Two headlines for the same offer may prompt different responses. An A/B test splits traffic between two campaigns and compares the result you choose. Change one thing at a time so the comparison is clear.
Prepare two campaigns
Create version A and copy it for version B. Change the selected element while keeping the remaining conditions comparable. Both campaigns must belong to the same account, and you need edit access to both.
Check compatible platforms and publication locations. A test cannot repair a form missing its submit button or an Embed without a matching placement target.
Create the A/B test
Select both campaigns in the tests section, name the comparison and assign traffic to A and B.
Each version needs more than zero and less than 100 percent. A 50/50 split is a useful starting configuration. The same campaign cannot fill both positions.
Check scope and activation
Review the test's active state, addresses and restrictions. A saved test and two ready designs are separate requirements.
The test assigns and remembers a version using browser information. Refreshing is not a reliable way to alternate between versions. Check both designs separately as well.
Compare the same outcome
Inspect views, clicks and submissions over the same period. If the goal is signup, a higher click-through rate alone does not identify the version with more valid submissions.
A small sample or short period can produce a temporary advantage. The dashboard does not replace a decision about an adequate sample. Avoid changing several parts during the comparison, and check which campaign remains active when you finish.

