What is Retention cohort?
A retention cohort is a group of customers who first purchased in the same period, tracked over time to see how many buy again and how much they spend.
Updated
Cohort analysis is how you find out whether you are building a business or renting revenue. A blended repeat rate averages your best customers with your worst and tells you very little. Grouping by acquisition month, and ideally by acquisition channel, tells you a great deal.
The pattern worth watching for is the discount cohort. Customers acquired during a heavy sale routinely show a first-month repeat rate close to the average and then fall off a cliff, because they bought a price rather than a brand. If a large share of your customers arrived that way, a lifetime value modelled on the blended average will overstate what they are worth, and every downstream decision built on it will be wrong.
For a young brand the honest move is to stop projecting lifetimes and read the 90-day and 180-day figures you actually have. Those are measurable, they map to how quickly cash returns, and they are enough to tell you whether your current CAC is affordable.
A worked example
1,000 customers first bought in January. Their repeat behaviour is tracked month by month.
- Month 1: 180 order again (18%)
- Month 3: a further 120 (cumulative 30%)
- Month 6: a further 90 (cumulative 39%)
39% of the January cohort had reordered within six months.
Common mistakes
- Blending all customers into one average, which hides the fact that cohorts acquired on discount rarely return.
- Judging a cohort before enough time has passed for a normal repurchase cycle to complete.
- Measuring repeat rate on orders rather than customers, so a handful of loyalists make retention look broad.
Want these numbers working for your brand?
A written audit in three working days, whether or not you hire us.