What is a cohort?
In web analytics, a cohort is a group of users who share a defining characteristic during a specific time window—most often the acquisition date (e.g., users who had their First Visit in July) or a first key Event such as signup or first purchase. Cohorts let you track how behavior changes over time within comparable groups, rather than blending everyone into a single average.
Why cohorts matter
Cohorts power retention curves, repeat behavior analysis, and downstream value metrics like LTV. Instead of asking “What’s my overall Conversion Rate?” you ask “How do July’s users convert over their first 8 weeks?” This isolates performance shifts due to product changes, seasonality, or acquisition quality (Source / Campaign / UTM).
Common cohort types
- Acquisition cohorts: grouped by first session or first Engaged Sessions.
- Behavior cohorts: grouped by completing a specific Conversion or Micro-Conversion on Day 0.
- Attribute cohorts: grouped by device, geo, or audience labels (often tied to a Client ID).
- Lifecycle cohorts: users entering a phase (trial start, first order) used to compare time-to-Macro-Conversion.
How cohorts are used
- Retention & churn: Measure the share of each cohort returning, engaging (see Engagement Time), or purchasing across weekly/monthly “buckets.”
- Attribution quality: Compare cohorts from different channels under a consistent Attribution Model.
- Product impact: Track how a release affects subsequent behavior of the affected cohort.
- Pathing: Follow cohort-specific journeys in User Flow.
Tooling notes
Cohorts are supported across GA4 as well as alternatives like Matomo, Plausible, and Simple Analytics. The mechanics are similar: define the cohort (e.g., “first purchase month”), choose the granularity (daily/weekly/monthly), then chart a metric (retention, revenue per user, conversion lag) over N periods. For reliable reads, keep cohorts mutually exclusive, avoid mixing session- and user-level metrics (Session vs. user KPIs), and compare like-for-like windows.