Create an A/B test in Experiments tool

The original 11 ad objectives in Meta Ads Manager have been simplified into 6 to better align with your business goals. All the same functionality is available through additional settings. Starting from January 2024, you can no longer create, duplicate or import campaigns, ad sets and ads previously created with the original objectives. As soon as you transition from original to simplified objectives, make sure that you pause the campaigns, ad sets and ads that use original objectives to avoid running identical campaigns in parallel. Learn more about this update to the ad objectives.

Reach and frequency will be renamed reservation starting from October 2023. Learn more about reservation.

You can create an A/B test in Experiments to compare two or more ad campaigns and see which one performs best for your chosen business objectives. You can use two or more existing campaigns or duplicate an existing one to create your test.

The Experiments tool can be useful for A/B testing when:

• You have active ad campaigns that you're trying to compare.
• You want to set up each ad campaign with specific parameters before starting your test.
• You want to select additional metrics in addition to your key metric.

To create this test:

1. Go to Experiments.
2. Select Create test in the left-hand side toolbar.
3. Find A/B testing and click Get started.
4. Choose how you want to create your A/B test. You can use two or more existing ad campaigns (Test another campaign) or duplicate an existing ad campaign (Duplicate an existing campaign).
5. If you're using existing ad campaigns for your A/B test, you can choose up to five to compare in your A/B test. Note: You can't use any campaigns that are already in another test or any campaigns using the reservation buying type.
6. Schedule the start date and end date of your test. This is different from the start and end date of the ad campaigns you're using for your A/B test, but should be set to cover the duration of your ad campaigns, or enough time to yield reliable results.
7. Enter a name for your A/B test.
8. Choose how you'd like to determine a winner for the A/B test.
9. Click Review test details to review your A/B test options and schedule.
10. Click Create test to submit it for setup.

If you're not sure about what kind of variable, budget or schedule to choose for your test, learn more about A/B test creation best practices.

• About A/B testing
• A/B tests types available on Meta technologies
• Choosing a variable for your A/B test