Cohort Analysis
Cohort analysis groups customers by the time they signed up (or another shared characteristic) and tracks their behavior over subsequent periods. Instead of blending all customers together, it reveals how each group retains, expands, or churns independently — showing whether your business is actually improving.
Why Cohort Analysis Matters for SaaS Companies
Blended metrics lie. Your overall churn rate might look stable at 4% monthly, but cohort analysis might reveal that recent cohorts are churning at 6% while older ones churn at 2%. Without cohort analysis, you are flying blind — you cannot tell if product changes, pricing adjustments, or onboarding improvements are actually working.
An Operator's Take
Every engagement I start with cohort analysis. At one company, blended NRR was 105% — looked fine. But when I built the cohort view, a clear pattern emerged: cohorts from the last 6 months had NRR of 88%, while older cohorts were at 115%. The product had not changed — what changed was the customer profile. A new marketing channel was bringing in lower-quality leads who converted but did not retain. We fixed the acquisition targeting, not the product.
Common Mistakes
What I see go wrong at Seed to Series B companies.
Only doing cohort analysis annually. Monthly cohorts reveal problems 10 months faster than annual cohorts.
Grouping cohorts only by signup month. Also analyze by acquisition channel, plan tier, and company size — these often reveal more actionable insights.
Looking at cohort charts without acting on the patterns. If recent cohorts retain worse than older ones, something changed — find out what.
What to Do This Week
Concrete steps you can take right now.
Build a cohort retention table: rows are signup months, columns are months since signup, cells are retention percentages.
Compare your 3 most recent cohorts to your 3 oldest. If retention is getting worse, investigate what changed in product, pricing, or acquisition.
Segment your best-performing cohort by acquisition channel. That channel is producing your best customers — double down on it.
Related Resources
Frequently Asked Questions
How do you do cohort analysis for SaaS?
Group customers by signup month. For each cohort, track the percentage that remain active in each subsequent month. Display as a table or chart. Look for two things: is retention improving in recent cohorts (good) or degrading (bad), and is there a consistent drop-off point (indicating an onboarding or value-realization problem).
What tools are best for cohort analysis?
Amplitude and Mixpanel have built-in cohort features. ChartMogul shows revenue cohorts. For most Seed-B companies, a spreadsheet works well — export customer signup dates and activity status, then pivot by month. The tool matters less than actually doing the analysis regularly.
Need Help With Churn & Retention?
Most Seed to Series B companies are leaving money on the table. Let's figure out where.