A/B Test Simulator - Conversion Rate Prediction

A/B Test Statistical Simulator is an analytical utility designed to evaluate the statistical significance of split testing campaigns. Split testing is standard for optimizing websites, email newsletters, and ad campaigns. However, raw conversion counts can be misleading due to random chance. This simulator uses robust Bayesian and frequentist statistical models to calculate conversion rates, relative uplifts, p-values, and statistical power, showing you whether your variant is truly a winner.

The simulator calculates conversions and sample sizes, then computes standard Z-scores and two-tailed p-values for frequentist significance. It also uses beta-binomial distributions to simulate the probability that Variant B is better than Control A. These complex mathematical simulations are performed instantly in the browser client.