SaaS Cohort Analysis Guide for Founders
Cohort analysis reveals what aggregate metrics hide. Three types of cohort analysis (retention, revenue, behavioural), the Bubble.io implementation that stores longitudinal data from day one, and how to read a cohort matrix.
Understanding Customer Behaviour Over Time
SaaS cohort analysis is a method of analysing customer groups (cohorts) that share a common characteristic — typically their start date (month or quarter of first subscription) — to understand how their behaviour changes over time. By tracking cohorts separately, SaaS founders can answer questions that aggregate metrics obscure: Are newer customers retaining better or worse than older customers? Is a specific acquisition channel producing better long-term retention? Did a specific product change improve or harm retention for customers who experienced it? Cohort analysis is the analytical tool that connects product decisions to long-term commercial outcomes.
Aggregate retention metrics hide critical information. A SaaS reporting 5 percent monthly churn might be improving (newer cohorts churn at 3 percent) or declining (newer cohorts churn at 8 percent) — and the aggregate metric tells you nothing. Cohort analysis reveals the trend inside the aggregate and is the analytical practice that distinguishes data-driven SaaS founders from founders who manage by feel.
🔗 Related reading on Simple Automation Solutions
SaaS Metrics That Actually Matter
The complete SaaS metrics framework including cohort tracking implementation.
What Each Reveals
Retention cohort analysis
Groups customers by start month and tracks what percentage remain active at 1, 3, 6, 12, and 24 months after subscription start. Reveals: whether retention is improving or declining over time, the natural retention curve shape for your product (does it flatten at a stable percentage?), and the lifetime value trajectory for each acquisition channel.
Revenue cohort analysis
Groups customers by start month and tracks the total MRR generated by each cohort over time, including expansion revenue. Reveals: whether cohorts grow in revenue over time (indicating strong expansion and NRR), which acquisition channels produce the highest lifetime revenue per customer, and the true payback period for customer acquisition cost by channel.
Behavioural cohort analysis
Groups customers by a specific action they took (activated within 7 days, used Feature X, invited a team member) and compares the retention and expansion of action-takers versus non-takers. Reveals: which specific product actions predict long-term retention (the activation event), which features drive expansion revenue, and where in the product journey customers who churn diverge from customers who stay.
The Practical Implementation
Cohort analysis requires storing a cohort identifier on the Workspace record at the time of creation. SA’s standard implementation: a cohort_month field (text, format ‘YYYY-MM’) set on Workspace creation to the current year and month. This enables querying all workspaces by cohort month and tracking their status at subsequent intervals.
A monthly MonthlyWorkspaceSnapshot data type stores: workspace, cohort_month, snapshot_month, is_active, current_mrr, features_used_count, team_member_count. A scheduled backend workflow runs on the first of each month and creates a snapshot for every active workspace. This creates the longitudinal dataset that enables all three cohort analysis types.
Cohort visualisation: a matrix with cohort months as rows and months-since-start (0, 1, 2, 3…) as columns. Each cell shows the retention rate for that cohort at that point in time. A cohort matrix where the numbers in each column are consistently higher than the column to the left indicates improving retention over time. A cohort matrix where numbers are declining indicates worsening retention.
🔗 Related reading on Simple Automation Solutions
Bubble.io Scalability — Can Your No-Code App Handle Real Growth?
How to architect Bubble.io for the data volume that cohort analysis generates at scale.
Free SaaS Tech Audit — 30 Minutes, No Cost
Athar Ahmad personally reviews your SaaS product: security vulnerabilities, billing architecture gaps, and performance anti-patterns identified before they cost you customers, deals, or investor confidence.
- Multi-tenant security and privacy rule assessment
- Stripe billing architecture review
- Performance bottleneck identification
- Written remediation roadmap within 24 hours
Q: How many cohorts do I need for meaningful cohort analysis?
Minimum 6 monthly cohorts to see any trend. 12 cohorts to identify seasonal patterns. 24 cohorts to understand long-term retention curves. Most SaaS products should start building cohort data from their first customer, even if the analysis only becomes meaningful 6-12 months into the product’s life.
Q: What does a healthy SaaS cohort retention curve look like?
A retention curve that drops steeply in months 1-3 (normal churn as non-ideal customers self-select out) and then flattens at a stable percentage from month 4 onwards (the retained core). The level at which the curve flattens is your product-market fit signal: a curve that flattens at 60-70 percent indicates strong PMF. A curve that continues declining without flattening indicates a fundamental retention problem.
Q: How do I use cohort analysis to improve retention?
Identify the cohorts with the best retention curves and compare them to cohorts with the worst. What was different about the high-retention cohorts? Were they acquired from a different channel? Did they have a higher activation rate? Did they use a specific feature more? Each difference is a hypothesis to test: can you replicate the conditions of the high-retention cohort for all new customers?
Build or Fix Your SaaS the Right Way
Free Tech Audit for SaaS products that need assessment. Discovery Sprint to scope new SaaS ideas correctly before building. Both lead to better commercial outcomes.
