Free A/B Test Significance Calculator

It means the difference between version A and version B is larger than you would usually expect from random chance alone. In practice, a result is called significant when the p-value is below 0.05. Significance says the gap is probably real, not that it’s big enough to matter.

1. Wait until the test reaches its planned sample size or duration. 2. Enter visitors and conversions for A and B. 3. Press Calculate. 4. If the p-value is below 0.05, the difference is significant at the 5% level; also check the 95% interval and lift.

If A and B truly had the same conversion rate, you would see a difference at least this large about 5% of the time. It isn’t the probability that B is better, and it isn’t a 95% chance the result is correct. Smaller p-values mean the data are harder to explain by chance.

It uses a fixed 0.05 threshold for the two-sided test and a 95% interval for the difference. The exact p-value is shown, so if your team uses a stricter cutoff such as 0.01, compare the p-value with that number yourself.

Absolute difference is B’s rate minus A’s, in percentage points: 6.5% − 5% = 1.5 points. Relative lift divides that by A’s rate: 1.5 ÷ 5 = 30%. Relative lift sounds bigger, so report both when you share results.

It depends on your baseline rate and the smallest lift you care about; lower rates and smaller lifts need far more visitors. Estimate it with a sample size calculator before the test starts. This tool checks finished results and needs at least 10 conversions and 10 non-conversions per group.

The test relies on a normal approximation, which needs whole-number counts and at least 10 conversions and 10 non-conversions in each group. Conversions also can’t exceed visitors. Collect more traffic or check for typos, then calculate again.

Yes. Any yes-or-no outcome works: enter impressions as visitors and clicks as conversions to compare two creatives’ click-through rates. Keep the groups independent, with each person or impression assigned to only one version.

No. The z-test runs in your browser, and the visitor and conversion counts you enter aren’t sent to SocQ or stored once you leave the page.

Yes. It’s free with no signup or account. You can check as many finished tests as you like and copy every result, including the equation.