When it comes to developing a product, most people focus on “why” a feature should be implemented. I would like to look at the “how” A/B testing can support such efforts in order to improve your users experience on your platform. Many companies lose a good amount of conversion and therefore money because they struggle to understand the A/B testing tools that they are using and how to interpret the test results.

We will take a look at why A/B testing experiment results using conventional approaches are often hard to interpret correctly and will explore an alternate, visually easy to understand approach that was implemented using Bayesian statistics. We will touch on the implementation, as well as the theory behind this approach.

Finally I will present a new library for A/B testing experiment evaluation that will help you to effectively increase your conversion rates and help your product managers to make the right decisions.