What Is Cohort Analysis?
Cohort analysis is an analytical method that groups users into segments — called cohorts — based on a shared characteristic or experience, then tracks how those groups behave over time. The most common cohort is acquisition month: all users who signed up in January form one cohort, February signups form another, and so on.
By tracking cohorts longitudinally rather than looking at aggregate user behavior at a single point in time, cohort analysis reveals patterns that aggregate metrics hide. It answers questions like: Are customers acquired through paid ads retaining at the same rate as those acquired through organic search? Did a product change in March improve six-month retention for users who signed up afterward? Is customer lifetime value improving or declining across consecutive acquisition cohorts?
The technique originated in medical research to track patient outcomes across groups with similar risk profiles. It entered the product and marketing lexicon through the SaaS growth movement of the 2010s, where retention and LTV became the defining metrics of business health.
Why Cohort Analysis Matters for Marketers
Aggregate retention metrics are often misleading. A company showing 85% monthly retention on aggregate may be masking the fact that cohorts acquired two years ago retain at 95% while new cohorts retain at 70% — a deterioration that predicts serious future revenue problems. Cohort analysis surfaces this before aggregate metrics do.
For growth marketers, cohort analysis answers the question of whether acquisition quality is improving or declining over time. If cohorts acquired during a high-spend period churn faster than those acquired organically, that is a signal about channel quality, not just cost efficiency.
LTV calculations are meaningless without cohort data. A customer's true value can only be estimated by watching what similar customers did over time — how long they retained, whether they expanded, and when they churned. Cohort analysis provides the historical foundation for accurate LTV modeling, which in turn drives sustainable CAC targets.
How to Implement Cohort Analysis
Define your cohort dimensions: what shared characteristic groups users, and what time period do you observe? Common cohort types include acquisition date cohort, channel cohort (users acquired via a specific source), product tier cohort, and geographic cohort.
Define your behavioral metrics: retention rate (what percentage of cohort users are still active after N days/months), revenue per user over time, feature adoption rates, and refund or churn rates.
Build cohort tables in your analytics platform. Google Analytics 4 has a native Cohort Exploration feature. Product analytics platforms like Mixpanel and Amplitude provide more flexible cohort builders with custom event definitions. For e-commerce, tools like Triple Whale and Klaviyo include built-in cohort LTV views.
Export raw event data to a data warehouse (BigQuery, Snowflake, Redshift) for custom cohort analysis if your platform's native tooling is limiting. SQL-based cohort tables give full control over cohort definition and metric calculation.
How to Measure Cohort Analysis
The standard output is a cohort retention table: rows representing cohorts (January, February, March…) and columns representing time periods after acquisition (Day 1, Day 7, Day 30, Month 3, Month 6). Each cell shows the percentage of the cohort still active at that interval.
Healthy patterns show retention curves that flatten over time rather than declining to zero — indicating a loyal core user base. Deteriorating cohorts (where curves decline faster than earlier cohorts) signal worsening product-market fit or declining acquisition quality.
Track cohort LTV curves: cumulative revenue per user over time for each cohort. When LTV curves for newer cohorts plateau below those of older cohorts, you are acquiring lower-value customers — a critical strategic warning.
Cohort Analysis and AI Search
AI search platforms are creating new cohort dynamics worth tracking. Customers acquired through AI-generated recommendations — those who typed a query into ChatGPT or Perplexity and followed through to a brand — often behave differently from those acquired through paid ads. Early evidence suggests AI-referred customers exhibit higher purchase intent and different retention profiles. Creating a cohort segment for users who arrived via brand search (a proxy for AI-influenced discovery) and comparing their LTV and retention against paid cohorts can reveal the downstream revenue impact of AI visibility investment.