A/B Test Sample Size Calculator: How Many Visitors Do You Need?

An A/B test sample size calculator tells you how many visitors each variation needs before you launch, not after you have already spent weeks collecting data. Enter your current baseline conversion rate, the smallest lift you actually care about catching, and your confidence and power settings, and the tool returns the visitor count you need per variation and in total.

Why sample size comes before the test, not after

Most testing mistakes happen because a team starts a test without knowing how much traffic it needs, then checks the result daily and stops the moment it looks good. Deciding the sample size in advance fixes both problems. It tells you roughly how long the test needs to run given your traffic, and it gives you a stopping point to commit to before the data can tempt you into an early call.

How the calculation works

The calculator uses the standard formula for comparing two proportions. It combines your baseline rate and your target rate, which is the baseline plus the minimum detectable effect, with z-scores for your chosen confidence level and statistical power. Smaller minimum detectable effects need dramatically larger samples, since a one point difference is far harder to separate from noise than a ten point difference. Higher confidence and higher power both push the required sample size up as well, since each guards against a different kind of wrong call.

Reading the result correctly

The number you get is a minimum, not a target to stop at the moment you cross it. Let the test run to at least that sample size and ideally through one to two full business cycles so weekday and weekend behavior evens out. Once you have collected enough traffic, run the result through an A/B test significance calculator to confirm whether the difference held up. Planning sample size properly is the same evidence-first habit behind good SEO and conversion work, deciding what proof you need before you go looking for it.