Input visitors (sample size) and conversions (successes) for your control group and at least one variant. For revenue tests, also enter total revenue per variant.
Adjust confidence level, hypothesis type, and variant count. Hover over the i icon on every metric to learn what it means and what range is ideal.
The Results tab shows significance verdict, lift, p-value, statistical power, Wilson confidence intervals, distribution chart, and an SRM (Sample Ratio Mismatch) check.
Power Analysis: Required sample sizes for various effect sizes, plus a visual power curve.
Monte Carlo Simulation: 5,000 randomized trials using true binomial sampling for probabilistic outcome estimation.
Insights: Plain-English recommendations with projected business impact.
Download a multi-sheet Excel report with Summary, Insights, Simulation, and Power Analysis tables.
Test: Two-proportion Z-test with signed z-statistic for one-sided tests.
CIs: Wilson interval (more accurate than Wald for small samples and extreme rates).
Power: Corrected single-term non-central normal formula.
Multiple comparisons: Bonferroni correction for 3+ variants.
SRM: Chi-square test for sample ratio mismatch detection.
Simulation: True binomial Monte Carlo, 5,000 trials.