Cohort Analysis

Cohort analysis is an analytical method that splits users into groups by a shared starting attribute and tracks them over time. Sign up month is a common cohort criterion, but first channel, first product or acquisition campaign are also valid. The method reveals behavioural differences that vanish inside aggregate views.

What the method reveals

Instead of viewing the total of all users, you see how different generations evolve. A cohort acquired in January might behave very differently from one acquired in July because product, onboarding or channel mix changed. Classical applications are retention curves, repurchase rates or the evolution of customer lifetime value across several months.

How a cohort table is structured

Rows stand for the individual cohorts, columns for the time elapsed since their start. Each cell holds the chosen metric such as active users, revenue or conversion rate. This makes it easy to see at a glance whether new cohorts outperform or underperform older ones and at which point the largest drop offs typically occur.

Practice note

Cohort analysis only works with cleanly structured data and a clearly defined start event. Statements for small cohorts suffer from random noise, so a minimum size is needed for defensible comparisons. Combined with segmentation by channel or product, the method often delivers the most honest view of the real effects of marketing moves.