Cohort Revenue Analysis
Compares cohorts at equal age rather than at a point in time, which is the error that makes every trend unreadable.
Free — no card needed
Create a free account
to open Cohort Revenue Analysis — and the other 364 prompts across 21 categories.
Already have an account?
Running it, start to finish
- Restructure the data by cohort age before reading anything.
- Find where the curve flattens, or state that no cohort is old enough.
- Compare recent to older cohorts at the same age to test whether changes worked.
What you get back
The output this produces, every time.
- Enforces equal-age comparison, without which older cohorts always look better for no reason.
- Identifies where the retention curve flattens, which is what any lifetime estimate must rest on.
- Reports revenue and logo retention separately, since they can move in opposite directions healthily.
Getting better results
Where this usually goes wrong, and how to avoid it.
- Check the cohorts are comparable. If channel or plan mix shifted between periods, the trend is a mix shift and not a retention change.
- Do not extrapolate an unflattened curve. It is the most common overstatement here, and lifetime value built on it is a guess with decimal places.
- Watch for the healthy-but-ugly pattern. Losing small customers while growing large ones looks like churn on one chart and is a good business.
More Finance prompts
All Finance →- Finance
Trading Journal Review
Reviews your own completed trade records for behavioural patterns — how you sized, when you deviated from plan — over history you already have.
ClaudeChatGPT002Open → - Finance
Three-Statement Model Review
Audits a financial model's structure and assumptions — where it double-counts, where it fails to tie, and which assumption the whole answer rests on.
ClaudeChatGPT000Open → - Finance
Annual Report to Plain English
Reads a published filing and reports what it actually says, including what changed in the language year on year — no view on the security.
ClaudeChatGPTGemini000Open →
Written for The AI University. Every prompt in this library is original work — authored, tested and revised here, not collected from elsewhere. 365 of them, free with an account.