Split audience 50/50 | Community
Skip to main content
New Participant
October 21, 2021
Solved

Split audience 50/50

  • October 21, 2021
  • 1 reply
  • 2444 views

Dear community,

 

I want to run an AB test for several weeks, where half of the users get version A every week and the other half version B. I want those users to not change versions throughout the whole time spam of the test. If I use random sample, I cannot assure that the same users will get the same version during this time. Do you have an idea how to set this up? 

 

Best,

Anna

This post is no longer active and is closed to new replies.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 Michael_Florin-2

If you want to make your sample cohorts persistent, you have a couple of options. The most common ones are:

1. Static Lists - Create two lists first, name them A and B, and distribute your recipients into these lists using the random sample flow step. After that, don't use the random sample flow step again, but use list membership instead.

2. A field value - Create a custom field in your database called "Random Cohort". Set up a Smart Campaign that assigns a value for "Random Cohort" based on the Random Sample Flow Step. Something like this:

Continue until every person in your database has a value in Random Cohort between 0 and 9. That gives you the option to create 10 persistent buckets of persons you can use for all kinds of ongoing, long-term splits.

1 reply

Michael_Florin-2
Michael_Florin-2Accepted solution
New Participant
October 21, 2021

If you want to make your sample cohorts persistent, you have a couple of options. The most common ones are:

1. Static Lists - Create two lists first, name them A and B, and distribute your recipients into these lists using the random sample flow step. After that, don't use the random sample flow step again, but use list membership instead.

2. A field value - Create a custom field in your database called "Random Cohort". Set up a Smart Campaign that assigns a value for "Random Cohort" based on the Random Sample Flow Step. Something like this:

Continue until every person in your database has a value in Random Cohort between 0 and 9. That gives you the option to create 10 persistent buckets of persons you can use for all kinds of ongoing, long-term splits.

anpa7992Author
New Participant
October 22, 2021

great thank you!!