Cohort retention measures how many members of a start-time group stay active, paying, or engaged at each subsequent interval.

How it works

Build cohorts by signup or first-payment month. For the January cohort, count 100% at month 0. At month 1, if 80 of 100 remain, retention is 80%; at month 6, if 55 remain, retention is 55%. Plot each cohort’s curve to see flattening (good) versus steady decline (bad). Definitions of “retained” vary: logged in, paid invoice, or used core feature. B2B may track logo retention (accounts) and dollar retention (ARR) separately. Mobile apps often report D1, D7, D30 retention by install cohort. Investors compare cohorts at the same age—month-6 for June versus month-6 for January—rather than different calendar months with unequal maturity.

Why it matters

• Founders: Retention curves tell you if onboarding and core value work. Improvements should show up as newer cohorts retaining better at the same lag.
• Investors: Retention is a leading indicator for LTV and expansion. Flattening curves near target benchmarks (varies by category) support scaling spend.
• Operators: Success and product teams set goals by cohort age—fix month-1 drop-off before chasing top-of-funnel volume.

Common mistake

Reporting retention on calendar months instead of cohort age. “March retention” without specifying months-since-start mixes immature and mature cohorts.

Related ideas

Cohort analysis, churn cohort, churn rate, gross retention, and net revenue retention complete the retention picture.