Experiment Calculator

Sample size before you launch, and a confidence interval after you stop. Both directions, with the assumptions written on the screen rather than buried in a footnote.

Loading…

About this tool

Two calculators that should always be used together and almost never are. Plan mode tells you how many users you need for the effect you care about. Analyse mode takes what you actually observed and gives you an interval — deliberately leading with the interval rather than a verdict, because "significant" is not a finding.

Worth knowing

If the required sample size is larger than your traffic, that is the finding. Do not run the test.
The interval is the answer. A p-value below 0.05 with an interval spanning −1% to +9% has told you almost nothing.
Peeking inflates your false positive rate. If you plan to stop early, plan the stopping rule first.