Template

Experiment Registration Form

An interactive preregistration form to capture hypothesis, metrics, population, instrumentation, analysis plan, risks, guardrails, and decision rules so experiments produce reproducible, actionable learning.

Interactive Tool

Experiment Registration

Use this preregistration form to record the essential design, measurement, and governance details before running an experiment or pilot. A well-filled registration reduces bias, makes results easier to analyze and compare, and helps teams turn short-term tests into lasting organizational learning. Required fields are clearly marked.

Practical tips: pre-specify your primary metric and analysis plan, set explicit decision rules and rollback criteria, power your test or note why a smaller pilot is acceptable, and confirm instrumentation and monitoring before starting.

Short descriptive title for the experiment (team-friendly)
State the cause-effect belief you're testing. Example: 'Showing price anchors will increase average order value by improving perceived value.'
The single outcome you will primarily use to judge success. Be specific about numerator/denominator and units.
Define the metric precisely (calculation, filters, window, units). Include expected baseline and meaningful minimum detectable change.
Which direction indicates improvement for the primary metric?
If you ran a power calculation, paste key assumptions (effect size, alpha, power). If not, note why (pilot, feasibility, resource limits).
Enter the estimated number of units (users, sessions, orders) needed or expected during the experiment.
Who will be included, how they're selected, and any exclusions (geography, device, customer cohort, time windows).
YYYY-MM-DD (use local timezone). Note: update if schedule changes.
Expected running time. Reassess if traffic or variance differs from assumptions.
Person responsible for execution, monitoring, and reporting.
Best contact for clarifying details or urgent issues.
Technical, analytics, design, approvals, budgets, or partner teams required to run the experiment.
List potential customer, legal, privacy, safety, or operational risks and how you'll mitigate them.
Define thresholds, monitoring alerts, and the exact conditions and steps for pausing or rolling back the experiment.
Confirm that these instrumentation items are in place before launch.
Describe how you'll analyze the primary metric (preprocessing, statistical test or model, covariates, segmentation, handling of outliers, multiple comparisons). Pre-specify any subgroup analyses or secondary endpoints.
Be explicit: what result leads to scaling, iterating, or stopping? Include thresholds and business criteria.
Does this experiment require explicit consent, IRB review, or special privacy safeguards?
Record any privacy controls, anonymization, or legal approvals needed.
Checking affirms you will not alter the pre-specified primary analysis after seeing interim results.
Link design docs, tracking plan, dashboards, code repos, or approval tickets.
Comma-separated tags to help search and group experiments (e.g., checkout, pricing, onboarding).
Record results, estimated effect, confidence intervals, p-values, decision taken, and next steps.
Owner attests that data, ethics, and governance requirements have been considered.
You can explore this tool now. Sign in or create an account to save your responses and return to them later.
Sign In to Save Create an Account
Make this tool part of your work

Save a personal copy, bring it to your team, or tailor the questions and workflow to fit what you are hungry to improve.

Member customization and team collaboration are coming soon.
Discuss Ask the Librarian

Discussion

Comments and conversation will live here.