Can I run multiple A/B tests at once, and can users be in multiple A/B tests?

Since it is possible to run multiple A/B tests at once, and a user can be in more than one concurrent A/B test, you should verify that your code does not allow collisions between multiple tests by checking that tweaks are not applied to the same variables and elements.

What if my tests don’t target the same users?

The Mixpanel UI allows you to run multiple A/B tests at once, but it is best practice to only run one A/B test at a time, even if your tests target different users. Creating individual, independent testing environments allows you to accurately isolate and gauge the effects of your variables. In other words, running multiple tests concurrently makes it difficult to pinpoint the driving cause for user behavior.

Instead of multiple A/B tests, consider one A/B test with many variants, which will allow you to better analyze the results of your testing.

How do I measure the effectiveness of my A/B tests?

If I set up two tests targeting the same users, which test will each user get?

Only one A/B test will be applied to a particular user at a time. If the user is targeted by multiple A/B tests from the same project, the user will only receive the test with the lower experiment number (i.e., the A/B test that was created first).

If one user is targeted by multiple A/B tests from different projects, the A/B test they see depends on several factors, including whether they’re on iOS or Android. If you’re configuring multiple A/B tests for users from different projects, reach out to support@mixpanel.com for guidance on best practices for your specific use case.