Meaningful Metrics: User Session Cohorts

Many startups in their quest for customers measure visits to their site or app. Great start. Then they measure how many come back again - the returning user. This is useful to a point.

A returning user can be returning on the same day, the next day or 90 days afterwards. And there lies the problem. The data is noisy.

More meaningful than the returning user is the user who has developed a repeat usage behaviour (stickiness).

A helpful way to measure stickiness is to group users by sessions. An example...

• 1st time user
• Repeat user 2-4 sessions
• Repeat user 5-10 sessions
• Repeat user 10+ sessions

These user session cohorts are much more interesting and useful numbers.

If you layer onto that the conversion rates for each group you can start to understand the impact repeat usage has on conversion. That is very useful indeed.

You can also analyse these user groupings on another dimension; the month they first used the app / site, it gets to be even more useful.