Reverzereverze

A/B Test Significance Calculator

Free Calculator

Enter visitors and conversions for each variant to see the statistical confidence, uplift and conversion rates β€” instantly and privately.

100% private

Runs in your browser β€” no numbers leave your device.

Your test data

People who saw the control (A).

How many of them converted.

People who saw the variation (B).

How many of them converted.

Statistical confidence

95.8%

The chance the difference is real, not noise (two-tailed). 95% is the usual bar to call a winner; below 90% keep testing.

Relative uplift (B vs A)

30.0%

How much B changes the conversion rate versus A.

Conversion rate A

5.00%

Control conversion rate.

Conversion rate B

6.50%

Variation conversion rate.

How it works

1

Enter variant A

Add A's visitors and conversions.

2

Enter variant B

Add B's visitors and conversions.

3

Read the confidence

See the two-tailed statistical confidence update live.

4

Decide

At 95%+ you can call a winner; below that, keep the test running.

Test screenshots that actually move the needle

Reverze generates A/B screenshot variants so you always have a stronger version to test.

How does an A/B test significance calculator work?

An A/B test significance calculator tells you whether the difference between two variants is real or just random noise. It runs a two-proportion z-test on your conversion data: it compares each variant's conversion rate, estimates the standard error from the pooled rate, and converts the resulting z-score into a confidence level. The higher the confidence, the less likely the gap happened by chance.

The industry convention is to require 95% confidence (a 5% significance level) before declaring a winner β€” that means there is only a 1-in-20 chance the result is a fluke. If your confidence is below 90%, the test is inconclusive: gather more data or run it longer. Very small sample sizes almost never reach significance, which is why big, bold creative changes β€” like the screenshots Reverze rebuilds β€” are easier to prove than tiny tweaks.

Frequently asked questions

How does an A/B test significance calculator work?
It runs a two-proportion z-test. It compares the two conversion rates, estimates the standard error from the pooled conversion rate and both sample sizes, computes a z-score, and turns that into a two-tailed confidence level. 95%+ confidence is the usual threshold to call a winner.
What confidence level do I need to declare a winner?
95% confidence (a 5% significance level) is the industry standard. It means there is only a 5% chance the difference is due to random noise. For lower-risk decisions some teams accept 90%, but below that a result is considered inconclusive.
Why is my A/B test not statistically significant?
Usually the sample is too small or the difference between variants is too tiny. Significance depends on both effect size and sample size β€” small samples rarely reach 95%. Run the test longer, send more traffic, or test bolder changes with a bigger expected effect.
What is relative uplift?
Relative uplift is how much variant B changes the conversion rate compared with A, as a percentage. For example, going from a 5% to a 6.5% conversion rate is a 30% relative uplift. It measures impact; the confidence level tells you how sure you can be that the uplift is real.