Photo by EranB
I was lucky enough to hear Scott Petry, Brad Feld and Ryan McIntyre telling the story of Postini, the company Scott founded, and Brad and Ryan funded. One phrase from Scott really resonated:
Data is necessary, but not sufficient, to make good decisions
I love and live by metrics, and I'm a big fan of the Customer Development philosophy, but I think there's dangers in taking it too far, and Scott crystallized the problem. Customer Development rocks because it forces startups to gather data, something most would otherwise neglect. It sucks if it prevents you from following paths that cause your metrics to dip in the short-term in return for a long-term win.
One fascinating example of this was profitability. Scott explained how reaching profitability early became a curse, because it became immensely important every quarter after that to avoid dipping into a loss, even when spending more money would be a long-term win. Irrational primates that we are, humans get a lot more worried by a small drop that causes the company to go from profit to loss than they do by the same drop when you're already making a loss, so the statistic becames their master not their servant.
I asked the guys to talk more about the metrics they used, and they talked about their in-depth tools for measuring customer satisfaction which helped them to an astonishing 96% customer renewal rate! But interestingly Brad and Scott both stressed the need to sometimes ignore or go against the evidence. It all reminded me of Tom Evslin's motto "nothing great has ever been accomplished without irrational exuberance" – the trick is knowing when to take that leap of faith. To do that, you have to know that you're taking a leap in the dark, which requires knowing the evidence in the first place and making a conscious decision to override it.
You should never ignore the data, but sometimes you have to listen and make a deliberate choice to go against it.