Business context
Why Feature Usage Frequency matters
A higher result means more feature events for the session volume. That can show deeper use, repeated steps within a workflow, or duplicated tracking.
- Business question
- How often is this feature used relative to the number of product sessions?
- Teams that use it
- Product, design, engineering, customer success, and analytics teams.
- Decisions it supports
- Feature investment, navigation, education, workflow design, and instrumentation.
Calculation
Feature Usage Frequency formula
(Feature Usage Events ÷ Total Sessions) × 100
Formula components
- Feature usage events
- Valid tracked events representing use of the selected feature.
- Total sessions
- All valid product sessions in the same scope and period.
- Per-100-session rate
- Multiplying the event-to-session ratio by 100 for readability.
- Event rule
- The exact action and deduplication logic used to record one feature use.
How to calculate Feature Usage Frequency
- Define one valid feature-use event and the applicable product sessions.
- Validate instrumentation and remove duplicate or test events.
- Count feature events and sessions in the same period and population.
- Divide events by sessions and multiply by 100; supplement with unique-user adoption.
Worked example
Feature Usage Frequency example
During a month, users generate 3,200 valid export-feature events across 8,000 product sessions.
- Feature usage events = 3,200.
- Total sessions = 8,000.
- Feature Usage Frequency = 3,200 ÷ 8,000 × 100 = 40.
Feature Usage Frequency is 40 events per 100 sessions.
The feature generates 0.4 usage events per session on average. The figure could exceed 100 if sessions commonly contain several feature events.
How to interpret the result
Read frequency with adoption breadth and user outcomes. A few power users can generate high frequency while most users never touch the feature.
The level depends on whether the feature is expected once per workflow or many times, as well as event design, session timeout, user role, and measurement window.
Common mistakes and limitations
- Calling the result a user percentage
- The numerator is events, so the result is events per 100 sessions and can exceed 100.
- Duplicate instrumentation
- One action may fire several events and falsely increase frequency.
- Session-rule changes
- Changing timeouts alters the denominator.
- Confusing frequency with adoption
- Frequency shows intensity; adoption shows distinct users reached.