What Stays in Vegas

by Adam Tanner

  “If they didn’t have a system of marketing to you, I’d never be there. I’d be at the Wynn or Bellagio,” Kostel said. “I would never be here if they were not offering—not some stupid reward like a bowling tournament and a banquet—but $850 or $1,000 in free play.” Although far from flawless, Caesars’ mining of personal data had worked. By knowing a lot about Kostel, they had kept his business.

  Opening Night

  On a late winter’s day in 2013, Loveman, Kanter, and other top officials traveled to the Midwest to attend the opening of the latest property in the Caesars family, Horseshoe Casino Cincinnati. Before the mayor and local dignitaries arrived to inaugurate a twenty-four-hour establishment that, short of natural disaster or emergency, would never close again, staffers anxiously prepared for zero hour. Uniformed poker dealers, most of whom had not worked in a casino before, crowded around tables to play one another in final practice rounds. They wagered play money.

  Inside the cavernous open casino floor, cleaning staff dusted off the two thousand slot machines, each costing as much as an economy car and collectively worth tens of millions of dollars. Eager patrons would soon spin them about seven times a minute, and, by the law of averages, 9 or 10 cents of each dollar bet would disappear into the machines. Melissa Price, the company-wide vice president of gaming, toured the floor. Experience told her each machine would bring in about $350 to $400 a day in revenue. The Horseshoe Cincinnati faced a major obstacle: Ohio bars smoking on casino floors, and many slot players like to smoke. Jamie Papp, who bought and set up the slot machines, said gambling, smoking, and drinking were all habits that went well together. He said slot revenues at casinos without cigarettes typically fall short of revenues at smoking casinos by 15 to 18 percent.

  Behind the scenes technicians checked the flow of data through more than three hundred miles of Ethernet cables—enough to stretch from Washington, DC, to New York and beyond—linking every slot machine spin and transaction to a basement IT server room. A long array of tubes and vents kept servers cool, and backup power stood at the ready, just in case. Even bartender stations were wired so staff could rush drinks ordered directly from slot machines.

  Security guards placed their thumbs onto an electronic reader to gain access to the surveillance room. There, four guards sat side by side behind four computer screens each. A supervisor at a fifth desk watched them watching the others, as well as his own cluster of computer monitors. They faced thirty large television screens on the wall in front of them, clustered nine at each end and twelve in the middle.

  One security guard peered into the action in the count room, where women broke open one stack of $20 bills after another and ran them through bill counting machines. If anything suspicious caught her eye, the guard could zoom the ceiling camera to focus so closely that she could read the quantity of bills spitting out of the counter. Another bank of monitors focused on card tables, ready to watch the action in a few hours’ time.

  A steady stream of staffers arrived through the back entrance, showed ID cards, then punched in to a time clock—all the movements were captured by the casino’s computer system. Some of the staff headed to an automated dry-cleaning rack. They entered in a number, and their cleanly pressed uniform appeared behind a slot large enough to pull out a suit on a hanger. Minutes later they emerged from changing rooms—male dealers in bow ties, waitresses in sleeveless tops pressed tightly against cleavage.

  Two employees punch in ahead of opening night at Horseshoe Casino Cincinnati. Source: Author photo.

  A supervisor called out the evening assignments for dealers lined up in the high-limit room. Most were excited, and some were a bit nervous.

  When the VIP opening hour came, the staff created a meandering gauntlet across the casino floor. As the guests arrived and made their way to a ballroom for the opening reception, the staff cheered wildly to welcome the visitors. Kanter, dressed in an orange tie to match the Total Rewards logo, wandered behind the line and applauded enthusiastically.

  After the invited guests enjoyed some drinks, food, and live jazz in the ballroom, management tried to hush the crowd to thank city officials, investors, and staffers who had made the opening of a new casino possible. Loveman used humor to quiet down his audience. “I want to remind you that I am a casino boss. You’ve seen all the movies and you know what I am capable of,” he said. “There are a lot of big guys with Italian names waiting for instructions.” To each side of the stage, a showgirl, imported from Las Vegas for the event and clad scantily in feathers, smiled perfectly throughout the presentation.

  A little later, I met Loveman in a small conference room off the high-limit room. For all the celebration, for all the customer data analytics Loveman had pioneered, his company continued to face tough times. A week earlier, the company had reported a 2012 loss of $1.5 billion. The year before that it had recorded a loss of nearly $700 million, and things would only get worse in the future. In both years the company paid more than $2 billion to service its debt alone. The Wall Street Journal dubbed Caesars “wobbly.” Even with its carefully tended client data, the heavy debt load threatened to become the storm that ravages all.

  Did data analytics represent an elegant sandcastle that a rising tide of debt would wipe out? “We’re not going bankrupt,” Loveman said. “We show a loss because of things like interest expense and write-downs on depreciated assets and things like that. But the operations of the company are highly profitable and very successful, so the company last year made right at $2 billion operating its casinos around the world. That’s an extraordinary number—very healthy number.” In 2013 that number was to stay largely the same, coming in at nearly $1.9 billion.

  A performer tethered to the ceiling pours champagne for guests on opening night at Horseshoe Casino Cincinnati. Source: Author photo.

  He noted the company had a low level of debt until it was privatized. “It was an overt decision by its new owners in 2008, and that’s what the burden has been. It’s a financing decision, not the operation of the business that’s the problem,” he said. “The measure of Total Rewards is how it competes against its local and regional competitors every month. What goes on the balance sheet is not indicative of that at all.”

  Loveman’s case was that there is a difference between the company’s overall loss under rules known as generally accepted accounting principles (GAAP) and the day-to-day profit it makes from running its casinos. He said the company makes money virtually every day virtually everywhere it operates. But the expense of the debt, write-downs of old buildings in Atlantic City, and other things make the GAAP-reported number a loss.

  Loveman said the company could ease its debt load by selling new shares to the public, but he did not want to do that. He said he could raise billions of dollars that way, which would lower debt and increase profits. “But the problem with that would be the existing shareholders, including myself, would be diluted terribly by the number of shares that we would have to give to do that and none of us wants to do that,” he said.2 Loveman does find it frustrating that the debt issue casts such a long shadow over his company and its innovations using customer data. “It’s a huge challenge. We’ve been working on it for a long time. The company was financed in a certain way because everybody thought the world was going to be better, better, better, and almost immediately after it was financed in this way, the world got worse, worse, worse,” he said.

  Gary Loveman at the opening of Horseshoe Casino in Cleveland, 2012. Source: Jason Miller, courtesy of Caesars Entertainment.

  And things got even worse in 2013, a year in which the company reported a $2.9 billion loss, nearly double the loss of 2012. By then, Caesars had not recorded an annual profit since 2009, which itself followed a year in which it lost more than $5 billion. By the start of 2014, Caesars faced a total debt of $23.6 billion—more than the annual GDP of dozens of smaller countries.3 Randy Fine, Loveman’s former Harvard Business School student, ended up working for him as the first vice president of T
otal Rewards. Now he is critical of his former boss. Fine expects that Caesars will one day have to declare bankruptcy. “There is no question that the company chose to take on too much leverage,” he says. “Debt can’t be serviced because revenues don’t support it, and revenue is driven by customer management.” Yet many investors disagreed, and the stock price of Caesars Entertainment had about doubled in the year after the opening of the new property in Cincinnati. Optimists thought that if anyone could extract themselves from such a debt, it would be Gary Loveman and his team.

  Visitors saw no clues of any of this uncertainty about the future on the casino floor that night in Cincinnati. As Loveman spoke, thousands of people lined up to try their luck in the new establishment. When they finally got in, many headed first not for the gaming tables or buffet restaurant, but for the Total Rewards lines to get their loyalty cards. In the Cincinnati area, about sixty thousand people already had rewards cards from visits to other Caesars casinos. A few weeks earlier management had ordered 197,000 entry-level Gold cards to distribute to new customers ready to share their personal data in exchange for rewards. On opening night 1,825 people signed up before swarming the casino floor. “You can get a free trip to the buffet,” said Edward Willis, sixty-six, a local who recently retired. “I’ll take free food.”

  Asked what he thought about sharing his personal data through the program, Calvin Daniels said he was anxious to collect his rewards and winnings. And besides, he added, “I’ve got nothing to hide.” By the end of the first week, 17,770 new people had volunteered to give away their data by signing up for Total Rewards.4 The opening night was not just a flash in the pan. Within the first year of operation, nearly five million people had visited the Cincinnati casino.5

  17

  Embracing Outside Data

  The Path to the Sunshine Test

  Tariq Shaukat inched up Las Vegas Boulevard in bumper-to-bumper traffic after leaving a meeting. As the chief marketing officer at Caesars, he worked out of a luxurious office in Caesars Palace near Gary Loveman. Yet much of his marketing team occupied space south of the airport, six miles away. He had come to know the route back to his office intimately, skirting past the airport, then the massive Mandalay Bay complex to his left, followed by the glass pyramid of Luxor. Next came the faux Big Apple skyline of New York-New York and the massive fountains in front of the Bellagio before he arrived at Caesars Palace.

  In all of Gary Loveman’s time there, Caesars had just once strayed from their policy of only using personal data they had collected with the customers’ consent. That had been a typical Loveman experiment. The company knew that many people visited Las Vegas who were not Total Rewards members. Perhaps if Caesars could identify some of those visitors, they could convert them into long-term clients? So they bought outside data, sent out promotions, and waited to see the results. Pretty soon, it became clear the effort was a dud. The failure loomed large in the internal corporate lore. Sometimes executives would float a new thought to supplement their information on clients by buying from data brokers such as Acxiom only for the past failure to be recalled, and the idea would be discarded, both on practical and philosophical grounds.

  Yet as the capture and interpretation of personal data continued to grow more sophisticated, Caesars risked falling behind as rivals gathered as much information as possible, from any possible source, often with great success. Early in Loveman’s tenure at the company, using outside data had seemed unfair, something clients did not expect. Since then the practice had become commonplace across many industries, even if most clients did not know about it.

  Even the way Caesars handled and processed their impressive amounts of data appeared a bit dated in the era of cloud computing. That’s why Joshua Kanter, joined on the line by the company’s IT director, called Shaukat that late summer day in 2012 as he was driving back from the airport-area office. They wanted to cast aside some of the past restrictions on how the company processed customer information, including a ban on storing data on external servers owned by other companies. Shaukat listened attentively. At one point he asked how the company could expand its understanding of customers by gathering information about them not provided directly to the company. When Kanter reiterated that Caesars did not allow any use of such third-party data, and did not approve of storing their own data on external or cloud-based servers, Shaukat asked simply, “Why not?”

  With that, Shaukat and Kanter started plotting a data revolution—a revolution at least by Caesars’ cautious standards. They would use outside data and find more efficient ways to process and store that information. Neither man had been around in the mid-1990s when the company first embraced its “no outside personal data” policy. Nor were they present when Harrah’s bought Caesars Palace and reconfirmed the restriction in the mid-2000s. But both felt it no longer made sense to embrace a blanket ban that predated even Loveman’s arrival at the company. Too much had changed in marketing to customers not to take a different approach. Also, as long as the customer data could be securely stored elsewhere, it did not necessarily have to stay in Vegas.

  A policy change on this scale required Loveman’s buy-in. Shaukat brought up the issue on the weekly senior management call. Loveman knew the marketing team wanted to beef up the company website and create a greater online presence. Shaukat referred to such efforts when he brought up the issue.

  “By the way, Gary, one of the things that we are looking at is third-party data. For example, as we get more into digital, what types of third-party data are we open to using and not?”

  Shaukat was new enough—a recruit from the world of consulting—that he could still question where the policy had come from in the first place without seeming unprepared. “There is a rumor at least that you said no third-party data under any circumstances can be used. Is that still true or not true?”

  He waited for Loveman’s response. The future of the company’s marketing direction rested on it. Loveman, whose deep voice gave a powerful impression even through the tinny speakers of a conference call, replied, “I don’t remember ever saying that, but we are absolutely not going to use data on personal, private information, such as information on assets, finances, and health care.”

  Many data brokers try to piece together clues about how well-off people are financially, what assets they own, and what they earn. That information was too personal, the CEO said. But he opened the door on other fronts: “We are absolutely not going to use that sort of data. But using data like some of the demographic data that’s out there, some of the social data that’s out there, I’m open to rethinking where the right line is.”

  Shaukat told Loveman that he would send suggestions on how the company should update how it used technology to handle customer data. They had moved the line.

  Shaukat, Kanter, and others, including from the legal team, started a formal revamp of the policy, emailing back and forth until they hammered out a document cautiously embracing outside data. Caesars could use third-party services like cloud-based computing to handle customer data. They could turn to outside data brokers like Acxiom for more insights on customers, provided they secured the right confidentiality agreements and security controls. The executives remained mindful that they should still approach outside data with sensitivity. The last thing they wanted to do was alienate loyal customers.

  “Data-driven marketing is so core to what we do that we have to make sure that we are preserving the trust and preserving our ability to actually use data to drive our relationships with customers. That’s the DNA of the company,” Shaukat later reflected. “We no longer have to reject out of hand, without even thinking about it, policy around using third-party data. But we will always want to be several steps on the right side, on the correct side of that line. . . . If there is a gray area we generally try to stay on the whiter side of that as opposed to getting into the gray.”

  Caesars needed to understand better how clients spent money beyond the casino floor, on things such as ent
ertainment and dining, increasingly important slices of the Las Vegas casino revenue pie. Figuring out who the potential gamblers are based on personal data has always proved rather difficult. The only broad personal data category predictive of potential gamblers is small businesses owners: entrepreneurs have a risk-reward profile similar to those willing to wage an all-or-nothing bet at a gaming table. But entertainment is another matter. Kanter started exploring partnerships with companies such as Ticketmaster, the ticket distribution and sales company, which knows what kinds of shows people like. If you have bought tickets in the past to concerts of ’80s rock bands, for example, Ticketmaster’s data may prompt the casino chain to recommend Caesars shows in that genre, much as Amazon and Netflix do for their users. Or maybe more information about dining habits would allow Caesars to better target their many culinary offerings. Part of the value of a partnership with a concert promoter was finding new customers, appealing to Ticketmaster customers who were not Caesars regulars.

  Behind closed doors, executives spent many hours considering how they might use personal data from social networks such as Facebook. They wanted to make the Total Rewards website more social so that guests could share with friends the news that they had just booked a room at the Flamingo Hotel in Las Vegas or that they were redeeming points for a gift in their merchandise store. When a customer signs in to the website via Facebook, the casino company can also see much of the data on the customer’s Facebook profile. As shown in earlier chapters, outsiders can infer a great deal of intimate information from Facebook, including from simple things such as “likes” and friends.