Evidence and reasoning
Cohort analysis
- Evidence and reasoning
Definition
Grouping customers or users by a shared starting point, such as the month of their first order, and comparing how each group behaves over the same periods afterwards.
For your role
Choose your role to see what this term means in your work.
Why it matters
Overall averages mix old and new customers. Cohorts show whether newer customers behave better or worse than earlier ones.
Where it is used
Retention analysis, product analytics and marketing reviews. Used by data analysts, product analysts, marketing analysts and customer success.
How to calculate it
Retention % for a cohort in period n = cohort members active in period n ÷ cohort size × 100
Worked example
Raket Parts B.V. customers grouped by the quarter of their first order, and the share ordering again in each later quarter:
First ordered Q4 2024 (229 customers): Q1 2025 97%, Q2 2025 95%, Q3 2025 97%
First ordered Q1 2025 (181 customers): Q2 2025 98%, Q3 2025 93%
Most "first orders" fall in the first two quarters because the records start in October 2024, not because those customers were new. A cohort table always needs that check.
Raket Parts B.V. is a fictional company created for Insyt.
Definitions vary. Cohorts can be defined by first order, sign-up, first payment or campaign, and "active" can mean ordered, logged in or paid. Check how your organisation calculates it.