Method
A clear sequence for application analytics work centred on user retention cohort analysis — so definitions stay stable and recommendations stay modest.
-
Agree definitions
We lock the signup event, return event, and return windows before any table is built. Ambiguous identifiers are named early.
-
Receive a safe extract
You share an anonymised export or warehouse extract. Live production access is not required for a Retention Cohort Review.
-
Build and annotate cohorts
Signup-week groups are mapped across agreed intervals. Releases, quiet weeks, and seasonal markers are annotated beside the pattern.
-
Write the brief
You receive a written reading with caveats and a short list of next checks — not a rewrite of your product strategy.
-
Walk through together
One working session clarifies questions with product, growth, or leadership. Further advisory hours are optional.
