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.
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.