feature-adoption

Feature Adoption

Track how users adopt new features — who's using it, how deeply, are they coming back, and what's preventing adoption.

When to Use

• "How many users have tried the new dashboard?"
• "Are users who tried the new search feature coming back to it?"
• "What's blocking adoption of the new onboarding flow?"
• "Compare feature usage between enterprise and free users"
• "Is the new feature replacing the old one, or are users using both?"
• "Show me sessions where users started using the new feature but stopped"

Mental Model

Feature adoption is a funnel:

1. Awareness: Users who saw the feature (visited the page, saw the announcement)
2. Trial: Users who tried it (clicked on it, interacted once)
3. Adoption: Users who used it repeatedly (came back, integrated into workflow)
4. Retention: Users who are still using it N days later

The drop-off between each stage tells you where the problem is.

Workflow

Step 1: Define the feature

Clarify what "using the feature" means:

• A page visit? (e.g., visited /dashboard/new)
• A click event? (e.g., clicked "New Report" button)
• A custom event? (e.g., feature_search_v2_used)
• A sequence? (e.g., visited /search → typed query → clicked result)

Step 2: Measure awareness

How many users encountered the feature?

fullstory:build_metric(
  query="users who visited /dashboard/new",
  output_type="single_number"
)

If this number is low, the problem is discovery, not the feature itself.

Step 3: Measure trial

Of those who were aware, how many tried it?

fullstory:build_metric(
  query="users who interacted with the new dashboard (clicked any element or spent >10 seconds)",
  output_type="single_number"
)

Trial rate = trial users / aware users. Low trial rate = the feature doesn't look compelling or its purpose isn't clear.

Step 4: Measure adoption (repeat usage)

Of those who tried it, how many came back?

fullstory:build_segment("users who used new dashboard at least once")
fullstory:build_metric(query="sessions with new dashboard usage per user in last 30 days", output_type="top_n")
→ "60% used it once, 25% used it 2-3 times, 15% used it 4+ times"