Gaining an Unfair Advantage with Big Data and Analytics

I’m surprised at how many organizations are still having trouble understanding the differences between business intelligence (BI) and advanced analytics. Well, Step 1 in the Big Data transformation is not only realizing that “Big BI” is not “Big Data,” but also understanding how they are different (see Figure 1) and how you can leverage Big Data to gain “unfair advantage” for your business.

Figure 1: Differences between BI and data science

The key differentiator between BI and data science is that BI is focused on retrospective and descriptive analysis—understanding and reporting on what happened in the past. However, data science is focused on not only understanding, or quantifying, why things happened in the past, but to leverage that insight to predict what is likely to happen in the future and deliver recommendations as to the best actions given the situation. I covered the difference between the two in detail in my recent blog, Business Analytics: Moving from Descriptive to Predictive Analytics.

Big Data Changes the Metrics that Predict Performance

As many of you know, I love the book Moneyball. But let’s be clear, the book and the movie are not the same thing. The movie is an interesting tale starring Brad Pitt (even my wife went to see the movie, primarily because of Brad Pitt). But the book Moneyball: The Art of Winning an Unfair Game is about statistics—it just happens to be staged around the story of baseball. One of the key take-aways from the book is that through advanced analytics (what they call sabermetrics in the world of baseball), we can uncover new metrics that are better predictors of performance.

Billy Beane, the General Manager of the Oakland Athletics, was the one of the first in baseball to leverage sabermetrics in the daily operations of a professional baseball team. As a result of his application of sabermetrics, he was able to identify metrics, such as on-base percentage and slugging percentage, that were better predictors of on-field performance, and as the title infers, was able to gain an unfair advantage in how much to pay for such performance. As a result, the Oakland A’s have regularly achieved a much lower cost per win then even the baseball powerhouses like the New York Yankees (see Figure 2).

Figure 2: Finding better predictors of performance

But as any of us who have been in sales knows, once you understand the metrics against which you will be measured or paid, you learn to “game” the system. And this has happened in the world of baseball as well. For example, one of the measures of fielding prowess (that can lead to a more lucrative paying contract) is fielding percentage. A fielder can actually improve his fielding percentage by not pursuing balls that are outside of his normal fielding range, since that could result in additional fielding errors. That, in turn, can result in a higher fielding percentage and likely a higher-paying contract. While this is good for the player’s personal stats, it is not good for the team (see Figure 3).

Figure 3: Gaming the system

Well, Big Data to the rescue! Most baseball stadiums are now equipped with video monitors that capture each and every movement of all the players on the playing field, at bat and on the base paths. New metrics are being developed that are better predictors of how effective given players are at fielding their position. One such metric is the Effective Fielding Range (see Figure 4).

Figure 4: Effective fielding range

The effective fielding range measures how effective the fielder is in covering his position and how much of the playing field the fielder can effectively cover. This rewards fielders who are not only making the “easy” plays, but also acknowledges their efforts to extend their fielding range in order to create more outs. This metric is not only a better predictor of the fielder’s fielding prowess, but is also a better measure of the fielder’s contribution to the overall team defense.

Like professional baseball, organizations in all industries need to constantly look for ways to exploit new sources of customer data, coupled with advanced data management and deep analytics, to uncover new metrics that are better predictors of performance and customer behaviors. Don’t be satisfied that your current metrics are the best you can do. Think Moneyball and seek those metrics that position you to exploit “the art of winning an unfair game.”

Summary

We don’t need to wait for new technologies to begin gaining insight and formulating recommendations needed to improve business performance and customer satisfaction. Existing sources of data on our customers—their behaviors, tendencies, propensities, interests and associations—can be mined today to uncover new insight that can be used to drive better performance and give us an “unfair” advantage over competitors (again, think Moneyball).