Why cohort analysis matters
Aggregate numbers hide the story: total revenue can grow while every new customer group behaves worse than the last. Cohorts separate the effects, showing whether retention is improving, which acquisition periods brought keepers, and how changes actually landed.
What cohorts reveal
- Retention curves: what share of a cohort returns by month
- Payback: when a cohort’s spend covers its acquisition cost
- Quality by source and season of acquisition
- Before-and-after effects of pricing or product changes
Questions
What’s the difference between cohorts and segments?
A segment groups by attribute at a point in time; a cohort groups by shared start and follows the group forward. Segments answer who, cohorts answer how it’s changing.
What does a healthy cohort curve look like?
A drop after the first purchase that flattens into a stable returning core, with newer cohorts flattening higher than older ones. Curves sliding toward zero mean the business is renting customers.