Read a retention cohort table
Prompt that explains a cohort table, compares cohorts at the same age and separates chance differences from real trends. Paste it into an AI tool of your choice, which then explains the table.
- Topic
- Spreadsheets & data
- Type
- Prompt
- Level
- Intermediate
- Time
- about 8 min
Results depend on the tool, the model and what you enter.
Result preview
Example: written by our editors, not real AI output.
Structure: rows = month of first purchase, columns = months since then, values = share of customers who bought again. Largest drop: from month 0 to month 1 (100% → 31 to 38%), similar in all cohorts. Same age compared: month 3 is 22% for the January cohort and 27% for the April cohort. Caution: the June cohort has only 84 customers; the last two columns are not complete yet. Answer: the new onboarding probably helped, because later cohorts keep more customers at the same age. The difference is small and based on a few months. More certain with: month 3 of the June and July cohorts.
Fill in
What you type stays in this browser window. We store and send nothing.
- month of first purchase
- week of sign-up
- month the subscription started
- first order via a campaign
- first app install
- did the new onboarding help
- when do most customers leave
- is it worth promoting the annual plan
- which acquisition month was best
- do discount customers stay
Your prompt
You have edited the prompt. Changes in the fields above no longer overwrite it.
This is an instruction for an AI, not the result itself.
Where this prompt works
These AI tools are a good fit for this prompt:
How to use it
- Fill in the fields or tap a suggestion. Leave out names, addresses and other personal data.
- Copy the finished prompt into an AI tool of your choice.
- Check the figures quoted in the answer against the table before you use them in a report.
Why does this work?
Comparing cohorts at the same age separates real improvement from simply looking at older groups. The fixed check for small cohorts stops chance differences being presented as trends.
Tips
- The most recent cohorts always look incomplete; compare them only up to the age they have actually reached.