Cohort Analysis is an analytical method that groups customers based on a shared characteristic — typically the month or week they made their first purchase — and tracks the behaviour of those groups over time. Instead of looking at the average of all customers, it shows how each “generation” of customers evolves in retention, purchase frequency and Lifetime Value (LTV).
In eCommerce, the most common cohorts are:
- Acquisition cohorts (customers who made their first purchase in a given month)
- Channel cohorts (customers from Google Ads, Meta Ads, email, organic, etc.)
- Product or first-purchase category cohorts
This analysis answers essential questions such as:
- Do customers acquired through discount campaigns return as often as those who bought at full price?
- Is retention improving or worsening month after month?
- What is the real LTV of each cohort and in which month do most customers stop buying?
- Which channels bring customers with higher long-term value?
Tools such as Google Analytics 4, CRM platforms, or specialised eCommerce solutions (Klaviyo, Triple Whale, Polar Analytics, etc.) enable these analyses. The real value appears when cohort analysis is combined with contribution margin and customer acquisition cost (CAC), to understand whether growth is profitable or merely volumetric.
A well-executed Cohort Analysis does not replace other metrics, but it prevents decisions based on misleading averages. Recent customers and long-standing customers behave differently; ignoring that difference can lead to incorrect interpretations of retention and profitability.