Foundation · 2 min
What you need to know
What is ecommerce cohort analysis?
Ecommerce cohort analysis groups customers by a shared starting event, such as first-purchase month, and compares their behavior at equal ages.
How can cohorts measure customer retention?
Cohorts show what share of customers return after the same amount of elapsed time from a shared start, such as first purchase. This separates a newer group’s limited observation time from a genuine retention difference.
What is customer churn in ecommerce?
Customer churn is the loss or inactivity of customers who were expected to keep purchasing. Ecommerce needs a product-aware inactivity definition.
How should ecommerce churn be measured?
Define churn as passing a product-aware inactivity threshold without a qualifying purchase or activity. Report the cohort, expected buying cycle and observation period because ecommerce rarely has a formal cancellation event.
Why are newer customer cohorts retaining worse?
Newer cohorts may come from different channels, promotions, products or seasons, or they may have worse delivery and product experiences. First confirm that cohorts are being compared at the same age and with the same metric definition.
How can retention be compared by acquisition channel?
Group customers by the channel credited for acquisition, then compare equal-age repeat behavior and contribution. Keep attribution rules consistent and separate channel quality from changes in product or promotion mix.
How can retention be compared by first product purchased?
Group customers by their first purchased product or product family and measure repeat behavior at equal cohort ages. This reveals whether certain entry products lead naturally to replenishment, accessories or no further need.
A quick store check
Create monthly first-purchase cohorts and compare repeat behavior at equal ages. Then segment by channel and first product.
Foundation · 2 min
Frequently asked questions
How long should a cohort be observed?
Long enough to cover meaningful repeat-purchase opportunities for the product, while comparing every cohort at the same elapsed age.
Can customers move between cohorts?
For acquisition analysis, keep the original starting cohort fixed; create separate behavioral segments for later changes.