Change the offer first
My first live ads weren't getting spend. I found the offer that was winning on the team dashboard and rebuilt the next batch around it.
01What the data said
My live ads were all one set: a $100,000 cash giveaway. Meta barely spent on them and they brought in no installs. Dozens of giveaway ads from other strategists were stuck the same way, so the problem was the offer, not the design.
The two statics with the most spend both ran one offer: your first pack can't lose. Your first paid pack is guaranteed to pull a card worth more than you paid.
02The insight
A giveaway asks people to hope. The first-pack offer removes the fear that stops them trying: "I'll just lose my money."
03The call
Drop the giveaway. Put the winning offer into new formats.
Offer comes before format in the order I test: angle, offer, persona, format. A new design won't fix an offer people don't want. So I fixed the offer first, then tested new formats on it. A straight copy of the winning static would just get grouped with it by Meta.
04Copy rules
- Say "your first pack is sweat free" or "can't lose". Never "guaranteed profit", and no dollar amount next to the offer. That reads as gambling.
- Every slab and app screen is a real upload, with labels and values exactly as the app shows them.
- Screenshot ads (Notes, iMessage) show only the screenshot. A logo on a fake text breaks the illusion.
05Iteration
Next I iterated on the client's winning "rip your first pack free" static, using characters each audience already loves: Luffy, Charizard and Zoro. Same offer, three ways in.
06Where it stands
No launch numbers for this batch yet, so no result is claimed. This case is about the decision.
Where AI fit: AI drafted image prompts from my concepts and checked renders against the real app screens. Reading the dashboard, dropping the giveaway and choosing the formats were my calls. Client spend numbers are left out on purpose.