Retention Curve

What Is a Retention Curve?

A retention curve plots the percentage of a cohort (users who signed up in the same period) who remain active at each subsequent time interval. Some curves decay and then flatten into a plateau; others keep declining over the available observation window. The shape is evidence about repeat behavior under the chosen activity definition, not a standalone verdict on product-market fit.

Also Known As

• Product teams: cohort retention chart
• Growth teams: N-day retention, survival curve
• Analytics teams: stickiness curve
• Finance teams: customer decay chart

How It Works

Illustrative example: A SaaS product plots weekly retention for its January cohort: Week 1 = 100%, Week 2 = 42%, Week 4 = 28%, Week 8 = 22%, Week 12 = 21%, Week 16 = 21%. Within that observation window, the curve plateaus near 21%. A later cohort plateaus near 27% after an onboarding change. The six-point difference is worth investigating, but a cohort comparison alone does not isolate the onboarding change from acquisition mix, seasonality, or other product changes.

Best Practices

• Do use cohort-based retention, not cross-sectional "retained users this month." Cross-sectional hides the leak.
• Do examine whether and where the curve flattens. Interpret the shape alongside use frequency, segment mix, and the length of the observation window.
• Do compare curves across cohorts to measure whether product changes actually moved retention.
• Don't confuse retention with engagement. A user can "retain" by logging in once a month and still not get value.
• Don't average retention across all cohorts. Average the per-period retention rate, not the absolute counts.

Common Mistakes

• Using a rolling 30-day active user count as a retention proxy. That conflates acquisition with retention.
• Ignoring early retention. Early behavior can be diagnostic, but the relevant window depends on the product's natural usage cycle.

Industry Context

Retention levels are not comparable without matching the activity definition, usage cadence, acquisition source, segment, and observation window. Consumer products may use daily or monthly activity, ecommerce teams often use repeat purchase, and B2B enterprise teams may focus on renewal or account-level usage.

The Behavioral Science Connection

Habit formation can contribute to retention in frequently used products, but retention can also reflect utility, switching costs, contracts, or workflow integration. A plateau may be consistent with a stable-use cohort; it does not by itself reveal the psychological mechanism behind the behavior.

Key Takeaway

Review retention before scaling acquisition, and diagnose the curve in the context of your business model. A declining curve is a prompt for investigation, not an automatic stop-spend rule.