Pre-launch planning
Estimate whether your traffic can support an experiment before scheduling it.
Estimate the required sample size per variant from baseline rate, minimum detectable effect, confidence, and power.
Planning sample size before you launch a test helps prevent underpowered experiments.
This calculator estimates the per-variant requirement using standard normal approximations.
Estimate whether your traffic can support an experiment before scheduling it.
This calculator expects minimum detectable effect as a relative uplift percentage.
Compare two variants using visitors and conversions to estimate uplift, z-score, p-value, and significance.
Compare three or more variants by visitors and conversions, review rates, and identify the current leader.