The 2019 Major League Baseball Home Run Derby (held on July 8, 2019) caught my attention this year because it was advertised as being powered by StatCast AI (Powered by AWS). As an avid baseball fan (Go Cubs Go…) I had seen games in the past that had used some of the StatCast AI technology. But the Home Run Derby intrigued me more than others because it’s a major national event, one of the most celebrated events of the Midsummer Classic. As an individual who is involved in building a business that is driven by data analysis, the thought excited me.
It’s not rocket science
The broadcasters began chatting about exit velocity. And how some of the competitors have a different exit velocity based on batting left- or right-handed (for those competitors that happen to be switch hitters). Exit velocity is a calculated value that can be measured and then used to predict whether the hit will result in a home run. During the derby, the calculation has to be done quickly, showing viewers the distance the last home run traveled and at what velocity. While this information is fun for the viewer, it also allows players to work towards an optimal Exit Velocity that would provide the result they are looking for, in this case a home run.
While the quickly calculated stats on exit velocity, launch angle, and distance traveled may seem impressive to the average viewer, they’re still just fairly simple calculations. My concern is that these calculations are being advertised as artificial intelligence–something not easily acquired. And when we think that way in the credit union world, we see data analysis as something outside our reach. Something we have to buy and seek out elsewhere.
How credit unions can use their data
In the credit union industry, we use statistics all the time and could use values from our members to make predictions. For example, a credit union may review the average age of a member who takes out a loan (the initial value) and then work towards attracting more members within that age range so that they have a better chance of lending to that individual. Seems simple right? That’s because it is. It is a simple statistic that is used to drive credit union executives towards a specific age range where they think they can succeed. Just like the baseball players who are striving to attain a specific exit velocity so they can achieve their desired result.
Credit unions need to analyze what are the key statistics that help them succeed, and they need to advertise and promote to members that help them get closer to those goals. There is no artificial intelligence or neural network that is necessary for success. There is just simple math that can be used repeatedly to succeed. Just like the average member age when acquiring a loan, credit unions can begin to build other statistics that make them successful. For example:
The average period it takes a member to pay off an auto loan
The longer a loan lives on the books, the more interest income a credit union earns. Therefore it makes sense for credit unions to find members who may be approaching the average period of time it takes a member to pay off an auto loan, and remind these members how the credit union provides excellent loans and how it can provide benefits to the member that no one else can.
The average amount of ACH deposits that members receive on a monthly basis
This can then be used to identify members who do not actually meet the monthly average, as these members may not be receiving all their payroll deposits via ACH at the credit union. I would lobby these members on the benefits of having all their employer payroll deposited to the credit union. The credit union will benefit because they will have the ability to earn interchange income as the member spend their payroll funds using their debit card.
Now once a credit union identifies the data that helps them succeed, they can adjust their strategies to ensure successes within that realm. In addition to adjusting strategies credit unions must continually reassess the statistics and assumptions that are important to them. Because the membership may not continue to act in the same manner that was originally seen when devising the strategy. This again is not Artificial Intelligence; it is a simple statistic that is used to predict an outcome.
The bottom line is there is simple math that can be used to help credit unions succeed. And it does not take a large investment to begin reviewing these statistics. That is why I believe what we think as being Artificial Intelligence is often just Artificially Intelligent. This is because we want to believe that there is something that we cannot attain, and something that we need to purchase, or something that we need to compete with the big dogs. But it’s not that hard. We can all start small and identify the performance indicators that help us to succeed.
