Cohort Retention

Cohort retention groups customers by the month they first paid (or first activated) and tracks what percentage of that group is still active 1, 3, 6, 12, and 24 months later. A January cohort of 100 customers showing 78 active in month 6 has 78 percent six-month retention. The reason this view matters more than blended churn is that it reveals whether the product is getting stickier over time, getting worse, or staying flat. Blended churn smears together customers acquired under different products, different pricing, and different ICP definitions, hiding the real signal. Cohort retention isolates them and lets the founder see which months actually fit.

For funded teams, cohort retention is the cleanest test of product-market fit available. The smile shape (retention rising over time as the product matures) signals genuine fit. The slide shape (retention dropping after month 3 and never recovering) signals the product solves a one-time problem rather than an ongoing one. The flat shape (retention dropping fast then leveling at a high plateau) signals a strong core of fit customers with poor targeting at the top of the funnel. Reading the shape of the curve is more diagnostic than any single retention number, and most funded teams under 50 have never plotted their cohorts past month 6 because the data work is genuinely hard.

The relationship between cohort retention and net dollar retention is causal, not coincidental. Net dollar retention is downstream of cohort retention because expansion can only happen on customers who are still active. If month-12 cohort retention is 60 percent, the absolute ceiling on NDR from that cohort is what 60 percent of the customers expand to, no matter how strong the expansion motion is. The AI Ops Department builds the cohort retention view continuously so the founder sees the leading indicator weekly instead of discovering an NDR collapse two quarters after the underlying retention started decaying.