Measure Feature Adoption in PostHog
Measure who could use the feature, who completed its meaningful action and who returned. Total event volume alone can rise because a small number of heavy users repeat the action.
Before you start: Have a feature eligibility definition, success event and stable identities. Record rollout dates and exclusions before building a launch report.
Worked example: A new report export is available to paid workspaces. Count eligible users separately from all signups, then inspect first successful export and later repeat export.
Step-by-step guide
Define eligibility
Describe who can see and use the feature. Include plan, role and rollout constraints so the denominator represents a real opportunity.
Separate exposure from success
Identify whether an event means the feature was shown, attempted or completed. Keep these stages distinct in the metric definition.
Measure unique adopters
Use an insight counting the intended people or groups, not only total events. Check that retries and automated jobs do not inflate adoption.
Inspect repeat value
Review repeat use within a reasonable cadence. A first-time trial and an enduring workflow are different outcomes.
Compare relevant segments
Break down by a few meaningful eligibility or acquisition dimensions. Keep observation time equal and avoid interpreting tiny groups as a broad trend.
Read the evidence together
Record reach, success and repeat use with the rollout context. Use replays or interviews to explain friction before proposing a redesign.
Common mistakes
All users as denominator
People without access cannot adopt. Define eligibility.
Clicks treated as completion
A failed attempt is not value. Validate the success event.
Totals treated as reach
One heavy user can create many events. Include unique adopters.
Tips
Keep the rollout audience in the report notes.
Inspect a failed attempt alongside a successful one.
Use the product’s natural repeat-use cadence.
Sources and verification
Based on product documentation checked on 2026-10-24. This is a documentation-based guide, not a hands-on benchmark. Features and availability can change.