What is Cohort Analysis? Definition, Examples, and Business Applications

Cohort analysis is the tool that reveals what aggregate metrics hide. When your monthly active user count is flat, it could mean retention is great and acquisition is slow — or retention is terrible and you're rapidly acquiring to compensate. Cohort analysis tells you which.

What is Cohort Analysis?

Cohort analysis groups users or customers by a shared attribute — most commonly the time period when they first became a customer (acquisition cohort) — and tracks how each group behaves over time.

For example, all customers who first purchased in January 2025 form a cohort. Cohort analysis tracks:

• What % of January 2025 customers purchased again in February?
• What % in March? April? December?

By comparing this retention curve across cohorts (Jan vs Feb vs Mar cohorts), you can see whether customer retention is improving, deteriorating, or stable over time.

Types of Cohorts

Acquisition Cohort (Most Common)

Groups users by the period they first became customers:

• "January 2025 cohort" — all customers whose first order was in January 2025
• Tracks retention, repeat purchase, and LTV for each monthly cohort

Behavioural Cohort

Groups users by an action they took:

• "Customers who used the mobile app in their first 7 days" vs those who didn't
• Shows how specific behaviours predict retention or churn

Size/Revenue Cohort

Groups customers by their first-order value or current account size:

• Enterprise cohort vs SMB cohort
• High-value cohort vs low-value cohort

What Cohort Analysis Reveals

Retention Trends

Cohort retention curves show the percentage of each cohort that remains active over time. This reveals:

• Whether retention is improving for newer cohorts vs older ones
• Which acquisition channels produce the most loyal customers
• At what time point customers typically disengage

Lifetime Value by Cohort

LTV cohort analysis calculates the total revenue generated by each cohort over time. This reveals:

• Whether newer customer LTV is higher or lower than older cohorts
• How long it takes to payback customer acquisition cost by cohort
• Which customer segments are most valuable over a full lifetime

Churn Patterns

When does each cohort start churning? At month 1? Month 3? Month 12? Knowing the typical churn inflection point tells you when to intervene with retention efforts.

Product Change Impact

After a product change, pricing change, or new onboarding flow, cohort analysis shows whether the cohorts acquired after the change behave differently from those before — isolating the causal impact.

Cohort Analysis Example: E-Commerce in India

An Indian D2C brand tracks monthly acquisition cohorts:

This shows improvement across all three cohorts — each successive cohort retains better at every stage. The brand's retention improvement initiatives are working.

Without cohort analysis, overall retention metrics might look flat if customer volume was also growing.

Cohort Analysis vs Average Metrics

The problem with averages: Average retention rate of 25% tells you nothing about whether retention is improving or declining over time, or whether early cohorts are dragging down the average.

What cohort analysis adds: By breaking performance into cohorts, you can see whether new customers behave better or worse than old ones — and whether changes you've made are having an effect.

How to Do Cohort Analysis

1. Define your cohort dimension — typically acquisition date (week or month)
2. Define your metric — retention rate, purchase frequency, revenue, churn
3. Define time periods — how many months/quarters to track each cohort
4. Build the cohort table — rows are cohorts, columns are time periods since acquisition
5. Add retention rate heatmap — colour code by performance to spot patterns visually
6. Compare cohorts over time — are newer cohorts retaining better?

For companies with Tally and CRM data, AI analytics platforms can automate cohort analysis by connecting customer acquisition dates to purchase history.

See customer lifetime value (LTV) for understanding how cohort analysis feeds LTV calculation.