Feb 3 2011

Don’t be average

I’ve written before about what I perceive to be an emerging, key trend among many new tech startups – the vast volumes of data companies now produce and the importance of having someone in your organization whose job it is to sift through these stacks of data and look for trends and patterns (I’ve even suggested to a few college students interested in startups and entrepreneurship that they make sure they’re taking plenty of math and stats classes as I see this as a great way for a young person with limited experience to pitch themselves to be quickly impactfull working with a start-up).

One word of caution about these data that I’ve been digging into a lot recently is the risk of using averages to represent what’s happening in your business. As the complexity of the trends businesses analyze increase, and as the volume of data produced increase along with this complexity, the value of looking at straight averages decreases. It’s a real danger in many cases to look at an average and think you know anything about what’s taking place inside a complex system (indeed, at one of the companies I work with we’ve instituted a “no averages” rule to force ourselves to do at least the next level of analysis of whatever it is we’re looking at).

I’m a huge fan of businesses collecting and analyzing the data they produce. And as the data get more complex, the need for more thoughtful analysis increases. So keep taking those stats classes…