Since the introduction of the first electronic player tracking systems nearly three decades ago, slot managers have been gathering ever increasing amounts of data pertaining to their guests.
What began with tracking how much players wager has grown to include when they wager, what games they play, when they play them, what restaurants they eat in, whether they stay in the hotels and use other services-virtually any piece of information that can help build a customer profile.
The information by itself can tell you what has happened. What will happen is another matter. That’s increasingly the focus as casino operations move into a brave new world of business intelligence, data warehousing and predictive analytics.
The “what will happen” can involve predicting casino needs in the short-term future.
“Right now a lot of products show you what’s happened in the past and what’s happening now,” said Rory Fagan, sales director for hospitality and gaming at SAS, who through its SAS Patron Value Optimization product brings casinos a suite of applications that includes data integration, data quality, data mining, business intelligence and campaign management. “Where SAS comes in is that we do the forecasting and predictive natures of what will happen. If a show lets out at 7:00 p.m., this is where we think the traffic will come, and this is what we think the projected revenue will be so they can have the proper staffing and optimize what the game values are.”
Or the “what will happen” can be farther down the road, with a payoff in customer development.
“Our gains to the property come in two forms, said Christy Joiner-Conglteon, president and CEO of Stics, which says its Predictive Analytics can help identify players who are likely to grow in value to the casino. “One is we find those players who are needles in a haystack if they were trying to look at it with less sophisticated tools. We pluck out all the good needles. We also are able to tell them that some of the people who appear to be good in a shallower analysis aren’t really good in a more detailed analysis. So we’re able to improve their response rates and the effectiveness of their programs and get more valuable people playing the slot machines because we avoid people who are not cost effective and bring in people who are cost effective.”
In formulating such predictions, analytical tools take into account not only information from the casino’s player rewards and hotel databases, but also demographic and socio-economic information, Joiner-Congleton added.
CLUSTERS OF PROGRESSThat enables slot managers to uncover some surprising trends, such as an association between different slots that are attractive to the same customers. Dr. Ralph Thomas, vice president strategic analytics and database marketing for Seminole Casinos in Florida, used SAS for an analysis of clusters that yielded unexpected results.
“Machine clusters, that’s the research and development side,” Thomas said. “We’re still working on that. That’s the high level stuff. We track every customer, every game they’re playing every day. So I can let the customers tell me what games are similar. It’s one thing to go to a slot guy and say, ‘Hey tell me how to group your games.’ And he might say, ‘I like Sex and the City because it looks like Hangover and I think the customers who like Sex in the City will like Hangover.’ And then you go back and do the analysis and it turns out there’s no crossplay between those games.
“The problem is it’s easy to take two games and find out if there’s crossplay. It’s hard to sift through all that data to find the relationships between the products based on the customer choices. We’ve had some progress. It told us the obvious. There are two games, Stinkin’ Rich and Bombay [both by IGT], and everybody knows the customers like both these games because it’s so obvious. And it came out in our analysis, yes, these two games are highly correlated, and have a high affinity to each other. Then we found a cluster of games that nobody knew were so similar. I can’t tell you this one, but it was really surprising and eye-opening to us. And then a third cluster just shocked me completely, and it has me thinking differently about how we optimize our slot floor. It’s completely opened my eyes to where I think about slots in a completely different way.”
The clusters, Thomas said, yield a peek into the future. For the present, SAS’ analytics help him model his customers’ behavior.
“The easiest and most reliable in terms of making you money are response models, and just saying look, when we have a mailer, these are the people we can leave out, save the postage and we’re not losing any money. That’s by far the quickest one,” Thomas said. “On the other side now we’re moving on forecasting customer behavior, besides just response. This gets trickier but we’re trying to see if we can find customers whose behavior would be different than we would have anticipated without the model. We do all sorts of segmentation. We have thousands of segments every month based on various criteria, and the question is when [the guests are] in the wrong bucket because their behavior is going to change. They’re not going to play like they did three months ago; they’re actually going to play differently. We have a program that recognizes our prediction that they’re going to play differently. That’s trickier business.”
IN GOOD COMPANYIn both the straightforward models and the trickier analysis, a number of companies are staking out their own claims for segments of the business, from slot floor analysis such as SAS and Stics to surveillance room applications from Pivot3 and the hotel and hospitality focus of Agilysys.
Among those touting their own business intelligence solutions:
Agilysys-With its hotel and hospitality focus, Agilysys streamlines processes with a suite of applications used throughout the lodging portion of the property. The Lodging Management System automates reservations, credit card processing and accounting, along with housekeeping and other aspects of the operation. It includes the LMS+ graphical user interface for easy operation, and the LMS ARTS guest activities planner.
The Lodging Management System was chosen recently by the Bear River Casino in Loleta, Calif., which will operate a 105-room hotel when it opens in March. Along with the LMS applications, Bear River will use Agilysys’ DataMagine signature capture solution to streamline the check-in procedure.
Teradata-Data warehousing and analytics under the Teradata name have a large casino presence, with gaming facilities using the application to manage data warehouses for analytics and business intelligence. The company also develops and markets customer relations management software under the Aprimo Relationship Manager brand. Teradata acquired the cloud-based integrated marketing software company Aprimo early in 2011.
“Gaming companies that use an enterprise data warehouse environment based on Teradata are accessing and utilizing their data to effectively grow their businesses,” said David Porter, director of hospitality solutions for Teradata Corporation. “Many of these companies are now evolving beyond pure analysis [what happened?] to predicting [what will happen and why?], rationalizing [what is happening right now?] and automated decision-making [what do I want to happen?].
“Teradata helps casino enterprises to sustain competitive advantage, improve service to slot and table game patrons, increase patron loyalty, optimize gaming revenue, achieve a greater return on promotions, better comply with gaming/ government regulations, reduce costs and improve operational efficiencies.”
“The BIS2 product works seamlessly over any existing schema,” said Andrew Cardno, chief technology offer at BIS2. “Quite simply this means BIS2 is able to install our software in hours and days into almost any environment. These environments range from the extremely sophisticated data warehouses to properties where all the data resides in the slot management system. This interoperability is a fundamental strength of the BIS2 technology and has enabling us to install over what will be over 100 properties in 2011.”
Marketing Results-The company’s Advanced Intelligence Marketing analysis tool is used in reporting and analysis, campaign management, player development and host management. Developers have more than 200 years of collective gaming experience for more than 110 casino operations. But AIM is not sold to everyone.
“AIM is licensed for use by Marketing Results Nevada Corporation and MRI clients but otherwise is not sold or licensed because it creates a competitive advantage for users,” MRI President and Founder Gary Border said. “In other words, to supply every casino with AIM’s proprietary tools would simply erode the advantage that AIM provides. We limit the AIM advantage to active clients.
“Slot operators attract more and better players more frequently because casino-branded messages are delivered with incentives. Many direct mail operations today simply distribute free play and value offers without building a relationship with players. AIM pinpoints trends in the database quickly to enable marketing management to exercise control over target segments and deliver services to the most profitable segments. AIM further enables digital communications with social networking benefits. Players open and maintain a dialogue with casino brands.”
Pivot3-The data necessary for running a slot floor today extends to the surveillance room. That’s where Pivot3 comes in.
“The amount of video data being stored by casinos is growing at more than 40 percent per year as more cameras are deployed, more megapixel cameras are introduced for resolution improvement and video is kept longer for liability and regulatory needs,” said Chief Strategy Officer Lee Caswell. “Our storage solution allows casinos to simply scale up the amount of video they can store by adding appliances to an existing installation. It also protects the video data against common problems, such as power outages, switch failures, or disk drive crashes so that the video is not affected. This is becoming more critical as the expectation is that video is never lost, especially in the casino environment where regulations require video data to be captured and stored for a predetermined period of time.”
GETTING REALWhere is all this taking us? One area is data mining and analysis in real time, SAS’ Fagan pointed out.
“We do have Advanced Analyst that can do modeling within the database and database processing within Teradata and some of those databases. We also have real-time ability to send out marketing offers. Most casinos, I think, are doing a lot of the analysis in batch processing after the fact. I do think we have a lot of casinos that are trending toward ‘Hey, I have someone here on property, I should be looking at their behavior and see how that’s changing.’ They all want to get to that real time marketing while the guest is on property as opposed to waiting until they get home to make the offer.”
The slot analytics market and the capabilities of data mining are on the rise said Stics’ Joiner-Congleton.
“We expect that predictive analytics will continue to get increased penetration in the market,” she said. “Predictive analytics as a scientific enterprise tends to take over the markets in which it is used. This is true in financial markets, it’s been gaining momentum in medical markets, and gaming is a very viable place to let the science seize the amount of effective work that a brain can do or that a lower level pool like sequel or a data warehouse can do by itself.
“They say that a brain can entertain only seven ideas at once, and I think mine can only entertain three,” she said with a laugh. “But statistics can entertain a limitless number of factors in making a decision. We see analytics as being able to understand the fundamental issues that cause people to gamble in the first place.”