Sample Size Calculator
Compute the sample size required to detect a given effect at 95% confidence and 80% power.
Inputs
Required Sample
—
Run to compute sample size
Variant rate target—
Sample per variant—
Total sample—
Why sample size matters
Calling a winner before you reach statistical significance is the most common A/B testing mistake. Most variants need 5,000-50,000 visitors per arm depending on baseline rate and detectable effect. Knowing the required sample before you launch lets you forecast how long the test will run.
Frequently Asked Questions
What MDE should I use?
5-10% relative lift is the standard. Below 5% requires huge sample sizes; above 20% probably doesn\u2019t need a test.
Want a CRO program with proper test methodology?
Riman Agency runs structured CRO programs.