How To Make Data Driven Decisions - For Casinos

You want me to what? Team members may not say that out loud, but we often think it when asked to make a data-driven decision. With all of the talk about analytics software, metrics and data-driven decision making, knowing how to sift through all of the noise to make the best decisions for your casino and your tribe’s holdings is difficult. The good news is you don’t have to be some sort of mathematical genius—or even have a degree—to figure it out. In fact, you can even be one of those people that has always been a little scared of numbers (gasp!) and still make better data-driven decisions than even some of the most seasoned casino execs.

What is data-driven decision making anyway? It isn’t as complicated as it sounds. It just means making a decision that can be backed up with verifiable data. So a non-data-driven decision to close the buffet on Mondays would sound like this: “The buffet doesn’t look very busy on Mondays, and it seems like we have too much staff and not enough guests—let’s close it on Mondays.” A data-driven decision on the same topic would sound like this: “We have 67% less covers on Monday in the buffet than our next-slowest day. We lose an average of $32,000 from that outlet every Monday, compared to the other days when we lock in an average profit of $3,700. We also only have 45 people coming to the buffet on Mondays, compared to an average of 300 on the other days—let’s close the buffet on Mondays.” Big difference, right? If you start using data to back the reasons that you think the team should do what you say, I guarantee you will boost your credibility, and decisions will start going your way. Kind of like magic—except it isn’t.

Why does it matter? Data and all of the beautiful reports in the world don’t mean anything if there isn’t a team to look at the reports, make sure to understand them, and use them to make decisions. In their 2012 feature on big data, Andrew McAfee and Erik Brynjolfsson describe the opportunity and report that “companies in the top third of their industry in the use of data-driven decision making were, on average, 5% more productive and 6% more profitable than their competitors,” even after accounting for several confounding factors. The good news is if you don’t have to make data-driven decisions, you aren’t alone. Most people don’t (even if they pretend like they do), and if you figure out how, it can give you and your casino a major competitive advantage.

Here is how to get started in three easy steps:

1. Pick the right metrics.

Most casinos get buried in the detail of 30 to 300 metrics and lose sight of the big picture. You would be shocked to know how many people get so caught up in looking at the details that they forget to check to see if they generated revenue and made a profit. At the end of the day, for the overall casino, that is what matters. One of the most important steps is to pick the right metrics to look at. If you aren’t sure what these are specific to your department, that’s okay. Get help from someone you know and trust. If you are the one expected to know, reach out to a friend that holds the same position at another property. Ask them, “If you could only pick three key data points on your reports to consider for operating your business, what would they be?” Don’t stop there. Ask, “Why those? What do they tell you? What is an example of those metrics raising a red flag? What is an example that shows you are operating your department well?” Be cognizant of the size of the casino you’re inquiring about in relation to your own. Make sure you pick numbers you have access to daily—you don’t want to wait until the middle of next month to know how your department is doing this month; it will be too late to adjust and do anything about it. The rule of thumb is to make sure your metrics are tracking something your business cares about—like that it is operating as efficiently as possible, that you’re getting the most out of your department (covers, slot fills, calls taken) while spending the least amount of money possible.

 2.  Ask the right questions.

Asking questions is a sign of brilliance and one of the critical steps in driving better results for your casino. Don’t let the analysts scare you! Here is a secret: Asking analysts questions often makes them uncomfortable, because they are worried you will spot a mistake in their numbers.  Push through the discomfort, be brave, and ask them questions until you understand what you need to. It is okay to push back on analysts’ conclusions. Here are some good questions to ask:

What is the source of your data? If the data isn’t accurate, then you shouldn’t be going “all in” on the decisions you make from it. Knowing where the data comes from will help you get a feel for how accurate the data is. For example, if the daily revenue number is pulled from your POS system, there is a high likelihood of accuracy. If it is pulled from printed receipts and tabulated by the accounting department because the power went out that day, you might need to omit that day from the data.

What assumptions are behind the analytics? Meaning, there may be conditions that might make the analytics invalid or misleading. It is important that you know. For example, if there was a blizzard last year so massive that it shut down the casino for three days, the group shouldn’t be celebrating the big year-over-year increase shown by the data and planning to execute the same marketing efforts next year as a result.

Does the data include outliers? If so, understand how the outliers affect the results. For example, if you had a player that hit Megabucks at your property, and it tanked revenue in the month of February, it is important to know if that loss is included in the analysis, or if it was removed, before you go celebrating this year’s results.

3. Is what I am seeing in the data legit?

 Most of us have heard the phrase “correlation is not causation,” but figuring out just what that implies when evaluating data isn’t that easy. Back to the weather example: if the property shuts down for three days and revenue plummets because of a blizzard, clearly the blizzard caused it (i.e., the relationship between the blizzard and the revenue drop is causation based). When the marketing team give away free lobster tails to players in March and see a 7% increase in revenue, it might be causation based or it might not. Be careful before you draw conclusions. Ask yourself, “What else could be the reason for the increase?” Is the economy up as a whole? Is a competitor’s casino temporarily shut down? The best way to know for sure is to maintain a control group when executing marketing programs—that is, a group you don’t market to—so you can get a feel for how many players would have come anyway. I understand this is easier said than done, so, at the vary least, consider that increases or decreases in revenue and profit can be caused by things other than the obvious, so list out what those causes might be for the team to consider before you make a final decision. In other words, when is it reasonable to act on the basis of a correlation?

Now you are armed with the basics of how to make a data-driven decision!