Cohort analysis
TERM
What is a cohort analysis?
COHORT ANALYSIS
A cohort analysis groups customers by when they started, for example per month or quarter, and tracks per group how usage, revenue and churn develop over time.
Averages across your whole customer base hide a lot. Stable churn can mean that older customers are loyal while new customers leave in droves. A cohort analysis makes that visible, because you compare customers who started in the same period. You see whether this year's customers perform better or worse than last year's, and what effect a new onboarding process, a price change or a different ICP has had.
You build a cohort analysis as a table. Each row is a cohort, for example all customers who started in January, and each column is a month after the start. In the cells you enter the percentage of customers or revenue still active. If you track revenue, you also see net revenue retention per cohort, including upsell. We recommend splitting cohorts by segment or channel as well, so you know which customers deliver the most value over the long term.
Example
A SaaS company selling project software saw annual churn stay stable and assumed everything was fine. A cohort analysis showed that customers from the last six months cancelled twice as often after three months as earlier cohorts. The cause was a shortened onboarding after a reorganisation. The team restored the kick-off and the thirty-day check-in, and the following cohorts performed as well as before.
Frequently asked questions
Why is a cohort analysis better than average churn?
Average churn mixes old and new customers together, so improvements or problems among recent customers disappear from view. A cohort analysis compares customers who started in the same period. That lets you see straight away whether a change in sales, onboarding or product affects retention.
How do you build a cohort analysis in Excel?
Export each customer's start date and monthly revenue or status from your CRM or billing system. Group customers by start month in rows and put the months after the start in the columns. Calculate per cell the percentage still active, and use conditional formatting to make patterns visible.
Related terms
Read more
How we help
GTM consultancy for technology companies
Frequently asked questions
Why is a cohort analysis better than average churn?
Average churn mixes old and new customers together, so improvements or problems among recent customers disappear from view. A cohort analysis compares customers who started in the same period. That lets you see straight away whether a change in sales, onboarding or product affects retention.
How do you build a cohort analysis in Excel?
Export each customer's start date and monthly revenue or status from your CRM or billing system. Group customers by start month in rows and put the months after the start in the columns. Calculate per cell the percentage still active, and use conditional formatting to make patterns visible.