A few months back, we discussed what may be the only casino upside to this pandemic: the ability to test floor optimization theories in real time. Prior to last March, convincing any operator to turn off a good portion of their floor to evaluate a data hypothesis would have been considered lunacy.
There’s long been a love-hate relationship between the BI/Analytical slot group and old-school slot managers. It’s been improving recently as new tools emerge, but there remains a good deal of skepticism. Each side was entrenched in their silos and armed with stories of epic fails generated from both camps.
But this unique COVID opportunity – or disaster, depending upon your point of view – has been a true win for the analytical team. The forced shutdown of 10% to 40% of casino floors has produced some major lessons that are applicable both now and into the future. Several casinos are actually making more revenue with fewer slots than before. The secret to doing so was surprisingly simple: increase the percentage of good games, regardless of your floor counts.
In our first pass on this topic, Nick Hogan of ReelMetrics postulated that “slot floors are grossly over-diversified” and horribly unbalanced. Earlier this month, Brendan Moore, an analyst with Eilers & Krejcik, went a bit further, issuing a quarterly column called “The Case to Reduce Slot Floor Diversification.”
While many of their arguments are similar, it is very significant that these data come from two different sources. Eilers & Krejcik is looking at data from 184 casinos and almost 200,000 slots (that is, when they’re all turned on again). ReelMetrics’s counts have similar numbers, and few of their metrics, if any, overlap with Eilers & Krejcik’s. While accurate projections and analytics can be obtained from small samples, this combined dataset is massive. Therefore, their conclusions are almost undeniable.
“It is time we take a hard look at what is not working for (us) on the casino floor,” Moore writes. “The typical casino had 219 distinct themes just for their Core Video Reels. Interestingly, only 3% of these unique games do at least 2x Zone averages. This make one wonder if operators are carrying enough of these high performers on their floor at the expense of floor share of themes that underperform.”
While both services stress the over-diversification of floors and the need to eliminate weak performers, Moore’s report also highlights the need to increase your share of the stronger games. He cites some specifics: “‘Buffalo Gold’ performs 2.2x versus Zone and 2.3x versus House. Despite unmistakable performance, about 50% of operators only carry one bank. This begs the question, are operators adequately matching slot machine supply with demand?”
This is an age-old question that has plagued operators for years. The reasons are varied and defy simple explanation. One element is surely the manufacturers themselves, who have a strange tendency to emphasize buying new titles over adding more of their existing winners. Another may be operator FOMO (fear of missing out) on the next big hit. And then there’s the failure to act on the obvious. Often the data is staring us in the face, as it still does with Buffalo Gold, Wicked Wheel Panda, Ultimate Fire, and Fortune Coin Boost. Doesn’t your data say to add more of those games? It should. So have you?
If you haven’t, it’s probably because there’s another simple theorem that states that increasing the game counts of an existing strong title will simply spread demand over more units. In other words, you if have five games doing $250 WPUPD (win per unit per day), adding five more will drop the WPUPD to $125 instead of generating incremental income.
In many instances, this is precisely what happens. Moore cites another strong Aristocrat game, More Chili, where the data show that pattern. In their recent samples, the game does a very respectable 1.3x. But, he says, “For each additional unit of More Chili we add to the floor, we predict unit level performance to fall.” Why does dilution affect this game, and not, for example, Buffalo Gold?
The answer requires even more data. The Eilers & Krejcik basic performance reports are free to contributing operators, but the company recently added a new fee-based service that offers greater insights, including heatmaps, Tableau dashboards, utilization trends and predictive possibilities on pricing, conversions, banking, and so forth. Utilization data is one of the keys to answering the questions outlined above about why one game can suffer from dilution and another can generate incremental revenue.
Many believe, this author included, that having performance index data from local, regional, and national sources is critical to making solid decisions. However, right now, neither of these services has access to as much player data as any individual operator can derive from their player tracking application. Each of these index vendors is working to add more of this data in the future, but right now operators have the huge advantage of being able to combine both critical metrics with ease.
Eilers & Krejci’s new performance dashboard
In the examples above, you can determine the session time or seat time of individual players, along with their repeat play patterns and individual utilization data on each slot. It becomes obvious after looking at those numbers why you need to add, or not add, more strong titles. Indeed, a game could be at 3x or higher, but you may not want to add more of that title. Why not? If the player data indicates weak repeat play, adding more games would simply mean the end is coming more quickly. Strong WPUPD on a game with weak repeat play indicates that once your players have all tried the theme, it will probably drop like a rock. Unless you have an unlimited source of newbies (in other words, the Las Vegas Strip), you should remain conservative and be prepared to quickly install new themes or cabinets.
Surprisingly, not that many operators are combining both their game performance stats and the individual player stats at the machine level in their analytics. Are you? If you haven’t done so, you’re missing a tremendous opportunity. That begs the question that if you have exclusive individual player data, do you still need a national/regional ReelMetrics or an Eilers & Krejcik service?
The answer is yes, and perhaps the best proof of concept is to look at our current overdiversification. If you don’t subscribe to one of these national databases, it can be argued that you don’t know what you don’t know. Every slot salesperson comes to your office armed with glowing reports of terrific results of their newest cabinet(s) or game(s) and mentions that ABC casino down the street just installed a bank of six games. This is perhaps the real reason why Moore’s report shows that disturbing number of 219 Core Video Reel themes.
Instead of relying on hearsay accounts, a quick review of ReelMetrics’ Star Scout report. or Eilers’ Top Indexing NEW Games, will provide the data you need to make better choices on introductory titles. They can also provide a comparison to the existing indexes to decide if you’d be better off adding more proven themes.
In addition to improving profitability, it stands to reason that the customer experience will be enriched if each player wandering your floor has a better chance of landing on a game that is both popular and enjoyable. Hopefully, they’ll avoid the disappointment of not being able to find an available popular game. While they may, or may not, have a winning experience on that game, at least their losses will be on a theme or cabinet they enjoyed playing. Property brand loyalty and repeat visits counts should therefore increase along with the WPUPD.
Heatmaps are another important tool for putting your data into proper spacial context on your floor. Compudigm, Casino Data Imaging and Bally Visualization were pioneers in these products. Viz Explorer later emerged as one of the leaders in heatmap analytics. All these vendors used a form of geo-mapping to display performance data in color-coded metrics. Their strength was showing anomalies and opportunities in zones or entire floors. Now both Eilers and ReelMetrics are incorporating heatmap features with their national indexes so operators “can have more information to make their decisions efficiently and effectively.”
Considering that Moore says the Eilers & Krejcik North American database contains almost 4,000 unique themes in the Core Video Reel segment, you can understand why these experts are united in their assessments that we’re over-diversified today. We should all be paying closer attention to our analytics – and theirs – and doing so as soon as possible to make the corrections necessary to boost our revenues and provide better customer experiences. Doing it now can help alleviate shutdown pain. It can also pave the way for stronger growth during the recovery.
It’s a shame that it took a pandemic to remind us of these simple lessons.
A new sample of a floor heatmap from Eilers & Krejcik
# # #
