[MUSIC] The second reason that analytics is a leadership problem is that analytics requires process and incentive changes. I want to give you an example from one of the most radical transformations towards big data and analytics that I know. And the remarkable thing about this transformation is that it started in 1998, more than 15 years before most firms seriously looked at data analytics. This is a story about Harrah's Entertainment, a casino gaming company. From 1998 to 2002, Harrahs went through an incredible period of growth. From about $400 million in revenue, to about a billion, but prior to 1998, Harrahs' prospects looked dim. You see, some years earlier, the famous Las Vegas entrepreneur Steve Wynn had built the Mirage. It became the new standard for luxury in Las Vegas, and it also led the transformation of Las Vegas from a somewhat seedy place to a Disneyland for adults. From Harrah's perspective, the question was, How can we compete against this new wave of competitors? Why would somebody come to Harrah's when new hotels like the Mirage have these opulent pools and shows like Siegfried and Roy's tigers? So, one way to compete was to upgrade facilities. But with 29 casinos across the United States, this turned out to be incredibly capital intensive. For example, The Mirage alone costs $630 million to build. Another way was to upgrade only casinos in markets like Las Vegas and Atlantic City, which is where the new competition was. But this would have created an inconsistent brand experience for Harris's customers. So instead, Harris decided to capitalize on the fact that they were much more geographically spread out than their main competitors. The other companies were concentrated in Las Vegas, Atlantic City, and a few other gaming markets. Harrah's, on the other hand, was all over the United States. The idea was this, let's leverage the fact that we have consumers in very geographically diverse markets. If we can create really good relationships with these consumers in the local markets where we don't face a lot of competition. When a consumer comes to Vegas or one of the other destination markets, we want to give them a choice. They can either go to the Mirage and look at the tigers and swim in the pool, but get treated as a nobody. Or they can come to Harrah's, which is not as nice, but they will be treated as kings, because Harrah's knows who they are. So what do you need to make the strategy work? You need three things, first, you need to identify consumers across markets. Second, you need to figure out who a good customer is. And third you need to figure out how to give these customers incentives to stay at Harrah's when they come to Las Vegas. That's the basic idea, pretty simple, but not so simple to implement. The problem was that most properties already had their own reward card and incentives. But each of those only applied locally. So you went to these South Lake Tahoe Harrah's and they would comp your room and give you a steak dinner, but only if you came back to South Lake Tahoe, not if you went to Vegas. In other words, the problem was that every property had its own profit and loss responsibility and resources. And as a result of that, the property manager saw consumers as belonging to their individual property and not to the corporation. It was as if the property managers were feudal lords who were running competing fiefdoms. So, what did Harris do, the answer, they implemented a massive organizational change. The hired a new chief operating officer. They demoted all the property managers to report into the new COO, and they created incentives for them to think of themselves as part of Harrah's, not just individual property managers. They put in total IT integration to track consumer behavior. And finally, they put in a loyalty card program called Total Rewards that gives consumers an incentive to visit Harrah's and use their card. Let me give you an example of what this did. Imagine you are a 55 year old woman who lives about 30 miles outside of Atlantic City, New Jersey. You are an experienced gamer, and therefore the first time you walk into a Harris casino, you first go to the total rewards office and you sign up for a total rewards card. This gives Harrah's information about who the woman is, where she lives and a few other things about herself. At this point, the woman walks into the high stakes slot room, where every pull costs $25. So the woman pulls once and looses $25, the woman pulls a second time and looses another $25. The woman pulls a third time loses $25 and you are perhaps waiting for a happy ending. But, no, the woman pulls a fourth time and loses $25 and she walks out of the casino and goes home. So in every other casino company at the time, somebody looks at this woman and asks is she a good customer? And the answer is not really, she wagered $100 with us which is actually not a lot. And so treating that woman really well is a little bit like an airline giving top status in their frequent flyer program after only one flight. Harrah's on the other hand, looks at this woman's profile and says let me see who, among all 30 to 40 million customer profiles look similar to this woman. In that they have a similar demographic profile and live at a similar distance from one of our core casinos. And then they say, who of those people started their gaming career in a manner that was similar to what this woman did when she walked into that high stakes slot room? And what they find is that customers with this profile have an incredibly high lifetime value for Harrah's. And so, a little while later, the woman receives a direct mail offer from Harrah's, which looks completely extravagant relative to the investment that she puts into the company. And she becomes a Harrah's customer for life, because she was treated like royalty. That's the idea, that's what they could do, and that's what the others couldn't do. What Harrah's did is to realize that analytics fundamentally requires a 360 degree view of the customer. But getting there is hard, and invariably it requires organizational and incentive changes. And the important thing about large scale organizational changes is not surprisingly, that leaders have to make the changes. Not data scientists and not IT specialists. And that is the second reason why analytics is a leadership problem. You as managers need to lead these process and incentive changes. [MUSIC]