Cohort Analysis

Cohort analysis groups customers by when they started, usually their signup month, and follows each group over time. A single average mixes new and old customers. A cohort view shows how each group behaves month by month.

How it works

1. Group customers by the month they joined. Each group is a cohort.
2. Track one value for each cohort over the following months, such as active customers, retained MRR, or revenue per customer.
3. Compare cohorts to see whether newer ones do better or worse than older ones.

A retention cohort might show that of the customers who joined in January, 90% are still active after one month, 80% after three, and 70% after six.

Why it matters

When you grow fast, new signups dominate the average. They haven't had time to cancel, so they can hide heavy churn in older groups. Cohort analysis separates the two and answers questions like these:

• Do newer cohorts retain better than older ones?
• At what point do customers usually cancel?
• Did a product or pricing change affect churn?

Cohorts are also the basis for accurate LTV estimates and for reading net revenue retention over a customer's lifetime. A revenue cohort that grows over time means expansion is outpacing churn in that group.