We are drowning in a sea of data. Data was supposed to buoy our businesses—deliver customer insights, savings opportunities, soaring profits. We were supposed to be masters of our data. But the terabytes of information—slot machine data, table game data, marketing redemption data, labor utilization data, point-of-sale data, hotel spend data, customer survey data, even valet parking data—can be an anchor that drags us under because the data is often not properly harnessed or is misunderstood.
Data for data’s sake is useless. Data should serve one primary purpose: Improve our business results by enhancing revenues, reducing costs, or keeping customers and employees happy and engaged. But this is not happening in practice. As many as 85 percent of modern data projects are estimated to fail, for many reasons, including internal politics, overambitious scale, and, most significantly, lack of the right data analysis skills.
You may think that you have mastered your data, but unless you can answer these four key questions, you almost certainly have not.
Do you know where your best customers are?
Every business wants to take great care of its best customers. We segment our customers into different tiers and give special perks to the top. But do you know where those customers are at any given time, and do you know what they really care about so you can reach them and keep them loyal?
At one of my properties, we were having a difficult time moving the service score needle for our highest tier customers. Our company was obsessed with service. We surveyed our customers every day, and the top tier scores were weighted at 7-10 times an average customer, so it was critical to make them happy.
We had what we thought was a great idea—provide an amazing experience in the high-limit room. We pulled out all the stops. We put our best employees in there. We prepared elaborate food menus with roast beef, shrimp and high-end drinks, and brought everything to them while they played. We had supervisors and managers come through to engage them, let them know we were at their service and ask if they needed anything.
People loved it. But our scores barely budged. Why? Because at any one time, we might have had eight to 12 top-tier slot players and a handful of table game customers in that high-limit room, but more than 100 of our top slot players were on the main floor playing their favorite lower-denomination game, and having to fight for service with a sea of other players.
We decided to dive deep into our survey data and found—based on a location code that was part of the survey—that there was one zone where a lot of our high-tier customers liked to play. Armed with this critical data, we attacked the problem anew. We increased staffing in that zone to reduce wait times, made sure we had extra attendants keep everything spotless, and reallocated some supervisor and manager time to that area. And guess what? Our scores skyrocketed within just a few weeks and set company records for score improvement. It turns out that knocking out the basics for a large number of customers made a much bigger difference than wowing a small number of these same valuable guests.
Today, we can track customers in real time. By integrating this data with behavioral insights we can anticipate customer needs, deliver more meaningful personal interactions, and free up their time to spend it how they want. It is this combination of data gathering, analysis, and execution that drives real loyalty and keeps your customers coming back.
Ask yourself: Are you digging past the obvious to not only find your best customers at any time but also learning what you can do to make their entire journey positive and memorable?
Do you know why your customers are coming in today?
Gifts can be a big draw for casinos. They were our lifeblood at two of our regional operations. We handed out everything from dining ware to bedding sets to bobbleheads of local sports heroes. The beauty of gifts is that the perceived value can far exceed the actual cost of the product. We typically bought at $14 to $18 per item but most gifts had a $50 to $100 “retail value” to our customers.
The problem we faced was that customers were also redeeming a lot of free slot play on gift days—Sundays. It seemed like a waste of money. Why should we let them “double dip?” So, we decided to black out Sundays and make the free play offers valid on other days to drive an incremental trip.
That Monday, with the new program in place, I eagerly looked at our free play redemption for Sunday, excited to see a reduction from the typical $120,000-$130,000. Sure enough, it was down significantly—by $40,000 or so.
Unfortunately, our revenue was also down over $100,000—not a good trade. Maybe this was a fluke, and one week does not a story make. But for the remaining Sundays that month, we saw the same pattern of lower free play but also lower revenue and profits. Why? To find out, we did some digging into the data.
It turns out that a lot of our Sunday customers who had been redeeming gifts were not coming for the gift. They came for the free play, and Sunday was the only day they had free during the week. Since they had the gift offer as well, they redeemed it—who doesn’t want a freebie?
We looked back at redemption patterns from these customers and saw that they were heavily weighted to Sundays, even on days when there were not gifts, so the data was there all along to figure this out, but we had not done the homework in advance.
By going with the obvious theory and plunging ahead, we solved one problem but created a bigger one. Without the free play offer, those Sunday customers did not have the right incentive to visit, as they were not motivated by the gift. We fixed the problem by restoring the Sunday redemption option and looked for other ways to reduce free play. But we cost ourselves by making assumptions based on the face value of the data.
Today, we can even better use data to understand the drivers of customer behavior. When doing our analyses, however, we need to have a keen eye, and we must look in advance to understand those motivations. Once you do, you can avoid costly mistakes and drive the behavior you want.
Ask yourself: Are there situations where your customers might be coming in for a different reason than you think—perhaps when you have some kind of promotion or sale? What are the implications for how you manage them?
Do you know the best measure of success for your products?
The No. 1 metric for slot machines in the gaming business is “win per unit per day” (WPU). We base many of our decisions on finding games with a higher WPU. We use our data to track the “house average,” and segment our games into tiers based on how they perform relative to house average.
But, surprisingly, WPU is not the best measure of a game’s value to your casino. Operators waste a lot of money and valuable time every year chasing higher WPU numbers.
Slot directors have all kinds of WPU-based metrics to judge their games. I have often heard things like, “For a leased game to get onto my floor, it needs to do at least three times house average,” or “every quarter we look to convert or replace the bottom 10 percent of our floor based on WPU.”
I have also had countless capital conversations with slot directors focused on finding games with a higher WPU to replace low-WPU games. The slot directors declared victory a few months later showing great new WPU numbers compared to the old ones.
Most often, though, the new games with the higher WPU simply reallocate coin-in that was going to the other games on the floor. They do not change your customers’ behavior—make them play longer, visit more often or drive new business—and your overall revenue goes nowhere.
How do you know a new game really helped you? To answer that, you have to dig into the data and look beyond the WPU numbers. Analyze the behavior of your customers who played the old games and see how it changed after you replaced those games. Did they start coming more often or spending more per trip? Who is playing those new games? Is it customers from segments that you were struggling to attract previously—maybe millennials or certain high-end customers? Or is it the typical demographics that most games are designed to attract already, that late 40s-to-mid-60s female customer?
Manufacturers aggravate the problem by producing too many games that appeal to the “middle.” They know that operators want games that will produce heady WPU numbers, so they chase the broad demographics, looking for a big hit—similar to how the old network TV model worked. Before cable, streaming, Netflix, Hulu and on-demand, TV shows had to score big ratings to drive high advertiser rates. The networks paid more attention to certain demographics, such as 18-to-45-year-olds, but if a show could not garner 10 million or more viewers, it would not be green-lit. And the result was a lot of relatively bland big network programming that did not take risks and was careful not to offend, lest it alienate any significant portion of the viewing audience.
Today, TV content is in a much more specialized and compelling place. There are more good shows than most of us have time to watch, and on multiple platforms. Shows that appeal to narrow audiences making it onto Netflix, Amazon or even YouTube never would have been produced 15-30 years ago.
The big broadcast networks can hardly compete on quality, garnering only a small percentage of Emmy nominations and suffering significant ratings declines. By providing high-quality content for viewers, the insurgent competitors like Netflix are winning. And you can do the same by knowing what each of your customers wants and providing it to them.
Casino floors are not that crowded most of the time. The core slot customer can almost always find a game they like. We need to deliver compelling games for other customer segments to bring in more new business and grow the total casino market.
Digging into detailed customer ratings data shows that there are clearly games that keep customers coming back, while others, even if they perform well from a WPU perspective, do not grow the business because they do not create the right customer experiences at the game.
Ask yourself: Is the highest revenue always the best outcome for a particular product? Are there longer-term considerations when thinking about revenues like sustainability and how to ensure repeat business? Hint: There are!
Do your customers really want what you think they want?
When we talk to our slot customers, the thing we hear the most is that they want “time on device” (TOD). Everyone wants their money to last and no one likes to lose everything quickly. Slot manufacturers know this, and they have often responded by making a lot of games designed to preserve the player’s bankroll.
There is a huge problem with this approach, however, when it comes to experienced slot customers. These games are boring. While some customers are OK with just killing time and some less experienced customers often gravitate to these “dribbler” games at first, customers tire of them and walk away eventually, so time on device actually goes down.
Most slot customers want that roller coaster—where they can win a meaningful jackpot, get ahead of the house at some point and make their decision whether to press on or cash out. Over time the success of manufacturers like Aristocrat—with their higher-volatility approach to game design—bears this out.
Average time on device is significantly higher for these more volatile games because people reinvest their winnings back in the game, hoping to win again and again. The “time on device” math models usually produce the opposite.
Armed with this insight from a rigorous look at our company data—because it is not just volatility, but other factors as well—we started quantifying the number of “big winners” and “small winners” each game tended to produce, and started allocating our slot capital to increase the number of winners on our casino floors across the company.
So, when you hear something from your customers like “I want time on device,” believe it, but do not assume that they necessarily know what that means. TOD is totally determined by the customer, and if the game is exciting, they will keep playing. If it does not have enough thrills for them, the majority of savvy customers will walk away. You will need to lay out your floors with the right mix of high and low volatility depending on what your customer mix looks like.
Ask yourself: Are there other instances where customers want a certain end result but the obvious answer is not the right answer—maybe fast service but an interaction that does not create engagement?
Critical To Success
You cannot run a successful enterprise today without leveraging data and using it to guide your key decisions. Properly harnessed, data can be an incredibly powerful tool to understand and predict customer behavior and drive efficiencies at every level. But it takes the right approach, talent and time.
Human beings are complicated, and their behavior often defies neat demographic labels. Proper leveraging of your data and digging deep to understand customer motivations will allow you to better predict future behavior and anticipate your customers’ needs, give you a significant advantage over your competitors, improve customer loyalty, obtain and retain new customers, and grow the overall market.