Facebook Ad Copy Variations

Why this prompt works

The five-framework constraint (PAS, AIDA, BAB, social proof, story-based) is what makes this prompt work. Without it, the model writes five variations that are stylistically interchangeable. Forcing each variant into a different copywriting framework produces output you can actually split-test, because the variables are the framework itself, not just word choice. The character caps on primary text, headline, and link description match Facebook's actual visible-portion limits, so what you get fits the placement. The 'avoid generic openers like Are you tired of' line cuts down on the standard cold-email energy that makes Facebook ads look like spam.

When to reach for it

• You're filling an ad account with starter creative for a new campaign and need diversity built in from day one.
• Your existing ads have plateaued and you want to test a fundamentally different angle, not just a new image.
• You're writing for a client and need to show the rationale (which framework, what hook) so they can pick rather than guess.
• You're a solo founder running ads and don't want to spend three days reading copywriting books before you launch.

How to customise it

Tone is the variable that shifts the output most. 'Casual' produces something that sounds like a friend recommending a product; 'edgy' leans on direct address and a touch of provocation; 'professional' is what you want for B2B placements. Set the price point honestly. The model handles a £30 product and a £3,000 product very differently, and giving it a fake number to anchor on produces ad copy that doesn't match reality. If your industry has a regulator (finance, health, supplements), add 'no health/income claims' to the constraint list. The model defaults to ambitious claim language otherwise.

What good output looks like

Five labelled blocks (PAS, AIDA, BAB, Social proof, Story), each with primary text, headline, link description, and CTA. The character counts are usually right at or just under the limits. The story-based variant is typically the most surprising output and the one most worth testing first, since it's the framework most B2B ad accounts under-use. The PAS and AIDA variants will look familiar; treat them as the baseline.

Watch out for

The model sometimes produces variations that all use the same emotional hook in different wrappers. Read the five outputs side-by-side and check that the underlying promise differs, not just the framework. If they all read like 'save time and feel calm', re-run with 'each variation must lead with a different specific outcome'.