The problem with averages
Aggregates blend together customers acquired at different times, prices, and behaviors. When your product or acquisition channel changes, averages can look stable while underlying cohorts deteriorate (or improve).
What is a cohort?
A cohort is a group of customers that share a start date (e.g., customers acquired in January). Cohort analysis tracks retention, expansion, and churn for that group over time.
When cohort analysis matters most
- When you change pricing, packaging, or onboarding.
- When you add a new acquisition channel or targeting strategy.
- When you're trying to forecast LTV or payback more accurately.
How to use cohorts with LTV and payback
- Compute retention/churn by cohort and segment (plan, channel, geo).
- Estimate LTV from retention curves rather than a single churn number.
- Compare payback by cohort to see whether acquisition quality is changing.