Plan the test before looking at the result
Enter visitors and conversions for control and variation. Use separate, independently assigned groups and the same definition of conversion. This calculator compares two proportions with a pooled, two-sided z-test.
The p-value measures how incompatible the observed difference is with the equal-rate model under the test assumptions. It is not the probability that the variation is better, nor a measure of the business value of the change.
Specify the sample size or stopping rule before the experiment. Repeatedly checking until p falls below a threshold can increase false positives. A sequential method needs a design that accounts for repeated checks.