Item — HN Reader

Yeah, me too man. There's also some frustration over here about what do these analytics actually mean. Like I have so much data coming at me from everywhere now, what exactly is it telling me. One neat presentation that helps provide some focus is Dave McClure's metrics 4 pirates. http://www.slideshare.net/dmc500hats/startup-metrics-4-pirat...

I've seen this a bunch of times, but I like that focus on Activation and Retention for new startups.

Ash Maurya's Lean Startup book is interesting in that it tries to quantify when you can be ready to launch a new product: when you have 40% user retention. But personally that seems high to me if your early users are full of random joes who've signed up to some email capture form that promises rainbows but they don't really know what they signed up for.

1 comments

I agree with both of you. I've been working for an incubator for the past 2 years building more than 6 startups and tried KISSmetrics on all them. In the end we would code our own analytics into the app or keep a text file of SQL queries we'd run once a week to get some real insights/learnings.

That's what led me to create [redacted]. I don't want this to be a plug, but one idea we're playing with is Release Cohorts. Instead of just looking at new users in cohorts, I believe we should also look at existing active users and existing inactive users, and group them based on your release cycle (given the users are experiencing a common feature set).

This comes from the idea that features are released for one of three reasons: 1) Improve new user activation, 2) Continue to engage existing users and 3) Reactive existing inactive users.