Template

Experiment Protocol — Interactive Template

A structured, saveable experiment protocol that standardizes design, measurement, instrumentation, safety, and decision criteria so teams can run faster, reduce risk, and produce trustworthy, actionable results.

Interactive Tool

Experiment Protocol Template

Use this interactive protocol to capture a complete, pre-registered plan for an experiment. A good protocol reduces ambiguity, prevents post-hoc changes, and makes results easier to interpret and act on. Fill the required fields, attach links to supporting documents, and save the protocol before you start the test.

Short, actionable title that communicates the change and target outcome. e.g. 'Reduce checkout friction to increase conversion'.
Person responsible for experiment execution and decisions (name and contact).
People or teams that should be informed of results or consulted during the experiment (names or roles).
One-paragraph summary: why this experiment matters, the customer benefit, and the core change being tested.
Use 'If [change], then [metric] will [direction] by [amount] because [reason]' format when possible.
Define the primary metric and exactly how it's measured (numerator/denominator, filters, window).
Other metrics to monitor (e.g., engagement, error rate, safety signals). Include units and measurement windows.
Current values for primary and secondary metrics used for context and sample-size calculations.
Optional: estimated number of units per variant. Leave blank if you will compute or iterate later.
Assumptions used for power calculations (minimum detectable effect, alpha, power, baseline), or links to the calculation.
Unit of randomization (user, session, account), method, stratification, blocking, and any exclusions. Explain why this plan avoids bias.
How traffic or units will be divided across arms.
Verify these items before starting the experiment.
Where data will be stored, naming conventions, data retention, and access permissions.
Statistical tests, time windows, handling of missing data, multiple comparisons plan, and any pre-specified subgroup analyses.
Pre-specified rule for deciding success, failure, or follow-up actions.
Conditions that should immediately pause or stop the experiment (e.g., safety signal, critical error rates, adverse customer feedback).
Privacy, compliance, potential harms, consent, and mitigation plans. Note any approvals required.
1 = low risk, 5 = high risk
1.0 10.0
YYYY-MM-DD
YYYY-MM-DD or estimated duration
Estimated run length to reach planned sample size or stable estimates.
Links to designs, PRDs, analysis scripts, dashboards, or approvals. Separate multiple links with commas.
How results will feed into prioritization, rollout, follow-up experiments, or process changes. Include owners for follow-up actions.
Record version, date, and brief summary of changes from prior versions.
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.