Cohort Analysis for SaaS: A Practical Guide

What Is Cohort Analysis?

Cohort analysis groups customers by a shared characteristic — usually signup month — and tracks how each group behaves over time. Instead of asking "what is our churn rate?" (a blended average that hides everything), cohort analysis asks "how does the January cohort's retention compare to the March cohort's?" That comparison reveals whether your product, onboarding, and pricing changes are actually working.

A blended 5% monthly churn could mean every cohort loses 5% monthly (steady decay) or that new cohorts churn at 10% while old cohorts churn at 1% (an onboarding problem). Only cohorts tell you which. If you are new to the underlying metrics, review What Is MRR and the Churn Rate Calculator first.

Building Your First Cohort Table

A classic cohort table has signup months as rows and "months since signup" as columns, with each cell showing the percentage of the cohort's original MRR (or customers) still retained:

• Step 1: assign every customer to the cohort of their signup month.
• Step 2: for each cohort, record starting MRR and starting customer count.
• Step 3: for each subsequent month, record remaining MRR and customers per cohort.
• Step 4: express each cell as a percentage of the cohort's starting value — this normalizes cohorts of different sizes.

Read the table two ways. Across a row shows one cohort's decay over its lifetime — the retention curve. Down a column compares different cohorts at the same age — "month-3 retention" for January vs. March signups. Column comparison is where insights live: if March's month-3 retention beats January's, something you changed between January and March improved retention.

Retention Curves: What Good Looks Like

Healthy SaaS retention curves share a shape: steep early decline (the first 1–3 months lose the poor-fit and never-activated customers), then flattening into a long tail of stable, loyal customers. The flatter the tail, the healthier the business.

• Revenue retention curves (MRR-based) can rise above 100% for strong cohorts when expansion outpaces churn — the hallmark of product-market fit in the installed base. See Net Revenue Retention for why this matters.
• Logo retention curves always decline; compare their tails across segments to see which customer profiles truly stick.
• Watch for curve shifts: if newer cohorts flatten faster (less early decay), onboarding improvements are working. If the tail steepens, something is eroding loyalty among mature customers.

Cohorts Beyond Signup Month

Signup cohorts are the starting point, but behavioral cohorts answer sharper questions:

• Plan cohorts: do annual-plan customers retain better than monthly? (They almost always do.)
• Channel cohorts: do customers from organic search retain better than paid-ad customers? This reveals true CAC quality.
• Feature-adoption cohorts: do customers who use a key feature within week one retain better? This identifies your activation metric.
• Segment cohorts: compare retention by company size, industry, or geography to focus your ideal customer profile.

Each of these turns a vague "improve retention" goal into a specific, testable hypothesis — which is the entire point of the exercise.

Using Cohorts to Forecast

Once you have 6–12 months of cohort data, the stabilized retention curves become your best forecasting input: apply the mature curve shape to young cohorts to predict their future revenue. This is the core of cohort-based MRR forecasting — it beats flat churn-rate math because it models the real decay shape (fast early, slow later) instead of assuming constant loss. Rebuild the curves quarterly; product and market changes gradually reshape them.

Conclusion

Cohort analysis replaces blended averages with truth: which customers stick, when they leave, and whether your changes are working. Build the table, compare columns, find your activation metric, and forecast from real curves. It is the single highest-leverage analytical habit a SaaS team can build — and it feeds directly into better MRR projections and smarter pricing decisions.