How to use Customer Retention Analyzer
- Provide the complete activity history with distinct event IDs, customers and dates.
- Set the as-of date through which activity has been observed.
- Review each acquisition cohort and its completed calendar-month return rates.
- Download the cohort report, keeping incomplete observations separate from zero returns.
Example: Customer Retention Analyzer
Of two January customers, one returns in February; March remains incomplete.
Supported inputs and limits
- Up to 2,000 activity rows, 80 columns, 2 MiB and 120 months of cohort history.
- First activity in the supplied history defines acquisition; omitted earlier history changes that definition.
- A customer counts once per calendar month.
- Every return rate uses the original cohort size.
- Month zero includes acquisition.
More details
The current as-of month is always marked incomplete with blank count/rate, including on its final day.
Future events are refused.
No causal retention explanation or predictive model is included.
Where your input is processed
This tool does not upload your inputs. See the limits for anything saved on your device.
Questions about Customer Retention Analyzer
Why does the current month show a blank instead of zero?
It is an incomplete observation period. Zero would suggest that a full month had elapsed without a return, so the count and rate remain blank.
Does retention use the previous month as its denominator?
Every completed age column uses the original acquisition cohort size. A customer can return after a missed month and still count in the later month.
Can I upload only the most recent three months?
You can analyze that supplied window, but first observed activity then means first within those rows. Supply complete earlier history when you need actual acquisition cohorts.