Bayesian A/B Test Calculator

Calculate your A/B test results using a Bayesian approach

Participants A

Conversions A

Participants B

Conversions B

Enter the number of participants and conversions for Variations A and B.
Adjust the prior distribution parameters if you have prior knowledge about conversion rates.
- Use the "Prior Distribution" tab to set Alpha and Beta values.
- A uniform prior (Alpha = Beta = 1) represents no prior knowledge.
- Higher values for Alpha and Beta represent stronger prior beliefs.
Review the Bayesian Analysis Results for a summary of your test outcomes.
Examine the Posterior Distributions chart to visualize the updated beliefs about conversion rates.
Check the Probability of Improvement chart for a clear representation of the likelihood that one variation outperforms the other.
Use the Highest Density Interval (HDI) to understand the range of plausible values for the true difference in conversion rates.

Prior Distribution: The initial belief about the conversion rates before seeing the data. It's defined by Alpha (successes) and Beta (failures) parameters.
Posterior Distribution: The updated belief about the conversion rates after observing the data. It combines the prior distribution with the observed data.
Highest Density Interval (HDI): The range of values containing a specified probability mass (usually 95%) of the posterior distribution. It represents the most credible values for the parameter.
Expected Lift: The estimated percentage improvement of one variation over another, based on the posterior distributions.
Probability B > A: The likelihood that Variation B outperforms Variation A, calculated from the posterior distributions.