Skip to content
Book a demo
Back to glossary

Retention

Cohort analysis

Grouping customers by when they started and tracking each group over time, so retention and expansion can be read without new customers hiding them.

Cohort analysis groups customers by the period they joined and follows each group forward. The January cohort is measured at month 1, month 2, month 3, and so on, separately from the February cohort.

It exists because aggregate metrics lie during growth. A business adding customers quickly can post improving overall retention while every individual cohort is getting worse, simply because young cohorts have not had time to churn yet.

Cohorts also make product and pricing changes legible: if the cohort that started after a repricing retains better than the one before it, you have evidence rather than a theory. See retention curve for what a healthy cohort looks like over time.

Related terms

  • Retention curve The share of a cohort still active plotted against months since they started. Whether it flattens is the single most important thing about it.
  • NRR (Net Revenue Retention) Revenue this period from a cohort of existing customers, divided by what that same cohort paid a year ago, including expansion, contraction and churn.
  • Customer churn The share of customers who cancelled in a period, counted as logos rather than as revenue.