Sample size calculator
Enter your baseline rate and the smallest lift you care about. We estimate the users and days you need before you start.
Two-sided test, 95% confidence, 80% power. Kerf runs this check for you before any experiment starts, and warns when traffic is too thin.
Understanding the inputs
What is the minimum detectable effect?
The MDE is the smallest change you would still act on. A smaller MDE needs many more users, so set it to the lift that would actually change your decision.
What is statistical power?
Power is the chance of detecting a real effect of at least the MDE. Eighty percent is a common default, which means one in five real wins will be missed.
What does confidence mean?
Confidence of 95 percent means that if there were no real difference, you would see a false positive about five times in a hundred tests.
How is run time estimated?
We divide the total users needed by your daily eligible traffic, then round up to whole weeks so every day of the week is represented.
Can I stop early?
Only with sequential testing. Peeking at a fixed sample test inflates false positives. Guardrail metrics are the safe way to watch a test while it runs, and you can read more on the guardrails page.