FA-301a · Module 1
The Retention Matrix
3 min read
The retention matrix — sometimes called a cohort triangle — is the single most powerful analytical tool in SaaS finance. It arranges cohorts by row and retention periods by column, creating a triangular grid where each cell shows the percentage of original revenue retained at that age. Reading down a column shows how different cohorts perform at the same age. Reading across a row shows how a single cohort degrades over time. Both directions tell a story.
Gross Revenue Retention by Cohort (% of Month-1 MRR):
────────────────────────────────────────────────────────
Cohort Mo 1 Mo 3 Mo 6 Mo 9 Mo 12 Mo 18
────────────────────────────────────────────────────────
Q1-25 100% 97% 93% 90% 87% 81%
Q2-25 100% 96% 91% 87% 83% —
Q3-25 100% 94% 88% 83% — —
Q4-25 100% 92% 85% — — —
Q1-26 100% 90% — — — —
────────────────────────────────────────────────────────
Read down the Mo 3 column: 97→96→94→92→90
Each successive cohort retains worse at the
same age. This is cohort degradation — the
most important signal in the matrix.
Do This
- Build retention matrices for both revenue and logo (customer count) retention
- Read down columns to detect cohort degradation trends
- Flag any column where the trend is consistently downward across 3+ cohorts
Avoid This
- Report only the blended retention number that averages all cohorts together
- Build the matrix once and never update it — retention behavior evolves
- Ignore early-stage signals because "the sample size is small" — act on the trend, validate later