Is it possible to run 2 different A/B tests on the same page? | Community
Skip to main content
I
isolda
Level 2
March 17, 2022
Solved

Is it possible to run 2 different A/B tests on the same page?

  • March 17, 2022
  • 2 replies
  • 1846 views

Hi!

 

We are trying to run two different experiences, both have to be A/B, pointing to the same page.

Nevertheless, we can't find a solution to get both custom scripts, as it only returns one of them.


Is it possible to run 2 different A/B tests on the same page at the same time?

If so, does anybody have a solution for it?

 

Many thanks.

    This post is no longer active and is closed to new replies. Need help? Start a new post to ask your question.
    Best answer by SundeepKatepally

    It is possible provided both have distinct segments i..e no clash on the traffic

    again it more depends on your segments i.e. are segments mutual exclusive in both of the AB Tests. If not , you will not see proper results.

    2 replies

    SundeepKatepally
    Community Advisor
    SundeepKatepallyCommunity AdvisorAccepted solution
    Community Advisor
    March 18, 2022

    It is possible provided both have distinct segments i..e no clash on the traffic

    again it more depends on your segments i.e. are segments mutual exclusive in both of the AB Tests. If not , you will not see proper results.

      kchaitanya
      kchaitanya
      Level 3
      March 21, 2022

      I would argue that the experiences should be mutually exclusive and not the segments. It is okay to have overlapping segments witnessing both the tests as long as the experience from one test does not influence the experience from another test. 

      Target splits visitors randomly to an A/B test. So each test would randomly split the traffic individually to qualify for its experiences. As long as the mutual exclusivity of the experiences are taken care of the results should be valid.

        Mycal_Tamariki
        Mycal_Tamariki
        Level 3
        March 22, 2022

        You could run the test as a MVT rather than 2 different AB's to determine the best combination, then if needed you could run an AB with the winning variant to test it's true performance vs. current/default.  Hard to tell if this is an acceptable solution without more detail, just a thought based on what information has been provided. 

          Terms & ConditionsAccessibility statement

          Loading footer...