SBJ/July 21-27, 2014/Research and Ratings

Big data can answer where and when but not why

There is a relatively new form of research that is already changing your life. Predictive analytics, often called “big data,” allows researchers to combine data from a wide variety of sources and predict what people will do, when they will do it, how much of it they will do, how much they pay for it, and how often. Most of the information in these vast databases is of little obvious value, but a handful of the seemingly random variables when linked together make it possible to predict how people will act.
  
In isolation, those variables are not powerful. Together, they are game changers.

Here are a couple of hypothetical examples. The first data source provides constant details about the weather at a stadium: temperature, cloud cover, humidity, precipitation, etc. The second data source comes from scanning tickets at the turnstile when a person comes to a game and when that person leaves, if they leave before the game is over. The third data source is constant concession sales from every sales source in the stadium. Imagine these data points are collected for every game. In time, a team would know that if it is cloudy at the beginning of the game but the sun came out before the second inning, you would still have a potential walk-up crowd. The radio announcers could say, “It’s turning into a great day for a game,” and you would get 112 more fans at the game. Conversely, if the temperature drops below 60 degrees during a night game, a team can sell more coffee or hot chocolate, and if it offers specials at the concession stand closest to the most popular exit, 23 percent of those who were thinking of leaving will buy a hot beverage instead and go back to their seats.

Powerful stuff. Predictive analytics will provide laser-like accuracy in determining which media to use, and with what weight, to increase attendance or viewership. It will help predict the number of items that will be sold at a point in time in response to myriad predictable behaviors beyond the weather. The bottom line is improved by more accurate product production and availability.

This is just the tip of the iceberg both in terms of what predictive analytics can do today and the incredible potential it has for the future. The reason the horizon is so bright, and the reason we have capability today, has nothing to do with new thinking on the part of researchers. We knew of these models when I was a student in the 1970s. But back then, we didn’t have computers powerful enough to handle the data. Now we do.

There is another side to predictive analytics. At this point, the vast majority of the information used in predictive analytic databases is “secondary data.” That is, data that were not directly or intentionally gathered from humans, or data that were collected for some reason other than how they are being used with predictive analytics. The temperature in the fifth inning, the moment a ticket is scanned, the time a cup of coffee is purchased, the price of that coffee — none of that
information comes from talking with a person.

Why does that matter? Predictive analytics allows us to predict consumer behavior more profitably, but it does nothing to tell us why people are there in the first place. That may not seem critical, and when you are talking about consumer transactions it is indeed less critical. Buying toothpaste is a transaction. Being a fan is a relationship. I never tell stories about how excited I am that I am almost out of toothpaste and will soon get to buy a new tube. But in the fan relationship, if we are not having experiences that inspire stories we want to tell, we are not having a relationship — and sooner or later, our love will wane and we won’t be in the stadium to be affected by predictive analytics.

This is the fifth anniversary of Up Next. In the first column, John Walsh from ESPN said he was concerned we were cashing in on the love of sports adults have now that was built on $5 tickets to games 30 years ago. That was forward thinking and is a caution for us today. Predictive analytics will allow us to make the most money, the most efficiently, at the lowest cost, but only from people who are in relationships with sports today. It does nothing to answer how we build fans for tomorrow.

The chart to the left shows what affects, starts or builds the love of the game for different demographic groups. Notice that for kids ages 12 to 17 years old it is all about play. Second most important is going to games. Watching games on TV is almost an afterthought. And yet, most of the money in sports now comes from the media value.

Predictive analytics won’t tell us the key to the future is kids playing the game. We can only know that by talking directly with people. So, full speed ahead with predictive analytics. It is great for business. But we are not toothpaste. We need to constantly refresh our understanding of the relationship fans feel with their sports, and the best way we can do that is through research that directly and intentionally asks them why they love the game.


Rich Luker (rich@lukerco.com) is the founder of Luker on Trends and the ESPN Sports Poll.


Return to top

Related Topics:

Research and Ratings

Video Powered By - Castfire CMS Powered By - Sitecore

Report a Bug