Check the experiment first
A small p-value cannot repair biased assignment, broken tracking, sample-ratio mismatch, novelty effects, or a metric selected after seeing the data. Validate experiment quality before interpreting the calculation.
Free calculator
A/B Test Statistical Significance Calculator
Compare two conversion rates with a two-proportion z-test and inspect lift, z-score and approximate two-sided p-value.
Observed result
A conversion4.00%
B conversion5.20%
Relative lift30.0%
Two-sided p-value0.2002
Not below the conventional 0.05 threshold.
Do not use statistical significance alone as a stopping rule or business decision. Check the test plan, effect size, sample size and experiment quality.
Method and interpretation
How to interpret statistical significance
Planning formula
z = (p₂−p₁) ÷ √[p̄(1−p̄)(1/n₁+1/n₂)]Inspect effect size
Statistical significance asks whether the observed difference is unusual under a no-difference model. It does not say whether the absolute or relative lift is large enough to matter commercially.
Respect the stopping plan
This calculator provides a conventional two-sided z-test approximation. It is not an always-valid sequential test. If results were checked repeatedly, use a method designed for continuous monitoring.
Worked example
If A records 40 conversions from 1,000 visitors and B records 52 from 1,000, the observed rates are 4.0% and 5.2%. That is a 30% relative lift, but the p-value must be interpreted alongside the planned sample, confidence interval, and test quality.