When you walk into a casino today, with every slot machine pull and roll of the dice, owners are collecting more data points in what is already one of the richest environments for tracking customer behavior. By analyzing these treasure troves of customer data, casinos can knowledgeably deliver a personalized experience to each and every customer.
It wasn’t always this way though. Before the advent of loyalty programs, the average player came and went with barely a free gin and tonic to show for the visit. There was value being left, quite literally, on the table.
So how did Caesars Entertainment go from an enterprise where operating decisions were based on intuition and experience to one built on data, analytics, and experimentation?
Step 1: Collect Data
When Gary Loveman moved from his post in academia at Harvard Business School to be the Chief Operating Officer of Caesars Entertainment in 1998, Las Vegas casino figureheads like Steve Wynn and Sheldon Adelson collectively mocked the decision. Everyone waited to see how long it would take before the Professor scuttled back to his post at Harvard.
With a PhD from MIT, Loveman understood that in order to play in the high-stakes world of Las Vegas casinos, he would need data and lots of it. In his first year at Caesars, Loveman introduced the Total Rewards customer loyalty program. By giving each customer an individualized loyalty card to track all of their gambling and service transactions, Loveman had given himself the necessary mechanism to collect an unprecedented set of personalized customer data.
Upon entering the cavernous interior of a Caesars property today, your Total Rewards account becomes the basis for all dining, gaming, entertainment, or hotel transactions. If you make a return visit to any of the 40 resorts and casinos which Caesars maintains worldwide, your Total Rewards account history gives Caesars an immediate knowledge of your prior preferences and spending habits. In addition to collecting data within the confines of their own properties, Caesars also maintains current relationships with Visa, Starwood Hotels, Hawaiian Airlines, and Norwegian Cruise Lines. The end result is a rich tapestry of consumer behavior patterns that allows Caesars to better predict what you want when you want it.
Today, Gary Loveman has moved into the CEO role at Caesars and claims over 45 million Total Rewards members worldwide, all of whom readily supply data about their personal habits to Caesars.
Step 2: Analyze Data
“There are three ways to get fired from Caesars: theft, sexual harassment and running an experiment without a control group.” –Gary Loveman, CEO Caesars Entertainment (MIT Technology Review)
To date, Loveman has taken his glut of data and applied it to problems including customer lifetime value and drink service ROI. In a prior interview with the MIT Technology Review, Loveman stated, “I’m most interested in experiments around pricing. We have a lot of perishable inventory—hotel rooms and the like—and I want to make sure we’re getting the best prices.”
Early in his Caesars tenure, Loveman conducted one such pricing experiment on slot machine hold (hold is the percentage of money a slot machine is expected to keep in the long run, 4-10% is a typical range). After Loveman overheard two customers chatting in an elevator about why Atlantic City slots were better than Las Vegas slots, he had a realization.
Loveman knew that the customers had it exactly backward. The hold in Atlantic was 7% while the hold in Las Vegas was 5%; they were losing more money on average in Atlantic City! However, the customers couldn’t tell the difference. Was this unique to these two customers or was this a more fundamental issue of small sample size and randomness?
Loveman ran the numbers. How long would it take a person playing a slot machine to discern the difference between a 5% hold and a 7% hold? Well, due to the stochastic nature of slot machine returns, Loveman determined that it takes approximately 40 hours for a customer to confidently know that the two machines had unequal payouts.
Now ask yourself, when was the last time you played a slot machine for 40 hours? Probably never.
With this knowledge in hand, it was of course time to run an experiment. Caesars set up identical slot machines on their casino floor, some with the current 5% hold and some with the increased 7% hold. Once a significant sample of data was collected, the results were clear; no customers could tell the difference. The slot machines with a 2% higher hold saw no decline in customer usage but a 40% increase in gross profit.
Step 3: Profit!
From this slot machine experiment finding, Loveman proceeded to raise the hold of slot machines throughout all of his casinos. While the rumor on the strip was that Caesars had lost touch with their customers, this simple change netted Caesars over $300 million in additional profit since it was rolled out over a decade ago.
Experimentation and data analysis continues to be a competitive advantage for Caesars. With data as the foundation, Caesars is now able to reshape their services and offerings in real-time on a per client level. Although framed as a means to deliver each customer with exactly what they want, Loveman has achieved, in economic terms, a means of near-perfect price discrimination. Through ultimate customization of offerings, Caesars can extract the entirety of a customer’s willingness-to-pay and collect the money that was previously left, quite literally, on the table.