Free tool
A/B Test Significance Calculator
Enter each variant's visitors and conversions to see whether your result is statistically significant — or plan how much traffic a future test needs.
Control (A)
Rate: —
Variant (B)
Rate: —
Two-tailed two-proportion z-test at 95% confidence. For low conversion counts (< ~20 per variant) treat the result with caution.
e.g. baseline 4% and lift 15% means detecting a move from 4% → 4.6%.
Per-variant sample for a two-tailed test at 95% confidence and 80% power.
How to read the result
The calculator runs a two-proportion z-test: it asks how likely it is that the difference between your two conversion rates would appear by pure chance if the variants were actually identical. When that probability (the p-value) drops below 5%, the result is conventionally called statistically significant at 95% confidence.
Two classic mistakes make A/B tests lie: stopping early the moment significance first appears (peeking), and running tests that were never big enough to detect a realistic lift — use the sample-size tab before you start. A significant result also isn't the same as an important one: judge the lift and its business value, not just the p-value.
ProveLift runs this analysis continuously on live experiments — with Bayesian probabilities alongside, sample-ratio-mismatch alerts, and guardrails that hold a winner until your test has enough runtime and data to be trusted.