Choosing the right winner in an A/B test is critical to maximizing lead conversion. Heyflow simplifies this process by statistically evaluating which version of your flow is most likely to perform best.
In this article, you’ll learn more about how Heyflow determines the winner of an A/B test, helping you confidently make data-driven decisions.
How We Determine the Winner
Heyflow uses a Bayesian statistics model to calculate the Probability of Being Best (PoBB) for each variant of your test. We compare views and submits of the Baseline and Variant to calculate which version is more likely to drive higher conversions.
The PoBB is displayed as a percentage on your Analytics Dashboard, offering an easy-to-understand way to assess performance.
🔎 If you want to learn more about Heyflow’s Analytics Dashboard, click here.
Understanding Result Categories
To help you interpret your A/B test results, Heyflow groups the Probability of Being Best (PoBB) into three categories:
< 75% PoBB: Collecting data
This is the starting point for all A/B tests. At this stage, the system is gathering data to begin evaluating which flow performs best. You need at least 10 submits for both the baseline and variant to see the probability to be best. Let the test continue running for more reliable results.
75-89% PoBB: Baseline or variant is leading
At this stage, one version (either the Baseline or Variant) has emerged as the leader. While not definitive, it provides early insights into trends.
≥ 90% PoBB: Baseline or variant has won
When the PoBB for the variant or baseline reaches 90% or higher, Heyflow identifies this variant as the winner. You can confidently conclude the test and use the results to implement the best-performing flow across your campaigns, or use it as the new Baseline for further testing.
Start a new A/B test to continue optimizing your flow’s performance!