Cohort analysis tracks groups defined by a common starting point—signup date, first purchase, or campaign—to see how metrics evolve for each vintage.

How it works

Define the cohort event and grain (weekly or monthly buckets). For each bucket, measure retention, revenue per user, activation rate, or payback at periods 0, 1, 3, 6, 12. Plot curves or heatmaps: rows are cohorts, columns are periods since start. Compare whether March signups retain better than August signups after a product launch. Consumer apps cohort by install week; B2B SaaS by contract start or go-live. Cohort analysis pairs naturally with churn cohort views and LTV modeling. Clean data requires stable user IDs and excluding test accounts. Seasonality can distort short cohorts; investors look for several months of vintages before drawing conclusions.

Why it matters

• Founders: Spot whether marketing channels deliver lasting customers or one-time buyers. Prioritize product work when newer cohorts underperform older ones at the same age.
• Investors: Cohort charts are standard in Series A+ diligence. Flat or improving retention cohorts support efficient growth claims; decaying cohorts trigger harder questions on CAC payback.
• Operators: Customer success can target interventions by cohort age—onboarding fixes help month-0 to month-3 curves most.

Common mistake

Cohorting on the wrong anchor event—using invoice date when usage starts at implementation—produces misleading retention and revenue timing.

Related ideas

Churn cohort, cohort retention, LTV/CAC, activation rate, and net revenue retention are standard follow-on metrics.