Predicting Success In Sports Sponsorship

BY ADAM GROSSMAN

IThe New Year often brings with it two things – resolutions and predictions. While almost everyone tests their resolutions almost no one tests their predictions. Why is that the case and why is that important for sports sponsorship?

When making New Year’s resolutions, many people look at the something in their past behavior or actions to determine something they want to change. For example, one of the most common resolutions is to work out more by going to a gym more frequently. One of the most common things that happens is that people do not keep their resolutions. A standard business model for gym owners is signing up more people than can possibly be accommodated at a gym location. These companies bet that many more people will sign up and establish recurring, monthly payments than will actually go to the gym.

While it is obviously not ideal (to put it mildly) that people do not keep resolutions, what is more ideal is the overall structure of resolutions. More specifically, people often make testable claims (i.e. I will go to the gym more often this year than last year) and with falsifiable outcomes (i.e. how many times did I actually go to the gym) that can be easily examined with testing.

 Resolutions by “normal” people typically use the best practices in scientific method that “predictions” by experts do not always follow. For example, experts often predict business trends that will happen in a certain industry over the following year in late December of the following year or early January of the current year. As many “wise men” ranging from quantum physicist Niels Bohr to Yankees Hall of Famer Yogi Berra have noted, prediction is very difficult, especially about the future.

As Nate Silver (among many others) have documented, experts can make claims that that are difficult to falsify, are often wrong, are rarely ever checked if they are wrong, and will continue to make predictions that are often just as likely to be wrong as “ordinary” people making resolutions. More surprisingly, experts will often not seek out new evidence and are less likely to change their predictions even in the face of evidence of their errors. 

The question becomes why make predictions at all if even experts (or more derisively pundits) “are roughly as accurate as a dart throwing chimp.” Annenberg School Professor at the University of Pennsylvania Philip Tetlock has studied this question most famously in his book, “Superforecasting: The Art and Science of Prediction” (quoted above). The primary answer is that people can learn a significant amount from being wrong.

In fact, Tetlock’s has identified a group of “Superforecasters” that perform better than 98% of the population (including experts). What he found is that Superforecasters basically do what experts do when making predictions. They make falsifiable claims, test those claims, and make adjustments based on the evidence. The quality that appears to make superforecasters so good at making predictions is that they are curious people that constantly seek out information. 

What does this have to with sports sponsorship? One of the key challenges to the future success of sports sponsorship is often perceived to have some of the same challenges as predictions. In particular, sponsorship inventory is sold based on industry experts’ non-falsifiable claims without having robust measurement to demonstrate if these claims are accurate or the curiosity to determine what actually works best for buyers or sellers of sports sponsorship.   

For example, one common claim is that sports teams create unique connections with fans different from other advertising or marketing channels do with fans that are difficult to “quantify”. There exists a “gut-level” association that people have with sports teams that is transferred to sports sponsors in ways that drive a return on investment (ROI) to their business.

Is this actually true? In the past, this claim has not often been tested because it was not thought that it could be tested. Developments in artificial intelligence and machine learning technology make this testing about how people feel and the impact on a business a testable hypothesis.

The Block Six Analytics (B6A) Corporate Asset Valuation Model examines how much increases in sentiment, awareness, and engagement will have on a company’s top-line annual revenue growth. Our Social Sentiment Analysis Platform (SAP) examines the text and images of social media posts to determine if conversation from both owned and earned accounts create lifts in a sponsor’s sentiment, awareness, and engagement for conversation relating to sports property with a company’s target demographics. We then can examine credit card transaction data, annual / quarterly statements, a client’s proprietary data, and / or B6A research to see if there is a correlation to increase revenue generation (we do often find a positive correlation).  

One potential criticism to our approach is that correlation does not equal causation. More specifically, it is impossible to say that the lifts in sentiment, awareness, and engagement are the direct cause of a company’s top-line revenue growth. One would need to run a double-blind research experiment that eliminates confounding variables (i.e. sponsorship is usually one component of a company’s overall marketing strategy) to make a causation claim.

That is actually a feature, rather than a bug, of our approach. In particular, our goal with our client is to determine which sponsorship opportunities will increase the likelihood of success. More specifically, which sponsorship opportunities will put companies in the best position to maximize their ROI and ROO goals for the coming year. We then can test our predictions quickly throughout the year using our machine learning technology and make adjustments to a sponsorship portfolio based on the results.

In fact, that is exactly what we did with Pepsi and the Dallas Cowboys with a new LED tunnel cover at AT&T Stadium. Pepsi used our technology to determine if the sign drove value to the organization (it did) and what was the best creative to use for this activation by testing different logos throughout the season. Pepsi was “quickly able to validate its decision to build the tunnel cover sign and optimize creative on a game-to-game basis” and this played a role in Pepsi renewing its larger deal with the Cowboys.

A critical New Year’s resolution should be to test predictions. Building in the ability to have falsifiable predictions that can be adjusted throughout a campaign, season, or year with clear metrics should help drive ROI and ROO for buyers and sellers of sports sponsorship.