Retention lift from a feature
Compares how long users stay when they adopt a given feature in their first week against users who never touch it.
Authored by Intempt Technologies
Recipe curators
Agentic workflow
Coming soon1 step Blu runs end to end. It drafts a plan and nothing runs until you approve it.
- 1
Compare adopters and non adopters
Coming soonBlu builds a step
About this recipe
When to use it
Power users retain better and also use more features - that doesn't mean the feature caused the retention. This puts retention curves for week-1 adopters directly against non-adopters, so the real lift (or lack of one) is visible.
What it delivers
Plots cohort retention curves side by side for users who adopted a target feature in their first week versus those who didn't, isolating the feature's real retention lift.
How you prompt this
Describe the outcome in your own words and Blu Agent maps it to this recipe, or run it by name. It drafts a plan first and nothing runs until you approve it.
Tell Blu Agent what you want
“Compares how long users stay when they adopt a given feature in their first week against users who never touch it.”
recipe.md preview
Built from the published catalog. Blu Agent's full step prompts load in your workspace when you run it.
Inputs & outputs
What the recipe needs before it runs, and what it leaves in your workspace.
Inputs
- Nothing declared beyond an Intempt workspace. Blu asks for anything else it needs as it goes.
Outputs
How the recipe runs
Blu follows the steps in order and shows you each one before anything goes live.
- 01Compare adopters and non adoptersCreate a Retention report called "Retention Lifted by Feature Adoption". Configuration: this recipe runs for a configurable target feature, identified by Click on.target_id (default: the most-clicked target_id matching the feature pattern in the last 30 days; user can specify). Cohort definitions: - Cohort A: users who emitted a Click on event with the target_id within their first 7 days after User created ("early adopters") - Cohort B: users who did NOT emit a Click on with the target_id within their first 7 days ("non-adopters") Anchor event: User created Return event: Session start Cohort granularity: Weekly (signup cohorts based on User created) Time range: Last 12 weeks Chart type: Two retention curves overlaid (Cohort A in one color, Cohort B in another) plus delta line showing absolute retention gap at each week For each week (W1, W2, W4, W8, W12), surface: - Cohort A retention rate - Cohort B retention rate - Absolute retention gap (A − B) in percentage points - Cohort sizes Annotations: - Flag the week where the retention gap is largest (the "magic moment"). - Flag if W4 retention gap is >15 percentage points (the feature is a strong retention lever). - Flag if W12 retention gap is <5 percentage points (feature isn't actually retention-driving). - Highlight the cohort size of "early adopters": if it's <30% of the base, the feature isn't getting enough first-week exposure. Surface whether the target feature is genuinely retention-correlated. This is the canonical "magic moment" / "north-star action" analysis. Taxonomy notes: - Click on.target_id is the canonical feature handle. User created and Session start are canonical. - Cohort A/B split is computed from the existence of a Click on event with the matching target_id within 7 days of User created.
Install Intempt and run it
- 1
Sign up
Create a free Intempt workspace.
- 2
Connect your data
Add the sources this recipe reads from.
- 3
Ask Blu
Type /retention-lifted-by-feature-adoption in Blu once this recipe is live there.
- 4
Review and approve
Blu builds each step and waits for your OK before anything goes live.
