Cohort Retention Template for Product Growth

Aggregate retention measures "of all users active last month, what percentage are active this month." It blends all cohorts together, making it impossible to see trends. Cohort retention tracks each signup group independently: "of users who signed up in January, what percentage are still active in February, March, April." Cohort analysis reveals whether your product is getting better or worse at retaining users over time. The [Product Analytics Handbook](/analytics-guide) explains when to use each.

Calendar months are simpler and align with business reporting. Rolling 30-day periods are more precise for products where signup date matters (e.g., monthly billing cycles). For most B2B SaaS products, calendar months are sufficient. If your product has usage patterns tied to billing cycles, use rolling periods.

Eight to twelve monthly cohorts gives you enough history to see trends while keeping the analysis manageable. For weekly cohorts, 12-16 weeks is sufficient. More cohorts provide more data points but make the retention table harder to read. Focus on the trend across cohorts rather than any single cohort.

For B2B SaaS, 30-40% flattening retention is good and 40%+ is excellent. This means 30-40% of users who sign up become long-term retained users. Consumer products typically flatten lower (10-20%). The absolute number matters less than the trend. A product that flattened at 25% last quarter and 30% this quarter is on a strong trajectory.

Look at where the biggest decay happens. If P0 to P1 has the steepest drop (common), the problem is likely in activation. Use the [activation funnel template](/templates/activation-funnel-template) to diagnose it. If P1 to P2 or later has unusual decay, the problem is in ongoing value delivery. Users activated successfully but did not find enough reason to return. ---