What it means
A store gains new customers every month, and looking only at total sales mixes recent buyers with long-time repeat buyers. Cohort revenue separates those groups to show how each one spends after joining.
Stripe describes cohort analysis as grouping customers who share a characteristic, such as sign-up month or first purchase period, then following their activity over time. Define the cohort before calculating.
January first-time buyers can form one row and February first-time buyers another, and a customer who buys again in April stays in the January row rather than becoming a new April customer. Use relative age as well: month zero is the acquisition month and month one is the next, so comparing January's month-six revenue with June's month-one revenue says little about their relative quality.
Choose period or cumulative revenue and label the table clearly. Period revenue shows what a cohort spent during each month after joining, while cumulative revenue adds the amounts from month zero through a given age.
Keep the original cohort size as the denominator for cumulative revenue per acquired customer: if 2,000 people joined and their cumulative eligible revenue reaches $600,000 by month six, it is $300 per original customer, whereas dividing by only the surviving active customers answers a different question. Use a consistent revenue basis, because gross orders, net sales after returns and recognised subscription revenue may differ.
Refunds and discounts can change the shape of the curve, and a new cohort may look strong in month zero because it received a large first-order discount while its later purchases are weak. Cohort revenue is also not lifetime value by itself, since a fully observed lifetime may span years while a cohort table often covers only the months available.
Use retention alongside spend. A cohort can earn more because more customers return, because active customers spend more, or both, so separate the retained-customer count and average spend per active customer to explain the result.
Compare contribution when costs vary, since a high-spending cohort could be expensive to acquire or serve, and beware of calendar seasonality when a holiday cohort is compared with a quiet-month cohort. A table with rows by acquisition month and columns by age makes missing future periods visible.
Do not fill future months with zero and call that poor retention, because those cells have not happened yet. For an owner, cohort revenue asks whether newly acquired customers become durable customers, but the right action still depends on margin, acquisition cost and why the pattern changed.
In practice
Real-world examples.
Example
A January cohort of 2,000 first-time buyers generates $600,000 in cumulative net revenue through month six. That is $300 per original buyer at that age.
Example
A July cohort's month-one revenue is compared with prior cohorts at month one, not with those older cohorts at month twelve. The comparison holds age constant so that newer groups are judged fairly.
Example
A promotion lifts month-zero cohort revenue but repeat revenue in months two and three is weak. The team reviews contribution rather than judging only first orders.
Formula
Calculation
Cumulative cohort revenue per original customer at age m = eligible revenue from that cohort from entry through month m / number of customers originally in the cohort. Example: $600,000 / 2,000 = $300 at month six. Period revenue for month six is a different measure.
To see the difference, suppose month six alone brought in $40,000 from the same cohort. Period revenue per original customer for that month is $40,000 / 2,000 = $20, while the cumulative figure of $300 includes every month from zero to six.Case study
Seen in the real world.
This entirely fictional case follows Canal Boxes, an invented subscription retailer. Total sales rose for three quarters, but a cohort table showed newer first-purchase groups spending less after their first month. Managers had initially praised a steep welcome offer.
The team separated acquisition channels, checked refunds and compared contribution at the same cohort age. It revised the offer and waited for later cohorts to mature before claiming an improvement. The company and outcome are invented.
Watch out
Common mistakes.
- Comparing a new cohort's month-one revenue with an older cohort's month-twelve total.
- Dividing cumulative revenue by only active survivors while labelling it per original customer.
- Treating future unobserved months as zero revenue or assuming a pattern proves causation.
Questions
People also ask.
How is a cohort usually defined?
Often by first purchase or first billing period, but the shared starting event must be stated.
Is cohort revenue cumulative?
It can be period-by-period or cumulative; label the measure and compare equal cohort ages.
Is cohort revenue the same as customer lifetime value?
Not necessarily. A cohort table records observed revenue; a lifetime estimate may extend beyond observed periods.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%