In the first 5 blog posts about Analytics we covered Vanity Metrics, the Pirate Metrics framework, the criteria of a good metric, the key activity, the key metric and cohorts. The final problem we need to overcome is one we wrote about in almost all the previous blogposts: The danger of having metrics that make you feel good. The premature feeling of “It’s working, it’s working!”

The power of analytics is that it can uncover the pain points in your startup. Not to make you give up or to make you feel bad, but to show you where the problems are on your way to a working business model.

When we first started tracking Analytics at Snowciety, my previous startup that showed you where your friends are on the slopes, we tracked Activation as the amount of signups divided by downloads. And it looked pretty good!

We had an Activation rate of up to 90% in winter. Logically the rate was lower in summer when people downloaded the app, but did not visit any ski resorts. After having a moment where we felt pretty good, we started to dig. Since Snowciety was not taking off like a rocket, there had to be a pain somewhere. When we looked at the users who performed our key activity (tracking their day on the slopes), another pattern emerged:

Only around 30% of all downloads resulted in a truly activated user. It hurt. A lot.. But it did show us the pain in our business model and an opportunity to fix it.

Another great example is by Andreas Klinger. Currently working for Producthunt, he previously had a startup called Lookk, a fashion discovery and shopping site. Their Activation was 90% and not just looking at signups, but based on one of their key activities. 90% of the new users voted on a designer. But.. there is always a but, after Andreas started digging deeper he found that most users only voted for friends instead of actually using the platform. So he changed the Activation moment to having voted for at least 2 different designers. Boom. PAIN. And again an opportunity to fix it.

Remember, pain is good. It is better to know what you don’t know and does not work, than to not know what you don’t know. Learn, Improve, Repeat. Iterate experiment by experiment towards a working business model.