Video Business Academy clients stayed two to three years on average. A retention curve shows how that happens, one intake month at a time.
A retention curve, or cohort chart, shows what share of the clients who joined in one month are still clients at each month after. It is the picture behind lifetime duration. One average hides the shape; the curve shows where clients leave.
What counts as still a client?
Paying, or inside a term they paid for, on any offer. A client who finished the six-month program and bought an extension is still a client. A client who finished and bought nothing has left, even if they can still open Skool.
This is where most coaching curves go wrong. Counting only the first program makes every curve fall to zero at the end of the term. That hides whether the back end works at all.
How do you build the table?
From your payment processor and the client board. Five steps, and a spreadsheet is enough. The board supplies the term end dates for clients who paid upfront and show only one payment.
- Export every payment from Stripe or Whop with customer, date and product.
- Give each customer a cohort: the month of their first payment.
- For each customer, mark the last month they were paying or inside a paid term.
- For each cohort, count the share still a client at months 3, 6, 9, 12, 18 and 24.
- Put cohorts in rows, months in columns and the shares in the cells.
Here is an example with made-up numbers, for a six-month program that has a next offer. It shows the shape only and is not a client's data.
| Cohort | Month 3 | Month 6 | Month 9 | Month 12 |
|---|---|---|---|---|
| January | 92% | 85% | 38% | 30% |
| April | 94% | 88% | 45% | 36% |
| July | 95% | 90% | 52% | Not yet |
How do you read a retention curve?
Across a row, you see one cohort's life. The big drop in the example comes right after month six, the end of the first term. That cliff is the back end: how many clients found a next offer waiting.
Down a column, you see whether the program is getting better. If the July cohort keeps more clients at month nine than January did, something you changed worked. If it keeps fewer, check what changed between those intake months.
Track over time, not only live. People who only see the current month never see whether they got better or worse three months ago. So they loop in the same place and never improve dramatically.
Where should the curve flatten?
After the first term, if the next offers are right. You never sell 100% of clients into the next offer. But extensions and next offers can stretch how long a client stays by another 30%, up to 50%, depending on the offer.
Some clients stay up to two years, and a few stay for years longer; those are the exception. Video Business Academy is the long end: clients stayed two to three years, and 30%+ of revenue came from resells.
Result from a named client with written permission. Typical clients of this cohort saw a range of outcomes; nothing here is a guarantee. How we handle claims.
How often should you update the curve?
Monthly, once the month's payments are in. Each update adds one new cohort row and one new cell to every older row, so the work is small once the sheet exists.
Keep every old version. The curve is for comparison over time, and a sheet overwritten every month loses the history that shows whether a change worked.
What do you change when the cliff is steep?
The timing of the next offer, first. If the resell conversation starts later than 60 days before the term ends, the extensions and resells are already missed. A steep cliff is often a calendar problem.
Then the result. Clients who reached the promised goal are the ones who buy the next chapter. A steep cliff next to a low client success rate is a delivery problem showing up at the end.
Further reading: Retention math: why 10% to 3% churn is 3.3x LTV · When to start the renewal conversation
Diego Miescher
