A few weeks back, while the Padres were in the middle of their latest and worst losing streak, our digital marketing agency got a bunch of tickets to the San Diego premiere of
Moneyball, the big Hollywood adaptation of Michael Lewis' best-selling 2003 book about the (literally) game-changing Oakland A's General Manager Billy Beane, starring Brad Pitt, which opens
I love the book and have read it several times, not only because I'm a big baseball fan but because it once was a matter close to my wallet. Earlier in my career, I worked for Linea
Baseball Consulting, compiling and analyzing advanced statistical research, and arranging interpreter services for Major League Baseball teams. Moneyball was an extremely influential book in
my professional development, and that baseball experience ultimately became a major factor for my entry into the world of Web analytics.
At the core of Moneyball is sabermetrics, a late 20th-century term used to define the objective, empirical evidence of baseball statistics, some of them well-known, like batting average and slugging percentage, while others are quite obscure but arguably far more revealing, like VORP, for "value over replacement player"; or BABIP, for "batting average on balls in play."
In the end, I did a personal statistical analysis and discovered there was more money to be made in the world of marketing analytics. But I still love how statistics bring order and explanation to baseball, far more than any other sport. Since leaving baseball's number crunching game, I've noticed how Web analytics and sabermetrics have similar goals and methods...but there are some critical differences.
Web Analytics Isn't a Zero Sum Game
In baseball, there can only ultimately be one winner every year. In practical terms, this means any statistical advantage gained by one team is made at the expense of all other teams. When A's GM Billy Beane exploited an inefficiency in the talent market through statistical analysis, he improved his team. But it came at the expense of the Yankees, Dodgers, Padres and every other franchise in the league.
In Web analytics, the goal is also optimization, but there can be multiple winners, even within the same verticals. If every company was diligent about analyzing and optimizing their digital properties, I guarantee that nearly all of them would increase goal conversions, customer satisfaction and repeat business. The only losers in the digital world are people who continually make mistakes and use foolish strategies because they don't use the mountain of data available to them!
Baseball Is A World Of Finite Possibilities
Another important distinction is that baseball has a limited set of outcomes. A single pitch can be a strike, a ball, an out or a hit. A game is either won or lost. Predicative analytics in baseball are easier -- not easy, but easier -- because, for better or worse, the game doesn't change much. It is possible to accurately determine which statistics translate well from year to year. The only factor that can't be accounted for is luck, but even taking that chaotic element into consideration, sabermetrics allows you to make smart baseball decisions using data.
On the Web, every site, tool, and app is different...and every person experiences them differently. Recent history has repeatedly proven that the Web can be unpredictable (Rebecca Black would stymie even the Billy Beane of the music biz, no doubt), which means that it is ever more important for companies to make use of skilled analysts, who understand that you must always be looking to optimize based on both qualitative and quantitative data.
Unlike baseball, it is difficult to write a mathematical formula to predict the performance of a new site, or
a new social media campaign; but it is possible to use data in a variety of other ways to make smart decisions and give yourself the greatest chance at success.
The Web Is Constantly Changing, Baseball Is Forever
As I pointed out, baseball is fundamentally the same game that has been played since the late 1800's. There are very few things that could be done to the sport to alter it in any statistically significant way. The game has gone through eras; the deadball days of the early 20th century; the pitcher's paradise of the '60s when Koufax, Spahn, and Gibson terrorized hitters on 18-inch mounds; the '90s when steroids fueled Bonds, Sosa and McGwire into the history* books. Still, through sabermetrics, statistics can be normalized and the fluctuations of any given period can be filtered out.
On the Internet, one great idea or product can change the game.
We have seen it so many times that it is hard to quantify:
with Google, Adobe, WebTrends, IBM and many others constantly releasing new features to try and keep up with the pace of innovation, but companies must still effectively deploy the brightest minds
available to make sense of this ever-growing mountain of data.
While sabermetricians continue to refine their methods and formulas to make them more accurate, the modern web analyst must be constantly evolving as the Web evolves. With no end in sight, that appears to be an enormous and exciting task for us to handle.