Cohort Analysis Explainer
Cohort analysis groups customers or users by a shared starting event and compares their behavior over time. A common cohort is the week or month of first purchase or sign-up. Each group is then tracked by age, such as week one, month three, or year one.
This structure separates lifecycle effects from calendar effects. A total active-user figure can rise because acquisition is strong even while newer customers retain poorly. A cohort table can reveal whether onboarding, product changes, pricing, or channel mix improved later groups.
Cohorts may also be defined by plan, region, campaign, device, or product, but comparisons require enough observations and consistent definitions. Attribution, delayed conversion, seasonality, survivorship, identity matching, and changes in tracking can create misleading patterns.
Use cohort analysis to ask a specific question, then connect retention or repeat behavior with revenue, margin, service cost, and customer experience. Do not treat correlation with a launch or campaign as proof of cause. Preserve privacy, minimize unnecessary data, and document the event and time logic.
Acquisition date, first purchase, or another event creates comparable lifecycle groups.
Define the event and time zone consistently.
Track each cohort by time since entry, not only overall totals.
This can reveal retention changes hidden by rapid acquisition.
Channel mix, seasonality, tracking changes, and small samples affect results.
Use experiments or deeper analysis before claiming causation.
Explore more "Explainers"
Discover additional explainers across politics, science, business, technology, and other fields. Each explainer breaks down a complex idea into clear, everyday language—helping you better understand how major concepts, systems, and debates shape the world around us.