacation experiences. Harvest Customer Feedback No other media even comes remotely close to the data measurement capacity that mobile offers, which begins with exposure to the advertisement, followed 238 THE PREDICTIVE CASINO by the persuasive effect of the advertisement and, finally, to the actual 16 purchase of a product. Today, in many cases, it is usually possible to tell if a comment, a “Like”, a vote, or a blog, is coming in through a mobile or a social media platform. Facebook, YouTube, Foursquare, Twitter, WeChat, QQ, and a whole host of other apps are available for the mobile platform and it is the content that is of the utmost importance, not the platform from where it came. Mobile analytics—the use of data collected as a visitor accesses a website from a mobile device—can effectively track unique visitors, as well as reveal a mobile user’s network, device and location. With site analysis added to a mobile analytics service, marketers can capture mobile metrics such as link tracking for campaign analysis and page tracking for site analysis. Data collected as part of mobile analytics typically includes information such as page views, length of visit as well as such mobile-specific information as mobile device, mobile network operator or carrier, country where the mobile user is calling from, language of the caller, and a unique user ID, which is required because http cookies and JavaScript do not work reliably on mobile browsers. As He et al. argue in their article Social Media Competitive Analysis and Text 229 Mining: A Case Study in the Pizza Industry , “The wide adoption of social media tools has generated a wealth of textual data, which contain hidden knowledge for businesses to leverage for a competitive advantage.” By digging through the vast amounts of unstructured social media data, businesses can 229 discover important brand information. “Decision makers can also use the findings to develop new products or services and make informed strategic and 229 operational decisions.” Text mining is “an emerging technology that attempts to extract meaningful 229 information from unstructured textual data.” It is a form of data mining that attempts to find patterns, models, and/or trends in either structured or unstructured data such as text files, HTML files, social media files as well as a whole host of other proprietary files. Solutions such as SAP’s Infinite Insights, SAS’s Enterprise Miner, SPSS Modeler, and R can be used in the text mining process and they “use sophisticated computer paradigms including decision tree construction, rule induction, clustering, logic programming, and statistical 229 algorithms to find insights and patterns from unstructured textual data.” The He et al. study “examined the social media sites of the three largest pizza chains and applied text mining to analyze unstructured text content on their 229 Facebook and Twitter sites.” The He et al. study attempted to answer the 229 following questions : • What patterns could be found from their Facebook sites respectively? 239 ANDREW PEARSON • • What patterns could be found from their Twitter sites respectively? What were the main differences in terms of their Facebook and Twitter patterns? The results of the study revealed that the three largest pizza chains—Pizza Hut, Domino’s and Papa John’s—are all active in social media and have committed 229 substantial resources for their social media efforts. The data showed that each pizza chain was committed to providing a delightful experience to its customers. For example, if customer questions could not be answered immediately, the pizza chain’s representatives quickly apologized and directed customers to a toll-free telephone number or customer service for further 229 assistance. Of the three chains, the one with the smallest market share—Domino’s Pizza— demonstrated a higher level of commitment and consumer engagement than 229 the other two through the number of social media posts and user comments. Perhaps the fact that they were so burned in social media previously also had a little to do with it? In particular, Domino’s Pizza responded to user comments more quickly, which, the authors believe, reflects their strong efforts in 229 monitoring and handling their social media activities. In addition, the study found that user engagement levels on Facebook were 229 much higher than on Twitter. Not only are there more Facebook fans than Twitter followers, but “the three pizza chains offered more promotional and 229 user engagement activities on Facebook than on Twitter. It was the differences in the platforms that was the key; “Facebook allows people to stay connected and supports more active user participation; Twitter is mainly used 229 for submitting concise updates and noteworthy information.” The study demonstrated the importance that social media had become in the field of customer engagement for each of these three pizza chains. “Specific staff members have been assigned to engage customers and monitor the 229 content that customers created in their social media applications.” Each chain has used social media as “an additional customer services and communication tool to gain insight into consumers’ needs, wants, concerns and 229 behaviors in order to serve them better.” These pizza chains are using social media to listen and engage with their customers, handling customer 229 suggestions and complaints. This is something every casino operator should be doing. Social media is also being used for competitor analysis. “Social media competitive analysis allows a business to gain possible business advantage by analyzing the publicly available social media data of a business and its competitors. A business can compare its social media data to the social media 229 data of their competitors to gain perspective on their performance.” He et al. conclude the paper by offering up the following recommendations 240 THE PREDICTIVE CASINO when it comes to companies establishing a social media monitoring and 229 competitive analysis strategy : • • • • Constantly monitor your own social media presence and your competitors’ social media presence. Establish competitive benchmarking. Mine the content of social media conversations. Analyze the impact of social media findings and events on your business. All-in-all social media has become a great place for companies to gain a real competitive advantage. “Correlation between social media findings (consumer sentiments and opinions) and events (e.g., price changes, rival’s promotional activities) and structured data like sales data need to be examined to understand how competition affects business and provide [sic] information for 230 decision making,” state Dey et al. Reputation Management Managing one’s reputation has always been a difficult and complex task, but it has become even more challenging because the Internet has made it so easy to 231 search for the reputation of a company or a person. As Alison Woodruff argues in her article Necessary, unpleasant, and disempowering: reputation 231 management in the Internet age , the Internet “creates a potentially permanent record of people’s alleged or actual actions that is readily accessible throughout much of the world. Online damage to one’s reputation can translate into offline harm, limiting an individual’s opportunities to find a job, attend college, or establish social relationships.” According to Wikipedia, reputation management is the 232 : “Understanding or influencing of an individual's or business's reputation. It was originally coined as a public relations term, but advancement in computing, the Internet and social media made it primarily an issue of search results. Some parts of reputation management are often associated with ethical grey areas, such as astroturfing review sites, censoring negative complaints or using SEO tactics to game the system and influence results. There are also ethical forms of reputation management, which are frequently used, such as responding to customer complaints, asking sites to take down incorrect information and using online feedback to influence product development.” Sites like Yelp, Angie’s List, and Ripoff Report have become critical tools for 231 consumers to choose a particular business or service. Reviews have become st the new advertisements in this 21 Century world. In some unfortunate cases, 241 ANDREW PEARSON reputation harm can even be dangerous. For example, when Korean hip hop artist Daniel Lee was wrongly accused of diploma falsification, outraged 233 netizens threatened to kill Lee and his family. Of course, this might also show the cultural differences between Asian and the US as I’m not sure diploma falsification would be reason enough to warrant death threats against a hip hop artist in America, it might actually be seen as a bragging right or a badge of honor. At minimum, reputational harm can cause emotional trauma; at worse, 234 it can even lead to suicide. Of course, uncovering nefarious individuals can 234 be a benefit to society as well. Casinos can utilize social media to better understand its customers and many companies do provide these kinds of social media services. 235 In his article A 5-step Guide to Reputation Management Using Social Media , Brynley-Jones lays out a guide to social media reputation management on a budget: 1. 2. 3. 4. 5. Decide what you want to track—define the “keywords” relating to your company that you want to track in online conversations, including: ◦ Company name ◦ Company website address ◦ Names of products ◦ Names of senior employees and directors ◦ Names of close competitors ◦ Common expressions—e.g., “[Company] sucks”, “company is great.” Set up accounts on free social media monitoring tools, including such sites as Google Alerts, Social Mention, Whostalkin.com, Hootsuite, Trackur, Viral Heat, Scout Labs and Vocus, BackType, Blogpulse, Monitter, Tweetbeep, Wholinkstome, BoardTracker, and Naymz. Set up your alerts and searches through services like Google Alerts or Netvibes, as well as RSS feeds that notify you when your keywords are mentioned. Set up your own social media accounts. A service like knowem.com allows businesses to search over 500 popular social networks, over 150 domain names, and the entire USPTO Trademark Database to instantly secure your brand. Engage—How you respond to comments and posts made about your company is purely up to you, but there are certain rules of thumb: ◦ Act quickly—take advantage of social media’s most important quality—its real-time accessibility. Misconceptions can be snubbed out instantly with quick and factual replies. Never expect things to just disappear. Blog posts and forum comments can 242 THE PREDICTIVE CASINO ◦ ◦ remain in search engine results forever, so you need to make sure your viewpoint is there too. Be nice—the first instinct might be to get defensive and emotional, but let clearer heads prevail. Getting angry and/or making threats will likely backfire. “Try and reason with detractors and understand where they are coming from. By showing that you’re listening, you’ll win respect and support from others” (Brynley-Jones, 2009). Be pro-active—when industry-specific discussions arise, get involved early and often. Offer your personal perspective as this can encourage promoters to back you, while also diffusing a potential detractor’s ire. There are several reputation management systems out there, including Klout, 236 which, according to Wikipedia uses : “Twitter, Facebook, Google+, LinkedIn, Foursquare, Wikipedia, and Instagram data to create Klout user profiles that are assigned a unique ‘Klout Score’. Klout scores range from 1 to 100, with higher scores corresponding to a higher ranking of the breadth and strength of one's online social influence. While all Twitter users are assigned a score, users who register at Klout can link multiple social networks, of which network data is then aggregated to influence the user's Klout Score.” In her article How Your Content Strategy Is Critical for Reputation Management, Rebecca Lieb argues that, “Monitoring—and addressing—online reputation issues boils down to search engine optimization. Creating, disseminating and promoting strong, credible, positive content is pretty much the only weapon at 237 a marketer’s disposal.” In (American) football there is saying, “The best offense is a good defense” and this should be kept in mind when it comes to online reputation management. Businesses should start with a content strategy and a content marketing plan 237 already in place. It’ll be too late to assemble the tools you needs to douse 237 the blaze once a fire has been set. “Rather, you not only want those tools in place, you want to have already constructed fortifications in the form of plenty of optimized content on the web in general, as well as on blogs, social media 237 and social networking sites,” Lieb contends. “It’s also critically important that all online content and digital communications be optimized for search. This includes PR, marketing, and investor relations, as well as any other digital content available on the web, anywhere. Optimized text, images, audio, and video results in more content in search engine results 237 pages.” Fake news was a huge problem in the 2016 U.S. Presidential election and that showed just how insidious it can be if not countered quickly with facts. Even then, the damage isn’t always mitigated quickly. 243 ANDREW PEARSON Businesses should also understand that content marketing is an ongoing 237 process and it should be budgeted for accordingly. “You’ll always have to 237 continue what journalists have long called ‘feeding the beast,’” but this can actually be a good thing as getting content into the hands of journalist could help disseminate news about the company into interesting and unique channels. Continuing with the sports analogies, even the best baseball player in history struck out more times than he hit home runs and so it will be with reviews; a business just can’t expect to have nothing but happy customers. “Online reputation management isn’t about obliterating any negative mention or association made with your organization, rather by mitigating those negative 237 results with strong, positive, visible, and consistent content,” the company will be able to build a strong and worthy reputation argues Lieb. Social Shopping According to Webopedia 238 : “social shopping is a slang phrase used to describe networked shopping. In electronic commerce it refers to consumers who use social networking services and sites to share their latest purchases, deals, coupons, product reviews, want lists, and other shopping finds. Some people may use affiliate links when they write about shopping and products on social spaces including Facebook, Twitter and other social networking services.” Companies can use services like Facebook, Twitter, Weibo and WeChat to talk directly to their customers; Twitter is a specifically useful tool because it can help a business quickly move perishable inventory. Twittering last minute airfares to its Twitter followers helps airlines such as United and JetBlue sell 239 their perishable inventory of seats. The same could be done with tickets for concerts, theater seats, sporting events, or even hotel rooms. By understanding up-to-the-minute supply, as well as predicting upcoming demand for its rooms, an IR can reach out to its social media followers to move rooms quicker than it currently would be able to. 240 In his article Top Tips for Creating an Online Retail Community , Phil Woodward argues that today's online shoppers are more socially active than ever before and they like to participate in conversations that allow them to share their knowledge as well as learn things from others. Today, brands are more willing to communicate with customers in an informal and open way as 240 well. “These off-site communities are powerful in that they give customers a platform to express their opinions, but the next step for retailers is having an 240 online community built directly into their site,” states Woodward. 244 THE PREDICTIVE CASINO Apple has created a community within its online store that allows shoppers and potential customers to ask questions about the brand and its products and it 240 has thousands of active conversations going on at any one time. This service enhances loyalty by giving customers a place to candidly talk about Apple’s 240 products with fellow enthusiasts. The first goal of building an online community should be to engage the audience and encourage repeat visits to the site. “To achieve this, retailers should ensure users can express themselves, that they can easily see what 240 others are up to and that they feel contributions are valued.” Users should be able to create profiles as this lets them feel as if they are a part 240 of a network of like-minded people. “Retailers should let users follow things they find interesting—this is a powerful way for them to create their own network of interests. Regular email updates about their interests will also bring 240 them back to the site.” “Online communities can be miniature societies where shoppers and staff members sit alongside one another to share their wisdom,” Woodward 240 notes. These social environments should be friendly, offering impartial and expert answers to questions from customers and potential customers, while 240 also making the brand seem responsive and approachable. Woodward advises that, “Retailers should ensure staff are available to answer any questions. Swift responses add to the customer's sense that they are participating in a conversation, and support their perception of a retailer's 240 product expertise.” The community can then be used to bridge the gap between online and bricks240 and-mortar stores as well as being an entre into the omni-channel. When a customer purchases something in-store, the retailer's staff are on-hand to assist with questions and/or problems and an online community should 240 replicate this experience. However, it also offers something that can't be replicated: the mix of expert staff and genuine customer opinion, which is “a strong combination that gives shoppers more reason to visit the retailer and 240 makes communities easier to develop,” advises Woodward. “Although the retailer's website is the hub of community activity, it shouldn't end there—spreading the community further allows a wider audience to become involved in the conversation, while also exposing potential new 240 customers to the brand.” Social media sites should be exploited to build and 240 support the community. “Letting users log in to the community with social profiles facilitates the sharing of information, and share buttons ensure users 240 can spread content across their social networks” with ease, contends Woodward. “Retailers should also actively feed content into their own social 240 media sites so that their off-site communities support their on-site one.” This creates a broad base of experience by drawing in insights and opinions from a 245 ANDREW PEARSON wider group of people, thereby creating a strong and knowledgeable network 240 of experts. Other social platforms besides Facebook showed some success in generating 241 orders, as well. “The antiques and collectibles industry was the best example, with Pinterest generating 74% of social orders. YouTube saw success with digital products, services and merchandise, grabbing 47%, 36% and 29% of 241 orders, respectively.” In conclusion, while Facebook was the clear leader, retailers would benefit from using other platforms like Wanelo.com, Fancy.com, Pinterest, Fab.com, Polyvore, Luvocracy, Opensky.com, Faveable, Ownza and Etsy, amongst many others. In China, Weibo and the ecommerce site Taobao have formed a synergistic partnership in which storeowners on Taobao create accounts on Weibo and utilize it as a channel to market their products and communicate with customers. It is reported that Taobao and Sina Weibo share 75% overlapping 242 users among the 500 million users that they have. “A linking of each of their accounts will allow Weibo users to log onto Taobao to make purchase[s] and payment[s] and Taobao users to log onto Weibo to view news and release products. Hundreds of millions of user will be able to perform the functions of social media, online shopping and making payment[s] on two platforms 242 without switching account[s]” A comparable situation in the US would be for users to be able to access and purchase products on eBay from their Twitter account with a PayPal account that was integrated into their Twitter account; all-in-all some pretty powerful stuff. “A deep integration of Sina Weibo and Taobao is about taking advantage of the four elements: account, marketing, data and mobile,” according to 242 Advangent. Since the partnership in April 2013, user-centric product recommendations have started to appear on Weibo, these allow users to see 242 Taobao products that might interest them. Weibo and Taobao have also considered other measures to keep their social shopping experience robust and engaging. On the one hand, they have launched discount promotions targeting the seller’s Weibo fans through a series of “Fan festivals” that happen to coincide with important events; while on the other hand, Taobao and Weibo are tracking the user’s comments very closely—too many negative reviews and 242 complaints will get sellers blacklisted and punished. This is almost a perfect scenario for sellers as it gives them a far-ranging platform to not only sell their wares on but also gain instant social media feedback—it’s almost word-of-mouth marketing on steroids. Casinos can take advantage of this market and sell hotel rooms and IR experiences through these channels, thereby connecting with Chinese consumers directly. The B2C marketing cloud company Emarsys believes that social commerce will 246 THE PREDICTIVE CASINO soon become a mainstay channel for consumers. Although many retailers don’t know exactly what to do with social commerce yet the social channels are boldly leading the way. In 2015, Pinterest launched the ‘Buy it’ button and Instagram created an expanded ad program called ‘Instagram Ads.’ At the same time, other social media channels such as Twitter and YouTube are exploring ‘Buy It Now’ features and a multitude of mobile apps are popping up that allow easy e-commerce purchases. And why not? It makes complete sense to simplify the purchasing process and Facebook is already jumping into the fray as a new service inside Facebook Messenger allows users to readily swap money with others. Casinos should embrace this bold new frontier and the Predictive Casino would simplify the purchasing process, as well as the purchasing behavior collection process. Social Media Analytics As Melville and Lawrence explain in their article Social Media Analytics: 243 Channeling the Power of the Blogosphere for Marketing Insight , social media analytics is “the practice of gathering data from blogs and social media websites and analyzing that data to make business decisions. The most common use of social media is to mine customer sentiment.” Social media analytics evolved out of the disciplines of social network analysis, machine learning, data mining, information retrieval (IR), and Natural Language Processing (NLP). According to Melville and Lawrence, the automotive analysis of blogs and other 243 social media sites raise the following intriguing marketing questions : 1. 2. 3. 4. Given the enormous size of the blogosphere, how can we identify the subset of blogs and forums that are discussing not only a specific product, but higher level concepts that are in some way relevant to this product? Having identified this subset of relevant blogs, how do we identify the most authoritative or influential bloggers in this space? How do we detect and characterize specific sentiment expressed about an entity (e.g., product) mentioned in a blog or a forum? How do we tease apart novel emerging topics of discussion from the constant chatter in the blogosphere? 244 As Margaret Rouse explains in her article Social Media Analytics , step one of a social media analytics initiative is “to determine which business goals the data that is gathered and analyzed will benefit. Typical objectives include increasing revenues, reducing customer service costs, getting feedback on products and services and improving public opinion of a particular product or business division.” Once these business goals have been identified, “key performance indicators (KPIs) for objectively evaluating the data should be 247 ANDREW PEARSON defined. For example, customer engagement might be measured by the numbers of followers for a Twitter account and number of retweets and 244 mentions of a company’s name,” states Rouse. Through social networks like Twitter and Weibo, organizations can pick up 126 customer satisfaction in real time. “Social media is enabling companies such as Coca-Cola, Starbucks, and Ford to go beyond standard customer satisfaction data gathering to innovate by setting up and participating in communities to 126 gain feedback from customers.” A good example is MyStarbucksIdea.com, it is a website where “Starbucks customers can relate their experiences and offer ideas about how to improve the Starbucks experience, from drinks to foods to 126 ambiance.” Which of the following objecƒves does your organizaƒon seek to achieve by implemenƒng customer analyƒcs technologies and methods with social media data? (Please select all that apply.) Gain deeper customer understanding 56% Idenƒfy customer paths to buying decision 31% Monitor and measure senƒment drivers 30% Determine value of social media engagement to markeƒng campaigns 29% Discover new audience segments 27% Gain insights for new product development 24% Analyze social networks, links, and graphs 22% Differenƒate influencers from followers in social media Increase engagement beyond passive social media monitoring 20% 19% Analyze compeƒƒon's "share of voice" 18% Monitor and analyze social acƒvity in real ƒme Improve "long-tail" analysis of buying by small groups of customers 14% 11% We do not analyze social media data 32% 0% 10% 20% 30% 40% 50% 60% Figure 12: Customer Analytics and Social Media Objectives Based on 1,546 respondents from 418 respondents; a bit more than three responses per respondent, on average. Source: TDWI Research 248 THE PREDICTIVE CASINO When looking at what objectives companies were seeking when implementing customer analytics technologies with social media data (see Figure 12), TDWI Research found that gaining a “deeper customer understanding” topped the 126 list at 56% “Social media listening can provide an unprecedented window into customer sentiment and the reception of an organization’s marketing, 126 brands, and services.” Besides the broad objective of gaining deeper customer understanding, nearly one-third (31%) seek to identify attribution, or paths to buying decisions, which can be done on a limited scale with services like Google Analytics as well as 126 other Web site analysis applications. 30% or respondents sought to discover customer sentiment, which is important because it helps companies discover positive and negative comments in social 126 media, customer comment and review sites. “Sentiment analysis often focuses on monitoring and measuring the “buzz” value, usually through volume 126 and frequency of comments around a topic.” Simply deciding which social media sites’ data to analyze can be one of the biggest challenges facing businesses going down the analytics path. “Organizations have to research where their customers are most likely to express themselves about brands and products. They need to spot influencers who have networks of contacts and take it upon themselves to play an 126 advocacy role.” “About 20% of respondents are interested in differentiating 126 influencers from followers in social media. “Link analytic tools and methods specialize in identifying relationships between users in social communities and 126 enabling organizations to measure users’ influence.” “With some tools, data scientists and analysts can test variables to help identify social communities as ‘segments’. Then, as they implement segmentation models for other data sources, they can integrate these insights with social media network analysis to 122 sharpen models and test new variables,” explains Stodder. A good example of the use of influencers to help in marketing is Ford’s “Fiesta 112 Movement” campaign. Ford generated buzz eighteen months before it reentered the US subcompact-car market with its Fiesta model by “giving 100 social-media influencers a European model of the car, having them complete ‘missions,’ and asking them to document their experiences on various social 112 channels.” The social media campaign included videos on YouTube, which generated more than 6.5 million views, and “Ford received 50,000 requests for 112 information about the vehicle, primarily from non-Ford drivers.” The overall 112 results: in late 2010, some 10,000 cars were sold in the first six days. Analytics are critical in helping organizations “make the right decisions about when, where, and how to participate in social media. It isn’t enough to just listen; organizations must insert themselves and become part of the 126 conversation.” When doing so, however, companies should keep in mind 249 ANDREW PEARSON 89 advice from The Cluetrain Manifesto —“Conversations among human beings sound human. They are conducted in a human voice,” as well as this: “When delivering information, opinion, perspectives, dissenting arguments or 89 humorous asides, the human voice is typically open, natural, uncontrived.” One interesting strategy is for a casino operator to start viral campaigns via Twitter, using hashtags for a topic; the campaign could be a part of a larger marketing strategy. Companies can then “monitor social media to see what people say and analyze how the campaign is playing among influencers and 62 across networks.” For example, Unilever’s Dove brand series of webintegrated commercials attempted to fundamentally redefine the brand and Unilever’s Sunsilk shampoo campaign that, according to the company, placed a “net seed” onto YouTube with the video titled “Bride Has Massive Hair Wig 62 Out.” The video, which showed a bride-to-be reacting in horror to her wedding day hair, contained no brand references and quickly accumulated 62 three million hits. Later, Unilever came forward to claim the ad, saying it was intended to plant the term “wig out” in the culture, a term that was to be used 62 in conventional advertising for Sunsilk products at a later date. Klear, a social media and social data platform that focuses on influencer marketing, offers a product that can help IRs understand the effects of their influencer marketing. Klear’s campaign reports contain the following 245 summaries : • • • • How many influencers participated in the campaign Number of updates the influencers posted during the campaign Engagements metrics Number of people who saw the content The report also includes a drill-down analysis for each and every influencer. For 245 each influencer the report will show : • • • • • • Who is the influencer Their expertise Fanbase across different social networks Top posts during the campaign Engagements for these updates A direct link to the influencer’s profile on Klear This is a paid service, but most of the information is publicly available and this is something an IR could build up in-house, should they want something customized. Influencer marketing taps directly into what Deighton and Kornfeld call the five paradigms of digital interactive marketing, i.e., social exchanges—building identities within virtual communities—and cultural exchanges—firms offering 62 culture products that will compete in buzz markets. This peer-to-peer 250 THE PREDICTIVE CASINO interactivity should motivate the desire to exchange and share information, 62 which should help market any IR event. Social Media Monitoring It is high time to revise Wanamaker’s oft-made quote that he didn’t know which half of his marketing spend was useful (as it is probably the most overquoted quote in the history of marketing) because now we not only have the ability to figure out which advertisement works for which customer, but we can also extrapolate how that advertising will work on customers similar to the ones we might want to target. Today, digital advertising should employ a multi-screen strategy that follows its audience throughout his or her digital day. As previously mentioned, successful mobile advertising requires three things—reach, purity and analytics. Analytics “involves matching users’ interests—implicit and explicit, context, preferences, 16 network and handset conditions—to ads and promotions in real time.” In this chapter, I will address the many reasons and ways a casino operator can utilize analytics to enhance its marketing, campaign management, sales, market research, fraud and risk/management, contact center operations, supply management, as well as for a whole host of other departments and channels. Throughout this chapter, I will also mention some of the analytics services that are offered to businesses, some are free, some are paid. 89 In 1999, The Cluetrain Manifesto warned: “Reviews are the new advertising.” Today, this is truer than ever before. There are a multitude of platforms that allow users to rate or comment on a restaurant, a retail establishment, a hotel, a casino property or even a local handyman or plumber. Used properly, reviews can be the new advertising currency for a casino’s marketing department. Companies such as Dell, Cicso, Salesforce.com, the American Red Cross, and Gatorade are creating Social Media command centers that monitor the social conversations about their companies. These social media centers enable company employees to monitor conversations from the social web on channels such as Twitter, Facebook, and YouTube, amongst others, in an attempt to keep track of the health of a company’s social brand. In December 2010, Dell became one of the first companies to launch a social media command center. Based at company HQ in Round Rock, TX, twelve fulltime employees monitor conversations about Dell and its products around the globe, responding via @DellCares or forwarding the post to the right internal 246 Dell team. Through Dell’s Social Media Listening and Command Center, Dell aggregates and culls through the 25,000 conversations about Dell every day 246 (more than 6 million every year). “We’re monitoring conversations in 11 languages 24/7, and each one is an 246 opportunity to reinforce our brand,” explains Karen Quintos, Dell CMO. 251 ANDREW PEARSON Quintos explains that 246 : “With the tremendous amount of information being generated, we can track basic demographics, reach, sentiment, subject matter of the discussions, the sites where conversations are happening, and more. We leverage these analytics to identify customer support needs as they happen, influence product development, insert ourselves into conversations with IT decision makers and connect with people having the most impact on these conversations.” 247 In his article Taking Back The Social-Media Command Center , Scott Gulbransen argues that, “To do the command-center model right, a setup has to envision a real-time workflow empowered to take action on all of the relevant content being analyzed, whether it be insights derived from real-time monitoring, opportunities to respond, or great discovered content to feature that elevates you and your fans.” Gulbransen recommends breaking down a 247 command center into the following critical functions : 1. 2. 3. 4. Identify trends and insights—track not only the key themes, but also how they evolve over time. Review the content—monitor a wide variety of terms that are meaningful to the brand and assign employees to sort through the responses, deciding which one warrants a response, and what might interest the community at large. Curate the best stuff—leverage the great content that is being said about the company as well as champion those great content providers. Listen and Respond—this is a two-way conversation, listen and respond quickly and accordingly. Macy and Thompson explain that, “Beginning in 2008, the terms sentiment and influence were introduced into the real-time ecosystem to describe how to begin measuring the different aspects of listening. In addition to listening to mentions, companies and brands started to realize that they needed to listen 91 to ‘influencers’, the leading voices whom others listen to.” Companies and brands quickly realized that there were influential voices that should be listened to because they held a lot of weight amongst other members of the 91 business community. Today’s consumers are data producing machines; mobile operators know every phone call an individual makes as well as how long each phone call lasts. If signed up for their rewards program, the supermarket where an individual shops knows what foods she likes to buy and what type of shopper she is. Visa, Mastercard, and American Express know where their cardholders like to drink and dine. Tivo and Netflix know which movies their subscribers prefer to watch. 252 THE PREDICTIVE CASINO Company databases across the world house vast amounts of data about their consumers and this data can help advertisers and marketers shape unique marketing campaigns to target specific individuals with individualized offers. All of this data falls under the umbrella of “customer analytics” and social 126 media is the new frontier for making sense of this data. Here, “customers are influencers, not just generators of sales transactions as seen through point-of126 sale and e-commerce systems.” By sharing their opinions, feelings and disappointments on social networks, customers can “influence each other by commenting on brands, reviewing products, reacting to marketing campaigns 126 and product or service introductions, and revealing shared buying interests.” Unlike casual conversations, comments, updates, likes and dislikes uploaded to social networks are collected and, therefore, analyzable and measureable. This results in “a data tsunami: the actions and content generated by participants in social media create ‘Big Data’ sources that are full of potential for tracking and 126 understanding behavior, trends, and sentiments.” Remember, this can be highly quantifiable data, especially when run through a data warehouse like Hadoop. Casino operators should be studying attribution analysis for its social media campaigns on platforms like Facebook, YouTube, Twitter, Weibo, etc., etc. Getting people to actually state their feelings and opinions about a product is paramount and it can help with attribution analysis, which can reveal such 122 things as what kinds of campaigns most influence customer behavior. In digital advertising, attribution is traditionally done at a user-specific level, where a consistent user identifier can be established across all analyzed events. In traditional media, attribution is generally done at the macro, user-group level, as there is no consistent user identifier available. 248 In its Social Media Analytics: Making Customer Insights Actionable , IBM believes that the “mistake many organizations make is to treat social media as distinct and separate from other customer data and divorced from revenue generating imperatives.” IBM recommends companies venturing into the social 248 media space do the following : • • • Integrate company-wide information from different data sources to drive the business through deeper consumer insight. Define the real value of the company’s brand—its equity, reputation and loyalty—at any moment in time, in any place in the world; and Understand emerging consumer trends, both globally and locally and apply predictive models to determine actions with the highest probability of increasing relevance and maximizing marketing campaign ROI. IBM recommends businesses ask the following questions when devising a social 248 media plan : 253 ANDREW PEARSON • • • Assess—also referred to as “listening,” at this stage a company should monitor social media to uncover sentiment about its products, services, marketing campaigns, employees and partners. The questions that need to be asked at this stage include: o What are you company objectives? Are you looking to: § Attract customers? § Increase the value of existing customer relationships? § Retain customers? o How do customers interact with you today? o What are they interested in? o Where and when do they use social media? o Are there significant influencers who speak to your brand or products? Measure—proactive analytics can uncover hidden patterns that can reveal “unknown unknowns” in the data. Questions that businesses need to ask at this stage include: o Who are you targeting with your social media initiatives and why? o What will you be measuring: § Share of voice § Activation § Brand sentiment § Influencers o Sales over the life of the customer relationship? Integrate—social media can give businesses both a broad view of their operations as well as a detailed and intimate view of their individual customers. Questions to ask at this stage include: o What is your vision for social media and its integration into the company’s operational marketing systems? o Do you have a profile of your customer advocates? Can you predict sentiment on products, services, campaigns? o How do you measure the effects of social media on brand equity and reputation, pipeline, and sales orders and margins? o How will you integrate social analytics into other customer analytics? Regardless of the sophistication and scope of any social media initiative, the end goal, IBM argues, should be in alignment with corporate imperatives and 248 goals as well as produce a measurable ROI. In the hospitality industry, Marriott has taken the lead in implementing a highly 249 social media-savvy and YouTuber-based marketing strategy. According to a 250 company press release , Marriott teamed up “with the Emmy-nominated, 254 THE PREDICTIVE CASINO multi-platform Internet and pop culture media brand What’s Trending and host Shira Lazar to produce and release content videos that showcase the power of 250 mobile check-in from Marriott.” Empowering travelers to experience more in cities around the world, the videos took a tongue-in-cheek approach to capturing the social influencer’s personal journeys thanks to Marriott’s mobile 250 check-in system. One of the YouTube videos, which has been viewed over 4 million times, chronicles a hotel lobby dance party surprising the millionth Marriott mobile 249 app check-in user. In another video, YouTuber Casey Neistat, shares his 249 Marriott travel experience. This resulted in a very organic and genuine 249 campaign that resonated with Casey’s audience. As opposed to a traditional advertising measures, marketing with Neistat ensured “a beautifully captured, 249 day-in-the-life of experience as opposed to a scripted commercial.” In an interview with David Beebe, the VP of Global Creative & Content Marketing for Marriott International, advises that “rather than dictating the topic and type of content you want from the influencers, let them come back to you with 249 a concept that plays to their strengths and social credibility.” The ad specifically showed how mobile friendly Marriott properties have become, even to the point of checking patrons in via mobile. What shouldn’t be lost on the reader is the fact that the use of mobile check-in is actually a winwin for both parties involved. The guest doesn’t have to bother with one of the industry’s big pain points—waiting in the check-in line and it reduces labor needs considerably. The Predictive Casino should be able to understand when rooms become available, i.e., reservation staff gets alerted once the room is cleaned and then the room can be doled out to the most valuable/important members of the hotel’s loyalty programs. If no loyal patrons are waiting, the people who aren’t in the loyalty program but who have provided the hotel with their mobile numbers can be serviced. This would be a recognized customer service upgrade by most patrons. This also allows the casino operator to track the speed at which its maids clean the rooms, which could prove valuable in its own right. This chapter will go into detail about customer analytics, social media analytics and it will conclude with a list of the tools that can be used to build up a casino’s social media monitoring plan. Social Media Monitoring Tools Twitter is an excellent resource to get real-time marketing information. It even has a search function that allows users to “See what’s happening–right now,” as its website claims. An excellent way to get honest and instant, unfiltered, real-time feedback from consumers, Twitter is a service that has never been available in the history of marketing and it is a potential goldmine for customer feedback. Table 11 lists the Social Media Tools and websites available to 255 ANDREW PEARSON business users to track engagement and customer feedback. Name Comments Addict-o-matic Addictomatic searches the best live sites on the web for the latest news, blog posts, videos and images. It’s a social media listening tool to keep up with the hottest topics, perform ego searches and get info on what’s up, what’s now or what other people are feeding on. You can personalize your results dashboard and keep coming back to your personalized results dashboard for that search. News pages provide the latest headlines on topics such as entertainment, politics, shopping, sports and more. Board Reader BoardReader allows users to search multiple message boards simultaneously, allowing users to share information in a truly global sense. Boardreader is focused on creating the largest repository of searchable information for our users. Users can find answers to their questions from others who share similar interests. Our goal is to allow our users to search the “human to human” discussions that exist on the Internet. Buffer Buffer makes your life easier with a smarter way to schedule the great content you find. Fill up your Buffer at one time in the day and Buffer automagically posts them for you through the day. Simply keep that Buffer topped up to have a consistent social media presence all day round, all week long. Get deeper analytics than if you just post to social networks directly. Buzzsumo Analyze what content performs best for any topic or competitor. Find the key influencers to promote your content: • • • • Discover the most shared content across all social networks and run detailed analysis reports. Find influencers in any topic area, review the content they share and amplify. Be the first to see content mentioning your keyword; or when an author or competitor publishes new content. Track your competitor’s content performance and do detailed comparisons. Commun.it Can help you organize, increase, and manage your followers, and can do so across multiple accounts and profiles. At a glance you can see different aspects of your community management, like the latest tweets from your stream and which new followers might appreciate a welcome message. Crowdfire Crowdfire is a powerful phone app and online website that helps you grow your Twitter and Instagram account reach. This tool has a variety of functions designed to understand your social analytics as well as manage your social publishing. Cyfe is an all-in-one dashboard software that helps you monitor and analyze data scattered across all your online services like Google Analytics, Salesforce, AdSense, MailChimp, Facebook, WordPress and more from one single location in real-time. Shows a variety of valuable information related to your Facebook page, such as growth, engagement, service and response time, and of course Karma (a weighted engagement value). FanKarma also provides insight Cyfe Fanpage Karma 256 THE PREDICTIVE CASINO Name Comments into Twitter and YouTube; the latter could be particularly valuable if you're creating a video marketing strategy. Followerwonk Google Alerts Google Trends Hootsuite HowSocialable Iconosquare Klear Klout Kred LikeAlyzer Followerwonk is a cool social media analytics tool thet lets you explore and grow your social graph. Dig deeper into Twitter analytics: followers, their locations, when do they tweet. Find and connect with influencers in your niche. Use visualizations to compare your social graph to competitors. Google Alerts are email updates of the latest relevant Google results (blogs, news, etc.) based on your searches. Enter the topic you wish to monitor, then click preview to see the type of results you’ll receive. Some handy uses of Google Alerts include: monitoring a developing news story and keeping current on a competitor or industry. Trends allows you to compare search terms and websites. With Google Trends you can get insights into the traffic and geographic visitation patterns of websites or keywords. You can compare data for up to five websites and view related sites and top searches for each one. Monitor and post to multiple social networks, including Facebook and Twitter. Create custom reports from over 30 individual report modules to share with clients and colleagues. Track brand sentiment, follower growth, plus incorporate Facebook Insights and Google analytics. Draft and schedule messages to send at a time your audience is most likely to be online. HootSuite has the dashboard for your iPhone, iPad, BlackBerry and Android. Monitor and post to multiple social networks, including Facebook and Twitter. Create custom reports from over 30 individual report modules to share with clients and colleagues. Track brand sentiment, follower growth, plus incorporate Facebook Insights and Google analytics. Draft and schedule messages to send at a time your audience is most likely to be online. HootSuite has the dashboard for your iPhone, iPad, BlackBerry and Android. Key metrics about your Instagram account. Number of likes received, your most liked photos ever, your average number of likes and comments per photo, your follower growth charts and more advanced analytics. Track lead conversations, send private message as on Twitter, and improve communication with your followers. Social media monitoring, analytics and reporting. Influencer marketing, find and create relationships with the top influencers in your sector and build your community. Competitive analysis tracks your social media landscape, see what’s working for them and develop your strategy. Klout’s mission is to help every individual understand and leverage their influence. Klout measures influence in Twitter to find the people the world listens to. It analyzes content to identify the top influencers. Kred is a social-media scoring system that seeks to measure a person’s online influence. Kred, which was created by the San Francisco-based social analytics firm PeopleBrowsr, attempts to also measure a person or company’s engagement, or as they call it, outreach. PeopleBrowsr hopes that that combination can offer a more informed metric for noncelebrities like entrepreneurs and those whom they follow and look to for advice. This Facebook analysis tool comes up with stats and insights into your page and begins every report with a list of recommendations. Keep track 257 ANDREW PEARSON Name Comments of where your Facebook page stands compared to other pages by following the comparison to average page rank, industry-specific page rank, and rank of similar brands. Mention Mention prides itself on “going beyond Google Alerts” to track absolutely anywhere your name or your company might be mentioned online. When you subscribe to Mention's daily email you get all these wayward hits right in your inbox, and the Web dashboard even flags certain mentions as high priority. Mentionmap Explore your Twitter network. Discover which people interact the most and what they’re talking about. It’s also a great way to find relevant people to follow. The visualization runs right in your browser and displays data from Twitter. Mentionmap loads user’s tweets and finds the people and hashtags they talked about the most. In this data visualization, mentions become connections and discussions between multiple users emerge as clusters. Built by the team at Sprout Social, Must Be Present searches your Twitter account to find how quickly you respond to mentions. Their engagement reports place you in a percentile based on other accounts so you can see how you stack up to the speed of others. Must Be Present NeedTagger A super-powered Twitter search tool, NeedTagger runs language filters and keyword searches to determine which Twitter users might need your products or services. The tool shows you real-time search results and sends a daily email digest of new finds. NutshellMail Collects your activity on Facebook, LinkedIn, and Twitter (and even places like Yelp and Foursquare) to provide an email overview of your accounts. You set how often and when you want to receive the recap emails. Put it to use: If you have a weekly metrics plan you can have NutshellMail send a message once a week with an overview of your accounts. You can then extract the data and insights straight into your weekly report. Omgili Omgili helps you find interesting and current discussions, news stories and blog posts. Direct access to live data from hundreds of thousands of forums, news and blogs. Very easy to use, no signup for web interface. Find out how many people are pinning from your website, seeing your pins, and clicking your content. Pick a time-frame to see how your numbers trend over time. Get better at creating Pins and boards with metrics from your Pinterest profile. Learn how people use the Pin It button on your site to add Pins. See how people interact with your Pins from whatever device they use. Get a glance at your all-time highestperforming Pins. Pluggio is a web-based social media tool to help marketers easily grow and manage their social media profiles (Facebook and Twitter). It includes a suite of tools to organize and keep track of multiple accounts, get more followers, and automate the finding and publishing of excellent targeted content. The full set of social media tools. Post content to over 10 social networks with one single click of a button. Get real time click-through statistics with your domain name. Measure and analyze the best results from your social posts. Monitor the social media conversations that are important Pinterest Analytics Pluggio Postific 258 THE PREDICTIVE CASINO Name Comments for your business. Quintly Sentiment SocialMention Quintly is the professional social media monitoring and analytics solution to track and compare the performance of your social media marketing activities. Whether you are using Facebook, Twitter or both, Quintly monitors and visualizes your social media marketing success. Benchmark your numbers against your competitors or best practice examples. Sentiment was born in 2007 and now boasts a team of bright enthusiastic people dedicated to provide the best social customer service and engagement platform for business. SocialMention tracks areas such as sentiment, passion, reach, and strength to not just tell you what's being said about your search but how those reactions feel. While you track your brand or yourself, you can also see how your sentiment changes over time. Social Rank Identifies your top 10 followers in three specific areas: Best Followers, Most Engaged, and Most Valuable. Your most engaged followers are those who interact with you most often (replies, retweets, and favorites); your most valuable followers are the influential accounts; and your best followers are a combination of the two. Social Rank will run the numbers for free and show you the results today, then follow-up each month with an email report. Social Oomph Schedule tweets, track keywords, extended Twitter profiles, save and reuse drafts, view @mentions and retweets, purge your DM inbox, personal status feed — your own tweet engine, unlimited accounts. Keeps track of your hashtag campaign or keyword on Twitter, Instagram, or Facebook with a full dashboard of analytics, demographics, and influencers. This tracking tool Tip Top Topsy TipTop Search is a Twitter-based search engine that helps you discover the best and most current advice, opinions, answers for any search, and also real people to directly engage and share experiences with. A search on any topic reveals people’s emotions and experiences about it, as well as other concepts that they are discussing in connection with the original search. A powerful search engine for Twitter content. Want to know how a certain term is being used on Twitter? You can search links, tweets, photos, videos, and influencers. Twazzup Offers real-time monitoring and analytics for Twitter on any name, keyword, or hashtag you choose. The Twazzup results page delivers interesting insights like the top influencers for your keyword and which top links are associated with your search. Tweepi Has a number of useful Twitter features, many of which fall into a couple categories: managing your followers and supercharging who you're following. For management, you can unfollow in batches those who don't follow you back, and you can bulk follow another account's complete list of followers or who they're following. Tweetcaster A Twitter management tool for iOS and Android devices and provides the basics of what you'd expect from a Twitter dashboard plus a few fun extras: enhanced search and lists, hiding unwanted tweets, and photo 259 ANDREW PEARSON Name Comments effects for your images. Tweetdeck Lets you track, organize, and engage with your followers through a customizable dashboard where you can see at a glance the activity from different lists, followers, hashtags, and more. TweetReach Shows you the reach and exposure of the tweets you send, collecting data on who retweets you and the influence of each. Identify which of your tweets has spread the furthest (and why) and then try to repeat the formula with future tweets. TwitterCounter Twitter Counter is the number one site to track your Twitter stats. Twitter Counter provides statistics of Twitter usage and tracks over 14 million users. Twitter Counter also offers a variety of widgets and buttons that people can add to their blogs, websites or social network profiles to show recent Twitter visitors and number of followers. Provides a snapshot of your Twitter profile and can even track Facebook and Instagram as well. Two of Twtrland's most helpful tools are a live count of how many followers are currently online and advanced search functionality that includes keywords, locations, and companies. Local companies can perform a location search to see which area accounts are most popular and potentially worth following. Twtrland SumAll ViralWoot SumAll is a powerful social media analytics tool that allows our customers to view all of their data in one simple, easy-to-use visualization. Social media, e-commerce, advertising, e-mail, and traffic data all come together to provide a complete view of your activity. Pin Alert feature lets you track what are people pinning from your website, who is pinning the most and what images from your website are trending on Pinterest. Thousands of social media marketers and agencies use Viralwoot for their clients. You can manage & grow multiple Pinterest accounts with a single Viralwoot account. WhosTalkin WhosTalkin is a social media monitoring tool that lets you search for conversations surrounding the topics that you care about most. Whether it be your favorite sports team, food, celebrity, or brand name; Whostalkin will help you find the conversations that are important to you. WhosTalkin search and sorting algorithms combine data taken from over 60 of the most popular social media sites. WhoUnfollowedMe Who.unfollowed.me is a service that helps you track your unfollowers, in real time, without waiting for a DM, or email. It allows you to check your unfollowers on your schedule, every 15 minutes, without waiting for an email or a direct message. Table 11: Social Media Tools Source: https://www.dreamgrow.com/69-free-social-media-monitoring-tools/ Conclusion Social media listening can provide an IR an ongoing real-time window into customer sentiment, as well as give the businesses verifiable information about 122 the company’s marketing campaigns, brands, and services. 260 THE PREDICTIVE CASINO In chapter one, I mentioned Paul Greenberg stated that social CRM was “a philosophy and a business strategy supported by a technology platform, business rules, workflow, processes, and social characteristics, designed to 111 engage and reach accordingly in a collaborative conversation” and the Predictive Casino must add social media elements to its CRM systems to give a complete personalization experience. Continuous customer engagement can be fostered through a multitude of channels and they are cheap to use, if not free, in some cases. The beauty of this system is that it can be a real win-win when it comes to company marketing as customers who are happy with a business’s products and/or services will often comment and blog about the products and/or services they like, while those who aren’t happy with it can be reached out to and, hopefully, converted into satisfied customers. Often, the simple act of responding to a customer’s comments can turn the tide of negativity and, as long as the remedies are constructive, can turn a hostile customer into a positive one, and, possibly, one who might even tout the company’s excellent customer service at a later date. Mullich offers the following tips on how to get the low-down on rivals 1. 2. 3. 4. 133 : Understand that day-to-day online chatter can be misleading, but, over time, An IR can find directional trends important to its business and industry. The deepest insights often come not from general sources, like Facebook and Twitter, but from blogs and forums that are specific to an industry. Think broadly about the nature of one’s “competitors”—sentiment analysis can help a business prepare for unexpected entries that might be preparing to take a piece of its business. Keyword search teams can help. The information you can gain online about competitors is limited, and often must be combined with your own internal data to bring actionable insights. There are, of course, limits to what competitive sentiment analysis can provide. “The challenges you might address, using your company’s own customer, product, and transactional data, are far more extensive than those you can tackle via available competitor data,” says Seth Grimes, an analyst who runs 133 the annual Social Analysis Symposium. “For instance, you’re not going to have access to your competitors’ contact-center notes and warranty claims, or to your competitors’ customer profiles and transaction records. But with your 133 own company’s, you can create some very rich analyses,” he adds. For the above reasons, competitive analysis is usually just one piece of the vast 133 data mosaic. For example, one company that noticed a drop in sales of its 261 ANDREW PEARSON flagship product analyzed online chatter and found customers were talking 133 enthusiastically about a new product a competitor had just released. “When the company analyzed its contact-center data, it found that returns correlated to discontent about an attribute its own product lacked, but the new 133 competing product offered.” The company was quickly able to identify the problem and by using a combination of competitive sentiment analysis, discovery from its own internal data, it was able to tweak its own product to 133 make it much more competitive. As Grimes says, “Sentiment analysis can help you understand how the market perceives you and your competitors’ products and services, but keep in mind that sentiment is only an indicator, useful in measuring and projecting market 133 impact, not a substitute for strong human judgment.” There is a dark side to all of this tracking as the case of IFA and Shopsense showed. Health Insurer IFA Insurance teamed up with Shopsense, a grocery 251 chain in Midwest America, and bought their loyalty card customer data. The insurance company discovered some intriguing patterns in the loyalty card data, such as the correlation between condom sales and HIV-related claims, for 251 example. It also discovered such things as households that buy cashews and 251 bananas quarterly are the least likely to develop symptoms of Alzheimer’s. Although this information did prove to be highly profitable for IFA, I believe it is a clear violation of customer trust and privacy. As Katherine Lemon explains in her article, How Can These Companies Leverage 252 the Customer Data Responsibly , “Customer analytics are effective precisely because firms do not violate customer trust. People believe that retail and other organizations will use their data wisely to enhance their experiences, not to harm them. Angry customers will certainly speak with their wallets if that trust is violated.” Another concern for consumers is what Lemon calls “battered customer 252 syndrome.” She explains that, “Market analytics allow companies to identify their best and worst customers and, consequently, to pay special attention to 252 those deemed to be the most valuable.” “Looked at another way, analytics enable firms to understand how poorly they can treat individuals or groups of customers before those people stop doing business with them. Unless you are in the top echelon of customers—those with the highest lifetime value, say—you may pay higher prices, get fewer 252 special offers, or receive less service than other consumers,” Lemon adds. “Despite the fact that alienating 75% to 90% of customers may not be the best idea in the long run, many retailers have adopted this ‘top tier’ approach to managing customer relationships. And many customers seem to be willing to live with it—perhaps with the unrealistic hope that they maybe reach the 252 upper echelon and reap the ensuing benefits.” 262 THE PREDICTIVE CASINO “Little research has been done on the negative consequences of using marketing approaches that discriminate against customer segments. Inevitably, however, customers will become savvier about analytics. They may become less tolerant and take their business (and information) elsewhere,” Lemon 252 warns. On August 13, 2014, Facebook announced a major step forward in the area of attribution analysis. It said that it “would start telling advertisers on what device people saw an ad and on what device they took an action, such as buying a product or signing up for a test drive, as a result of seeing that ad. That means Facebook will be able to credit mobile ads that lead to desktop 253 sales and desktop ads that result in mobile purchases.” Peterson notes that, “Advertisers can already track conversions through Facebook on desktop and on mobile, but to date Facebook hasn't broken out conversions by device type for advertisers to see. For example, advertisers have been able to see if their desktop and mobile ads lead to conversions, but 253 they didn't know on which device type those conversions were taking place.” However, Facebook’s new cross-device conversion measurement only works for advertisers who place specific Facebook trackers on their websites and 253 mobile apps. “Without sharing users' personal information with the advertiser, those trackers can see that a Facebook user is checking out the advertisers' site or app and whether they've converted in the advertiser253 specified fashion.” If the person does convert, “Facebook's trackers can trace back to see if that person has seen an ad from that advertiser on Facebook, 253 which may have directly or indirectly led to the conversion.” Of course, nothing is 100% certain when it comes to attribution analysis, but this is a big step in the right direction. “A smart combination of listening to the online conversation already taking place, learning what people want, and then providing what they are open to receive from the brand constitutes the winning ticket,” advises Macy and 91 Thompson. One company that proved quite adept at this was the Alamo Drafthouse Cinema in Austin, Texas, although in a rather inverse way. The Alamo Drafthouse Cinema is one of the most respected arthouse chains in the country 237 and they have one rule: “If you talk or text during a movie, we kick you out.” “The Alamo Drafthouse ejected—without a refund—a customer who, despite a posted rule, was texting on her cellphone during a film. Incensed, the customer left a long, vitriolic and obscenity-laden voicemail message on the theatre’s 237 voicemail.” 254 The Alamo converted the message into a YouTube video , which quickly went viral and garnered more than three million hits and thousands of positive 237 responses from cinema-goers everywhere. The story made CNN and other 263 ANDREW PEARSON major media outlets as well. 237 “This customer’s attempt to censor the Alamo for enforcing its policy instead turned the cinema into a hero—when it recognized the opportunity to turn a complaint into content that spoke to the frustrations of filmgoers everywhere 237 who are faced with rude behavior.” Oftentimes, as Lieb so succinctly points out, “reputation management is as simple a matter as turning lemons into 237 lemonade.” The above is an example of good brand and anti-brand management all rolled into one and, when it comes to anti-brand management, businesses should be well aware of the threat this unique problem presents. The Internet has allowed companies to reach consumers in very cheap and easy ways, but the downside is that it can also give a loud and reverberating voice to unhappy customers very cheaply, too. Because humans are, by nature, more attuned to negative messages than positive ones (think about how a raised voice in a crowded restaurant gains instant attention) businesses need to react very quickly to negative stories in the social media world. Although anti-brand sites are usually created to attempt to hurt brands, smart companies can use these websites to their advantage as well. Kucuk offers four basic strategies for companies to counter these anti-brand Websites; work with experts; monitor symbolic haters; talk to complainers; and combat 210 opportunists. A word of warning here: don’t set up fake accounts to try to get your marketing message out. There are people out in the blogosphere who would love nothing more than to uncover the latest social media scam, especially in China, and they probably have access to some of the most sophisticated tools to ferret out dishonest behavior. Don’t pick fights with customers on social media either. Threatening customers with legal action is never a good idea. The old rule that you shouldn’t whip out a gun unless you’re willing to pull the trigger should be kept in mind when it comes to social media as well. Case in point: Greenpeace vs. Nestlé. “In 2010, Greenpeace launched a campaign consisting of a web site and a Facebook page protesting Nestlé’s sourcing of the palm oil used in Kit-Kat candy bars. The oil was being harvested from—and destroying—Indonesian rain forests, pushing orangutans further toward the brink of extinction and threatening the 237 livelihoods of local residents.” “Destroying rain forests and driving species into extinction is no one’s idea of corporate responsibility, but Nestlé was ill-prepared for the angry onslaught of Facebook protesters, particularly when people started to replace their profile 237 pictures with a Kit-Kat logo modified to read ‘Killer.’” This is nothing new, people have been scrawling bastardizations of company names on city walls or wearing twisted or ironic company slogans on T-shirts for decades, but upon 264 THE PREDICTIVE CASINO seeing their altered logo, Nestlé, amid a substantive discussion about 237 sustainability, started to hand out take-down notices. Nestlé stock plummeted, and the story—as these stories always seem to do— 237 hit the mainstream media. “A few days into the debacle, Nestlé changed the corporate statement on its Facebook page to read: ‘Social media: as you can 237 see we’re learning as we go. Thanks for the comments.’” The main lesson here is: never underestimate an opponent who has the echo chamber of social media on their side. Greenpeace is an incredibly wellorganized, media and web savvy organization and companies “must be wellprepared with content, PR-savvy, and the ability to create the right kind of 237 content and deliver it in the appropriate voice.” Threatening to sue your customer is, just about always, a losing strategy, so is nitpicking over a logo or 237 defending copyright issues when lives are—literally—at stake. The stupidity here is that Nestlé could probably have anticipated this campaign as Unilever and Kraft had already stopped using the troubling Indonesian 237 supplier at Greenpeace’s behest. “So in addition to not leaving their social media communications and content to unseasoned and junior staff, they could have rehearsed scenarios they must have known were coming, even if they 237 didn’t exactly know the form they would take.” Proactivity is almost a necessity in this day and age of instant access and the 24/7 news cycle. “While Nestlé eventually did capitulate on its palm oil sourcing, reportedly the company had been investigating new sources all along, and planned to stop 237 buying from the Indonesian supplier.” This is news that would have kept the angry and baying dogs of social media from its door, but, something they, strangely, decided to keep mum about. Nestlé could have backed this story up with powerful content, such as with “blog entries from company executives, with documentation of the research it was conducting, with interviews with experts on sustainability, supply chains, 237 product sourcing.” All of this would have countered the negativity, if not have provided a “wellspring of substantive content the company could have used in its very public discussions with protesters. Instead, the memo from 237 legal took precedence, with unnecessarily disastrous results.” In many cases, companies seem to bring social media crises upon themselves through careless thinking and a lack of channel monitoring. In her article The top 10 corporate social media disasters, Rebecca Burn-Calandar mentions six social media crises—the “JP Morgan Snarkpocalyse,” “the British Gatastrophe” (British Gas), “Mayday for British Airways,” “Twitter shouts itself hoarse at Tesco,” “The Quantas Bashtag,” “Home Depot” (the monkey picture fiasco)— that could be considered self-inflicted wounds. A seventh, “Disobeying his master’s voice,” could probably have been avoided with stricter administrative controls at HMV as well. 265 ANDREW PEARSON The lesson here is to be humble when venturing into social media and understand there could be a lot of people out there who are just itching to vent their frustration at you, especially if you are working in a business like banking, oil and gas, or even gaming. If you’re a company like JPMorgan, better to really think long and hard before you set up a Q&A Twitter session asking users to send in questions for one of their executives. One has to wonder about JPMorgan’s hubris when something like this was pitched. How could they not have known they’d be ridiculed with questions like: “Quick, you’re in a room with no key, a chair, two paperclips, and a light bulb. How do you defraud investors?” Or snarky comments like: “But really, we better stop making fun of #askJPM before they find a way to monetize the hillarity and charge us for our enjoyment at 33% APR.” Take to heart author Paul Gillin’s warning that, “Transparency may be the most disruptive and far-reaching innovation to come out of social media.” Social media has really upped the ante when it comes to customer engagement because, through it, customers cannot only connect with a brand, but also with a brand’s consumers. “What is new is that customer engagement is not just a brand's connection with the customer. It is also the customers' engagement 91 with one another in the horizontal, viral aspects.” It is these horizontal and viral aspects that can be so important—and potentially lucrative. Conversely, Macy and Thompson warn that a “lack of engagement limits brand leadership 91 effectiveness and ultimately defeats the purpose of the medium.” “A smart combination of listening to the online conversation already taking place, learning what people want, and then providing what they are open to receive from the brand constitutes the winning ticket,” recommend Macy and 91 Thompson. 266 THE PREDICTIVE CASINO Tips • • • • • • • • • • • • • • • • • Keep your user name identical on all social networking sites to allow for easy cross-posting and cross-referencing. Offer something to make people want to come back. It doesn’t have to be something tangible, it just has to be something interesting. Choose social media applications based on where a target audience congregates. Either join an existing social media platform to take advantage of its popularity and built-in user base or build a presence from scratch; both have their own unique advantages and disadvantages. Businesses should ensure that all of their employees have access to all of the company’s social media platforms, but administrator rights should be doled out selectively. When stepping into the social media world, be active, interesting, humble, unprofessional (to a certain extent), and honest. When it comes to brand management, companies should keep an eye on e-complaint websites, as well as on consumer blogs to monitor the type and duration of problems that flare up. Open an innovation hub that allows users to post their ideas to the company. Set up Google Alerts for your company name and any of your company’s products or services For anti-brand management, businesses should buy potentially negative Internet domain names that could target their brand in a negative way. Buy up website names that are mistypes of your company name, thereby countering typo-squatters. There are four basic strategies for companies to counter anti-brand websites; work with experts; monitor symbolic haters; talk to complainers; and combat opportunists. Counter negative stories with timely positive and credible ones, with as many links to reputable sources as possible. When appropriate, email upset customers, perhaps attaching discount coupons or gift cards to show your remorse for poor service. Create an online community for your products rather than simply building an ecommerce site. Engagement sells like nothing else. Keep your community engaged by injecting your personality into every engagement (as long as it is not a boring personality, that is) and act as if you are interacting with friends. Know the interests of your consumers and build upon those interests. 267 ANDREW PEARSON • • • • • • • When it comes to crisis management, it is best to tackle a problem head-on. Using disarming humor and/or clever irreverence can stop crisis-like situations from turning into full-blown social media disasters. Write and publish content with embedded links to other content, pictures, and videos to interest the potential online audience. Use Tweeter Spy to help determine which tweets result in the most traffic back to your site. Use untweetme.com to find former Twitter followers who might have left because of poor service. Use Viralwoot to gain Instagram followers. Use ShareRoot’s suite of free tools to get more followers on sites like Pinterest. Use Klear to help measure influencer engagement. 197 Outing, S. (2007, September). Enabling the Social Company. Enthusiast Group. Retrieved from Steveouting.com: http://www.steveouting.com/files/social_company.pdf 198 Business.com. (2010, November 8). Top Tools to measure your social media success. Retrieved from Business.com: http://www.business.com/info/social-media-monitoringtools 199 Honigman, B. (2012, November 29). 100 fascinating social media statistics and figures from 2012. Retrieved from Thehuffingtonpost.com: http://huffingtonpost/brianhonigman/100-fascinating-social-me_b_2185281.html 200 Hierarchy of Social Marketing. John Jantsch. Duct Tape Marketing. https://www.ducttapemarketing.com/the-hierarchy-of-social-marketing/ 201 http://www.simplypsychology.org/maslow.html 202 Li, C. a. (2008). Groundswell: Winning in a World Transformed by Social Technologie. Harvard Business Press. 203 Ramirez, A. (2009). The Effect of Interactivity on Initial Interactions: The Influence of Information Seeking Role on Computer-Mediated Interaction. Western Journal of Communication, 300–325. 204 Bates, P. (2011). Social Media Theory—the 1:9:90 rule. Retrieved from Yell: http://marketing.yell.com/web-design/social-media-theory-the-1-9-90-rule/ 205 Burn-Calander, R. (2013, November 27). The top 10 corporate social media disasters. Retrieved from The Telegraph: http://www.businessinsider.com/the-top-10corporate-social-media-disasters-2013-11 206 Shamoon, S. a. (2011). Brand Management: What Next? Interdisciplinary Journal Of Contemporary Research In Business, 435-44. Retrieved from Business Source Complete. 207 http://www.managementstudyguide.com/brand-management.htm (Accessed 1/1/2017) 208 Jones, L. a. (2011). Building Brand Communities Through Online Interaction: The Case of Social Media. (pp. 17-19). Orlando: 2nd International Colloquium on ConsumerBrand Relationships. 268 THE PREDICTIVE CASINO 209 http://journals.ama.org/doi/abs/10.1509/jmkg.73.5.30?code=amma-site 210 Kucuk, S. U. (2008). Negative Double Jeopardy: The role of anti-brand sites on the Internet. Brand Management, Volume 15, No. 3, 209-222. 211 http://money.cnn.com/2015/02/27/technology/security/iran-hack-casino/ (Accessed 1/5/17) 212 Burkitt, L. (2010, January 1). Need to build a community? Learn from Threadless. Retrieved from Forbes.com: http://www.forbes.com/2010/01/06/threadless-t-shirtcommunity-crowdsourcing-cmo-network-threadless.html 213 Hodge, A. (2012, October 22). How Social Media Is Changing Brand Management. Retrieved from Mondaq.com: http://www.mondaq.com/x/202746/Social+Media/How+Social+Media+Is+Changing+Br and+Management 214 Simply Zesty. (2012, November). How to increase your social media following. Retrieved from simplyzesty.com: http://www.simplyzesty.com/Blog/Article/November2012/How-To-Increase-Your-Social-Media-Following 215 Malshe, A. (n.d.). A Typology of Social Media Crises. Retrieved from Academia.edu: http://www.academia.edu/1476436/A_Typology_of_Social_Media_Crises. 216 Gundel, S. (2005). Towards a New Typology of Crises. Journal of Contingencies and Crisis Management, Volume 13 (3), pages 106 – 115. 217 http://www.youtube.com/watch?v=5YGc4zOqozo (Retrieved: February 15, 2017). 218 Hale, J. E., Dulek, R. E., & Hale, D. P. „Crisis Response Communication Challenges Building Theory From Qualitative Data”. Journal of Business Communication, 2005, 42 (2), 112-134 219 Korolov, M. (2010, August 11). Why virtual worlds suck for business—and some solutions. Retrieved from Hypergrid Business: http://www.hypergridbusiness.com/2010/08/why-virtual-worlds-suck-for-business-andsome-solutions/ 220 Goldenberg, D. (2013, September 5). Virtual roses and the rise of yy.com. Retrieved from Newyorker.com: http://www.newyorker.com/online/blogs/currency/2013/09/virtual-roses-and-the-riseof-yy-music-china.html 221 https://www.stageit.com/site/landing 222 Grassegger, H., Krogerus, M. December 3, 2016. I Just Showed That The Bomb Was There. Das Magazin https://www.dasmagazin.ch/2016/12/03/ich-habe-nur-gezeigtdass-es-die-bombe-gibt/. 223 Bifet, A. a. (2010). Sentiment knowledge discovery in twitter streaming data. Retrieved from University of Waikato, Hamilton, New Zealand: http://www.cs.waikato.ac.nz/~ml/publications/2010/Twitter-crc.pdf 224 https://www.rivaliq.com 225 http://buzzsumo.com 226 . O’Connor, B., Balasubramanyan, R., Routledge, B. R. and Smith, N. A. From tweets to polls: Linking text sentiment to public opinion time series. In Proceedings of the International AAAI Conference on Weblogs and Social Media, pages 122–129, 2010. 227 Haden, J. (2014, April 23). 60 Awesome Social-Media Tools for Entrepreneurs. Retrieved from Inc.com: http://www.inc.com/jeff-haden/60-awesome-social-mediatools-for-entrepreneurs.html 269 ANDREW PEARSON 228 Landau, J. (2014, July 22). GoPro's Viral Video Marketing Campaign Turns It Into Top YouTube Brand in the World. Retrieved from NY Daily News: http://www.nydailynews.com/news/national/gopro-marketing-turns-top-youtubebrand-article-1.1875573 229 He, Wu, Zha, Shenghua, Li, Ling. Social media competitive analysis and text mining. A case study in the pizza industry. International Journal of Information Management 33 (2013) 464-472. http://saharbread.sahargroup.ir/Uploads/28460.pdf 230 Dey L., H. S. (2011). Acquiring Competitive Intelligence from Social Media. Proceedings of the 2011 Joint Workshop on Multilingual OCR and Analytics for Noisy Unstructured Text Data. 231 Woodruff, A. (2014, April 26). Necessary, unpleasant, and disempowering: reputation management in the Internet age. Retrieved from www.allisonwoodruff.com: http://www.allisonwoodruff.com/publications/2014-Woodruff-CHI2014ReputationManagement.pdf 232 https://en.wikipedia.org/wiki/Reputation_management 233 Davis, J. (2012, April 24). The stalking of Korean hip hop superstar Daniel Lee. Retrieved from Wired.com: www.wired.com/2012/04/ff_koreanrapper/all/1 234 Maag, C. (2007, November 28). A hoax turned fatal draws anger but no charges. New York Times. 235 Brynley-Jones, L. (2009, October 4). A 5-Step Guide to Reputation Management Using Social Media. Retrieved from oursocialtimes: http://oursocialtimes.com/a-5-stepguide-to-reputation-management-using-social-media/ 236 https://en.wikipedia.org/wiki/Klout 237 Lieb, R. (2012, July 10). How Your Content Strategy Is Critical for Reputation Management. Retrieved from Marketingland.com: http://marketingland.com/howyour-content-strategy-is-critical-for-reputation-management-16073 238 http://www.webopedia.com/TERM/S/social_shopping.html 239 http://www.usatoday.com/travel/flights/2009-08-02-jetblue-united-twitterairfares_N.htm (Retrieved: July 21, 2016). 240 Woodward, P. (2013, June 25). Top Tips for Creating an Online Retail Community. Retrieved from The Guardian: http://www.theguardian.com/media-network/medianetwork-blog/2013/jun/25/top-tips-online-retail-community 241 eMarketer. (2014, March 31). Facebook Is No. 1 for Social Commerce. Retrieved from emarketer.com: http://www.emarketer.com/Article.aspx?R=1010721&RewroteTitle=1 242 Advangent. (2013, August 13). ina Weibo and Alibaba announced a strategic alliance to cooperate in social shopping. Retrieved from advangent: http://www.advangent.com/zh/2013/08/13/sina-weibo-taobao-social-commerce-2-0/ 243 Melville, P. &. (2009). Social Media Analytics : Channeling the Power of the Blogosphere for Marketing Insight. Retrieved from citeseerx.ist.psu.edu: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.157.3485&rep=rep1&type= pdf 244 Rouse, M. (n.d.). Social Media Analytics. Retrieved from techtarget.com: http://searchbusinessanalytics.techtarget.com/definition/social-media-analytics 245 https://klear.com/ 270 THE PREDICTIVE CASINO 246 salesforce.com. (2013). 10 Examples of Social Media Command Centers. Retrieved from Salesforce Marketing Cloud: http://www.salesforcemarketingcloud.com/resources/ebooks/10-examples-of-socialmedia-command-centers/ 247 Gulbransen, Scott. January 22, 2014. Taking Back The Social-Media Command Center, Scott Gulbransen. Forbes. http://www.forbes.com/sites/onmarketing/2014/01/22/taking-back-the-social-mediacommand-center/#3c283a5d6513 248 IBM. (2013, February). Social Media Analytics: Making Customer Insights Actionable. Retrieved from IBM.com: http://www-01.ibm.com/common/ssi/cgibin/ssialias?infotype=SA&subtype=WH&htmlfid=YTW03168USEN 249 How Travel & Hospitality Brands Are Marketing With Social Media Influencers http://mediakix.com/2015/08/travel-hospitality-brands-marketing-with-social-mediainfluencers/#gs.rw8hqKw 250 Marriott International Partners with What’s Trending and YouTube Stars to Develop Original Content Series. February 6, 2015. Marriott.com. http://news.marriott.com/2015/02/marriott-international-partners-with-whatstrending-and-youtube-stars-to-develop-original-contentseries/?aff=MARUS&affname=10l1110&co=US&nt=PH 251 Davenport, T. (2006). Competing on Analytics. Harvard Business Review. 252 Lemon, K. (2007, May). How Can These Companies Leverage the Customer Data Responsibly. Retrieved from http://blog.hansacequity.com: http://blog.hansacequity.com/Portals/11224/docs/article%20on%20Analytics.pdf 253 Peterson, T. (2014, August 13). Facebook Now Tells Whether Mobile Ads Lead to Desktop Purchases. Retrieved from AdAge: http://adage.com/article/digital/facebookmakes-link-mobile-ads-desktop-purchases/294568/ 254 https://www.youtube.com/watch?v=1L3eeC2lJZs 271 272 THE PREDICTIVE CASINO CHAPTER SIX Casino Operations “Smart homes and other connected products won't just be aimed at home life. They'll also have a major impact on business. And just like any company that blissfully ignored the Internet at the turn of the century, the ones that dismiss the Internet of Things risk getting left behind.” ~Jared Newman Fast Company Overview The rise in popularity and the rapid growth of the gaming industry has created a highly competitive environment for casino companies worldwide. Industryleading gaming companies have expressed the need to identify and develop their clientele so that they can enhance their guest's experience, as well as to increase customer loyalty and generate new business leads, whether they are in Las Vegas, Macau, Singapore, Vietnam, Cambodia, Australia, or a whole host of other worldwide gaming destinations. The casino industry is exceptionally robust and, throughout the Far East, casinos are either currently being built in places like South Korea, Vietnam, The Philippines, and Russia or close to receiving parliamentary approval in countries like Japan and Taiwan, so the competition is only going to get stronger. Cruise line companies are also targeting this gaming market and these ships that cost upwards of half-a-billion U.S. dollars, in some cases, contain more slots and tables than the massive integrated resorts in Macau. Genting Dream, th which sailed into Hong Kong on November 12 , 2016 and will use that city as one of its homeports, has been described as a “Floating, luxury, integrated 255 resort.” Besides the abundant table games and slots aboard, Genting Dream also offers the FSG Mobile platform, which contains live baccarat and live roulette games, so customers can wager on any casino table game in any area of the ship that is Wi-Fi enabled at any time (at least when it is in international waters). At the start of this book, I spoke about Amazon receiving an anticipatory patent so it could deliver packages before customers even thought about ordering them and this concept should infuse the Predictive Casino because casino patrons today will expect this type of anticipatory service from every company 273 ANDREW PEARSON they choose to deal with from now on. These patrons will be highly sophisticated customers who expect information about what they want to buy to be, literally, at their fingertips. With anticipatory shipping, Amazon wants to have stock availability close to where the actual orders are coming in from so that they can be shipped and arrive at the customer’s house shortly after they have been ordered. This concept works for a casino company, too. Understanding customer needs on such an intimate level so that every facet of a casino’s operation is ready to deliver what a patron wants when he or she wants it is the Predictive Casino’s goal. Currently, casino hosts will move heaven and earth for a high-rolling VIP, but why can’t that level of service be brought down to a lower level of guest? The Predictive Casino can deliver it. Bringing VIP-type service down to other important guests may sound difficult, but it is possible if the casino can both quickly capture data and get it into the hands of those whom can utilize it immediately. Casino operators probably have what they need, but making the intelligence actionable is the problem. Today, casino patrons want to be able to pull up their points balances on their mobile phones and soon, they will want to pay for things throughout an IR on their phones as well, possibly with Bitcoins. They will soon expect the casino to know when they’ve walked through the front door. They’ll want their favorite drink and dealer awaiting them once they approach their lucky baccarat table. They’ll want to be able to walk straight to their room without bothering to stand in line at the check-in desk, and they’ll want an experience so personalized that they will feel like returning home to a known family. This can all be handled by the Predictive Casino as the remainder of the book will show. Smart Operations The term “Smart” has become synonymous with a vision of technology integrating with multiple information and communication platforms, including Internet of Things (IoT) solutions to enhance quality, performance and interactivity of services, as well as to reduce costs and resource consumption and to improve contact between different parties, such as between citizens and governments. The Predictive Casino has answers for the following questions: • • • • • • • How can a casino operator utilize IoT technology? How can a casino save on it resource use? How can a casino improve the management of patron movements? How can an IR best utilize its transportation fleet? How can a casino reduce its labor needs? How can an IR cut down on waste? How can a casino ensure its security. 274 THE PREDICTIVE CASINO These are all important questions that need to be asked. Today’s IRs are massive structures and sensors have become so small and cheap they can be put almost anywhere. IoT sensors can be used for smart parking, smart lighting, as well as part of a mini smart grid. They can also be used for silo stock calculation—measuring the emptiness level and weight of goods, as well as waste management, tracking movements on the casino floor, and perimeter access control. For the casino’s logistics department, IoT aids quality of shipment conditions, item location, storage incompatibility detection, and fleet tracking. IoT sensors can even be installed to ensure a buildings structural health, as well as part of a swimming pool remote management system. For a casino, the CCTV security systems could even be set up to alert a host or a VIP manager when a particular high roller steps aboard the casino bus on the Macau-China border or on the casino floor. The CCTV system could also be used to capture problem gamblers or cheats on the casino floor. Edge video analytics might also be helpful in places like Singapore, where IRs are required to check the passports of each incoming guest. With cameras that are able to compare the faces of entering customers against the casino’s patron database, these patron records would be onscreen for the security personnel to quickly approve or reject. Rather than waiting for a record to be pulled up, the security personnel would simply confirm or deny the patron in half the time it would normally take them to pull up the record. This would cut down on the long lines at the front of the casino that are really costly choke points. The casino makes money when gamblers are gambling, not when they are standing in line, waiting to get in. As Figure 13 shows, analytics can be used to boost services quality, reduce operating costs, as well as build ROI. As the IBM Thought Leadership White Paper Descriptive, predictive, prescriptive: Transforming asset and facilities 256 management with analytics states, “As facilities and assets become more ITlike—instrumented, intelligent and interconnected—the convergence of physical and digital infrastructures makes their management increasingly complex. And in a physical world outfitted with millions of networked sensors, vast amounts of facilities- and asset-generated data make extracting meaning 256 increasingly difficult.“ When it comes to capital planning, real estate lease management, operations, facilities maintenance and energy consumption, IRs must tame the three Vs of big data—volume, variety, and velocity—and analytics are the best way to do so. 275 ANDREW PEARSON Figure 13: Understanding analytics, definitions, applications Source: IBM 256 Utilizing the right analytics in the right place can yield impressive ROI results. As 256 IBM states : “From basic to advanced capabilities, analytics can yield dramatic results. One study found that an organization that uses basic automation to expand its reporting capabilities can improve its return on investment (ROI) by 188 percent. But adding additional capabilities such as data management, metadata to ensure uniform data interpretation, and the ability to gather and analyze data from outside the organization, can boost ROI to as high as 1,209 percent.” IBM’s specific examples include: • • “Effective facilities and asset management uses data analytics to proactively manage facilities and maintain equipment, optimize utilization, prevent breakdowns, lower occupancy and operational 256 costs, and extend asset life.” “Utilizing analytics to monitor energy-intensive equipment across the facilities portfolio, identify operating anomalies in real time, and 276 THE PREDICTIVE CASINO • generate corrective work orders can dramatically reduce energy 256 consumption.” “To help mitigate risks of failure in facilities and assets, analytics can detect even minor anomalies and failure patterns. Identifying issues early helps organizations deploy limited maintenance resources more cost-effectively, maximize equipment uptime and improve customer 256 service levels.” Inventory Optimization Inventory optimization—the ability to have the right inventory, in the right quantities and at the right locations, at the right time to meet the supply and 257 demand of parts and materials in the enterprise —is an important part of the Predictive Casino. As Tristan O’Gorman explains in his blog 5 Inventory 257 Analytics Best Practices To Achieve Inventory Optimization , “Significant benefits exist for organizations that optimize their inventory by reducing inventory items and stock levels, thus avoiding associated carrying costs and obsolescence write-downs.” O’Gorman adds that, “Indirectly, organizations can generate savings by using time formerly spent on inventory management to ensure physical assets’ reliability and availability.” “Spare parts, an essential component of the availability of any system, have intermittent consumption patterns and usually have only one specific use, and organizations can often source them only from the system manufacturer. For these reasons, many organizations overstock spare parts to avoid costly system downtime. Unfortunately, overstocking incurs its own costs—and they can be 257 significant,” warns O’Gorman. Often, organizations use unreliable manufacturer lifetime and degradation information to stock spares, but such data can be imperfect and, therefore, 257 unreliable. IRs can “use inventory analytics to identify items that are trending toward being out of stock, providing a means of stock management more reliable than supplier data. In addition, research has shown that monitoring 257 technology can reduce the need for spare part inventory.” As O’Gorman explains, to “maintain ideal inventory stock levels, organizations must accurately classify inventory. ABC analysis, a particularly popular classification system, classifies inventory by the relative priority of each item against other items in the inventory. A-classed items, the most important, typically make up 5 to 10 percent of inventory. B-classed items typically make up the next 15 to 25 percent of inventory. C-classed items, the least important, 257 make up the remaining 65 to 80 percent of inventory.” However, classifying inventory like this isn’t as easy as it seems; ABC analysis 257 requires ongoing review and revision to achieve optimal item distribution. In addition, many organizations rely on corrective maintenance procedures that 277 ANDREW PEARSON 257 lead to unscheduled inventory demands. O’Gorman recommends that 257 organizations implement forecasting practices such as the following : • • • • • • Forecasting demand and planning supply. Communicating, cooperating and collaborating. Eliminating islands of analysis. Using tools wisely. Emphasizing forecasting. Measuring everything. With these practices in place, inventory analytics and predictive asset maintenance can make use of the wealth of data that an IR generates, which will help it capitalize on an analytics solution designed for asset inventory 257 management. As O’Gorman argues, “Obsolescence is an unavoidable fact of inventory management, but unfortunately, many organizations manage it reactively. To manage obsolescence proactively, organizations must be able to answer questions such as the following: How can we anticipate obsolescence? What contingency plans have we in place? What are our most important needs? 257 Should we maintain items, or replace them? How do we ensure safety?” Gorman advises IRs to: “Develop an obsolescence risk assessment process, helping inventory managers assess the probability that items will become obsolete and flag the items that are at greatest risk of becoming so. To augment the process, inventory managers can use analytics to identify already obsolete items, identify items that the organization can afford to manage reactively and reduce the manual effort involved in computing the 257 probability of obsolescence. Adding analytics into the obsolescence process gives IRs actionable data that can be used to strengthen their bargaining position when negotiating supplier 257 agreements, systems upgrades, or it could help risk mitigation purchases. By optimizing reorder points for inventory items, organizations can avoid excess stock conditions. IRs must juggle the need between having sufficient inventory to handle fluctuating demand, while also keeping costs down and the potential benefits are huge, especially for buildings like the Galaxy Macau and the Venetian Macao, two properties that have some of the biggest building 257 footprints in the world. In his article, Why the food and beverage industry needs to start using internet 258 of Things technology , Carolyn Heneghan argues that, “One of the most important applications for IoT in a food and beverage industry context is 278 THE PREDICTIVE CASINO temperature tracking and control. Using sensors in the product, manufacturers can track shipments from afar, both where it is and how cold it is.” The data captured will let the IR know if products have been spoiled and possibly why, 258 like if there's a problem with delivery truck refrigeration. “It's not really the temperature of the truck that matters, but the temperature of the food,” said Jim Cerra, cofounder and CEO of PlanetTogether. “If they're opening up the truck to drop off material at various stores and warehouses, the temperatures are fluctuating in the truck. They can be very specific with the 258 French fries or lettuce if they're violating the limits.” 258 Light can also affect quality and safety as well. Too much light can hasten bacteria growth. “That’s particularly important for companies embracing the clear packaging trend, or for producers like brewers that traditionally use glass 258 or see-through materials as packaging,” Heneghan argues. “IoT technology can detect light, and can send data about when a package is first opened or 258 how much sunlight hits a product during shipping,” Heneghan explains. Waste Management According to the Food and Agriculture Organization (FAO) of the United 259 Nations, annual food wastage is estimated at 1.6 billion tonnes. The economic impact of this waste is priced at over $750 billion, a number that 259 doesn’t include fish and seafood. The FAO warns that it is not only the loss of money that is consequential, but the “environment is also damaged by an increased carbon footprint, additional landfill deposits and higher demand for agricultural land. Using more land for agriculture, in turn, poses a threat to 259 various plant and animal species.” In his article Smart Food Management Utilizes IoT to reduce Cost, Waste and 260 Pollution , Shay Adar explains that multi-sensored smart devices can monitor perishable goods in storage and in transit and relay location, temperature, and humidity data to the cloud. “This provides real-time information on the precise conditions of each unit of goods and thus enables corrective actions to reduce 260 waste and save money,” Adar adds. An IR can utilize this data to ensure both the quality and the amount of its purchased product is being received in a timely manner. The transition to Ag3.0 means farmers are no longer looking to the clouds for weather patterns, but rather to the “cloud” for its big data and analytics capabilities, hoping to maximize the life and quality of their produce and livestock. As Adar explains 260 : “A cold chain is the entire process that perishable goods go through, from production to market, including storage and 279 ANDREW PEARSON various methods of transport. During this entire chain, temperature and other factors must be maintained to safeguard the quality of the product. Traditional monitoring systems are either stationary wired sensors, or portable temperature data loggers. The former are only suitable for warehouses, while the latter are only good for hindsight. This is due to the fact that the logged information can only be read once the goods have arrived at their destination. At that point, the only decision to make is whether to place the goods on the shelf or in the trash.” Because cold chain monitoring occurs in real-time, corrective action can be 260 taken immediately, which could, potentially, reduce spoilage and waste. Once an IR department knows that the goods it is expecting will most likely arrive spoiled, the goods can be rejected and replacements can be found. In some cases, a failed delivery means items won’t make the menu, or, in rare cases, a restaurant has to be closed, which means a wait staff might not be needed. Admittedly, this is a rare occurrence, but there will be certain shipments that, if they are not received an appropriate set of labor needs will also not be required. A “reduction of superfluous trips delivering spoiled goods, can lead to a significant decrease in fuel and shipping costs, as well as the carbon footprint that comes with them,” adds Adar. The perishability of a product can also be tracked. A sensored cold chain adds 260 precision into the supply chain. “By placing sensors on each pallet of food, the differences in environment conditions between pallets in the same shipment can be measured and collected. This data can then be used to calculate the remaining shelf-life of the product. With this information, the traditional FIFO (First In First Out) inventory system can be replaced with the more effective FEFO (First Expired First Out),” Adar recommends. ML models can also be built to discover the optimized delivery methods or these models can help spot outliers in the supply chain. GPS location can be built into products, so that an IR knows exactly where 261 every item it is buying is located in the supply chain. Every object will have its own unique identifier, and the IR will be able to pin-point each and every item or piece of equipment the company has. This can “greatly decrease the number of lost or stolen products, increase management of stock shortages and 261 overstocks, and better identify inefficiencies.” For example, the IR can be alerted about shipments that get stuck in traffic at a certain time of day, items that sell better at certain locations, when and, potentially, where items get 261 damaged, how and perhaps why food arrives spoiled, amongst other things. The ability to view, track, and monitor inventory of all kind will improve 261 enormously through integration with the IoT. Inefficiencies that went unnoticed before will not only become actionable intelligence—and definable 280 THE PREDICTIVE CASINO ROI—can help an IR make the case for an investment in IoT. 261 Data Center 150 In his article Machine Learning Applications for Data Center Optimization , Jim Goa explains that the modern data center (DC) is a complex interaction of multiple mechanical, electrical and controls systems, coupled with all kinds of operational software. “The sheer number of possible operating configurations and nonlinear interdependencies make it difficult to understand and optimize 150 energy efficiency,” Goa warns. However, Google has developed a neural network framework that learns from actual operations data to model plant performance and predict Power Usage Effectiveness (PUE) “within a range of 0.004 +/ 0.005 (mean absolute error +/ 1 standard deviation), or 0.4% error for 150 a PUE of 1.1.” The model, which has been extensively tested and validated at Google data centers, demonstrates that “machine learning is an effective way of leveraging existing sensor data to model DC performance and improve 150 energy efficiency.” Goa explains the process as such: “Google Machine learning is well suited for the DC environment given the complexity of plant operations and the abundance of existing monitoring data. The modern largescale DC has a wide variety of mechanical and electrical equipment, along with their associated setpoints and control schemes. The interactions between these systems and various feedback loops make it difficult to accurately predict DC 150 efficiency using traditional engineering formulas.” For example, “a simple change to the cold aisle temperature setpoint will produce load variations in the cooling infrastructure (chillers, cooling towers, heat exchangers, pumps, etc.), which in turn cause nonlinear changes in 150 equipment efficiency.” Everything from ambient weather conditions to equipment controls to the age of the hardware and complexity of the software 150 can affect the DC’s efficiency. Gao argues that, “Using standard formulas for predictive modeling often produces large errors because they fail to capture 150 such complex interdependencies.” Furthermore, “the sheer number of possible equipment combinations and their setpoint values makes it difficult to determine where the optimal efficiency 150 lies.” To address this challenges, Goa selected a neural network as the 150 mathematical framework for training the DC energy efficiency models. According to Gao, the neural network utilized “5 hidden layers, 50 nodes per 150 hidden layer and 0.001 as the regularization parameter.” The training dataset contained “19 normalized input variables and one normalized output variable (the DC PUE), each spanning 184,435 time samples at 5 minute 281 ANDREW PEARSON 150 resolution (approximately 2 years of operational data).” 70% of the dataset was used for training, while the remaining 30% was for cross-validation and 150 testing. In conclusion, Gao found by using the machine learning framework developed in his paper, he was able to predict DC PUE within 0.004 +/ 0.005, 150 approximately 0.4% error for a PUE of 1.1. Actual testing on Google DCs indicated that, “machine learning is an effective method of using existing sensor data to model DC energy efficiency, and can yield significant cost savings. Model applications include DC simulation to evaluate new plant configurations, assessing energy efficiency performance, and identifying 150 optimization opportunities.” This is a model that can also be implemented with an IR. Smart Parking In his article The Future Of Smart Parking Is Integration With Automated 262 Technology , Ryan Citron argues that “The future of the smart parking market is expected to be significantly influenced by the arrival of automated vehicles (AVs). Several cities around the world are already beginning to trial self-parking vehicles, specialized AV parking lots, and robotic parking valets. For example, in Boulder, Colorado, ParkPlus is working on deploying a fully automated parking garage in the Western United States through Boulder’s PearlWest mixed-use development. “The company’s automated parking system uses lasers to scan cars and a robotic valet to park the vehicles,” Citron 262 explains. Since the cars can be packed close together, up to four times as 262 many can be parked in the same area as a traditional garage. The cars are 262 delivered within two minutes of a retrieval request as well. The city of Somerville, Massachusetts, has also partnered with global 262 automaker Audi to develop self-driving and self-parking cars. In 2018, the city plans to have a small fleet of cars with piloted parking technology deployed 262 to test self-parking capabilities with a specialized nearby parking garage. The hope is that these self-parking cars will help improve traffic congestion by dropping riders off in front of their destination and then the car would park 262 itself, thereby minimizing the time spent taking up space on the road. Drivers won’t be circling blocks trying to find a space. In 2020, “phase two of the project is expected to commence with the deployment of a full fleet of selfparking Audi cars. By 2030, the self-parking garage is targeting availability to 262 the broader AV market.” Parking can, of course, also be a great predictor of busyness within a casino and/or retail mall. License plates can also be scanned as cars are pulling into an IR’s parking lot and, should VIPs be recognized, alerts can be sent to the appropriate departments. If the license plates of problem gamblers are 282 THE PREDICTIVE CASINO spotted, security can be notified as well. Smart Energy Today’s massive IRs also have huge carbon footprints and green construction, social innovation, and sustainability should be part of the IR’s property 263 management discussion. In Hitachi’s Enabling Energy Resilience white paper, the Japanese conglomerate argues that: “The role of energy in our lives is changing. We are increasingly relying on energy-intensive information and communication technology to connect with each other and move our economy. As these changes meet head on with increasingly severe weather events, fluctuating energy price, and an ageing energy infrastructure, threats to energy security are no longer just a national problem—they have become a local one. Organizations from local government to hospitals to military bases are looking carefully at strategies to ensure their own energy resilience. While energy independence is rarely a practical approach, organizations can use a combination of proven management strategies and advanced technology to increase their energy resilience and protect their core mission.” Although their study was a few years ago, the American Society of Civil 264 Engineers (ASCE) 2009 Report Card for America’s Infrastructure gave the US energy infrastructure a “D-“ or poor grade. Its findings included: • • • • • Electricity demand has increased by about 25% since 1990 while construction of transmission facilities decreased about 30%. 70% of U.S. transmission lines and power transformers are 25 years or older. 60% of circuit breakers are more than 30 years old. Equipment failure will increasingly lead to intermittent failures in power quality and availability. The limited capacity of older equipment is resulting in an increase in congestion points in the grid—leading to increased brownouts and blackouts. The report adds that “unless substantial amounts of capital are invested over the next several decades in new generation, transmission, and distribution 264 facilities, service quality will degrade and costs will go up.” For the foreseeable future, the grid itself will become less stable, even though U.S. 264 utility companies are pushing infrastructure and smart grid technologies. This is certainly not a uniquely American problem, I would argue that the grid in many Asian countries, including China, face similar problems of ageing and 283 ANDREW PEARSON instability. As Ciabattoni et al. explain in their article A Smart Lighting System for Industrial 265 and Domestic Use , “lighting control is an efficient way to reduce energy consumption and to prevent energy waste, and it can be effectively used in conjunction with LED systems.” Successful lighting controls make use of motion 265 sensors, occupancy sensors and photosensors. An IR can install motion and occupancy sensors that detect activity within a certain area. Energy is saved because lights are turned off automatically when someone leaves a room or area. Photosensors can prevent outdoor lights from operating during daylight hours, and they offer the potential to control lights inside a mall or a meeting 265 hall by measuring ambient lighting levels. 255 http://www.traveldailymedia.com/243679/genting-dream-arrives-home-in-hongkong/ (Accessed 25 November 2016) 256 IBM Software Thought Leadership White Paper. Descriptive, predictive, prescriptive: Transforming asset and facilities management with analytics. October 2013. https://static.ibmserviceengage.com/TIW14162USEN.PDF 257 O’Gorman, Tristan. September 14, 2015. 5 Inventory analytics best practice to achieve inventory optimization. IBM Big Data & Analytics Hub. http://www.ibmbigdatahub.com/blog/5-inventory-analytics-best-practices-achieveinventory-optimization 258 Heneghan, Carolyn. September 6, 2016. Why the food and beverage industry needs to start using Internet of Things technology. www.fooddive.com. http://www.fooddive.com/news/why-the-food-and-beverage-industry-needs-to-startusing-internet-of-things/425693/ 259 http://www.fao.org/news/story/en/item/196402/icode/ 260 Adar, Shay. April 10, 2016. Smart Food Management Utilizes IoT to Reduce Cost, Waste and Pollution. CEVA. http://blog.ceva-dsp.com/smart-food-management-utilizesiot-to-reduce-cost-waste-and-pollution/ 261 How the Internet of Things (IT) Will Transform Inventory Management. www.clearspider.com http://www.clearspider.com/internet-of-things-inventorymanagement/ 262 Citron, Ryan. January 26, 2017. The Future of Smart Parking Is Integration With Automated Technology. Forbes. http://www.forbes.com/sites/pikeresearch/2017/01/26/smart-parking/#1a8d3c9e4961 263 https://www.hitachiinsightgroup.com/en-us/pdf/solution-profile/hitachi-solutionprofile-enabling-energy-resilience.pdf 264 http://www.infrastructurereportcard.org/2009/sites/default/files/RC2009_full_report.p df (Accessed 11 January 2017) 265 Ciabattoni, L., Freddi, A., Ippoliti, G., Marcantonic, M., Marchei, D., Monteriu, A., Pirro, M. A Smart Lighting System for Industrial and Domestic Use. Mechatronics (ICM), 2013 IEEE International Conference on http://202.171.252.18/s3.amazonaws.com/academia.edu.documents/44315995/A_sma 284 THE PREDICTIVE CASINO rt_lighting_system_for_industrial_a20160401-17463vaug5x.pdf?AWSAccessKeyId=AKIAIWOWYYGZ2Y53UL3A&Expires=1486352178&Signat ure=WKHkk0Lb1wzvDVEmy3iKu4MZHKg%3D&response-contentdisposition=inline%3B%20filename%3DA_smart_lighting_system_for_industrial_a.pdf 285 286 THE PREDICTIVE CASINO CHAPTER SEVEN The Predictive Casino “A large portion of marketing dollars are wasted when the wrong offers are made to the wrong people at the wrong place and time… The IoT will generate an enormous, truly unprecedented amount of precise information about buyers and their needs. It’s a marketer’s dream come true.” ~Jon Gettinger Sr. Vice President of Marketing Aria Systems The Customer Journey The customer journey starts a long time before the customer even enters the casino. It begins the moment a potential customer browses to a casino operator’s webpage or notices an advertisement for a casino on television, or on the Internet, or in print, or on a billboard thousands of miles away from the casino. It can even be while browsing a casino’s website, connecting with its social media accounts, or even the moment the customer actually enters the casino. From a few browser clickstrokes, a casino operator’s ecommerce department can create a click path analysis that reveals customer interactions on the casino’s websites. Descriptive analytical functionalities can then provide a deeper understanding of the customer journey. Column dependencies (standard in most of today’s Data Integration software tools) can visually display the strength of a relationship between attributes within any dataset. This helps users better understand the characteristics of their data and is often used to help target further analytics. A recommendation engine can help predict a person’s interest based on historical data from many users. This is useful in increasing client engagement, recommending more relevant choices and increasing customer satisfaction. For example, recommendations can predict interest in casino games, products, hotel rooms, and IR services. From checking into the hotel, to gambling (or even learning how to gamble), to eating in the IR’s restaurants, or drinking in its clubs or bars, or enjoying a show or an event, through to the moment the patron leaves the property, every interaction’s data should be collected. Throughout this entire customer 287 ANDREW PEARSON journey, the casino’s DW can help collect, analyze, visualize, and then, potentially, live stream recommendation content to those who need it. Like the proverbial butterfly who flaps its tiny wings in Brazil, and sets off a typhoon in Manila, any customer who shows any inkling towards visiting the property can be quantified and analyzed so that not only is their trip a rewarding one, but also that the customer touch points are reduced as much as possible so the labor needs of the IR are kept to a minimum. Once a customers has arrived in Macau, Las Vegas, Manila, or any casino town, he or she will leave a trial that can be picked up in a multitude of ways, including via social media, geo-fencing applications, facial recognition technology, or simply by checking into a hotel room. The swipe of a patron card in a slot machine or at a baccarat table should set off a whole sequence of events, including the building of an analytical model that takes into account how this one individual fits into an overall table games revenue management model. Rapid advancements in facial-recognition technology have reached the point where a single face can be compared against 36 million others in about one 266 second. A system made by Hitachi Kokusai Electric and reported by DigInfo TV shown at a security trade show recently was able to achieve this blazing speed by not wasting time on image processing—it takes visual data directly 266 from the camera to compare the face in real time. The software also groups faces with similar features, so it is able to narrow down the field of choices very quickly. The usefulness to the casino’s security enforcement is pretty obvious, but it can be used by multiple departments; facial recognition technology can be set up to send alerts to hosts, pit bosses, retail store clerks, or just about anyone necessary. Once a face has been recognized, alerts can be sent to casino and/or retail personnel through a mobile app or an SMS message. A screen can display the shopper’s or casino patron’s name, or a photo just taken from the video feed. Shopping preferences, and other details, like a customer’s average daily Theo or ADT, can also be surfaces onto a mobile device. Mobile Advertising Framework In their article Building a Big Data Analytics Service Framework for Mobile 140 Advertising and Marketing , Deng et al. discuss a domain-specific big data 140 service platform for mobile advertising and marketing. The system leverages the latest open source products like Hadoop and Apache Spark to create a big 140 data processing platform. “The core recommendation engine provides a training predictive model on a training set by using ML algorithms, such as 140 collaborative filtering, clustering and classification,” Deng et al. note. Lovelock and Wirtz’s “Wheel of Loyalty” concept and its three sequential 288 THE PREDICTIVE CASINO steps—building a foundation for loyalty, creating loyalty bonds, and identifying 109 and reducing factors that result in churn —should be kept in mind when building up the foundation of the Predictive Casino’s CRM system. The most important part of the second step is the cross-selling and bundling of products and a real-time stream processing recommendation engine could certainly help with that. System Architecture 17 Earlier in this book, I mentioned that Forrester Research considered the creation of a single repository containing structured and unstructured data about a consumer was one of the difficulties of personalization marketing, but the system below might be a good start. Deng et al’s system “enables location-based adverting to engage with its target customers by studying their profiles and dynamic behavior patterns. Unlike other data analytics engines, their system provides a holistic advertising recommendation approach for mobile users by providing a real-time, Big Data based solution for precise marketing and analysis. Their system uses state of the art Big Data technologies, such as MongoDB and Spark, over a cloud ”140 infrastructure. The outcome of this project consists of three parts: • • • Precise market advertising and analysis algorithms. Recommendation analytics algorithms. A prototype system that implements the proposed algorithms and 140 solutions based on location-based advertising solutions. The system includes the following function components to capture 1. 2. 3. 4. 5. 140 : Device Location User Profile and Interests, including gender, age, address, profession, interest, etc. Ad Publisher Products Item Sets. A recommendation engine that would predict which relevant advertisements would be provided to the end user. Customer-Oriented Requirements: This provides qualified ad recommendations based on user preferences, behavior, and insights from social profiles. The system tracks the geolocation of the user and, based on the location ID and location category, the recommendation engine would push appropriate ads to the end user. Figure 14 reveals how the Spark Streaming service would work, with inputs from Kafka, Flume, HDFS, Kinesis, and Twitter leading into a Spark database. As Deng et al. explain, “Streaming implements a data flow model in which data (time series facts) flows continuously through a topology (a network of transformation entities). The slice of data being analyzed at any moment in an aggregate function is specified by a sliding window, a concept in Complex Event 289 ANDREW PEARSON Processing (CEP). A sliding window may be as low as ‘last minute’, or ‘last 60 140 minutes’, which is constantly shifting over time.” Figure 14: Apache Spark streaming processes Source: Apache Spark As Deng et al. note, the recommendation engine contains an off-line training system that produces pre-aggregation for the ad recommendations that get 140 pushed out to the end users. “This real-time recommendation system will load in-stream data as training datasets. Moreover, real-time recommendation system can leverage pre-aggregation results produced by off-line batch mode 140 trained machine learning models,” Deng et al. explain. To fully support location-based mobile advertising capabilities, the “system needs several fundamental profile datasets, such as geography information for 140 describing several aspects of location.” There is also important information for targeting and ad mapping, such as application profiles, merchant profiles as 140 well as user profiles that need to be included. The ad related information is stored in MongoDB, a persistent data repository which is constantly being 140 updated. On the client-side, the system is built with browser-based HTML5 140 technology. Figure 15 shows a schematic of how the system would work. That is followed by Table 12, which shows a break down of Amazon’s AWS streaming service. 290 THE PREDICTIVE CASINO Ads UI Virtualization: VMS Cluster Workflow Navigation Customer Displayer Graph Displayer End-User Displayer Ads RESTful Service Internet Internet Ads RESTful API Loading balancer Ads Mobile Data Server Ads Recommendation Engine Spark Spark M-Lib streaming clustering (InColmemory) filtering Spark DB query Spark in-memory / Distributed system Pre-Processor Ads Data Cleaning Normalization Human Intervention Interface Ads Post-Processor Ads Data Reorganizer Optimization Caller Ads Sampling Engine Statistics Algorithms ORM Ads data caching Distributed File System (HDFS) Virtualization: VMS Cluster Figure 15: Database and streaming services with Hadoop and MongoDB data base Source: www.intelligencia.co Amazon Example Analytics Option Amazon Redshift Amazon Kinesis Description Used for analyzing global sales across product mixes, add clicks and impressions, social trends, along with the storage of historical stock market data and the aggregation of gaming data. Measures are provided or quality, operational efficiency and financial performance. Amazon Redshift is also compatible with many business intelligence systems and is designed for data warehouse workloads with structured data. Unstructured data may be prepared and structured for Amazon Redshift through the use of Amazon Elastic MapReduce. Used in processing real-time stream data for analysis. Stream data may be rapidly moved from data sources and 291 ANDREW PEARSON Amazon Elastic MapReduce Amazon DynamoDB Amazon Machine Learning continuously processed. Data may be transformed and redistributed, analyzed real-time, or decomposed and aggregated across data streams. Amazon Kinesis enables realtime analytics such as customer engagement and website clickstream. Since data is not batched application logs can be pushed directly to an Amazon Kinesis stream for processing. Data processed by Amazon Kinesis may be used in the extraction of metrics and generation of key performance indicators that feed real-time reports and dashboards. Data may be moved and stored through Amazon S3, Amazon Glacier, Amazon Redshift, or Amazon DynamoDB. Uses Apache Hadoop for providing a framework for running big data processing and analytics through the distribution of data sets across compute notes in a Hadoop cluster. The capability is typically used in risk modeling and analytics for threats, ad targeting and click stream, genomics, prediction, and ad-hoc data mining. Stores and retrieves large amounts of data with millisecond latency and is integrated with other services. This capability is commonly used for mobile apps, gaming, digital ads, sensor networks, online shopping carts and managing web sessions. Uses algorithms to find patterns in data for creating machine learning models used in predictive analytics. Predictions may be real-time or scaled. This technology can be used to build predictive models for detecting fraud, recommendations to customers based on prior actions, targeted market campaigns, automatically structuring information, identifying customer attrition risks and mitigations, and a variety of automated solution recommendations. Table 12: Amazon Web Services (AWS) for Big Data Analytics Source: Journal of Innovation Management Once the customer has arrived on property, CRM strategies should kick in. Bain 116 & Company recommend the following CRM strategies : 1. 2. 3. 4. “Define strategic ‘pain points’ in the CRM cycle. Focus on problems that have a large impact on customer satisfaction and loyalty, where solutions would lead to superior financial rewards and competitive advantage. Evaluate whether—and what kind of—CRM data can fix those pain points. Calculate the value that such information would bring the company. Select the appropriate technology platform, and calculate the cost of implementing it and training employees to use it. Assess whether the benefits of the CRM information outweigh the expense involved. 292 THE PREDICTIVE CASINO 5. 6. Design incentive programs to ensure that personnel are encouraged to participate in the CRM program. Many companies have discovered that realigning the organization away from product groups and toward a customer-centered structure improves the success of CRM. Measure CRM progress and impact. Aggressively monitor participation of key personnel in the CRM program. In addition, put measurement systems in place to track the improvement in customer profitability with the use of CRM. Once the data is collected, share the information widely with employees to encourage further participation in the 267 program.” 268 In their article “Knowing What to Sell, When, and to Whom ,” authors V. Kumar, R. Venkatesan, and W. Reinartz showed how, by simply understanding and tweaking behavioral patterns, they could increase the hit rate for offers and promotions to consumers, which then had an immediate impact on revenue. By applying statistical models based on the work of Nobel prize-winning economist Daniel McFadden, researchers accurately predicted not only a specific person’s purchasing habits, but also the specific time of the purchase to 268 an accuracy of 80%. Obviously, the potential to market to an individual when he or she is primed to accept the advertising is advantageous for both parties involved; marketers don’t waste time advertising to consumers when they aren’t primed to accept the advertisements, but do market to consumers when and where they might want to use the advertisements. Predictive modeling is only useful if it is deployed and it creates an action. Taking advantage of the more powerful, statistically based segmentation methods, customers can be segmented not only by dollar values, but also on all known information, which can include behavioral information gleaned from resort activities, as well as the patron’s simple demographic information. This more detailed segmentation allows for more targeted and customer-focused marketing campaigns. Models can be evaluated and reports generated on multiple statistical measures, such as neural networks, decision trees, genetic algorithms, the nearest neighbor method, rule induction, and lift and gains charts. Once built, scores can be generated in a variety of ways to facilitate quick and easy implementation. The projects themselves can be re-used and shared to facilitate faster model development and knowledge transfer. 269 In his paper Predictive Analytics , Wayne Eckerson advises creating predictive models by using the following six steps: 1. Define the business objectives and desired outcomes for the project and then translate them into predictive analytic objectives and tasks. 293 ANDREW PEARSON 2. 3. 4. 5. 6. Explore and analyze the source data to determine the most appropriate data and model building approach and then scope the effort. Prepare the data by selecting, extracting, and transforming the data, which will be the basis for the models. Build the models, as well as test and validate them. Deploy the models by applying them to the business decisions and processes. Manage and update the models accordingly. By utilizing data from past campaigns and measures generated by the predictive modeling process, casino operators can track actual campaign responses versus expected campaign responses, which can often prove wildly divergent. Additionally, casino operators can generate upper and lower “control” limits that can be used to automatically alert campaign managers when a campaign is over or underperforming, letting them focus on campaigns that specifically require attention. One of the benefits of automating campaigns is that offers based on either stated or inferred preferences of patrons can be developed. Analysis can identify which customers may be more responsive to a food/beverage offer, a room offer, and/or a free chip offer. The result: more individualized offers are sent out to the casino's patrons and, because these offers tap into a customer’s wants, desires, needs and expectations, they are more likely to be used; more offers used mean more successful campaigns. By understanding what type of patron is on its property, why they are there, and what they like to do while they are there, a casino operator can individualize its marketing campaigns so that they can be more effective, thereby increasing the casino property's ROI. With predictive analytics, casino operators can even predict which low-tier and mid-tier customers are likely to become the next high rollers. In so doing, casinos can afford to be more generous in their offers as they know that there is a high likelihood that these customers will appreciate the personalized attention and therefore become long term—and, hopefully, highly profitable— patrons. Once the patron leaves the property, the marketing cycle begins anew. RFM models can project the time at which a patron is likely to return and social media should be checked for any comments, likes or uploads, left by guests. All of a patron’s captured information can now become part of the Master Marketing Profile that will be the basis for future marketing efforts. Combing the daily, weekly and monthly Master Marketing Profiles will also allow the casino to develop insightful macro views of its data, views that could help with facilities, labor management and vendor needs. 294 THE PREDICTIVE CASINO Table Games Revenue Management As a casino executive explained to me one time, “we sell widgets, just like every other business in the world, it’s just that our widgets come in the shape cards, most of which have numbers on them.” The expenses incurred to make these widgets include the cost of the casino buildings, direct labor costs (dealers and supervisors) and equipment (cards, shuffler leases, etc.). 270 Historically, table games departments have been the heart of casinos and even though there has been a steady decline in table games revenue since 271 2000 , this is only true for non-Asian casinos. In places like Macau, the table 272 games/slot ratio is closer to 90/10 , so table games are a much more important revenue generator there compared to in the US. Since the highest-betting gamblers are estimated to worth 20 to 50 times more than their lower-betting counterparts, casinos want to ensure that these highbetting players always have a spot at a table, even though it would mean that 273 lower-betting gamblers may not get a chance to play. With so much money at stake it is rather amazing that most casinos base their table game minimums on the judgment of floor managers. Bill Zender warns in his article Table Game Management for the Small 274 Casino : “I can’t emphasize enough how important game pace is to maximize an operation’s revenue potential. The more hands you deal, the more decisions you achieve. The more decisions you achieve, the more revenue you earn. It’s as simple as that. Many executives seem to focus on outcomes. I’ve worked for executives who believed in dealing more slowly when you were losing. This theory is so far off base it’s not even funny. Even in a smaller table games operation, if you can achieve one more round per hour on every open table game, just one more round, you can increase your table game revenue by at least $50,000 annually.” Table 13 illustrates Zender’s point that “the revenue potential of a blackjack game based on several variables: average bet per wagered hand; hands wagered per round; mathematical house advantage based on decks, rules, and player error in percentage; rounds dealt per hour adjusted for the number of wagering positions on the table; and hourly labor expense estimated by adding the dealer’s hourly wage and benefits cost with one-fourth of the hourly cost 274 for the floor supervisor.” These variables are used to determine hourly revenue potential known as theoretical win (T-Win) and net return, also known as hourly profit. The net return is calculated before additional operating costs, such as cost of equipment, complimentary table service, table game promotions and customer reinvestments (comps). 295 ANDREW PEARSON Hands per Rd H/A% Rds Per Hr T-Win Labor Expense Net Return $1 7 1.5% 54 $5.67 $25 ($19) $1 1 1.5% 220 $3.30 $25 ($22) $2 7 1.5% 54 $11.34 $25 ($14) $2 1 1.5% 220 $6.60 $25 ($18) $3 7 1.5% 54 $17.01 $25 ($8) $3 1 1.5% 220 $9.90 $25 ($15) $4 7 1.5% 54 $22.68 $25 ($2) $4 1 1.5% 220 $13.20 $25 ($12) $5 7 1.5% 54 $28.35 $25 $3 $5 1 1.5% 220 $16.50 $25 ($9) $6 7 1.5% 54 $34.02 $25 $9 $6 1 1.5% 220 $19.80 $25 ($5) $7 7 1.5% 54 $36.69 $25 $15 $7 1 1.5% 220 $23.10 $25 ($2) $8 7 1.5% 54 $45.36 $25 $20 $8 1 1.5% 220 $26.40 $25 $1 Avg Bet Table 13: Bet Return Based n Average Played Wager in Blackjack Source: Table Game Management for the Small Casino 274 The results are quite stark, only if you have table minimums in the range of $6$8, with full tables will the casino start making money. As Zender argues, “Optimally, your game needs to generate an average bet of $8 per player before you can expect to overcome your labor cost and start paying for any complimentary service and player reinvestment. Heaven help the dead 274 games.” Several other studies have looked at the TGRM problem, including Daniel 275 Boykin’s Table Games Revenue Management: A Bayesian Approach , Chen et 276 al’s A revenue management model for casino table games , Haley & Inge’s 277 Revenue Management—It Really Should Be Called Profit Management and 278 Clayton Peister’s Table-games revenue management , not including the above-mentioned Zender article. For Boykin, “Revenue management (RM) is a complex process for optimizing revenue from a fixed inventory which has applications in various industries. It has its origins in the Airlines industry. By looking at the common practices of 296 THE PREDICTIVE CASINO RM in other industries, it is possible to develop a RM system that is applicable 275 to a table games department.” As Haley and Inge explain, “Affinia Hospitality saw revenues increase 17 percent over the prior year in the first month after implementing manual RM processes in a new central reservations office. The Millennium Bostonian Hotel paid back all of their start-up costs and more in the first month after converting from manual RM processes to an ASP-based service. Harrah’s Entertainment credited its RM system (installed in 2001) with increasing room and gaming 277 revenues 13 percent for the year in 2002.” This is real money and highly quantifiable ROI. Haley and Inge also note that, “Anyone selling a perishable product knows that you need to flex your pricing in accordance with demand, lead time, competitors and a host of other factors. Hotel rooms, airplane seats, advertising time, fresh produce and winter clothing are all subject to revenue 277 management tactics.” Table game seats fit well within this paradigm. Boykin states that or any RM system, there must be three set conditions 1. 2. 3. 275 : Differences in customers Variation in demand Product perishability and a fixed production “If all the customers of a business or industry are uniform then there is very little to maximize,” Boykin explains. “The airlines utilize price differencing between business travelers and recreational travelers. This includes price differencing for days of the week, time of the year, even time of day. Hotels 275 also take advantage of variations among customers,” Boykin adds. Casinos should be able to as well. Besides the “variations between businesses vs. recreational travelers, even within these broad groupings there are a significant 275 number of variations based upon demographics.” As Boykin argues, “In areas where you have greater variations, there is a greater potential to exploit that 275 variation and therefore a greater potential to maximize revenues.” TGRM issues include: • • • • The problem of which tables to open on the casino floor, and for each table what bet minimums and bet maximums to assign. The trade off is between having limits that are too high, thus excluding some customers who would otherwise have bet, and limits that are too low, thus leaving some money in the customer’s pocket. The optimization problem is a simple maximization of revenue, given knowledge of the demand for each bet size. Labor management needs must be added to the model to understand dealer needs. For an IR, customers can be divided into gaming play segments, i.e., type of 297 ANDREW PEARSON game, style of play, and minimum price per hand play. Boykin explains, “Variations in demand is evident through observation over a period. The limited demand and perishability of the product is inherent to the nature of a 275 game.” The second condition—a variation in demand—certainly affects the casino 275 industry just as it does the airline and hotel industry. “If the amount of demand is known, accurately and consistently, then there is no need for a sophisticated tool to maximize the revenue from that demand,” but unknown demand provides RM opportunity. Just like the airline industry, the casino industry is dependent on fluctuating seasonal and sometimes even daily 275 demand. “The greater the inability to forecast demand accurately, the larger 275 the risk of management not maximizing revenues,” says Boykin. The third condition—product perishability and fixed production—is certainly true about the casino industry, as it cannot simply add new tables or double 275 book a seat to satisfy excessive demand. Likewise, the perishability of the “product”, e.g., one poker hand, one dice roll, or one spin of a roulette wheel, 275 is obvious as well. Once the event occurs, there is no opportunity to recover 278 the bet that was not placed. Boykin states that, “The aspects of table games that call for the use of a revenue management system are the variable demand of the games and the 275 variable betting threshold of each player.” Previous research has also looked at other variables such as win per available seat hour, or length of each play 276 278 session; the Chen et al and Peister studies looked into the development of table games revenue management systems, but neither used the existing 275 counts and player database to determine the minimums. Boykin argues that “By predicting the next time period’s expected demand in terms of head count, then applying the percentage of players at each average bet level, the number of players expected at each betting level can be predicted. After optimizing and considering overall house advantage, we can 275 maximize profit for the next time period in a proactive manner.” As Peister argues, “Since table games generate revenue through a house advantage built into the game itself on each wager placed, maximizing the wagers placed in an hour is a critical component of any revenue management 278 system for table games.” In his paper, Peister established the win per available seat hour (WPASH), and 278 looked to maximize casino win per seat hour. By manipulating the table minimums and number of open games, Peister created a distribution that sacrifices a few seats to increase the number of hands dealt at a table, while 278 maximizing the casino win. Peister also identified a major data issue for any potential revenue manager—actual demand is censored because when demand exceeds capacity it is impossible to tell how many players the casino 298 THE PREDICTIVE CASINO and, therefore, how much money the casino loses due to a player’s inability to 278 find a seat. By utilizing the Cox survival regression, Peister was able to predict the survival of each seat per hour, i.e., the likelihood of it remaining 278 vacant throughout the entire hour. The main weakness of this model was that the mathematical calculations were complicated and would prove difficult to anyone who didn’t have extensive 275 mathematical and statistical training. Peister acknowledged and accepted these limitation, but argued that one of the primary reasons the Cox Regression model was used was because of an unknown underlying distribution of the 275 players. Chen et al. took a different approach, choosing to measure Theo win rather 276 than gross win and they utilized Croston’s method instead. With this 276 method, Chen et al. were able to forecast intermittent demand. The authors differed from Peister in that they compared their simulated results to actual 276 revenue numbers from a casino. Then they estimated game demand at any 276 given hour through a ratio of the two equations. Once demand was determined, a maximization equation was then applied “to determine the maximum house advantage for the given demand by adjusting the spots per 275 table, minimum bet, average wager, and table limit.” A table-opening plan for the shift manager was developed from this, which was based upon the 275 forecasted demand, for the maximization of house advantage. In the Chen et al. study, the casino could potentially bring in more than sixteen thousand dollars in theoretical incremental revenue at the blackjack tables on any given 276 day, which would represent a considerable increase in potential revenue. 275 Like Peister method, Chen et al.’s study also has some drawbacks. Primarily, in their simulations, they assumed uniform distribution of betting between 276 table minimum and table maximum , which would rarely, if ever, happen in 275 real life. “In fact, most Shift Managers would look upon results based on this 275 assumption as highly suspicious,” Boykin warns. Another drawback is that Chen et al.’s method would require extensive data collection and it would be labor intensive. Boykin argues that “Even though this is a simpler method, this weakness still leaves room for improvement in a table games revenue 275 management system.” 275 For Boykin, analysis of the data begins with the analysis of demand. “Hourly demand data is inherently a time series collection issue, Boykin states, arguing 275 that, “a time series analysis would be a logical plan for forecasting demand.” “However, after looking at the data, the realization that the miscellaneous variables to produce a reliable enough prediction model through time series analysis would be cumbersome and limited in scope,” Boykin warned. “This would not suit the needs of creating an operationally efficient model, which would require the ability to update quickly and with flexibility. Additionally, 275 time series regression requires data to be consecutive.” “This either requires 299 ANDREW PEARSON the casino to start tracking hourly head counts or have a large block of 275 consecutive hourly head counts in a recent time period.” Tracking head counts would actually be a good idea and it is certainly possible with video analytics technology, but that option wasn’t available to Boykin so he chose an 275 alternative method. Boykin argues that, “A Bayesian approach allows one to utilize expert opinion and prior knowledge of a system, and is quickly and easily adaptable by using historical data and prior information to predict the demand for the next time period and each subsequent observation, updating the model and prediction 275 using the next set of observed data. According to Boykin, “This allows for a 275 very flexible model that would adapt itself based on recent observations.” The model itself was set to predict expected demand for each part of the day in 275 conjunction with player distribution. As Boykin explains : “Once the player betting data is segmented, the percentage of players in each player cluster is then calculated. Since this is representative of the entire population of players for the casino, the assumption is that any random sample of sufficient size would have the same player distribution. Therefore, the percentages in each cluster can be applied to the expected number of players for each day part. This would give the Shift Manager the expected number of players for each betting level at each day part. The Shift Manager could then determine if tables would need to be opened to accommodate future players, and at what level the table minimums should be set.” Boykin concluded that, “By using Bayesian techniques, it is possible to develop a revenue management system that would reduce uncertainty in the Shift Manager’s estimations of future business. By reducing this uncertainty, the department can maximize revenues in a proactive manner, and the system can be used as a tool to assist casino managers in the better management of the 275 table games department.” The revenue management model could also be improved “by using observational data gathered by observing players as they enter the property and noting what attracts certain players to certain games. This could be incorporated into the database so the model could then classify players to each 275 game type,” recommended Boykin. Once again, video analytics would be perfect for this. Social Media As I have stated throughout this book, a casino operator is only as strong as its weakest customer relationship and casino operators should look to foster 300 THE PREDICTIVE CASINO stronger relationships with their current and future customers by taking advantage of the marketing opportunities that mobile and social media offers. One should also keep in mind Lovelock and Wirtz’s Wheel of Loyalty when 109 developing CRM systems and goals. These include building a foundation for loyalty, creating loyalty bonds as well as identifying and reducing the factors 109 that result in customer churn. Customer satisfaction is the foundation of true customer loyalty, while customer dissatisfaction is the key factor that drives 109 customers away. This may sound obvious, but its importance cannot be overstated. Figure 16 shows how a casino would engage its customers in a loyalty platform that utilizes social media as an important part of the process. Rules Engine Define and apply rules for rewards economy, geofencing, frequency, influence and more Listening Define and listen for triggers such as photos, hashtags, keywords, topics, likes, video views, pins, comments, links, sharing, check-ins, geo-posts, and more on social and messaging channels Moderation Rewards & Content Moderate posts for UGC quality, consumer profiles, and pick rewards before conversion Deliver points, discounts, reminders, digital content, sweeps, contest entries and more, frictionlessly in real-time Automation Messaging Data & Analytics Automate responses, rewards and conversion at scale and in the moment Smart email, social and messaging responses from your brand based on campaign and user criteria as well as language translation Acquire social identity tied to customer records. Measure reach, impressions, conversion rates and more Figure 16: Casino Engagement and Loyalty Platform Source: chirpify.com Through mobile and social media analytics, casino companies can create a single customer view that helps produce one-to-one, personalized marketing, which many would consider the Holy Grail of advertising. Marketing to the “customer of one” is one of the major slogans being bandied about by software companies these days and, although it might sound simple, it is anything but. When a casino operator decides to go social, it should remember Kaplan and 23 Haenlein's five specific points of social media; be active, be interesting, be humble, don't be afraid to be unprofessional and, most of all, be honest. I would agree with all of those, although “unprofessionalism” has its time and place. Dishonesty can be discovered quickly and it can reverberate through the social media world with devastating consequences so, in this case, honesty is always the best policy. 301 ANDREW PEARSON Casino operators have jumped into social media, but only in a limited way and the rest of this book contains ideas on how to utilize at least five of the six platforms 23 that Kaplan and Haenlein discuss. When blogging and microblogging, companies should follow the eight best practices Twitter recommends when businesses use its service to build a following, increase a reputation, and raise customer trust. When building a social media marketing campaign, Kaplan and Haenlein’s four steps of social media—listen, join, participate and create—should be kept in 171 mind. Throughout the rest of this chapter, I will gives ideas on what types of topics and/or influencers should be courted for these specific industries, where—and if—they are appropriate. 66 In her book, The Mobile Marketing Handbook , Kim Dushinski lists eight types of advertising campaigns that a mobile marketer can engage in—voice, text messaging, the mobile web, mobile search, mobile advertising, mobile publicity, social networking, and proximity marketing. To these eight, I would add another two—OTT and mobile apps marketing. Search is always going to be an important part of a company’s marketing plan and, for that reason alone, creating a Google+ presence is imperative. “Putting a little time and energy into you’re a casino’s Google Plus page can lead to improved local search capability. Pages that do well on Google Plus receive a higher index on Google search. And that’s not all, Google Plus content— meaning the posts you share on your page–can show up in search results in 279 instances where your website may not.” Casino operators should take advantage of the golden opportunity Google has offered them—“prime placement on the right-hand side of search results, with photos and 280 promotional posts.” This presents a very good opportunity for brands to get in on the ground floor and get prime advertising placement right now. Search engines are constantly looking for updated information on websites and adding things like blogs and customer forums is a cheap and effective way to get customers and/or clients to generate new content for you. Social networks empower users to create communities of shared interest and building communities where customers can talk about a casino’s events, games and/or services is a very cost effective way to market a casino and/or an integrated resort. This will also keep one’s ranking high in the search engines. Search advertising falls into two main types—natural search results, and paid 281 sponsorship based on keywords. “Natural search requires high-quality, constantly updated content and search engine optimization (SEO). Paid search requires work to optimize keyword choice and messaging, but can be 281 phenomenally expensive.” When not optimized for conversion, this can be a 281 very pricey channel to use, with low conversion rates as well. 302 THE PREDICTIVE CASINO Advertising on social media, blogging, microblogging, and social networks has the advantage of being highly targetable, but, as this is new technology, best 281 practices are still emerging. “Advertising on microblogging channels such as Twitter works much like an ongoing conversation. Conversations can be initiated based on users’ interests or questions, and monitored with set 281 searches on keywords.” Advertising on social media usually includes higher conversion rates and lower costs, although volumes tend to be smaller than 281 with other channels. . On its website, Twitter recommends building a following, increasing a businesses' reputation, and raising a customer's trust by following these best 186 practices : 1. 2. 3. 4. 5. 6. 7. 8. Share: disseminate photos and behind the scenes info about your business. Even better, give a glimpse of developing projects and events. Users come to Twitter to get and share the latest, so give it to them! Listen: regularly monitor the comments about your company, brand, and products. Ask: question your followers to glean valuable insights and show them that you are listening. Respond: reply to compliments and feedback in real time. Reward: Tweet updates about special offers, discounts and timesensitive deals. Demonstrate wider leadership and know-how: Reference articles and links about the bigger picture as it relates to your business. Champion your stakeholders: Retweet and publicly reply to great tweets posted by your followers and customers. Establish the right voice: Twitter users tend to prefer a direct, genuine, and, of course, likable tone from your business, but think about your voice as you Tweet. How do you want your business to appear to the Twitter community? As per mediakix’ 9 Ways Brands Are Advertising On Snapchat best ways to advertise on Snapchat: 1. 2. 3. 4. 5. 6. Snap Ads Geofilters Lenses Discover Channels Live Channels With Influencers a. Takeover b. Promotion c. Unboxing d. Product Placements. 303 282 , here are the ANDREW PEARSON Social Media Influencers In its article The 5 Different Types of Influencer Marketing Campaigns, Mediakix claims that there are probably a limitless amount of ways for brands to create effective influencer marketing campaign, but in general, these campaigns fall 283 within one of the following five subcategories : 1. 2. 3. 4. 5. Product Placement—this involves incorporating a company's product, services, or logo into a digital influencer's content just as it has been done in the film industry for decades. Just like actors in films, social media stars have earned the trust of their followers, so product placements are an excellent opportunity for brands to gain valuable exposure to millions of engaged consumers through the influencer’s YouTuber, yy.com, Instagram, Viner, or Snapchat account. Contests, Giveaways, Sweepstakes—Hosting social media contests like giveaways, sweepstakes, or best-of contests, such as best photograph, video, or blog competitions can generate buzz about a casino brand, as well as foster goodwill among consumers. These contests compel social media users to take a specific action (like following the brand’s channel or increasing company exposure by using branded hashtags). Aligning with a social media influencer, an IR can promote a contest that will leverage the social media star's large follower base and ensure that consumers participate in the campaign. Theme/Hashtag Campaign—Hashtags are great ways to build a theme around a campaign. Focusing each influencer marketing campaign around a central theme or hashtag that is leveraged throughout all of the social channels helps build cohesion and encourages consumers to get involved by using the brand's hashtag in their own content. As Mediakix recommends, “Developing and implementing an influencer marketing campaign around a memorable branded hashtag is one of the best ways brands can facilitate a genuine social conversation and increase brand exposure, especially if the hashtag happens to go 283 viral.” Creative Influencer Campaign–These give the social media star much more freedom to create content and these campaigns usually center around a specific concept or idea. Done right, these campaigns allow the digital influencer to interpret themes to create unique brandsponsored content, leading to increased levels of engagement from the social media influencer's followers and/or subscribers. Campaign to Build Social Followers—casino brands can invite social media influencers to expose new audiences to their brand's social media accounts. Snapchat Takeovers—having a social media influencer "take over" a brand's Snapchat account for a set period of time—is one of the most effective ways for businesses to reach 304 THE PREDICTIVE CASINO thousands or millions of new followers as well as organically grow their own Snapchat follower base. The Integrated Resort Casino Casino and hotel operators can use social media to manage their brand, enhance brand loyalty, as well as engage both their current customers and their potential customers. The social media world is also the perfect place to harvest customer feedback, provide real-time customer service, build fanbases and drive traffic to a casino’s website. Blogs and micro-blogging sites are also important mobile and social media channels and casinos should monitor Twitter feeds for both their satisfied and 284 unsatisfied customers. In his article Casinos Saving Face Online , L. Benson states that “Social media Web sites such as Facebook and Twitter are changing the face of customer relations at major Las Vegas hotels.” “Resorts are setting up their own fan pages where executives can monitor customer suggestions on how to improve business, bask in guests’ kudos, offer immediate assistance to customers in distress—and cringe when unhappy patrons post critical remarks 284 that ding their companies.” As part of the ongoing dialogue that a casino should foster, the casino patrons are doing their part, “with their comments and reviews not only reaching casino managers but an untold number of other customers and potential 284 customers over whom they can now wield influence.” Casino executives can no longer carefully craft resort marketing messages, but, as the following example from Benson shows, they have allies in their patron ranks coming to 284 their defense : “A gambler ranted on South Point’s Facebook page last week, “Please folks ... do N-O-T gamble in this casino. They run some of the TIGHTEST machines in Las Vegas. I LOSE almost E-V-ER-Y time I try playing at South Point.” South Point managers read every word—and let the comment ride. Before they could engage the man in a public debate about the competitiveness of their slot machines, another customer came to the property’s defense minutes later: “If you don’t like the South Point that much—then just don’t go there. But the rest of us LOVE the place ... better luck next time.” The tweeter who came to the defense of the casino could have been rewarded for his or her loyalty. Perhaps tweeting him or her some free matching slot play may have been seen as too blatant an act of quid-pro-quo, but there is no 305 ANDREW PEARSON reason why (if the casino can link this gambler with his or her patron profile) South Point didn’t offer a nice reward for such loyalty in a future marketing campaign. Casino companies should also feel compelled to reward their customers through Facebook, Twitter, WeChat, and Weibo or any number of blogging and micro-blogging services. The beauty of using these channels is the ability of the customer to share these awards or stories of these awards with friends and contacts. It wouldn’t be that hard to do, either, as a casino can ask patrons for their social media accounts upon sign up. Jones and Sasser warn that, “Extremely dissatisfied customers can turn into ‘terrorists,’ providing an abundance of negative feedback about the service 114 provider.” Through social media channels, negative feedback can reverberate around the world within seconds. Today, more than ever, casinos must spot dissatisfied customers and approach them before they do irreparable harm to the company’s image and reputation and social media is one of the best channels in which to engage them. Casinos need to empower their patrons to post on Facebook or WeChat or Twitter or comment about their experience and, hopefully, turn them into apostles. In Jones and Sasser’s zone of affection, satisfaction levels are high and “customers may have such high attitudinal loyalty that they don’t look for 114 alternative service.” It is within this group that “Apostles”—members who praise the firm in public—reside and this is the group that is responsible for 115 improved future business performance. A simple search of the Twitter feed on the multiple services I mentioned in the previous chapter will probably reveal a list of patrons who could be courted for marketing purposes. Dovetailing the above, another example from Benson’s article is of a woman who posted on her Twitter page that she had “just touched down” in Las 284 Vegas. “Because Twitter posts can be monitored by keywords, a Wynn Las Vegas employee was able to immediately respond: ‘Welcome! Come on over to 284 our resort to explore and discover. You won’t be disappointed.’” Having a social media monitoring command center is a must these days. Searching for keywords like “Las Vegas” or “Macao” or “baccarat” or any of a hundred other 284 iterations that reference gambling could be a good start. In Macau, the Chinese government’s restrictions on gambling wouldn’t come in to play as Twitter is not officially available in China, but it is used considerably in Hong Kong and throughout the ASEAN region. As for building fanbases, “Big brands—including casinos—that don’t develop social media programs do so at their peril, said Jennifer Van Grove, an associate 284 editor at Mashable.com.” With over 8,000 Twitter followers, Van Grove makes the point that if she posts something, some of her followers are going to 284 reply and may share her Tweet with their followers. As she so succinctly 306 THE PREDICTIVE CASINO warns, “You could have a whole chain of comments based on one incident. 284 These hotels have to be involved.” Real time technology gives hotels and casinos the ability to see—and know— what is going on in real time around them, and this allows them to easily counter negative perceptions instantly. As was the case with the negative South Point Casino and Resort diatribe, the countering positive comments would be considered more trustworthy as they came from an actual user, rather than a press release from the criticized company, which are usually, and, understandably, viewed with skepticism. “There is a great upside for 284 companies that go about it the right way, Van Grove argues. “Social media can hold hotels more accountable to their customers, fix problems, correct 284 misconceptions and build loyalty,” she adds. Social media marketing makes good economic sense as well. Given the explosive growth of social media sites, “these might become more cost284 effective than using traditional advertising and marketing methods.” Social media is also universal, for every Facebook in the US, there is a corresponding RenRen or WeChat in Asia, yet there is no reason why a casino in Macau isn’t on Facebook today; in reality, most of them already are. To maintain credibility with customers, casino companies shouldn’t remove negative comments or constructive criticism from these social media sites unless the person posting the comment uses foul language or says something 284 offensive to others. “Everyone’s entitled to their opinion,” Sally Gaughan, South Point’s director of Internet marketing, said about the negative slot 284 machine comment. “We wanted to give people a place to talk about the 284 South Point and we wanted it to be genuine.” Facebook should be a part of every casino’s social and mobile media marketing plan, but simply putting up a Facebook page won’t cut it these days; creativity and uniqueness are needed to get noticed in today’s highly competitive social media market. Gamification is also a good way to stand out from the crowd. Facebook bots can also ad customer service channel that can answer common customer questions. A few years ago, California's Pechanga Resort & Casino used Facebook to increase brand awareness in a unique and innovative way. Pechanga created Slot Wars™, an interactive Facebook game that allowed fans to “battle” with each other on competing slot machines. Players could customize their slot machines by uploading personal images and these became part of the slot reels. At its peak, Slot Wars™ had over 10,000 active players from Southern California alone. Pechanga also saw significant increases in its on-property slot play, which was, the casino believed, directly attributable to its Facebook presence. Mohegan Sun also used Facebook to attract patrons. Through the Facebook 307 ANDREW PEARSON Connect feature, Mohegan Sun pulled a user's Facebook profile into its Shine Maker app, where a customized experience was created for him or her. By asking users to rate their desires according to a series of six scales, a customized video catered specifically to the individual was built. The individual could then distribute the video to his or her social media friends and acquaintances, spreading word of the casino’s brand far and wide. Perhaps as more of a marketing gimmick than anything else, in August of 2013, The Casino at The Empire in London, performed the world’s first ever casino 285 crowdsourcing experiment. Leveraging the powers of social media, The Casino at The Empire gave Dave Sargeant (the lucky punter who earned his role as a social casino lab rat via a Facebook contest) “£1K with which to wager, 285 with any winnings going straight into his pocket.” “The kicker was his every wagering decision—whether to hit, stay, pick red or black or play a three card 285 poker hand—would be crowdsourced via the Empire’s Twitter followers.” “It took three hours for Sargeant to go through his stack of chips, making a total of 63 wagers on the advice of his digital backseat drivers. The net result was a £932.50 payday for Sargeant, who said, ‘All that tweeting was hard work’ but 285 he wasn’t complaining.” By giving customers instant access to the information they need when they need it most, a casino can enhance their patron’s on-property experience. Whether its patrons are on-property to play baccarat, poker, blackjack, slots, bingo, or if they want to gamble in the sports book, these instant messaging services can provide a patron with instant information that can not only enhance their experience but, potentially, shape it. Bars, Coffee Shops, Nightclubs, and Restaurants The continuous mission for every bar, coffee shop, nightclub and restaurant owner is to constantly increase sales. With so many similar drinking and dining establishments vying for a customer’s attention, effective marketing measures are key not only to attracting new patrons, but also to increasing loyalty from current patrons. Effective marketing helps foster an establishment's identity and enhances customer loyalty. Word-of-mouth marketing is the most cost-efficient way to increase sales of a bar, coffee shop, nightclub, or restaurant, but this kind of marketing works both ways: negative reviews often travel faster than positive ones so vigilance is necessary in the social media arena. Pictures of food, along with babies and animals, are probably the most shared images on social media, especially on Chinese blogging and social media sites. Restaurants, in particular, should set up blogs that showcase their operations and their upcoming events. Chefs today have become rock stars and it probably wouldn’t take much to convince a restaurant’s executive chef to blog about his 308 THE PREDICTIVE CASINO or her work process. Blogging does, once again, create new content for the search engines and this could be an important differentiator. Facebook is a great place for bars and dining establishments to build a presence as well. “Facebook Places Pages are for businesses that actually have a physical location and, unlike other business pages, they contain an area that 286 allows users to ‘check-in’ at the businesses’ physical location.” “Check-ins” inform a customer’s friends of his or her location. This means that the business’ customer base are virally spreading the word about the venue and since this taps into the second most trusted form of advertising around–customer reviews—these recommendations have a decent chance of being acted 286 upon. However, do be aware that a “places” page may currently exist for your business because, when someone checks into a place that doesn't already 286 have a Page, a new Page gets created to represent the location as well. Hashtags can be used in tweets or on Instagram posts to connect a person’s 287 social media behavior directly with a bar. This will get a bar’s customers 287 promoting the drinking establishment. With the correct hashtag, a bar can reach customers who are not currently finding them on Twitter, Snapchat, or 287 Instagram. Getting people to use a bar’s name in its messaging will amplify 287 awareness of the establishment. On the OTT front, bars, coffee shops, nightclubs, and restaurants can use WeChat QR codes to encourage customers to sign up for branded WeChat 193 accounts, which can help collect contacts for the brand’s CRM system. For a restaurant or café, this could mean sending out a flash alert to followers about a promotion taking place in the next hour for a free coffee, or a two-for-one discount. Once users come to the restaurant or café, they can pay on location, provided they have the payments function set up on their WeChat account. It makes the retail process much simpler. Brands can take pre-payments or set up 193 small sales without cash registers. 288 A common objection to QR codes is that they are rather ugly. “It is true that the standard out-of-the-box QR code isn’t particularly aesthetically pleasing, but brands can improve the appearance with clever design tricks that ensure scalability and beauty. By going with a custom-designed QR code, restaurants can represent their style, while also reassuring customers that the restaurant 288 has put some thought into the experience.” As Cliff Kuang explains in his article Disney’s $1 billion bet on a magical 24 wristband , “If you want to imagine how the world will look in just a few years, once our cell phones become the keepers of both our money and identity, skip Silicon Valley and book a ticket to Orlando. Go to Disney World. Then, reserve a meal at a restaurant called Be Our Guest, using the Disney World app to order your food in advance.” “If you’re wearing your Disney MagicBand and you’ve made a reservation, a 309 ANDREW PEARSON host will greet you at the drawbridge and already know your name—Welcome Mr. Tanner! She’ll be followed by another smiling person—Sit anywhere you like! Neither will mention that, by some mysterious power, your food will find 24 you,” explains Kuang. How did all of this work? Kuang reveals that the MagicBands on the guest’s wrists are tech-studded wristbands that feature a long-range radio that can 24 transmit more than 40 feet in every direction. The host received an alert on 24 her modified iPhone that the family was just a few paces away. The kitchen 24 was also notified so they queued up the order. When the family sat down at their chosen table, “a radio receiver in the table picked up the signals from their MagicBands and triangulated their location using another receiver in the 24 ceiling.” “The server—as in waitperson, not computer array—knew what they ordered before they even approached the restaurant and knew where they 24 were sitting,” explains Kuang. It all works so perfectly, it is almost as if it were magic—and that’s exactly what Disney want you to think it is. What’s lost in all of this wonderful service is the fact that Disney has been able to streamline and optimize the entire process, getting the restaurant orders hours before it needs to be made, which could help with inventory control. This also cuts out the need for waiters to take the orders, reducing labor needs. All-in-all, the process is producing a friction-free environment that works on multiple levels, not just wowing the guests. Retailers The majority of retailers worldwide now realize that mobile is an important marketing and selling channel for them and mobile is really the only channel that helps consumers move from online-to-offline seamlessly. “Research shows retailers will generate approximately $689 billion through mobile-influenced sales in 2016. Currently, 58 percent of smartphone owners use mobile for in289 store related shopping and are more likely to convert in-store as a result.” In her article Engage Customers and Gain Advocates Through Social Media and 290 Social Networking , Wendy Neuberger argues that: “Social commerce is about making a retailer’s brand a destination. Retailers really need to listen to what their customers are saying. Customers can provide valuable input and feedback that can be used to make more informed assortment decisions, changes to Website features and enhancements to the shopping experience.” “When customers feel their voice is being heard, they feel a stronger 290 connection to the retailer and are more likely to become advocates.” Neuberger claims it is important for retailers to identify and engage with the key influencers for several reasons, the two most important being: “to empower their advocacy or capabilities, which helps build and foster a sense of community among brand loyalists, and empowers those loyalists to better 310 THE PREDICTIVE CASINO advocate on behalf of a brand, product and/or service.” 290 Neuberger recommends that retailers use the following social media 290 platforms : • • • • • • Blogs: retailers can provide additional product or category information here as well as post how-to information in the form of text, photos and/or videos. Retailers should also provide space for customers to add feedback and/or comments about their retailing experience. Micro-blogging: coupons, sales and promotions can be offered through these channels. Retailers can “‘tweet press releases, provide exclusive tips and tricks to customers, and ask for customer feedback, suggestions or ideas for improvements. Some retailers even use 290 Twitter as a customer service mechanism.” Co-Shopping: this is a form of social shopping and it enables two people—a customer and sales associate or two shoppers in different locations—to share a joint shopping session using live instant messaging such as Skype, WeChat or any number of other OTT services. Widgets: these are tiny applications that can be embedded into a website, blog or social network that are portable and relatively inexpensive to create. Social Bridging: anyone who has signed into a website using their Facebook, Pinterest or Twitter account knows what social bridging is. “This level of authentication provides enough credentials to participate in the social elements of the site. Additional authentication is required to complete a shopping transaction due to the sensitivity of the content included in a shopper’s account. Social bridging can be used to drive traffic and engage existing and new customers. It can access a user's identity, their social graph, and stream activities such as purchases and other social participation on the retailer’s site” says 290 Neuberger. In-Store Kiosks and Flat Panels can be provided to enable customers to 290 use social networking tools from within a store. Mobile and social media marketing has the potential to revolutionize the paradigms of retailing from ones in which the customer must physically enter the retailing environment to one in which a retailer can enter the consumer’s 291 environment through anytime, anywhere devices such as the mobile device. Instead of filling newspapers and magazines with advertisements, a retailer should create databases of opted-in customers who have the propensity to purchase goods in the future. By tracking OTT, SMS opt-in use, a retailer can get immediate and highly quantifiable data on who has signed up for their mobile coupons, who has opted-in to their mobile campaigns, who has used 311 ANDREW PEARSON their mobile coupons, and who might be planning to attend certain advertised events. 292 In her article Retailers Doing It Right in Social Media , Cherise Luter advises that whether done on Wordpress or Tumblr, blogging is an important part of a 292 retailer’s social media strategy. Luter adds : “Of all the places a company can build a lifestyle around their brand, a blog is the best suited. On Tumblr, Labrotatoria's Musing On... and The Classroom's Stay Classy Houston are great examples. They both share inspirational style images, product info and behind the scenes posts. CakewalkStyle Shop has three blogs, yes three. Lifes a Cakewalk is its main blog, Influencers focuses on bloggers and taste makers, and Style Guide shares its latest store items and trends.” The best part about being on Twitter or Weibo or any of the other instant 292 messaging services is the ability to interact with a customer in real time. This can help “Resolve customer service issues, get a pat on the back, or valuable 292 feedback all from the comfort of home or the store backroom.” “The retailer Lilly Rain is really great about interacting with its customers online. They retweet and reply to the messages and even re-post blogger websites that show them love. They also have an active Google+ page, which is great for 292 Google searches,” Luter explains. Luter also advises that “Twitter should not be ignored. Use Twitter to tweet out ‘instant sales.’ Get the word out about leftover products, new products in limited supply, or last-minute sales. Believe it or not—your customers ARE using Twitter. And if they see something great at your store—they WILL 292 retweet.” A customer's retail experience could be improved by giving the customer instant access to her information when and where she needs it. For example, QR Codes could be attached to an item's price tag and, once a customer takes a picture of the tag, their mobile phone could connect to the retailer's CRM system and a return text with very specific product information could be sent to them. This information could include which celebrities wear the label, what other colors the item is available in and what other accessories might match the outfit. Neuberger argues that it is very important to monitor the market conversation to understand what the marketplace is (or isn’t) saying about a retailer (their 290 brand, products, services, etc.). “Retailers need to understand the tone and impact of the conversation and begin to identify areas of opportunity for helping shape that conversation and gather valuable market intelligence,” says 290 Neuberger. 312 THE PREDICTIVE CASINO For Neuberger, “Social media metrics include sentiment, activity, share-ofvoice, and thematic content of online conversations. Trends and key influencers (“mavens”) and the most active sites/blogs are identified and tracked. By understanding the impact, retailers will have a way of identifying measurable progress, quantifying the return on social media investment, and 290 enabling benchmarking against future efforts.” Several retailers are also venturing into video-casting. For example Estée Lauder’s Clinique brand launched the 40 episode drama series, Sufei’s Diary, on 196 a dedicated Web site that broadcast daily. “While skin care was part of the story line and products were prominently featured, Sufei’s Diary was seen as entertainment—not a Clinique advertisement—and has been viewed online more than 21 million times. Clinique’s online brand awareness is now 27 percent higher than that of its competitors, although social-media content 196 costs significantly less than a traditional advertising campaign.” Another good example of Big Data use in retail is profiled in Bernard Marr’s Big 7 Data in Practice. In it, Marr describes the story of Pendleton & Son, a local 7 butcher based in north-west London. Established in 1996, the butcher shop had enjoyed a steady customer base and good reputation for years, but in 2014, a supermarket chain store moved onto the same street, and it affected 7 overall footfall and hit revenues hard. As Marr explains, “While founder Tom Pendleton was certain his shop offered superior quality and choice compared to the supermarket, the trouble was conveying this message to the public and getting customers through the door. Trying to compete on price wasn’t working and, with falling income, son Aaron 7 Pendleton turned to data to help keep the business afloat.” A Big Data consultant suggested “installing simple, inexpensive sensors inside the store window to monitor footfall and measure the impact of window 7 displays and promotions.” Using this sensor data as well as internal data such as transaction and stock data, the butcher shop was able to “measure how many people walked past the shop, how many stopped to look at the window display and sandwich board sign and how many people then came into the 7 store as a result.” Armed with this knowledge, the butcher shop was able to 7 refine its displays and messaging based on what most interested customers. The insight on one’s current business was almost eclipsed by sensor data that 7 pointed to an unexpected and potential new revenue stream. “As two popular pubs were located at the end of the street, the hours of 9 p.m. to midnight proved particularly busy in terms of passers-by—almost as many as the busy lunchtime period. So the Pendletons decided to trial opening at night and serving premium hot dogs and burgers to hungry folk making their way home 7 from the pub.” Analytics was even used to decide menu items; “In order to decide on what 313 ANDREW PEARSON products to offer at night, Aaron analyzed trend data from Google Trends to see what food items were particularly popular. This led to the creation of their 7 pulled pork burger with chorizo.” Going forward, the Pendletons were hoping to expand their use of data in 7 order to increase their knowledge of customers even further. Weather data is now included to predict demand even more accurately and a customer loyalty 7 app was in the works. Pendleton & Sons are a perfect example of how a company can step into analytics and they are, literally, being richly rewarded 7 for their leap of faith. The data revealed some interesting behavioral information on their customers 7 as well. “In short, the Pendletons found that local customers favoured inspiration and ideas over cheap deals, which were available every day in the supermarket. They were able to use this insight to improve their messaging and get more people through the door—and those who entered the shop were far 7 more likely to make a purchase as a result.” In addition, the late-night openings proved enormously popular and the company decided to make this a permanent feature on Friday and Saturday nights. Not only did this provide much-needed additional revenue, it also introduced the company and their products to a whole new set of customers. When the IR knows its customer base so well, it has the power to manipulate the demography of its clientele so that a concept like a pop-up store might be much more profitable than a stand-alone retailer. For example, if the IR was hosting an eSports tournament one week, it would make sense to have several stores filled with eSports products, i.e., video games, hardware items like computers, and gaming paraphernalia, but the following week might be a musical act from Korea, so typical K-Pop merchandise would be fitting. The Future The Predictive Casino is a casino that takes into account how a patron who walks through the door of the casino, or onto a bus, or is even driving down I15 toward Vegas will affect every facet of a casino property. It follows a customer before arrival, through his or her entire stay, then keeps tabs on them once they leave. As Kahle states, eventually we're going to set a time 122 frame on the sales funnel that never expires. From the moment of first contact, when the casino’s systems capture an IP Address, through capturing the social ID, to understanding the social activity, all the way through to the patron card sign up process so that the casino can understand gaming and commerce behavior (see Figure 16). The only thing remaining is to capture post-transaction information if and when it comes in. 314 THE PREDICTIVE CASINO Once all of this data is captured, geofencing, location-based services, and location-based advertising offer some unique ways to reach consumers when they are in the all-important “decision mode.” Personally, I am not a fan of checking into locations for something as illusory as a virtual badge, but that is only my opinion, many others seem to enjoy the gamification aspects of this process. There are plenty of people who are more than happy to give away their personal information for something as ethereal a virtual badge or an opportunity to win something of value so why not take it. A casino company that is able to offer highly specific advertisements to customers who might just need a little extra nudge to make that purchase should find an investment in geofencing applications very profitable. Geofencing marketing does raise the issue of privacy, but I believe the day will come when mobile users who enter a mall, a casino, or a shopping area will view messages that ask for their permission to accept a location-based ad as harmlessly as they currently view television and radio advertising. It is very possible that they will embrace this form of advertising because of its immediacy. Regarding privacy issues, Kuang does have a point when he claims that, “No matter how often we say we’re creeped out by technology, we tend to acclimate quickly if it delivers what we 24 want before we want it,” particularly if it has to do with context-aware 24 technology. “Just consider how little anyone seems to mind now that the Google Maps app mines your Gmail. Today, Google Maps is studded with your location searches, events you’ve arranged with friends, and landmarks you’ve chatted about. It’s delightful, and it took hold faster than the goosebumps could. The utility seems so obvious, your consent has simply been assumed,” 24 notes Kuang. Figure 17: Customer funnel Geotrigger services will probably be the next iteration of geofencing applications. These send targeted location-based messages to app users who 315 ANDREW PEARSON either enter or leave a geofenced area. Geotriggers can send the right message to the right person at the right time in the right place, which should increase their use as well as prove ROI positive. In the Predictive Casino, a patron can sign into the casino WeChat or a standard casino-branded app and be able to pull up his player card points balance. He or she can receive and use coupons at onsite restaurants and bars, as well as to sign up for gaming tournaments. Several casinos currently have branded games and/or apps that let patrons play for free or for money and trade in game points for real rooms or free food. By giving customers instant access to the information they need when they need it most, a casino can enhance their patron’s on-property experience. Whether the casino’s patrons are on-property to play baccarat, poker, blackjack, slots, bingo, or if they want to gamble in the sports book, these instant messaging services can provide a patron with instant information that can not only enhance their experience but, potentially, shape it. As previously mentioned, perhaps one of the best uses of location-based services is in the MICE space. The massive size of some exhibition halls can make finding a particular booth or floor section a daunting proposition. Indoor mobile communication technology with location awareness technology can 52 help conference-goers navigate a vast conference floor. Before arriving at a conference, a mobile user would be able to register his personal preferences and, once he enters the exhibition hall, a route map would be sent to his mobile phone. Vendor appointments could even be set up so that they are located near each other so that the conference-goer wouldn’t have to run around frantically trying to make meetings that are spread out all over the 52 convention floor. At 10:00 am on November 3, 2016, the first skill-based gaming machine, 293 Danger Arena, went live at Harrah’s Resort in Atlantic City and it ushered in a whole new concept of gaming. Danger Arena is the first offering from New York-based company GameCo and its three three-seat carousels have been 293 given prominent spaces on casino’s gaming floor. As the executives at the event stated, Danger Arena represented the first step in the next evolution of the gaming floor, and the hope was it would appeal to younger audiences who 293 weren't showing up to play slots — the millennial and Generation X gamers. In her article Atlantic City welcomes GameCo’s skill-based games, Clare Fitzgerald states that, “Despite being the newest thing on the gambling scene, Danger Arena's arcade-style cabinet and comic book graphics give it a charmingly retro appearance, and the black controller, with its two thumbsticks for navigation and aim, will be instantly familiar to anyone who's played on a 293 PS2, X-box or, frankly, almost any modern two-handed console system.” The game itself is a pretty straightforward “first-person action” game, where 316 THE PREDICTIVE CASINO 293 you try to kill as many robots as you can in 90 seconds, explains Fitzgerald. “Some of the robots are invincible, and shooting them only stuns them, making the minimum six kills needed to cash just a bit more difficult to achieve. As the player, however, you're also invincible, so the robots can't kill you before the 293 clock runs out,” states Fitzgerald. There’s also a “chance to hit a randomly generated bonus payout, which takes the form of a power ball lurking within 293 some of the maps,” so even in this skill-based game there are elements of luck involved in beating it. Only time will tell whether skill-based games are the future of gaming, but one thing is certain, with the younger generation showing little interest in slot machines, something new has to be brought to the gaming floor to keep fresh blood coming through the doors. Perhaps as a sign of things to come in real-time marketing, when MTV announced the nominees of its 2014 Video Music Awards, it—no surprise here—leaned on social media to create buzz. What was surprising, however, was its choice of platform; MTV didn’t choose to let its 50,000,000 Facebook fans or its 11,000,000 Twitter followers know first, but rather it selected a 294 social media channel that had only 150,000 MTV followers—Snapchat. Perhaps this isn’t that surprising as MTV must be constantly creating a sense of cachet and cutting-edge exclusivity about itself to keep its younger fans engaged and Snapchat would have been considered the hip channel of that moment. “When going to a new place like Snapchat where we have a smaller community that we want to grow, we reward people for being the first followers there,” said Tom Fishman, MTV’s VP of content marketing and fan 294 engagement. Personally, I see this as a sign of things to come. “MTV isn’t alone in turning to smaller social platforms for real-time marketing. Hyundia’s World Cup strategy focused on Tumblr, while ESPN recently made use of Pinterest for posting pictures from the red carpet of the ESPYs. However, like an image on Snapchat, fame is fleeting in the social media sphere and there are already thousands of startups toiling away in obscurity in many corners of the world, working through the night on what they hope to be next year’s next big thing. It is just 294 the nature of that business. And it is constantly changing. Pinterest is currently beta testing for paid media, 294 which will sell promoted pins for events. “Snapchat does not carry paid ads yet, but industry insiders expect an amplification product from the platform 294 within the year.” “Real-time marketing cannot just be about owned and earned—it needs to harness paid,” argues Jordan Bitterman, chief strategy 294 officer at WPP’s Mindshare. Besides Apple and Google, Facebook is now entering the acquisition business full bore. Along with its Instagram buy, the Oculus Rift purchase further shows 317 ANDREW PEARSON that Facebook is obsessed with staying relevant by buying the next big thing argues Paul Berry, founder and CEO of New York City-based social publishing 295 platform RebelMouse. “Through this and other acquisitions, Berry thinks Facebook will become a brand-holding company in the future, similar to 295 Viacom or Hearst.” “I see them, better than anyone else, using their market capitalization to create even bigger market cap for the Instagrams or 295 WhatsApps,” he says. The purchase of Occulus Rift should help Facebook grab a large piece of the 295 advertising market. In five years, Arvind Bhatia, managing director of equity research at Sterne Agee, expects Facebook’s “graph search to become bigger, 295 and the company to make more inroads in e-commerce.” “Then, with its ancillary networks like WhatsApp and Instagram, Facebook will be able to run 295 its own platform, rather than operate through Android and iOS.” “They are being bypassed on the mobile platform,” Bhatia says. “They want to be the next Android, and so the only way to do that was to start from scratch, and 295 that’s what they’re doing with this virtual reality technology.” Just as Facebook might morph into a complex conglomeration of services and 295 brands, Twitter should continue to differentiate itself through its simplicity. First, the network has to focus on monetization, and in “Berry’s view, over the 295 next five years Twitter’s monetization will be faster than its user adoption.” “They are so simple and so public, and it works so much data around what you are interested in—the fact that it’s all public is perfect for advertising,” Berry 295 says. “I think it achieves being the short, public, status updates of the 295 Internet,” Berry adds. Meanwhile, Twitter’s decelerating growth worries Bhatia. “While he thinks the network will continue to grow, he believes it won’t equal Facebook’s user 295 295 base.” “And already Instagram is catching up,” he says. “While Facebook is going to become one of the four horsemen of technology, Twitter will be an 295 interesting, decent-sized company, but not a mainstay, if you will.” What 295 really sets Twitter apart is its openness. “As data science continues to boom, this openness will give Twitter increasing utility, and because of this he sees the long-shot possibility that the service gets bought out by a giant organization, like the United Nations, for dissemination and collection of data on a scale 295 never imagined.” 89 With their Cluetrain Manifesto , Rick Levine, Christopher Lock, Doc Searle, and David Weinberger warned that not only are markets conversations but the Internet is revolutionizing the way businesses communicate with their customers and if businesses don't adapt and treat their customers with respect, their customers will desert them. What better way to treat them with respect than to listen to them and respond accordingly, which mobile and social media channels do better than any other form of advertising. 318 THE PREDICTIVE CASINO In Greek mythology, once Pandora opened the box and let out all of the evils of the world, the one thing remaining inside was hope. Now, hope is not exactly what the casino industry is selling, except maybe to a small minority, but entertainment is. Today, more and more casino companies are referring to themselves as “entertainment companies” rather than gaming companies, not only because of the stigma attached to gambling and wagering, but also because of the new reality of the business; i.e., it is a multi-faceted, multi-offering business, and gambling is only one piece of the pie. Admittedly, a substantial piece of the pie, especially in places like Macau, but it’s not the entire raison d'être. More and more, gambling is becoming entertainment and casino companies have to deal with customers interested in a multitude of entertainment possibilities besides gambling. I wrote this book hoping that it would help clients and, possibly, potential clients discover the treasure that is hidden within their data, so that they could understand that what seems like a daunting set of challenges should actually not be feared and that there is incredible technology out there that can make every patron’s visit a personalized adventure that will make them return and return and return, which should bring a smile to every casino executive’s face. 266 Bea, Francis. March 25, 2012. Goodbye, anonymity: latest surveillance tech can search up to 36 million faces per second. www.digitaltrends.com http://www.digitaltrends.com/cool-tech/goodbye-anonymity-latest-surveillance-techcan-search-up-to-36-million-faces-per-second/ 267 Bain & Company. Customer Relationship Management. June 10, 2015. http://www.bain.com/publications/articles/management-tools-customer-relationshipmanagement.aspx 268 Kumar, V. V. (2006). Knowing What to Sell, When, and to Whom. Harvard Business Review. 269 Eckerson, Wayne. 2007. Predictive Analytics, Extending the Value of Your Data Warehouse Investment. TDWI Best Practices Report. https://www.sas.com/events/cm/174390/assets/102892_0107.pdf 270 Lucas, A. F., & Kilby, J. (2012). Introduction to casino management (1st ed.). Escondido, CA: Okie International, Inc. 271 Schwartz, D. G. (2013). Nevada gaming revenues: Long-term trends. Retrieved From: University of Nevada Las Vegas; Center for Gaming Research. 272 http://www.dicj.gov.mo/web/en/information/DadosEstat/2016/content.html#n1 273 Ferguson, Mark E., Richard Metters, and Carolyn R. Crystal. 2008. The “killer application” of revenue management: Harrah’s Cherokee Casino & Hotel. March 13, 2009. http://smartech.gatech.edu/handle/1853/18983 274 Zender, Bill. 2013. Table Game Management for the Small Casino. Billzender.com. http://www.billzender.com/os/resources/media/table_game_mgmt_small_casino_12_ 2013_p-3.pdf 319 ANDREW PEARSON 275 Boykin, Daryl. Table Games Revenue Management: A Bayesian Approach. May 1, 2014. UNLV Theses. http://digitalscholarship.unlv.edu/cgi/viewcontent.cgi?article=3062&context=thesesdiss ertations 276 Chen, M., Tsai, H., & McCain, S. C. (2012). A revenue management model for casino table games. Cornell Hospitality Quarterly, 53(2), 144-153. 277 Haley, M. & Inge, Jon. 2004. Revenue Management - It Really Should Be Called Profit Management. Hospitality Upgrade. http://www.hospitalityupgrade.com/_magazine/magazine_Detail.asp?ID=194. 278 Peister, C. (2007). Table-games revenue management. Cornell Hotel & Restaurant Administration Quarterly, 48(1), 70-87. 279 The Bright Blue Wave Team. (2014, May 22). Why Small Businesses Need Google Plus. Retrieved from brightbluewave.com: http://brightbluewave.com/blog/smallbusinesses-need-google-plus/ 280 Miller, C. C. (2014, February 14). The Plus in Google Plus? It’s Mostly for Google. Retrieved from New York Times: http://www.nytimes.com/2014/02/15/technology/theplus-in-google-plus-its-mostly-for-google.html?_r= 281 Vindicia. (2014). Digital Age/Digital Goods. Retrieved from vindicia: http://info.vindicia.com/White-Paper---Digital-Age-Digital-Goods9essentials_for_acquiring_subscription_and_recurring_revenue_customers.html 282 How to Advertise on Snapchat: A Brand Guide. http://mediakix.com/2016/06/howto-advertise-on-snapchat-marketing-infographic/#gs.IXCsq2I 283 The 5 Different Types of Influencer Marketing Campaigns. March 30, 2016. Mediakix.com. http://mediakix.com/2016/03/influencer-marketing-campaigns-5examples/#gs.Lz0k6B4 284 Benson, L. (2009, October 26). Casinos saving face online. Retrieved from Las Vegas Sun: www.lasvegassun.com/news/2009/oct/26/saving-face-online/ 285 Amsel, P. (2013, August 14). Zynga Slingo death watch; Empire’s social casino experiment a net win. Retrieved from Calvinayre.com: http://calvinayre.com/2013/08/14/business/empire-social-casino-crowdsourcingexperiment/ 286 Miller, J. (2012, August 3). 5 Top Tips For Marketing Your Bar Or Restaurant On Facebook. Retrieved from beeliked.com: http://beeliked.com/index.php/social-mediabuzz/5-top-tips-for-marketing-your-bar-or-restaurant-onfacebook/#QeE9gA38BZgxHYfb.99 287 Moran, G. (2013, February 18). Craft Beer Marketing | How To Use Social Media To Promote Your Craft Beer Bar. Retrieved from Marketingthing.com: http://marketingthink.com/craft-beer-marketing-how-to-use-social-media-to-promoteyour-craft-beer-bar/ 288 Chan, H. (2012, August 3). 4 Ways Restaurants Should Use QR Codes. Retrieved from mashable.com: http://mashable.com/2012/03/08/qr-codes-restaurants/ 289 Abramovich, G. (2013, 11 June). 5 retailers that get mobile. Retrieved from Digiday: http://digiday.com/brands/5-retailers-mobile/ 290 Neuberger, W. (n.d.). Engage Customers and gain advocates through social media and social networking. January 24, 2013. 320 THE PREDICTIVE CASINO ftp://public.dhe.ibm.com/software/solutions/soa/newsletter/2010/newsletter-mar10article_social_media.pdf 291 Shankar, V. V. (2010). Mobile Marketing in the Retailing Environment: Current Insightes and Future Research Avenues. The Journal of Interactive Marketing. 292 Luter, C. (2014, March 10). Retailers Doing It Right On Social Media. Retrieved from HoustonPressBlogs: http://blogs.houstonpress.com/artattack/2014/03/retailers_doing_it_right_on_so.php? page=2 293 Fitzgerald, Claire. November 3, 2016. Atlantic City welcomes GameCo’s skill-based games. Online.casinocity.com. http://online.casinocity.com/article/atlantic-citywelcomes-gamecos-skill-based-games-125055 294 Sloane, G. (2013, July 28). Real-Time Marketing Isn't Just About Twitter. Retrieved from adweek.com: http://www.adweek.com/news/technology/real-time-marketingisnt-just-about-twitter-159140 295 Pullen, J. P. (2014, April 2). What Will Social Media's Giants Look Like in 5 or 10 years? Retrieved from fortune.com: http://fortune.com/2014/04/02/what-will-social-mediasgiants-look-like-in-5-or-10-years/ 321 322 ABOUT THE AUTHOR ANDREW PEARSON was born in Pakistan, grew up in Singapore and was educated in England and America. With a degree in psychology from UCLA, Pearson has had a varied career in IT, marketing, mobile technology, social media and entertainment. In 2011, Pearson relocated to Hong Kong to open Qualex Asia Limited, bringing its parent company's experience into the ASEAN region. Pearson is currently the Managing Director of Intelligencia Limited, a leading implementer of BI, CI, data warehousing, data modeling, predictive analytics, data visualization, digital marketing, mobile, social media and cloud solutions for the cruise line, gaming, healthcare, telecommunication, hospitality and retail industries. Pearson has also leveraged Intelligencia’s expertise to implement software solutions from such vendors as SAS, Hitachi, SAP, Qlik, and Microsoft at some of the world's biggest casino operators and lottery companies. In 2010, Pearson wrote The Mobile Revolution and it was published by Qualex Publishing. In 2013, Pearson was invited to write a chapter in Global Mobile: Applications and Innovations for the Worldwide Mobile Ecosystem, a book on mobile technology. The book, which was co-authored by several of the mobile industry's leading figures, was published in July 2013. This is Pearson’s fourth book. Pearson is also a noted columnist, writing on topics such as mobile media, social media, predictive analytics and cloud technology for such publications as ComputerWorld HK, The Journal of Mobile and Social Media Marketing and The Mobile Marketer. Pearson is also the president of the Advanced Analytics Association of Macau. An avid traveler, Pearson is a sought-after speaker on such disparate topics as casino and mobile marketing, data analytics, social media, gaming, and interactivity. If he's not pounding the pavements of Hollywood, he's probably wandering the labyrinthine streets of Hong Kong's Lang Kwai Fong, or tearing up useless betting slips at Happy Valley (perhaps the most perfectly named racecourse in the world (for some)), or dining at a hawker center in Singapore, or doubling down at the gaming tables in Macau. Basically, Pearson's trying to find the next great story that the world doesn't yet know that it desperately wants to see. Social Media LinkedIn: andrew-pearson-96513a3 Twitter: intelligenciaMD Academia: PearsonAndrew Blog: medium.com/@intelligentsiaf www.intelligencia.co 323 ANDREW PEARSON 324 THE PREDICTIVE CASINO INDEX Behavioral patterns, 293 Betfair, 198 Bitcoin, 46, 47, 90 Blockchain, 3, 47 Blogging, 171, 192, 237 Blogosphere, 191, 214, 237 Blogroll, 190 Blogs, 26, 171, 207, 236 Bluetooth, 68 advertising, 315 Bookmarking, 171 Bookmarking websites Delicious.com, 171 Brand management, 183, 264 Apostles, 103, 112, 221, 306 Buckley, Ralf, 29, 32 Bwin.party, 35 5 51.com, 176 56.com, 85 A Activity tracing, 66 Adobe, 109, 110, 134, 135, 136, 167, 172, 189 Advertising banner ads, 61 definition, 61 Amazon, 1, 11, 12, 31, 74, 78, 82, 97, 197, 273, 274, 290, 291, 292 American Express, 252 American Marketing Association, 61 Analytics descriptive analytics, 4, 6, 139 diagnostic analytics, 4, 6, 139 edge analytics, 6, 58, 91, 154 predictive analytics, 1, 4, 6, 28, 95, 117, 118, 126, 132, 135, 139, 140, 154, 292, 294, 323 prescriptive analytics, 4, 6, 58, 139, 140 Text analytics, 127 Apple, 67, 69, 85 Artificial Intelligence, 36, 38 Ashton, Kevin, 8, 9, 31, 227 Astra beer, 50 Attribution analysis, 128, 129 Augmented Reality, 2, 40, 41, 42, 44, 45, 89, 90 wearable, 42 C Caesars, 3, 31, 108, 109, 110, 115, 133, 134 Carroll, Dave, 226 Chauhan, Alok S., 151 China Confucian society, 178 mobile subscriber count, 177 social media subscribers, 178 Chinese mindset, 179 Cicso, 251 Classification of Social Media, 174 Clickstream, 14, 65, 89, 117, 126, 127, 132, 133, 139, 291 Clickstream analysis, 16, 132 Click-thrus, 61 Cloud computing, 61 Cntv.cn, 85 Collaborative projects, 26 Common Short Codes, 67 Concept of self-presentation, 174 Consumer-Generated Media, 172 B Baidu, 189 Bebo, 171 Behavioral information, 293 325 ANDREW PEARSON blogs, 172 digital video, 173 mobile phone photography, 173 news, 172 online encyclopedias, 173 podcasting, 173 user reviews, 173 Content communities, 26 Content management, 171 Cookies, 239 CRM, 4, 13, 16, 21, 36, 64, 69, 72, 77, 96, 98, 99, 101, 103, 104, 105, 110, 113, 114, 140, 161, 183, 196, 197, 208, 261, 289, 292, 293, 301, 309, 312 CrowdOptic, 40, 44, 45 Cryptocurrencies, 46, 47 CSC, 20, 67 Cultural exchanges, 66, 88 Customer Acquisition Model, 156 Customer analytics, 19, 62, 63, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 248, 253, 262 Customer behavior, 14, 19, 63, 104, 123 Customer Centric Relationship Management, 103 Customer churn, 14, 17, 101, 103, 112, 117, 123, 124, 126, 129, 164, 167, 183, 289, 301 Customer Churn Model, 163 Customer Conversion Model, 159 Customer dissatisfaction, 102, 112, 221, 301 Customer Loyalty, 111, 123, 128, 129, 167 Customer satisfaction, 102, 112, 221, 301 Customer segmentation, 117, 155 Dell, 212, 251 Digital Nomads Campaign, 212 social media command center, 251 Deloitte, 42, 89 Deng, Xiaoping, 179 Digital Interactive Marketing: The Five Paradigms, 65 Disney, 27, 28, 32, 49 MagicBand, 27 Domino’s Social Media Crisis, 228 Drury, Glen, 171 Dushinski, Kim, 67, 302 E Eckerson, Wayne, 293 Emarsys, 21, 22, 32, 246 F Facebook, 4, 14, 22, 26, 48, 49, 50, 51, 55, 62, 66, 69, 71, 77, 78, 81, 85, 88, 92, 99, 101, 102, 108, 109, 113, 117, 127, 128, 132, 171, 173, 175, 176, 177, 181, 183, 184, 185, 189, 192, 194, 195, 198, 199, 200, 202, 203, 208, 212, 214, 219, 220, 221, 222, 223, 224, 232, 233, 234, 235, 237, 239, 240, 243, 244, 246, 247, 251, 253, 256, 257, 258, 259, 260, 261, 263, 264, 265, 270, 271, 305, 306, 307, 308, 309, 311, 317, 318, 320 Mohegan Sun’s use of, 307 Pechanga Resort’s use of, 307 Facebook Connect, 307 Facial recognition, 17, 47, 48, 49, 90 Flickr, 171 folksonomy, 187 Ford, 249 Forrester Research, 19, 20 Four Steps of Social Media, The, 184 create, 186 join, 185 listening, 184 participating, 186 Foursquare, 51, 52, 55, 77, 88, 130, 177, 182, 194, 239, 243, 258 D Data lake, 4, 5, 10, 12, 13, 14, 28, 96 Decision tree construction of, 142 Deighton, John A., 62, 65, 66, 88 Delicious.com, 188 326 THE PREDICTIVE CASINO G J Gartner, 2, 3, 10, 11, 31, 41, 45, 56, 68, 89, 90, 91 Hype Cycle for Emerging Technologies, 2 Geofencing, 51, 52, 90, 177 Giaglis, George, 52 Global Trust in Advertising, 113 Google, 54, 62, 65, 67, 69, 74, 78, 79, 85, 192, 200, 214, 224, 235, 312 Greenberg, Paul, 99 Gulbransen, Scott, 252 Jantsch's Hierarchy of Social Media, 209 JavaScript, 239 JetBlue Airlines, 244 Jiepang, 26, 52, 198 Jones, Thomas, 102, 112 Juniper Research, 173 K Kaixin, 176, 198 Kaplan, Andreas, 301 k-nearest neighbors, 146 Knowing What to Sell, When, and to Whom, 293 Kornfeld, Leora, 62, 65, 66, 88 Ku6, 26, 85 Kumar, R., 293 H Hadoop, 5, 11, 12, 15, 28, 32, 40, 79, 80, 81, 82, 127, 128, 253, 288, 291, 292 Haenlein, Michael, 301 Halbherr, Michael, 35 Hashtags, 309 Hexum, 26 Hierarchy of Social Marketing, 209 High rollers, 294 Hitachi, iii, 18, 82, 83, 283, 288, 323 L Lafayette de Mete, Boye, 178 Letv.com, 85 Lift and gains charts, 293 Lighthouse Signal Systems, 53 LinkedIn, 22, 99, 127, 128, 173, 185, 186, 194, 198, 212, 214, 223, 235, 237 Location analytics, 136, 137 Location-aware advertising, 74 Location-aware technology applications, 55 museum applications, 55 Location-based services, 55, 315, 316 advantages for businesses, 55 advertising, 315 Lovelock, Christopher, 98, 111 Loyalty, 15, 22, 27, 37, 48, 51, 52, 64, 69, 70, 96, 101, 102, 103, 104, 106, 109, 111, 112, 119, 121, 122, 123, 124, 128, 129, 133, 134, 137, 182, 183, 196, 208, 215, 220, 221, 245, 253, 255, 262, 273, 289, 292, 301, 305, 306, 307, 308, 314 I iBeacon, 74, 75, 92 IBM, 2, 10, 11, 18, 31, 35, 36, 64, 81, 82, 83, 87, 90, 110, 117, 123, 128, 129, 136, 141, 167, 253, 254, 271, 275, 276, 284 InfiniteInsights, 128 In-memory computing, 127 Instagram, 194, 198, 214, 222, 243, 259, 260, 304, 317, 318 Internet, 66 Internet of Things, 8, 56 Intromercials, 62 Inventory optimization, 277 IoT, iii, 4, 9, 10, 12, 13, 28, 36, 37, 56, 57, 58, 59, 60, 86, 87, 91, 93, 154, 155, 274, 275, 278, 279, 280, 284 Iqivi.com, 85 iTunes store, 74 327 ANDREW PEARSON mobile advertising, 67 mobile publicity, 68, 302 mobile search, 67 mobile web, 67 opting-in, 67 persona planning, 88 privacy, 315 proximity marketing, 68 push media model, 67 social networking, 68 text messaging, 67 types of advertising campaigns, 67, 302 voice, 67 Mobile Payments, 68, 91 Mobile positioning technology kinds of, 52 Mobile search, 78 Mobile TV, 85 Mobile value chain, 61 Mtime.com, 85 MySpace, 62, 127, 171, 185 M Machine Learning, 1, 3, 36, 38, 151, 168, 247, 281, 292 Maker’s Mark, 199, 204 Malshe, Ashwin, 223, 224, 225, 226, 227, 228, 229 MapReduce, 128 Market segmentation, 105, 106, 108, 153 Marketing campaigns, 19, 63, 118, 253, 293 automating campaigns, 294 Marketing promotions, 14, 104 Marr, Bernard, 7, 313 Mastercard, 252 Media Richness Theory, 174 Message boards, 171 Metcalf model, 25 Metcalf, Bob, 25 Micro-blog, 26, 192 Micro-blogging, 171 Microsoft, 35, 110 MMS, 20, 54 Mobile digital advertising, 61 Mobile advertising broad-based brand advertising campaign, 62, 88 campaigns, 62 Interactive, direct response campaign, 62, 88 targeted search advertising, 62, 88 Mobile Advertising Framework, 288 Mobile Advertising, the book, 60, 61, 62, 63, 77, 88 Mobile analytics, 118, 239 what it is, 239 Mobile commerce personalization, 19, 95 Mobile coupons, 311, 312 Mobile Marketing, 60, 67, 78, 91, 302, 321 advantages, 67 Digital Interactive Marketing: The Five Paradigms, 65 location-specific marketing messages, 68 marketing campaigns, 67 N Napster, 66 Net seed, 250 Netflix, 252 Neural networks types of training, 152 Neural Networks in Data Mining, 151 O Omni-commerce, 20 Optimizing Offers, 164 Oracle, 35, 110 OTT, 18, 20, 54, 68, 72, 302, 309, 311 P Page tracking, 239 Page views, 239 Patron Worth Model, 162 Pengyou, 198 Pentaho, 5, 18 Periscope, 84 Pinterest, 51, 117, 194, 198, 200, 202, 328 THE PREDICTIVE CASINO 208, 214, 222, 223, 237, 246, 247, 258, 260, 268, 311, 317 Podcasts, 73, 74, 171 Pokémon Go, 45 Post-roll video, 62 Predictive analytics k-Means cluster, 143 k-nearest neighbors, 145 Predictive analytics, 128 A/B Testing, 148 Decision Trees, 142 Discriminant Analysis, 152 Logistic Regression, 147 Neural networks, 151 Survival or Duration Analysis, 153, 154 Time Series Model, 150 Predictive asset maintenance, 39, 59, 154, 278 Predictive modeling, 294 predictive models, 293 segmentation methods, 293 Predictive models six steps of creating them, 293 Procter & Gamble, 61 Propensity to Respond Model, 158 Property exchanges, 66, 88 Proximity marketing, 68, 74 Push technology, 21 Pyra Labs, 190 Python, 6, 40, 117, 128, 138 Reichheld, Frederick, 111, 220 Reinartz, V.K., 293 Reinforcement learning, 39 RenRen, 176, 182, 198 RetailNext, 136, 137 RFM Models, 157 S Salesforce.com, 110 SAP, 35, 110, 117, 128 Sarnoff network, 25 Sarnoff, David, 25 Sasser, W. Earl, 102, 111, 112, 306 SCRM, 96, 99, 113, 183 Second Life, 173, 175, 230 Sentiment Analysis, 129, 130, 167 SEOMoz’s, 188 Sharma, Chetan, 60, 61 Short Message Service, 67 Sina Weibo, 176, 182 Singh, Dr. Yashpal, 151 Siri, 67 Site analysis, 239 Sitecore, 63, 64 Six Types of Social Media, 187 Blogs, 189 Collaborative Projects, 187, 188 Content Communities, 193 Social Networks, 194 Slideshare, 186, 215, 223 Smart Energy, 283 Smart Parking, 282 Smith, P.R., 24, 25, 32, 136, 269 SMS, 20, 22, 51, 54, 67, 237, 311 Snapchat, 175, 198 Social bookmarking, 171, 187, 188, 189, 202, 210 Social exchanges, 66, 88 Social marketing potential, 199 Social media, 68, 171, 172, 198, 199 collaborative and dynamic communication model, 171 The six types of, 26 two-way dialogue, 171 Social media analytics, 247, 253 Social Media Influencers, 304 Q Qieke, 26, 198 QlikView, 117 QQ, 26 QQ Bookmarks, 189 QR codes, 18, 20, 61 Qzone, 176, 198 R Real-time marketing, 18, 35, 76, 77, 255, 317 Real-time Technology, 76 Reed network, 25 Reed, David, 25 329 ANDREW PEARSON Social media monitoring, 99 Social Media Monitoring Tools, 255 Social Media Theory the 1:9:90 rule, 211 Social Media uses Add Interactivity to a Website, 214 Brand and Anti-Brand management, 215 Brand Loyalty Enhancement, 220 Build Fanbases, 221 Crisis Management, 223 Develop a Virtual Social World Presence, 229 Discover Important Brand Trends, 60, 235 Engage Customers and Potential Customers, 236 Harvest Customer Feedback, 238 Reputation Management, 241 Social Media Monitoring, 251 Social Shopping, 244 Social network analysis, 247 Social Network Sites Definition, 194 Social networking, 68, 171, 192, 302 Social networking sites, 26, 68 Social networks, 171 Social presence, 174 Source-Message-Channel-Receiver model, 171 Spark, 11, 18, 40, 82, 288, 289, 290 SPSS, 128 Starbucks, 51, 101, 216, 221, 248 Stodder, David, 118, 123, 132 Stream Analytics, 79 Stream Processing, 15, 32, 79, 81 comparison of services, 81 Streaming content, 84 Stumbleupon, 188 SugarCRM, 110 Supervised learning, 38 Sutton, Scott, 37, 89, 149, 158, 159, 160, 161, 162, 163, 164, 165 295 Tableau, 10, 11, 12, 141 TDWI Research, 129 Tencent, 176, 181 Tencent Weibo, 176 Text analytics, 129 The Chinese Mind, 178 The Cluetrain Manifesto, 76, 250, 251 The Economist, 190 The Mobile Marketing Handbook, 67 Thought tracing, 65 Threadless, 220 TIBCO StreamBase, 83 Tivo, 252 Toyota, 129, 130, 167 Tracinski, Rob, 30, 33 Tudou, 85 Tv.sohu.com, 85 Twitter, 22, 25, 53, 87, 99, 102, 113, 117, 127, 128, 132, 176, 181, 182, 183, 185, 189, 192, 193, 198, 202, 208, 210, 213, 214, 216, 220, 221, 222, 223, 224, 225, 227, 235, 237, 239, 244, 251, 265, 267, 302, 303, 305, 306, 308, 311, 312 Promoted Trends, 193 search, 255 selling perishable inventory, 244 Tweets, 192 what is it?, 192 Twitter Revolution, 23 Typology of crises model, 224 Typology of Social Media Crisis, A, 223 U Ulanoff, Lance, 1, 31 Unilever, 250 United Airlines, 244 United Breaks Guitars, 226 Unsupervised learning, 38 User Generated Content, 172, 175 categories, 173 mobile dating, 173 personal content distribution, 173 social networking, 173 Ushi, 26, 177 T Table Games Revenue Management, 330 THE PREDICTIVE CASINO WhatsApp, 22 Wheel of Loyalty, 101, 111, 288 Wikipedia, 62, 73, 76, 84, 171, 173, 174, 175, 187, 188, 194 Wirtz, Jochen, 98, 101, 111, 288 V VanBoskirk, Shar, 88 Venkatesan, Rajkumar, 293 Video pre-rolls, 62 video.sina.com, 85 Video-casting, 171 Videocasting, 84 Virtual game worlds, 26 Virtual Reality, 2, 40 Virtual social worlds, 26 Virtual worlds, 171 Visa, 252 Vlogs, 171, 207 X Xunlei.com, 85 Y Yahoo!, 61 YouKu, 14, 22, 26, 81, 84 YouTube, 14, 22, 55, 62, 66, 84, 88, 175, 177, 185, 188, 194, 212, 213, 214, 215, 224, 226, 228, 237, 238, 239, 246, 247, 249, 250, 251, 253, 255, 256, 263, 270, 271 YY.com, 25, 84, 177, 230, 231, 269, 304 W Wähner, Kai, 15, 32, 79, 80, 81, 83 Wanamaker, John, 207 Waste Management, 279 Wearables, 85, 87 Webopedia, 41, 189 WeChat, 4, 14, 22, 25, 26, 48, 51, 54, 55, 71, 72, 77, 81, 88, 92, 102, 117, 127, 133, 181, 183, 194, 195, 196, 197, 198, 204, 239, 306, 307, 309, 311, 316 Weibo, 26, 117, 181 Z Zedong, Mao, 178, 180 Zone of affection, 102, 112, 221, 306 Zone of defection, 102, 112 Zone of indifference, 112 Zook, Ze, 24, 25 331