What is A/B test?
An A/B test is a controlled comparison of two or more versions of something, such as an ad, that differ in one variable. In Meta Ads Manager, the A/B test shows each version to a separate segment of the audience so nobody sees both, then names the version with the lowest cost per result as the winner and reports a confidence percentage. Tests can compare creative, audiences, placements and other variables.
Meta recommends testing one variable at a time, giving each version the same budget and running a test for at least seven days, up to a maximum of 30. An A/B test differs from running several ads in one ad set, where Meta's delivery system shifts spend towards the ads it predicts will do best.
Updated
Sources
- 1.Meta's A/B test changes one variable, shows each version to a separate audience segment so nobody sees both, recommends equal budgets and at least seven days, and allows up to 30 days. Meta Business Help Centre, About A/B testing. Checked 14 October 2026.
- 2.The winner is the version with the lowest cost per result, reported with a confidence percentage. Meta Business Help Centre, Viewing and understanding A/B test results. Checked 14 October 2026.
- 3.Within an ad set, an ad earning more engagement runs more often, so it gets more impressions than the others. Meta Business Help Centre, How to fix it: One ad in your ad set is getting more impressions than the rest. Checked 14 October 2026.