Use this template to document experiments before launching. Include: hypothesis, primary metric, sample size, success criteria, and risks. Pre-registering tests prevents p-hacking and ensures you have a clear plan. Copy the template below or download as Markdown/Notion.
Why Use an Experiment Brief?
Prevents p-hacking
Commit to metrics and sample size before seeing results
Forces clear thinking
Articulate why you expect a change to work
Creates documentation
Build institutional knowledge of what works
Enables learning
Compare hypothesis to results, improve over time
The Template
Experiment Name
Clear, descriptive name (e.g., "Pricing Page - 3 vs 4 Plans")
Hypothesis
If we [change], then [metric] will [improve] because [reason]
Primary Metric
The ONE metric you're optimizing for
Secondary Metrics
Other metrics to monitor (guardrails)
Target Audience
Who will see this test? (all traffic, mobile only, etc.)
Sample Size
Calculated required sample per variant
Expected Duration
How long the test will run
Success Criteria
When will you call a winner?
Risks
What could go wrong?
Rollout Plan
How will you implement the winner?
Example: Pricing Page Test
Experiment Name
Pricing Page - 3 Plans vs 4 Plans
Hypothesis
If we add a 4th "Enterprise" plan above our current top tier, then conversion rate will increase by 8% because price anchoring will make the middle plan seem more affordable.
Primary Metric
Conversion rate (trial signups from pricing page)
Secondary Metrics
Plan distribution (% choosing each plan), time on page
Sample Size
8,500 visitors per variant (calculated for 8% MDE, 95% confidence)
Expected Duration
3 weeks (based on 6,000 pricing page visitors/week)
Download Template
Get the template in your preferred format:
Run Better Tests
Use this template to plan your experiments, then run them with ExperimentHQ.