Bayesian A/B Test Calculator
Probability-to-beat-control, expected loss, and credible intervals — not just a p-value. Paste your test results and get a decision in seconds.
Enter the raw counts from your testing platform (Optimizely, VWO, Convert, etc.)
Control (A)
Variant (B)
Results will appear here after you click Calculate.
Why Bayesian over Frequentist?
Classical (frequentist) testing gives you a p-value — the probability of seeing your data if there was no effect. That's not what you actually want to know. You want to know: what's the probability that B is better than A?
Bayesian testing answers that directly. It also lets you stop tests early without inflating false-positive rates, and it quantifies the expected loss of making the wrong decision — giving you a risk-adjusted framework for shipping winners.
How to read the results
The probability that Variant B has a higher true conversion rate than Control A. Above 95% is typically a strong signal to ship.
If you ship the winning variant and you're wrong, how much conversion rate do you expect to lose? Below 0.5% is considered safe to ship.
The observed percentage improvement in conversion rate. Remember: observed uplift ≠ true uplift. Use this alongside probability-to-beat-control.
Want someone to interpret your test results?
Book a free strategy call. Our CRO team will review your test data and tell you exactly what to do next.