WebmasterID logoWebmasterID
Reports & dashboards

Cohort exploration

Cohort exploration groups users by a shared starting event (the cohort inclusion criterion) and follows a return criterion across time windows. Unlike the fixed retention report, you choose the inclusion event, return event, granularity, and calculation — so the same data yields very different curves depending on those choices.

Verified against primary sources

What this means

Cohort exploration defines a cohort by an inclusion event (for example, first_open or first purchase) within a period, then tracks a return event across subsequent daily, weekly, or monthly windows, showing retention as a grid or curve.

Four choices that change the answer

Four settings define the cohort: the inclusion criterion (what makes a user part of the cohort), the return criterion (what counts as coming back), the granularity (daily/weekly/monthly), and the calculation (standard, rolling, or cumulative). Change any one and the curve changes. That flexibility is the point versus the fixed retention report — but it means two cohort explorations are only comparable when all four match.

How it appears in analytics and logs

A cohort cell shows the share or count of an original cohort meeting the return criterion in a later window. Curves are only comparable when inclusion criterion, return criterion, granularity, and calculation are held constant.

Diagnostic use case

Measure tailored retention — e.g. users who first purchased, returning to purchase again weekly — by defining the inclusion and return events yourself rather than accepting the standard report's defaults.

What WebmasterID can help detect

WebmasterID measures returning first-party users by event, so you can reason about cohort behavior without cross-site identifiers.

Common mistakes

Privacy and accuracy notes

Cohort exploration aggregates users into groups and may apply thresholds. It tracks cohort behavior, not identifiable individuals.

Related pages

Sources and verification notes

Last reviewed 2026-06-24. Facts are checked against primary/official sources where available; uncertain specifics are marked “Data not yet verified” rather than guessed.