CAMBRIDGE, Mass.— There are still surprises that bring a smile to Gary Loveman’s face, even after 10 years as CEO at Caesars Entertainment. He beamed telling the story of an employee award given to a housekeeper named Jadwiga Maruszewska. Hotel guests ask specifically to stay on her floor when making reservations.
Maruszewska not only knows the names of repeat guests at the Flamingo Las Vegas, she knows the names and habits of the guests’ children. It’s the kind of knowledge and customer relationship building that a database can’t match.
The story of personal service is a complement to the data-driven systems guiding decisions at Caesars that cultivate customer loyalty and determine, for example, which customers deserve a free dinner, a hotel stay or forgiveness on gambling losses after a bad day at the casino.
At a daylong big data management conference on Dec. 12 at the MIT Sloan School’s Center for Digital Business, Loveman spoke with Andrew McAfee, associate director of the Center, in front of about 300 people, about how data drives Caesars’ culture of constantly testing offers to customers, sampling and tracking customer reactions.
“There has to be what I refer to as an aggressively collegial approach,” said Loveman, a former Harvard Business School professor. “A person in the market has to reflect on conditions in the market and if things aren’t working they have to scream about it and say ‘Snap out of it knuckleheads.’ They can’t be deferential.”
People hate casinos when they lose money, Loveman said, and Caesars’ can’t simply reward losing gamblers with a steak dinner and hope they will come back again. There needs to be more. So at Caesars the Total Rewards program has developed timely, targeted benefits to loyal visitors to avoid overpaying for no change in customer behavior. That ensures a bottom-line gain and growing knowledge of the company’s customers.
“The player with big losses gets a steak and hotel room—that can ease the pain of the losing, without affecting the gaming for anyone else,” he said.
Every day, Caesars employees will make decisions to distribute benefits to customers informed by the data in the Total Rewards program, and based on staff members’ experience dealing with customers. “That will happen 50,000 times tonight before you go to bed,” Loveman said.
Analytics tracking the program identify any overly generous offers to prevent fraud among the 75,000 Caesars employees. Loveman said he studied the rapid and repetitive testing methods of companies such as Capital One when exploring ways to make customers into repeat visitors.
Total Rewards tracks activity within a casino from the moment a guest arrives, he said. Cameras can record a gambler’s choices at the card table and decisions to identify a player as a disciplined player or someone who guesses, makes wrong choices and is more likely to lose money. Rather than rewarding only purchases—the way airlines dispense frequent buyer points—the Total Rewards program builds loyalty through detailed profiles and the service or reaction to each experience.
“The $1 million loser is easier to identify,” he said. “But the guy who normally spends $150 and lost $400 – that’s where the mass of the business resides.”
Companies need to look at customers and lifetime value rather than counting profits simply by product line or geography, Loveman said. When McAfee asked what other industries are targets for reinvention, Loveman said airlines, cable television and mobile phone companies all could use a reboot.
“Imagine a reverse auction for a cable repair call,” Loveman said, adding he envisioned a consumer willing to pay more for a quick repair after a service outage.
The conference also featured Hal Varian, chief economist at Google, who pointed out that since he took his job in 2002 other companies including Yahoo, Microsoft and Amazon have hired chief economists, while growing numbers of companies are seeking data scientists.
“Lots of companies have data but what they lack is the knowledge or experience to know what do with it,” Varian said. They need to learn how to conduct experiments with customers using data. Each log-in to a website could involve dozens of tests, he said; the key is to get quick feedback, analyze the interactions and kill ideas that don’t deliver positive results.
David J. Wallace, a freelance writer and communications consultant, has written for The New York Times and Knowledge@Wharton. Follow him on Twitter: @gamechange.