by Adam Tanner
“If our hosts get access to that information, for example, as they are dealing with a customer, does the customer sort of recoil with horror that they actually know that?” Shaukat asks. “Or do they say, ‘I logged in using my Facebook account—of course they know that’?”
For now, the company has decided to limit how much Facebook data it will use. It will steer clear of certain personal information such as sexual orientation, even if such data might help sell tickets to certain shows. “The last thing I want is for customers to say, ‘You know what, you guys are creepy, you’re invading my privacy,’” Shaukat says. “I draw the line when the stuff is very personal.” He also emphasizes that Caesars protects privacy by not selling data about Total Rewards clients to others, although it sometimes sends promotions from other companies. Such promotion partnerships often make good sense, such as when an airline wants to fill its seats to Las Vegas by advertising deals on flights.
As the head of Total Rewards, Kanter was on the front lines of data-gathering about customers. He wanted to know all he could about his clients. But he did not want to strip them naked through the most aggressive, cutting-edge possibilities of data aggregation. He sought to bring more customers through the door but also hold his head high and believe that he was doing the right thing. As he considered new data, he applied what he called the Sunshine Test: “If all the information were out there in the light of day, would our customers understand and be okay with what we are doing, and is there some kind of commensurate value that they are getting?”
Of course, anyone can gamble or visit a casino hotel without being tracked if they decline to join the Caesars loyalty program. But Loveman believes customers will continue to share data if they get something in return. “If you prove to the guest that you use that information productively, they tend to be quite generous with it,” he says. “If you prove to them on the other hand that you are a dope and you don’t listen to them, then they are going to get pretty frustrated with it. For example, if you go to a grocery store and all your grocery purchases are scanned and you’re a lifelong vegetarian but every week you get an ad in the mail for meat, that pretty much convinces you that nobody is listening.”
Joshua Kanter of Caesars advocates a “Sunshine Test” when it comes to using personal data. Source: Author photo.
Data’s Blind Spot
For all their power and potential, big data and data analytics, whether about specific individuals or broader business trends, are not all-powerful and do not always produce the right answer. For Gary Loveman, data’s limitations in forecasting the future failed him when he most needed insight.
In 2006, rival casino innovator Steve Wynn offered to sell him a subconcession to operate in Macau for $900 million. Harrah’s had already passed on an opportunity to buy into Macau in 2002. That year, the government there ended a decades-old casino monopoly and auctioned off three licenses, opening the market to foreign operators. Executives involved in the decision say Harrah’s worried that association with Macau’s unsavory reputation could complicate their standing with US gaming regulators. Several years later Harrah’s had a second chance when three more sublicenses went on the market. By then it was clear that gambling in Macau was growing dramatically every year and nearing Las Vegas in gaming revenue.
Loveman, his chief financial officer, and others delved into their spreadsheets and conducted their usual vigorous analysis. Could their mathematical models justify such a high price tag just for a license to operate in the former Portuguese colony? Looking at statistics from 2005, the math nerds saw that all of Macau earned casino revenue of $5.7 billion in 2005, up from $5.1 billion the year before.1
A few weeks later, Loveman called Wynn with his conclusion: “Steve, it’s too expensive. We don’t believe that we can sustain that number.”
“That’s my price,” Wynn replied. “If you don’t want to pay it, I’ll find someone else who does.”2
Wynn found a buyer at his full asking price almost immediately. The buyer’s confidence in untested potential trumped rational analysis. Macau has since become the world capital of gambling, dwarfing Las Vegas, with about seven times more gambling revenue than Las Vegas in 2013. Chinese players have emerged as the world’s biggest whales—mega-gamblers who wager millions of dollars a year, and they love Macau. Loveman looks back on that decision as the biggest mistake of his career.
“No one had ever paid $900 million just for a license—not a building or a business,” Loveman says. “The tremendous accumulation of wealth in coastal China, the desire of folks in that market to gamble, the quality of the facilities themselves to be built—there was no precedent ever in the history of the industry that could anticipate such a thing. So as a result the methods that we used, the kind of conventional analytic methods that we used—they just can’t foresee that kind of thing adequately. So as a result we underestimated the value. . . . Even the most optimistic people—including Steve, who I consider to be wildly optimistic—could never have imagined that that market would be $35 billion in 2012, and that’s exactly what happened.” By 2007, Macau was nearing Las Vegas in overall casino revenue; it surpassed Vegas in 2008. Revenue continued to soar in the following years as Vegas went into a deep slump, and in 2010 Macau made $23.5 billion, more than double the casino revenues in all of Nevada.3 Total Macau casino business had about doubled again by 2013, when it recorded $45.2 billion in revenue.4
Wynn and Sheldon Adelson, the owner of the Las Vegas Sands Corporation, whose properties include the Venetian, both jumped at the chance to expand into Macau when licenses were first awarded in 2002, cementing their reputations as casino visionaries. Conventional wisdom had underestimated both men before. Many predicted Wynn would go bankrupt after he opened up the $630 million Mirage casino in 1989. “He’s such a genius that I’m not sure any analytics could trump his intuition, so he’s actually a little bit of an anomaly, I think,” says Patti Hart, CEO of slot machine maker IGT. “Steve has a very sophisticated analytics business. The problem is his ideas are always better than the analytics because he has just great intuition.” Conventional wisdom also expected Adelson to fail after he opened the Venetian in 1999. Both instinctual investors defied the skeptics.
Another example of the value of instinct played to the company’s advantage over the long run. After Harrah’s reached a deal to buy Caesars Entertainment in 2004 for $9.4 billion, Caesars Palace executives briefed Loveman and top managers on their deal with singer Celine Dion. She had agreed to anchor their new four-thousand-seat Colosseum starting in 2003, performing two hundred shows a year for three years. Many at the time thought a single headliner could not command big crowds like larger productions did.
Caesars Palace officials told Loveman and marketing chief Rich Mirman that the Celine contract had proved vital to reviving the Caesars Palace brand. The Harrah’s financial team had looked at how much they were going to pay her and thought the deal made no sense. But in the first months after the show opened, on days when Dion performed, the hotel and casino showed additional income of $200,000 from more gambling and food and drink revenue.5 “You know, we never in a million years would have done the Celine Dion deal, not in a million years. But yet it was the key to their success,” Loveman told Mirman as they left the meeting. It was another case where a company experiences intangible benefits that did not show up in the numbers. Celine Dion remains a leading performer at Caesars Palace.
Caesars’ Influential Shadow
Even using extensive data from its Total Rewards system, Caesars can sometimes be clueless in reading the intentions of their clients. David Schwartz, director of the Center for Gaming Research at the University of Nevada Las Vegas, is a Total Rewards member who receives a stream of what he considers mistargeted offers. The author of three books on gambling, Schwartz cites as examples discounted “insider” rates to stay at the company’s hotels in Las Vegas ranging from $20 a night at the Quad to $94 at Caesars Palace, under email headlines such as “Wis
h You Were Here . . .”6 He gets offers to visit Atlantic City, an unlikely lure for someone living in Las Vegas. And to attend an overnight party celebrating the 2009 comedy film The Hangover. Schwartz is married with children and says the latter offer does not appeal to him.
“What was the reason that the ultra-sophisticated Total Rewards system is sending a Las Vegas resident emails telling him that ‘the Sin City sunshine is calling your name again?’” he asks. He is skeptical about the extent to which there is any loyalty in the gaming business and says customers just go where they get the best deal.
Total Rewards head Kanter says the company is not mindlessly spamming local residents. The company knows that if a local does not have to drive home, he or she might gamble longer, generating more revenue. “He may not appreciate that logic, but it’s naïve to think that we did not intelligently arrive at the decision. Out of a hundred people that look like him, we’ll get a few of them who will respond and at the segment level it’s a good decision,” Kanter says.
Harrah’s Kansas City manager Tom Cook says the company often falls short of using the personal data on gamblers to make very specific targeted offers. Loveman is surprisingly receptive to such criticism. “We may not catch that as well as I claim to catch it. He’s quite right,” Loveman says. “The problem is that when you go onto the system and try to do it, it’s really hard and time-consuming. So pulling the data list and segmenting the data list geographically turns out to be a bitch, and so we don’t do it very well, in fact, so we wind up messing that up.”
The CEO says the company does not react quickly enough to details of a customer’s visit, and sometimes does not anticipate well enough what customers are doing. But overall, Caesars’ deep dive into personal data has long paid off. “Certainly by any measure it is tremendously effective,” Loveman says.7
Over time, clever use of personal data has dramatically improved Caesars’ bottom line in the Loveman years. When he joined Harrah’s, clients spent just 36 percent of their gambling budget with them. Today, customers devote more than 60 percent of their gaming spending to Caesars. VIPs spend nearly 80 percent with the company.8 Microtargeting clients made his company a leader in the industry and cemented Loveman’s rise to CEO in 2003. During his regnum Caesars became the world’s biggest casino company, running more than fifty casinos,9 up from fourteen, in seven countries.10 When Loveman arrived they operated only in the United States. Now they employ sixty-eight thousand people, up from fourteen thousand.11
Rival casinos have embraced loyalty programs and gather granular data on wagering and other spending. Some gamblers belong to so many programs that they keep a stack of loyalty cards on a keychain ring. The forgetful among them attach a string from their pants to the ring as a reminder to remove the plastic slab before getting up from a slot machine. “What our marketing is intended to do is to give us an extraordinary share compared with what our position would normally provide, and that’s how we measure ourselves,” Loveman says. “And we exceed our fair share in every market but one, I think, across the country.”
For years, casino rivals ridiculed Loveman and his management team as “casino nerds.” Go ahead, call him a nerd, propellerhead, or geek—he doesn’t mind. “There are traditionalists in the business who view that as a pejorative notion. They say that with disdain,” he says. “But the industry has really come along in this direction.”
As Caesars embrace outside supplementary customer data, it is likely that other companies inside and outside the world of casinos will continue to do so. You may not visit casinos at all. But rest assured that those credit card offers you get in the mail, the frequent flyer and loyalty club offers, as well as other solicitations are based on sophisticated guesses about what kind of customer you are and what you might be interested in. Even businesses as diverse as the Walt Disney Company embrace the data-driven approach championed by Loveman and speak in similar terms. “What we say at Disney is that there is never too much data,” says Leon Gantt, manager of Disney World’s information technology. “We want to have enough information to understand what they want beforehand and anticipate it.” The beloved company faces the same puzzle managers face across the economy: “We have these mountains of data . . . how can we use it without alienating clients while maximizing revenue?”
Opening the Secret Files
Things become more complicated with third-party data. From a consumer standpoint, the problem with such information has long been that these files constitute a black box. As with the Stasi files during the days of East Germany, the ordinary person has no hope of seeing the totality of what commercial data brokers such as Acxiom and Experian have assembled.
For some years, people could request to see just that part of their Acxiom file gathered from public documents, showing less interesting information such as their address and phone number. Yet almost no one was able to surmount the firm’s onerous requirements to see even this fragment. The process required sending Acxiom a Social Security number, date of birth, driver’s license number, current address, phone number, and email address, as well as a $5 check—and then waiting two weeks. From 2009 to mid-2012, between seventy-seven and 342 people per year had asked to see their files, with just two to sixteen annually providing enough information to get access, the company told a congressional panel.
Yet even then, the commercial dossier remained off-limits. That part of the file includes a description of one’s general family and financial situation. It might list race, ethnicity, religious affiliation, education, political affiliation, occupation, and hobbies. Acxiom, as well as other leading data brokers, might know what credit cards you use, as well as some health topics of interest to you such as diabetes or arthritis. Knowing that all of that information exists in unseen vaults at companies people had no relationship with made a lot of them—at least those who knew about it—rather uncomfortable. And the mounds of data continue to grow: in a typical week, Acxiom processes a trillion transactions—twenty times the number of searches conducted by Google.12
In 2011, Acxiom embraced a change in direction by hiring a new CEO, Scott Howe, a former Microsoft executive whom Gary Loveman had taught at Harvard Business School. In interviewing for the position, Howe preached a new direction for Acxiom. The world was moving toward increased transparency and consumer control, and companies that ignore this trend would do so at their peril, he said.13 He got the job leading 6,200 people and moved slowly but deliberately toward cracking open Acxiom’s doors to the public. After about a year and a half of internal reorganization Howe kicked off plans to allow people to see a part of their files instantly online. “Long before I came to Acxiom it had always bugged me that I did not know what data was collected about me,” he said. “I think it is bad business that companies lie or exploit or obfuscate the truth from their customers. That’s not the kind of company that any of us want to work at.”
Some of the old guard at Acxiom scoffed at Howe’s push.14 They recognized the pressure from privacy advocates and the threat of future government regulation. They had experienced similar spurts in the past and thought the scrutiny would pass. They saw no reason to change how things had always been done. But Howe pressed on. “I just could not work at a company, quite frankly, that it felt like we were not doing what was ethically right,” he said. “I couldn’t feel good about coming to work every day and being branded as the biggest company that no one’s ever heard of or, you know, the commercial equivalent of NSA or the supersecret spy guys.”
Scott Howe, CEO of Acxiom.
Source: Author photo.
In September 2013, Acxiom launched AboutTheData.com, a web interface that allows the general public to look up their data instantly. Howe knew that even after Acxiom spent millions of dollars on the project, the site could still be improved. But he thought it was best to go live quickly rather than wait years for engineers to design a perfect system.
Thousands went online to meet their digital doubles. What many consumers fo
und was often not an all-knowing Big Brother but a world of imperfect replicas, sometimes as odd as the flawed Bizarro world of Htrae (read that backward) of Superman comics, where things are the opposite of Earth.
Rich Mirman, the first data guru hired by Gary Loveman, remembered how Acxiom had long tried to sell data during his time at the company. He and others at the time were never convinced of the utility of such information, and seeing his own file did not change his outlook. Acxiom had accurately listed many of the kinds of things his household buys, but they completely misread who he is, saying he is single with no kids and a dial-up modem harkening to the Internet’s earliest days. “How many single men without kids own a minivan?” he wondered. “They should have figured out that I was married with kids—the dial-up modem is just silly.”
Joshua Kanter found that Acxiom knew about his hobbies, purchases, and home, but a lot of the small details were wrong. It knew that he owned his beloved 2011 Audi A5, but it was way off on his income. Overall, he thought, his profile looked as it might have two decades earlier, when he was running his college painting business.
Acxiom thought Kyle Prall, founder of the controversial site bustedmugshots.com, was still a student and that he had a child. In reality, he had left school more than a decade before and had not brought any offspring into the world. Claudia Perlich, the East German–born data scientist who was too young when the Berlin Wall fell to have had a Stasi file, looked herself up only to receive a response that “we were unable to verify you.” I found that same response on my own file.
For casino entrepreneur John Acres, who wants to reinvent gambling by using personal information to make games more responsive to individual preferences, Acxiom knew nearly the correct square footage of his house, but his file was wrong on his household income, his vehicles, and the value of his home. “Nothing in there was very informative,” he said. “It was still pretty worthless.”