What to test first
- AudienceBroad vs retargeting vs lookalike
- FormatVideo vs image vs carousel
- OfferFree estimate vs discount vs fast booking
- Creative angleProof vs price vs speed vs outcome
The angle sits at the base because it shapes everything above it. A strong angle can make an average audience work. A weak one fails everywhere.
Run a clean test
- One changeKeep everything else the same.
- SplitUse Meta's A/B test tool for fair groups.
- WaitCollect enough leads on each side.
- DecideJudge by cost per lead, not likes.
Common mistakes
- Ending the test after a day or two
- Judging by clicks or likes instead of leads
- Testing tiny changes, like one word
- Testing on a budget too small for clear results
Quick tests without the tool
Putting several ads in one ad set is not a true A/B test. Meta shifts spend toward early favorites, so one ad may barely run. It still helps find promising ideas fast. Confirm the winners with a proper test.
Good example. A dog groomer tests a video of a real grooming against a price graphic. The video brings leads at half the cost.
Bad example. A groomer changes the image, text, audience and budget in one week. Results improve, and nobody knows why.
Read what makes social ad creative work. Our Facebook and Instagram ads management tests a new angle every month.