Put the difference into words
Before anyone designs a mockup, write down the hypothesis: is it more important for the audience to see the process, the result or a specific use case? If you can't explain the difference without the image, the test may only be about a decorative detail.
Don't change the audience, offer, format and landing page all at once unless you have to. Otherwise it's hard to tell what actually drove the result. Agree on what stays fixed and which change you're testing. Not sure which channel to start with? See Google Ads vs paid social.
One layout, one message
Check readability at the real size it will appear in the feed. A long line in small type can look fine in the designer's file and still be unreadable on a phone. The main point should be clear without studying the details.
The image and the copy shouldn't promise different things. If you use an illustrative image, don't present it as a real client result. After the click, the page should continue that exact offer.
The winner isn't always the one with the most clicks
Lots of attention can come with irrelevant inquiries. Judge what happens next against a shared definition of a lead. Don't draw conclusions from a random gap on a very small number of events.
In paid social ads, a test should lead to a decision: develop the argument, refine the offer or stop the variant. A consistent brand identity keeps you recognizable while the content of your creatives is tested and changed.
