FA-201a · Module 3

Cohort Revenue Modeling

3 min read

Aggregate retention numbers lie by averaging. A 92% gross retention rate could mean that enterprise customers retain at 97% while SMB customers retain at 78%. If your growth plan depends on scaling SMB, the aggregate number is hiding a cohort-specific crisis. Cohort-based revenue modeling segments customers by acquisition period, size, channel, or product and tracks each cohort's revenue behavior independently. It is more work. It is also the only way to model revenue honestly.

Monthly Revenue Retention by Acquisition Cohort:
─────────────────────────────────────────────────────────
         Month 1  Month 3  Month 6  Month 12  Month 24
─────────────────────────────────────────────────────────
Q1-25     100%     96%      91%      84%       72%
Q2-25     100%     94%      88%      79%        —
Q3-25     100%     92%      85%       —         —
Q4-25     100%     90%       —        —         —
─────────────────────────────────────────────────────────

Pattern: Each successive cohort retains worse
at the same age. This is a product-market fit
degradation signal — invisible in aggregate data.

Do This

  • Segment retention by cohort to detect product-market fit degradation early
  • Model revenue by cohort age — each cohort has its own retention curve
  • Weight future revenue projections by the retention profile of recent cohorts, not early ones

Avoid This

  • Use blended retention rates that average high-performing legacy cohorts with degrading new ones
  • Assume new cohorts will retain like early cohorts — product-market dynamics change
  • Ignore cohort size — a 95% retention rate in a cohort of 10 is noise, not signal