Free tool

Experiment significance calculator

Find out whether the difference between your control and test is real or just noise. Works for A/B tests on conversion rates and on averages such as revenue per visitor.

Metric type

Confidence level

Hypothesis

Variant A (control)

Variant B (test)

Enter the numbers for both variants to see your result.

Decide sample size and duration before you start, and don't stop the test the moment it turns significant. Repeated peeking inflates false positives.

Need help running experiments?