Free experiment calculators
Plan a test, read the result and see what a lift could mean for revenue. Enter your own numbers and get an answer you can explain to your team.
See your result right away, without signing up. Your numbers stay in your browser.
Enter visitors and conversions for each variant to see what happened.
Your result
Your result will appear here after you calculate it.
All calculations run in your browser.
Before launch, estimate how many visitors you need. Once results come in, check the traffic split and conversion lift. Add your traffic and revenue to see what the lift could be worth. Each result explains the assumptions behind it.
Check tracking and traffic changes before you decide whether to roll out a variant.
Enter visitors and conversions for control and variant. The calculator shows a two-sided p-value, observed conversion lift and Bayesian chance that the variant beats control. A p-value below 0.05 indicates a difference under the z-test assumptions, but does not prove the change caused it.
It depends on your baseline rate, the smallest lift worth detecting, desired power and confidence. Enter those values in Plan a test to estimate visitors per variant. Smaller effects need much larger samples.
Divide the total required sample by eligible daily traffic for a rough duration. Allow at least one full business cycle and investigate traffic changes or tracking problems before deciding.
Sample ratio mismatch means the observed traffic split differs more than expected from the planned allocation. It can signal assignment or tracking problems. Check the split before interpreting conversion results.
The calculator multiplies baseline revenue per visitor by eligible monthly traffic, 12 months and relative uplift, then applies your effect-retention and seasonality assumptions. It is a scenario estimate, not a forecast.
Yes. Your result appears when you calculate it. You can use every calculator without creating an account.