top of page

How to calculate the number of variants in an A/B Test

Updated: Dec 13

To calculate the number of variants in an A/B test, for example when running A/B-testing using Emerse Labs in programmatic campaigns, you first need to understand the components being tested and the variations of each component. An A/B test typically involves two or more versions of a specific element being compared to determine which one performs better.


Here's a simple method to calculate the number of variants in an A/B test:


  1. Identify the components being tested: List all the elements you want to test in your experiment, such as headlines, images, call-to-action (CTA) buttons, or layouts.

  2. Determine the number of variations for each component: For each element you are testing, count the number of variations you want to include in the test. Make sure to count the original version as one of the variations.

  3. Multiply the number of variations of each component: To calculate the total number of variants, simply multiply the number of variations for each component. For example, if you're testing three different headlines and two different CTA button colors, the total number of variants would be: Total Variants = Variations of Headlines * Variations of CTA Button Colors Total Variants = 3 * 2 Total Variants = 6 In this case, you would have six different variants in your A/B test.


Please note that the method above assumes a full-factorial design, where all possible combinations of variations are tested. In some cases, you may choose to test only specific combinations or use a fractional factorial design to reduce the number of variants, especially when testing multiple components simultaneously.


To learn more about how Emerse can offer A/B-testing as a service, please contact us.

bottom of page