A/B Test Sample Size Calculator

Why decide the sample size first

The 5% false-positive rate a significance test promises only holds if you look once, at a sample size fixed in advance. Watch the dashboard and stop the moment it goes green and you will declare winners on tests where nothing happened. This calculator gives you the finish line to commit to before you start.

Minimum detectable effect

The MDE is the smallest change you actually care about catching. Smaller effects need dramatically more traffic — halving the effect roughly quadruples the sample. Set it to the smallest lift that would change a decision, not the smallest number you can type.

Absolute vs relative effect

“+2 percentage points” (10% → 12%) is an absolute effect; “+20%” of a 10% baseline is the same target expressed relatively. They describe the same test — this tool accepts either so you can use whichever your stakeholders speak in.