A/B test planning — sample size, MDE and duration
Enter baseline conversion, expected uplift, power, confidence and daily traffic.
Planned and detectable effect
| Scenario | Baseline | Target | n per variant |
|---|---|---|---|
| Fill in the fields above — the result is computed on this page. | |||
How to use this page
- Enter baseline conversion from 0 to 1.
- Set the smallest relative effect that matters.
- Review sample size, MDE and duration, then set a complete schedule.
What this page does not do: The formula uses a normal approximation, equal allocation and independent users. It does not account for repeated peeking, clustering or multiple metrics.
Further reading
The planner estimates sample size for a two-sided comparison of two independent proportions. MDE is the smallest relative uplift detectable by the available sample at the chosen power.
Duration is total required sample divided by A+B daily traffic, rounded up. In practice, cover full weekly cycles and use a pre-specified horizon.
Frequently asked questions
Does 0.1 mean 10%?
Yes. Enter conversion and relative uplift as proportions.
Can I stop the test early?
Only under a pre-planned sequential method. Ordinary repeated peeking inflates false-positive risk.
See also
A/B test calculator
Compare two proportions in an A/B test. Calculate conversion rates, difference, uplift, z score, p value and c
ACF and PACF autocorrelation calculator
Calculate ACF and PACF across time-series lags. Review the chart, reference bounds and coefficient table.
Adjusted R-squared calculator
Convert R² to adjusted R² using the number of observations and predictors in a regression model.