AI-Driven Casino Marketing: Achieving Better Results than Old-School RFM

Database marketing based on the Recency-Frequency-Monetary (RFM) approach has been the standard in casino marketing for many decades. Unfortunately, RFM-based marketing has several limitations for casinos. Potentially profitable players are overlooked, while current players may be trained to continually expect a discount. Meanwhile, a proliferation of alternative entertainment options have emerged and are changing customer behavior, presenting risks to casinos if they fail to change their marketing behavior. In talking with casino marketing executives, most are interested in becoming AI-driven and would like to be able to make the right offer to the right individual at the right time, but don’t know how to even start. Based on DataRobot’s experience with AI-driven marketing, we can recommend practical and effective approaches.

 

What are the key problems with RFM-based marketing? 

The first set of problems comes from the fact that it is backward-looking and focused only on players who have recently played. This methodology neglects two promising sources of revenue. By only looking at past play, this kind of marketing fails to identify potential high-value customers and develop them effectively. Additionally, by prioritizing recency, this marketing fails to make offers to infrequent players who may be ready for another trip to a casino. More troubling is that there is a lot of evidence in other industries that repeated discounting or marketing offers lose effectiveness over time. By making offers to regular players, it is likely that those players come to constantly expect the offer and it no longer drives incremental play.

Beyond casino marketing, we’ve seen an explosion in entertainment options - from streaming content to mobile gaming and esports -  and a corresponding change in customer behavior. There are more choices and each is working to drive customer engagement at the expense of other entertainment options. As Doug Bentz, Vice President at the San Jose Sharks, said, “Our biggest competitor is Netflix.” Companies like Netflix, Amazon, YouTube, and TikTok are already using AI not only to drive their marketing but even their decisions of what entertainment to develop and how to show it to their customers. To survive, casinos need to adopt AI-driven marketing.

 

How casinos can become AI-driven

The ideal approach individually targets each customer with an offer at the precise moment that will drive maximum additional revenue from that customer. To achieve this, casinos can use AI to create predictions that tell us if we make this offer to this customer at this time based on the customer's demographics and behavior, we will drive this additional revenue. However, because of past reliance on comping based on segmentation, customers with similar behavior tended to receive the same offer. Without a historical variation of offers, AI is not going to be able to predict which offer to make to which customer.

What should a casino do first then? Experimentation around the existing RFM model making different offers to similar customers and tracking their behavior would enable the casino to develop AI that can identify how different offers impact different players and ultimately predict the right offer for the right player. 

In the meantime, there is still a lot of other value AI can bring to the marketing. AI can be used to predict, based on past behavior, how much a player will be worth in the next month, quarter or year. With this insight, you can begin marketing to players who will be worth the most going forward instead of simply rewarding players for their past play. This approach has the advantage that the AI can target those players who have not played recently but are likely to resume play in the desired time frame. It can also identify those customers who are presently not high-value but are likely to become high-value in that time frame, and those players who have been worth a lot in the recent past but are on a declining trajectory.

Beyond using AI to begin marketing based on future value, you can use AI to shift your marketing efforts even more by targeting specific populations. AI can identify those customers most at risk of churning, and illuminate the precise reasons why at the individual level. Then marketing campaigns can target these risky customers in a personalized way and work to keep them playing at your casino. AI can also identify those customers who are most likely to abuse your freeplay offers, either by not playing all of the freeplay, or replacing some of their spend with the freeplay and actual playing less than they otherwise would have. With these predictions, you can deploy your marketing dollars to target more valuable customers.

Casinos face many problems in their marketing today, caused both by the effects of the way they market and by external factors. To survive and thrive, casinos need to embrace AI and learn to target their customers more effectively, generating value and minimizing wasted resources. The ideal solution will be to understand how a given offer drives a customer to spend and make the offer that drives the most value for the casino although, at this time, most casinos do not have the data necessary to build this solution. While developing this information, AI can still provide great returns by helping casinos understand the future value of their customers and redeploying their marketing spend to the most profitable customers.

 

New Call-to-action

 

About the Author:

Andrew Engel is General Manager for Sports and Gaming at DataRobot. He works with DataRobot customers across sports and casinos, including several Major League Baseball, National Basketball League and National Hockey League teams. He has been working as a data scientist and leading teams of data scientists for over ten years in a wide variety of domains from fraud prediction to marketing analytics. Andrew received his Ph.D. in Systems and Industrial Engineering with a focus on optimization and stochastic modeling. He has worked for Towson University, SAS Institute, the US Navy, Websense (now ForcePoint), Stics, and HP before joining DataRobot in February of 2016.