Implementation worksheet · 2 min read
An onboarding experiment exposure log template
An onboarding experiment exposure log connects assignment, actual component display and the later value event. Store these as separate stages. A user assigned to a tour may never see it, and a user who sees it may not activate. Keeping the stages distinct makes implementation failures and outcome analysis easier to diagnose.
Suppose a growth team compares a checklist with a guided tour. The experiment needs a stable assignment rule and a definition of exposure that both implementations satisfy. This worksheet describes the logging contract; it does not claim that either experience improves activation.
Put it into practice
1. Record assignment first
Store experiment ID, version, assignment unit and variant before the experience is shown. Keep assignment stable across refreshes. If teammates can influence each other, consider the account boundary when choosing the unit.
2. Define actual exposure
Specify when a component counts as displayed. A downloaded script or mounted hidden component may not be a meaningful exposure. Make the rule observable and apply it consistently across variants.
3. Deduplicate repeated display
Choose whether the analysis counts first exposure, session exposure or repeated views. Use stable identifiers and preserve raw events where appropriate. Do not let a re-render inflate the count.
4. Join the value event
Use the activation contract and observation window agreed before the experiment. Track completion of the underlying product action separately from tour completion, button clicks and checklist interaction.
5. Validate the full timeline
Test assigned-but-not-exposed, exposed-but-dismissed and exposed-then-activated accounts. Check variant allocation and missing identities. Report implementation gaps before interpreting the outcome difference.
Exposure log schema
Copy this structure into your review document and record your observed result for each row.
| Stage | Required fields | Meaning |
|---|---|---|
| Assignment | Experiment, unit, variant, time | Intended treatment |
| Exposure | Component, version, time | Actually displayed |
| Interaction | Action and task ID | User response |
| Value event | Outcome contract and time | Product value reached |
| Exception | Reason and affected stage | Measurement limitation |
A failure worth checking
Analyzing only people who completed a tour selects for users who were already willing to engage. Preserve original assignment for the main experiment comparison and use exposure or completion groups as diagnostic views with their limitations stated.
Common questions
Is component mount a good exposure event?
Only if mounting reliably means the user could experience it. Hidden or offscreen components often need a more explicit display rule.
Should assignment and exposure be one event?
Keep them distinct when exposure can fail or be delayed. That separation helps explain missing treatment and broken targeting.
Basis and scope
This is a proposed implementation method using illustrative examples, not a measured benchmark or a customer case study. Prepared with AI assistance. Validate product-specific behavior against current documentation and your own test environment.