By Ted Keenan, Vice President of Product Management, Aristocrat Interactive
Twenty years ago, the database marketing problem was relatively straightforward. One of our primary decisions was how much free play to offer. Today, casinos and resorts have far more ways to create value for guests — gaming, restaurants, hotels, concerts, spas, sporting events, and nightlife among them.
At the same time, the amount of information available about our players has exploded.
When I talk about moving from campaigns to conversations, I’m not talking about replacing free play with AI. I’m talking about the fact that the database marketing problem has fundamentally changed. AI can help us determine which experience is the right one for the right guest at the right time.
From Legacy Metrics to a Multidimensional View
Casino marketers know ADT. We’ve used it for years to segment databases and determine offers. But the information available to marketers today goes well beyond what happens during a gaming session.
We’ve moved from what I think of as an era of legacy metrics into an environment where player data is multidimensional. Gaming systems, point-of-sale systems, financial transactions, lodging, and other sources can all contribute to a broader picture of a player. The challenge is capturing that information, analyzing it, and figuring out what to do with it.
That leads to some important questions. What if every campaign could be personalized at the individual level rather than the segment level? What if reinvestment could be right-sized to true value? What if marketing could react mid-cycle instead of waiting for reports? And what if we could predict the value of a new player much earlier in the relationship?
Consider a player who routinely visits on Tuesday after work and spends $100. Then one Friday, that same player arrives, spends $100 at the slots, has a steak dinner and drinks, and puts $250 into a table game.
In a traditional reporting cycle, we might not recognize that change in behavior until Monday morning.
If systems can analyze activity as it happens, there is an opportunity to recognize that behavior while the player is still on property. A team member might respond personally, or some interactions could eventually be automated. Either way, we can recognize the behavior while there is still an opportunity to engage the player.
Don’t Wait for a Player to Prove Their Value
I learned a great deal about acquisition, retention, and lifetime player value during my years in online gaming.
When new players came into those businesses, we looked at the sequence of behaviors they exhibited. Over time, we could begin to identify patterns that indicated the likelihood of a player becoming loyal.
The same thinking can apply to casino player development. If a relatively new player is already exhibiting behaviors similar to our historically loyal players. Why wait six months for that player to prove their value before we begin building the relationship?
Technology can help us recognize those patterns earlier and get closer to making the right offer to the right player at the right time.
It can also help player development teams prioritize. If a host has 30 VIP players on the floor, there’s no practical way to interact with everyone at once. Data and AI can help determine which interaction may have the greatest impact and who the host should act on next.
Finding Growth in What You’re Already Spending
Better use of data isn’t only about creating more offers. Sometimes the opportunity is identifying offers we never needed to make.
Think about a player who arrives on a Friday night with friends and discovers $100 in free play waiting for him. He appreciates it, but did that offer influence his decision to visit? Did it cause him to choose your property instead of a competitor?
If the answer is no, we may have spent $100 without changing his behavior.
One business case involved a property with a 22% reinvestment rate. By looking more closely at where reinvestment was being redeemed, by whom and when, the property identified waste and reduced the rate from 22% to 15% with zero disruption in player trip frequency. The reported annual margin savings were $3.2 million.
Those savings don’t necessarily have to disappear from marketing. They can create an opportunity to invest in acquisition, retention, or other areas that can drive growth.
Building the Data Flywheel
For this to work, the technology has to do more than produce another report.
I think about it as a data flywheel: capture information, analyze and activate it, engage the player, refine what we learn, and feed that information back into the process.
Then do it again.
The data can come from gaming, point of sale, financial transactions, lodging, and other sources. Analytics and AI can help us personalize the experience. The final piece is what I call the “last mile”: getting the experience or offer to the player, whether that’s at the game, through mobile, or through another point of engagement.
That requires interoperability. Systems need to be able to take information in and make information available to other systems. We shouldn’t have to decide that one system must be the center of the universe. The data needs to move.
Mobile is part of that last mile, but mobile doesn’t automatically have to mean an app. Looking at other industries and other parts of the world, there are increasingly more ways to make it easy for customers to engage without requiring them to download something first. The goal is to reduce friction between the player and the experience.
More Data is Only Useful If You Can Act on It
One four-property business case involving 4,900 gaming positions started with a familiar problem. Raw CMS data was being downloaded and manually restructured in Excel, and there was no consistent enterprise view across the four properties.
Automating that process created a common view of the data that could be used for executive reporting, floor optimization, vendor accountability, and program integrity. Over two years, the properties reported a 15 percent increase in win per unit.
AI can help with this work, but I don’t think we need to make AI more mysterious than it is.
Much of this analysis can be done manually. It’s not rocket science; it’s just burdensome.
There is a tremendous amount of time involved in pulling reports, analyzing information, and deciding what to do with it. That’s the kind of work AI can be very good at.
The opportunity isn’t technology for technology’s sake. It’s using the information we already have — along with the information that continues to become available — to understand our players better and respond more effectively.
Knowing what a player did yesterday is useful.
Recognizing what that player is doing right now, understanding how that behavior fits into a larger pattern, and having the ability to act responsibly on it gives us an opportunity to build a stronger relationship while driving profitable growth.
