What are Traffic Allocation and Traffic Split?

Learn how Traffic Allocation and Traffic Split work in Fibr to control test exposure and distribute visitors across A/B variants—enabling safer, data-driven experimentation.

Traffic Allocation

Traffic Allocation in Fibr refers to the percentage of visitors coming to your landing page that you want to include in an A/B test or experiment. This feature allows you to control the exposure of your test variations, ensuring that only a specified portion of visitors will see a variation while the rest see the default or control version.

For example, if 100 users land on your page and the traffic allocation is set to 80%, only 80 visitors will be included in the test. The remaining 20 visitors will experience your landing page's original, unchanged version. Traffic allocation helps limit the number of visitors exposed to changes, reducing risk and providing flexibility for gradual rollouts of new features or designs.

This feature is particularly useful when you want to limit the exposure of unproven designs or content to a smaller audience before fully committing to a change based on performance data.

Traffic Split

Traffic Split is the distribution of the allocated traffic (within the percentage defined by traffic allocation) across the various test variants, including the control. This ensures that the visitors included in the test are shown different versions of your page in specified proportions.

For example, let’s say traffic allocation is set to 80%, meaning 80 users are part of the test. You can then use traffic split to determine how those 80 users are divided among the variations. If you have two variations and an original, you might split the traffic as follows:

• 40% of the 80 visitors see the original (32 visitors),

• 30% see variant A (24 visitors),

• 30% see variant B (24 visitors).

The sum of the traffic split percentages across all variants must always equal 100%. This ensures proper distribution of visitors and consistent results across all test variants.

Together, traffic allocation and traffic split provide granular control over how your visitors experience your A/B tests, helping you balance risk, exposure, and the ability to gather reliable data for decision-making.