Statistical significance between two variants

Product and marketing teams can use it after an A/B test to decide whether the gap between variants is strong enough to guide a choice.

Judge whether the observed difference between two variants is statistically significant, giving your decision a measured basis instead of a hunch.

How to use

Enter the number of visitors for variant A, then the number of conversions, each value on its own line.
Do the same for variant B, keeping the same order: visitors first, conversions second.
Run the analysis to compare both sets of figures and get a reading of the observed gap.
Review the result while keeping your sample size and the real test duration in mind.
Clear the values if you do not want them left in the browser.

Example

The sample contains four lines: visitors and conversions for A, then visitors and conversions for B. After running it, look at how the comparison is presented and remember that any gap still needs careful interpretation before it guides a decision.

Input

Limits and formats

The tool does not choose which variant to ship; it only helps you read the numbers you provide.
Miscounted visitors or conversions distort the outcome, especially when both variants were not measured in the same way.
A non-significant gap does not prove the variants are equal; it means the data cannot settle the question.
The calculation cannot replace a plan set before the test, with a duration, a goal and criteria chosen in advance.

Where processing happens

The whole calculation runs inside your browser. The figures you type remain on your device and are not sent to any outside service. You can rerun the analysis or clear the fields without creating an account or leaving a copy with the tool.

Frequently asked questions

Can I enter percentages instead of raw numbers?

Better to use real visitor and conversion counts. A percentage alone cannot reveal how many cases sit behind the measure, and the tool compares counts first.

What should I do if my test ran for several weeks?

Use the consolidated totals from the period you planned, and avoid stopping when the numbers look best. The tool does not account for duration or traffic changes over time.

Is the result enough to make a marketing decision?

No. It supports your reading of the gap, but other factors matter: cost, quality, seasonality or effects on other metrics. Also check that the test ran without technical problems.