The Winning Ad Checklist: 15 Things the Top Ads Have in Common

How we built the checklist

Every ad on the Ad Radar board already cleared $50k in estimated spend, so the board is a list of winners. To find what separates the best of them, we compared two groups on October 9, 2026:

• Top tier: 235 ads in Ad Radar's top 10% by winner score (live 21+ days, ranked on signals like days live, variants and engagement growth).
• The rest: the other 2,208 winning ads.

For each trait we measured how often it shows up in each group. Some traits are common to nearly all winners (they're on the list because missing them is a red flag). Others clearly separate the top tier. We note which is which.

The 15 points

The variant rows come with a caveat: variant count is one input of the winner score, so the gap is partly built in. We keep it at the top anyway, because it describes the single most consistent behavior behind top ads: the brand kept working on them.

What each point means in practice

1 and 2. It runs in more than one version

Two-thirds of top-tier ads carry at least two versions in the Ad Library, and half carry three or more. Among other winners, only 9% reach three. A winning ad is rarely one file. It is a family of openers, cuts and formats around one body. See iterations vs new concepts.

3. The opener is short

Median first line: 8 words in the top tier, 9 in the rest. Not a big gap, but both are short. If your opener needs a comma and a second clause, it's probably two openers.

4 and 5. It has a number in it

32% of top-tier openers contain a number (digits or words), against 26% for the rest. Ads tagged with the Specific Number Opener pattern make up 23% of the top tier and 16% of the rest. Numbers make a claim specific: "save up to 66%", "four shoes", "40 bags". More in number hooks.

6. It doesn't default to a question

Question openers are 15% of other winners but 11% of the top tier. Questions aren't bad, but they're the easiest opener to write and often the vaguest. Write the question version last, after a number version and a story version.

7. It names the problem before the product

About 6 in 10 winners on both sides use a problem-to-solution bridge as the hook mechanism. This is a baseline, not a differentiator: if your ad leads with the product and never names a problem, you are outside the norm for winning ads.

8. It knows who it's talking to

46% of top-tier ads talk to problem-aware or solution-aware readers, against 40% for the rest. These are viewers who know they have the problem, so the ad can skip the lecture and explain why this product. See awareness levels in ads.

9 and 10. It shows some proof

Two-thirds of winners on both sides carry at least medium proof (a demo, a specific result, a credential). Strong proof (a study, a test on camera) is rare everywhere but nearly twice as common in the top tier: 7.2% against 3.9%. The sample is small (17 top-tier ads), so treat it as a direction. See social proof ads.

11. There is a static version

Images are 19% of the rest and 26% of the top tier. A static image of a winning line is cheap to make and has a better hit rate than most teams assume.

12. Video length sits in the middle

47% of top-tier ads are videos between 31 and 120 seconds (44% for the rest). Median length is 62.5 seconds in the top tier and 64 seconds for other videos. Long is fine; just don't cut a story to 15 seconds by default. See best video ad length.

13. The click goes somewhere built for the ad

11% of top-tier ads send traffic to a listicle, quiz or advertorial, against 7% of the rest. Product pages are still the most common destination, but a pre-sell page shows up more often among the best ads.

14. It asks for one thing

95% of winners on the board use a buy CTA. No split goals, no "follow us". One action.

15. The offer supports, it doesn't lead

Offer-first openers are slightly less common in the top tier (9%) than in the rest (11%), and guarantee language appears in 4% of top-tier copy against 8% elsewhere. Offers matter, but the top ads tend to earn attention first and bring the offer in later.

What made no difference

Just as useful: things that showed almost no gap between the top tier and the rest.

If your team debates any of these, the data says spend the time elsewhere.

The checklist, ready to use

WINNING AD CHECKLIST (pre-launch)                    Ad: ____________

OPENER
[ ] One sentence, ~8-10 words
[ ] Contains a specific number (price, count, time, result)
[ ] Written in 3 patterns (number / story / problem); question version last
[ ] First frame shows the problem or the product in use

ARGUMENT
[ ] Problem named before the product
[ ] Awareness level chosen: ________ (problem- or solution-aware by default)
[ ] At least one proof moment shown, not described
[ ] Claims checked (compliance is on us, not the creator)

FORMAT
[ ] Video 31-120s full version + 15s/30s cuts planned
[ ] Static version of the strongest line planned

FUNNEL
[ ] One action (buy)
[ ] Offer present but not the opener (unless retargeting)
[ ] Destination matches the opener; pre-sell page considered

AFTER LAUNCH
[ ] Read window written: ___ x target CPA
[ ] If it wins: 3+ versions within 2 weeks

Use it before launch to catch weak drafts, and after launch as a reminder that the last box is the one top ads tick most often. To see the 235 top-tier ads themselves, filtered by your niche, the Ad Radar board shows each one with its tags, versions and days live. For the testing process around this checklist, see the creative testing framework.

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.