A/B Test Sample Size Calculator
Estimate A/B test sample size, traffic allocation and run time from baseline conversion rate and minimum detectable effect.
Tool workspace
visitors
visitors
adjusted for split
Experiment readout
How to use this tool
Formula used
Two-proportion test size estimate using baseline rate, target rate, alpha, power and traffic allocation.
Use when
Use before starting an experiment to decide whether the page has enough traffic for the effect you care about.
Use when comparing run time tradeoffs between MDE, power, alpha and traffic split.
Be careful when
Do not treat this as a final experiment readout; it plans sample size before the test runs.
Do not run very short tests that ignore weekday mix, seasonality or traffic-quality shifts.
Worked example
If baseline conversion is 5% and the MDE is 15%, the target rate is 5.75%; the tool estimates visitors per arm and expected days.