Interactive calculator

Find out if your A/B test result is real or just noise

This tool works out whether the difference between two versions of an ad, landing page or email, your A and B variants, is likely to be real or could just as easily have happened by chance. Enter the visitors and conversions for each variant and it returns a significance level you can use to judge how much confidence to place in the result. The maths behind it is standard, but the answer is only as good as the numbers you feed in, and a small sample size can easily produce a misleading result.

How-To Guide

A/B Test Significance Calculator: Telling a Real Result from a Lucky One

This calculator checks whether the difference between two versions of an ad, landing page or email is large enough to trust, or small enough to be chance. It's built for anyone running paid campaigns or on-site tests who wants a straight answer before committing budget to a winner. Feed it your visitor and conversion numbers for each variant, and it does the statistical work that would otherwise mean building a spreadsheet formula from scratch.

Level: Beginner
Best for: Marketers and business owners testing ads, landing pages or emails
  1. Add your test numbers

    Enter the number of visitors and conversions for each version of your ad, page or email so the calculator has something to compare.

  2. Run the calculation

    The calculator works out the conversion rate for each variant and checks how likely that difference is to be down to chance.

  3. Interpret the result

    Read the significance result alongside your sample size before deciding whether to roll a variant out or keep testing.

Is your A/B test result statistically significant?

This calculator compares the visitor and conversion numbers for two versions of an ad, page or email, and works out whether the difference between them is likely to be real. Enter the figures for each variant and get the result straight away, which suits anyone testing ad creative or landing pages without a statistician to hand.

Enter Your Test Data

Control

Variant

Confidence Level
Test Type

Results

Fill in the visitor and conversion numbers for both variants above, then press Calculate to see whether the difference between them holds up statistically.
Relative Uplift
vs. control
P-Value
Lower = more significant

This result is a statistical estimate, not a guarantee, and the calculator can give a misleading answer if either sample is small or the test hasn't run for long enough. A significant result means the difference is unlikely to be down to chance, but whether it's worth acting on depends on the size of that difference and your own testing strategy.

Free Tools

See whether your test result is real or just noise

The calculator takes the visitor numbers and conversion counts from your test and works out whether the difference between two versions is statistically significant or just the kind of variation you'd expect by chance. It's built for anyone running tests on ads, landing pages or email subject lines without a data analyst to hand. Treat the output as a steer for your next decision, since small sample sizes can still produce a misleading result.

No sign-up needed · Free to use · Works on any device