Glossary: Cohort Retention Curve

What is a Cohort Retention Curve?

A cohort retention curve plots the percentage of a user cohort that remains active over time since joining, showing how quickly users decay and whether a stable core of retained users forms.

Cohort Retention Curve Explained

Take everyone who joined in January and check back every week: how many are still active? Plot those percentages over time and you get a curve.

The shape tells the story. Every curve falls at first, that is normal. What matters is what happens next: a healthy curve flattens, settling at a core of users who stay. An unhealthy curve keeps sliding to zero.

A flattening curve is one of the clearest signals of product-market fit. It means some real group of users found durable value, and you can grow by stacking new cohorts on top of a stable base.

What a Cohort Retention Curve Means For

Examples

1. A protocol's weekly cohorts flatten at 18% retained wallets after week four, evidence of a durable core user base.

2. A team compares the airdrop cohort's curve, decaying to near zero in three weeks, against organic cohorts that flatten at 20%.

3. Each monthly cohort's curve sits above the last, showing onboarding changes are genuinely improving retention.

4. An analyst overlays curves by acquisition channel and finds community-sourced wallets flatten at triple the rate of paid ones.

Related reading

Explore Cohort Analysis, Retention Rate, User Retention, and Wallet Segmentation.

FAQs

How do you build a cohort retention curve?

Group users by their start period, measure the share active at regular intervals afterward, and plot the percentages over time since joining.

What can a flattening retention curve indicate?

It means the measured share returning has stabilized over the observed periods. That can indicate a retained group, but it does not by itself prove product-market fit; check cohort size, activity definition, incentives, and observation length.

What can a retention curve approaching zero indicate?

Few members of the measured cohort are returning under the chosen activity rule. Check cohort size, event coverage, and usage frequency before interpreting it as product failure; incentives and acquisition mix can also affect the curve.

How do retention curves work for crypto apps?

Define the cohort by an event or start period, then measure a specified return action in each later interval. The action may be an app event or onchain interaction; state whether the unit is a person, account, or address and how incentives are treated.

How should curves be compared?

Across sequential cohorts to measure product improvement, and across channels or campaigns to measure acquisition quality, always with the same activity definition.