Probably the most fundamental shift in the business of baseball over the past 10 years has been the rise of “sabermetrics,” as documented in Michael Lewis’ Moneyball. The theory is that in an overcrowded market where many teams are at an inherent disadvantage due to inferior budgets of the main players, smaller-market teams look to gain an edge by using advanced statistical measures. The rise of sabermetrics helped the underfunded Oakland A’s team be a perennial division champion in the decade’s first half, and led the Boston Red Sox to win two World Series after deploying sabermetrics with a larger budget.
Fundamental to sabermetrics is the shifting of priorities in what was considered a primary baseball statistic. For decades, the mark of a good baseball hitter was his batting average (BA), or hits per at bat. A lifetime batting average over .300 could lead a hitter to the Hall of Fame, whereas comparable hitters with lower batting averages could barely make an All-Star team. Starting with the A’s, the shift turned from batting average (hits per at bat) to on-base percentage (OBP), or (hits + walks + times hit by pitch) per plate appearance. While on base percentage had been a statistic for almost as long as batting average, the shift in priorities was much more drastic: the goal of a hitter was no longer not to get a hit every time at bat, but not to get out. With only 27 outs per team in a game, getting on base proved to be a lot more important than just getting a hit when you plugged in the numbers. Hence, players like Juan Pierre (career BA: .301, career OBP: .348) came to be seen as lot less valuable than players like Kevin Youkilis (career BA: .292, career OBP: .391).
I feel like a similar shift needs to happen when it comes to monitoring web analytics. For the past 10 years, most businesses have remained obsessed with raw traffic. This is a holdover from the ’90s, when the loosely defined “hit” of a web page could mean the difference between a multi-million dollar startup and a non-entity for venture capitalists. Those statistics have changed to “visits” and “page views,” but the value of these metrics is no less vague and short-sighted. Furthermore, some major web players have an inherent advantage in traffic that newer web businesses are often too intimidated to try to match.
But if you were to invest in a website, which of the following would you be more likely to spend money on?
Website A: 100,000 visits/month, 80% bounce rate.
Website B. 40,000 visits/month, 20% bounce rate.
Using this metric, 100,000 visits suddenly turns into only 20,000 visits where the user doesn’t immediately turn away. On the other hand, site B keeps 32,000 of its visitors on the site for an extended period of time. Despite the huge disparity in visit numbers, the number of visitors who actually spend time on the site is over 50% higher on site B, despite having less than half the raw traffic. Website B would be the better investment.
Of course, without the bounce rate available, Website A is the clearly superior site, which is why so many websites obsess with quick fixes for traffic like prominent social media links or a rapid on-page SEO overhaul. Yet, for smart monitors of web analytics, I think the time has come to value bounce rates a lot more highly than most have been doing recently. The fundamental shift here is not just to get visitors to hit your site, but to get visitors on your site who don’t leave.
