Ad Copywriting Formulas Checked Against Winning Ads

The formulas, scored

Copywriting formulas are structures for ordering an ad: what comes first, what comes next, where the product and the offer go. Every copywriting course teaches a few. We wanted to know which ones actually show up among winners, and whether any of them work better than the others.

Ad Radar doesn't tag ads by formula. It tags them on ten dimensions, including the hook pattern, the hook mechanism and the narrative angle. Most formulas map cleanly onto a combination of those tags, so we used the combinations as stand-ins. On the October 9, 2026 snapshot of 2,443 live Meta ads with $50k+ in estimated spend:

The board average top-tier rate is about 10%. The matches are approximations: a tag combination is not the same as a copywriter deliberately following a formula. But the groups are large enough (except the last two) to show direction.

The ranking runs almost opposite to how often each formula is taught. The most common structures (AIDA, PAS) are at or below average. The less common ones (Hook-Story-Offer, contrarian reveals) reach the top tier about twice as often.

AIDA: true of everything, predicts nothing

Attention, Interest, Desire, Action. It is the oldest of these formulas, and it describes almost every ad ever made: 95% of the winners on our board end by asking for a purchase, and almost all the rest ask for a lead. Their top-tier rate is the board average, because they are the board.

That's the problem with AIDA as a writing tool. It tells you an ad should get attention and end with an action, which you knew. It doesn't tell you how to get attention or what builds desire. For that you need something more specific, which is where the other formulas come in.

PAS: the default, and it needs its middle

Problem, Agitate, Solution. Name the problem, make it hurt, present the fix. It's the backbone of direct response, and on our board it's the single most common shape: 622 ads open on a problem or benefit and use the Problem-to-Solution Bridge as their hook mechanism.

As a group, they reach the top tier 8.8% of the time, below the board. The split by copy length explains why:

The "agitate" step is what most short PAS ads skip. They name a problem and jump to the product in two lines. With room to describe what the problem costs, how it feels, what the reader already tried, PAS reaches the top tier half again as often as the board average.

The agitation here is a single, specific line: harsh contrast that ages the eyes. It names a cost the reader may not have considered and makes their current habit the cause. The product claims that follow are the advertiser's own. More on this kind of copy in long-form ad copy and on the reader it targets in problem-aware ads.

BAB: before-after openers are rare, transformation stories aren't

Before, After, Bridge. Show the reader's life now, show it after, then present the product as the bridge between the two.

In its literal form, opening on a before-and-after, it's nearly absent from the board: only 4 ads use a Before/After Reveal opener. Too few to measure, though all four have estimated spends above $1M.

The broader version, a personal transformation story where the narrator's life changes, is more common. Ads with a Story/Case Study angle and a Personal Story hook mechanism (91 ads) reach the top tier 11.0% of the time with a $291K median estimated spend, one of the highest in the table. BAB works when the "before" belongs to a real person the viewer can recognize. See story and case-study ads.

FAB: the feature list that keeps running

Features, Advantages, Benefits. List what the product has, what that does, and what it means for the reader. In ads, it usually shows up as a bulleted list in the primary text: checkmarks, emoji bullets, short lines.

397 winners carry a bulleted list in their primary text. They reach the top tier 11.6% of the time, slightly above the board, and run a median of 161 days, 41 days longer than the board median. Feature lists are evergreen. They don't reference a moment, so the ad doesn't age.

FAB works best as the second half of an ad: the creative or the first lines make the reader want the product, and the list answers "what exactly do I get?" A list as the opener rarely does anything. Examples of how winners lay out these lists are in Facebook ad copy examples.

Hook-Story-Offer: the strongest formula on the board

Hook, Story, Offer. Stop the scroll, tell a story that builds belief, close with an offer. We matched it as ads with a story angle or first-person opener whose hook mechanism is Offer Pressure. Only 56 ads fit, and 21.4% of them reach the top tier, more than double the board.

The hook is a domestic moment, not a pitch. The story is a couple comparing the new cave to the one their cats already love, which doubles as a product demo and a testimonial. The offer arrives after the viewer is already watching cats play. Each part does one job.

The data elsewhere on the board says the same thing in different ways. Offer-led ads that open with a first-person story reach the top tier 26.5% of the time, against 8.6% for offer-led ads that open with the offer itself. Story plus offer is the strongest pairing we measure. More in offers and urgency in ads.

The formula the textbooks skip: the contrarian reveal

None of the classic formulas starts with "everything you believe about this is wrong." On our board, 32 ads use a Contrarian Reveal as their hook mechanism, and 21.9% reach the top tier, the highest rate of any hook mechanism with more than 20 ads.

The structure is: belief, contradiction, explanation, alternative. "I know they feel incredible in the store. That's exactly the problem." The reader's own experience is turned into evidence. It's close to PAS, except the problem is the thing the reader thought was the solution. The specific claims about named products are the advertiser's own and its responsibility. We cover the opener in contrarian hooks.

The bridge isn't the whole ad

One more finding complicates the classic formulas. Most of them end with a bridge from problem to product. On our board, 64% of winners use the Problem-to-Solution Bridge as their hook mechanism. Those ads reach the top tier 9.0% of the time. The 821 ads that use a different mechanism (a story, an expert, an offer, a contrarian reveal) reach it 10.8% of the time.

The bridge works. It's also what everyone else is doing. The formulas that win more often on the board are the ones that put something before the bridge: a story, a contradiction, a consequence.

Which formula for which reader

The board's data on awareness levels suggests a rough match:

The full breakdown is in the five awareness levels.

How to use formulas without sounding like one

1. Pick the formula by the reader, not by habit. If every ad in your account is PAS, you're writing for one awareness level.
2. Never skip PAS's middle. If your ad goes from problem to product in one line, add the cost of the problem, the failed fix, the moment it got worse.
3. Write one Hook-Story-Offer version of every concept. It's the rarest structure on the board and the most successful.
4. Try a contrarian opener on your best PAS ad. Same problem, but start with the belief you're about to overturn.
5. Use FAB lists after the hook, not as the hook. They keep ads running; they don't stop the scroll.
6. Let a person carry the formula. The structures that win (story, contrarian reveal, BAB) all have someone in them.

If you write with Claude, you can give it these structures and real examples to work from; see Claude prompts for Facebook ads. The Ad Radar board lets you filter winners by hook mechanism and angle, which is the quickest way to pull ten real examples of any formula in your niche.

Figures marked as estimated spend come from Ad Radar's model of engagement on public Meta Ad Library ads. They are estimates, labeled as such, and are best used to rank ads against each other.