The short answer
Warning and shock openers are only 6.4% of 2,443 winning Meta ads, but they carry an above-average median estimated spend ($250K) and reach the top tier at 10.3%. Warnings aimed at solution-aware buyers reach the top tier 16.3% of the time; warnings that open on the viewer's own symptom ("If your...") put 0 of 12 there.
How many winners open on a warning
Warning and shock openers are a minority on the Ad Radar board. On October 9, 2026, 156 of 2,443 live Meta ads (each with $50k+ in estimated spend) opened this way, from 79 brands.
156
winning ads open on a warning or shock line
10.3%
reach the top tier (board: ~10%)
$250K
median estimated spend (board: ~$223K)
92%
are video
"Top tier" is the top 10% of the board by winner score, among ads live 21 days or more. Warnings sit right at the average on that measure but above it on estimated spend, and they run long: the median warning ad has been live 129 days. Pet brands use them most (40 of the 156), followed by joint, skin, cleaning and wellness products.
Who the warning is aimed at changes everything
Split by the reader's awareness level (the Schwartz scale we tag every ad with), warnings behave very differently.
| Audience awareness | Warning ads | Reach top tier | Median est. spend |
|---|---|---|---|
| Unaware | 28 | 10.7% | $250K |
| Problem-aware | 41 | 7.3% | $347K |
| Solution-aware | 43 | 16.3% | $272K |
The best results come from warnings to solution-aware buyers: people who already use some fix and are told it's the wrong one. That is a warning with somewhere to go. The viewer is already spending money on the problem, so "you're doing it wrong" is news.
Warnings to problem-aware buyers spend more at the median but reach the top tier less often. These are mostly openers that describe a symptom the viewer already knows about, which brings us to the weakest group.
The warning that diagnoses the viewer
Twelve warning ads open with "If you..." or "If your..." followed by a symptom or a situation. None of them reaches the top tier.
That shape has two problems. It tells the viewer something they already know (they have the symptom), so there's no new information. And in health it brushes against Meta's personal attributes policy, which says ads must not assert or imply knowledge of a person's physical or mental health, including medical conditions. Ads that imply "we know you have this" risk rejection.
The warnings that work point at an object, a habit or a category, not at the viewer's body.
So you've started baking your own bread.
- Est. spend
- $988K
- Days live
- 334
- Format
- Video, 52s
The opener congratulates the viewer on a habit. The warning comes in line two ("Please do not store it in plastic"), aimed at a storage choice, not at the person. Any health claims later in the script are the advertiser's responsibility.
Four warning shapes on the board
The test that exposes an industry
The speaker runs an experiment and the result is the warning.
So basically, I took the same car, same problem, three different mechanics, and I let them tell me what was wrong.
- Est. spend
- $627K
- Days live
- 117
- Format
- Video, 76s
No threat is stated. The setup implies it: three mechanics will give three answers. "And I filmed it because I knew" turns the viewer into a witness.
The same advertiser's bigger ad opens on the threat directly: "This one easy mistake will cost you thousands when buying a used car." (est. $3.27M, 116 days, top tier, 8 variants). Both work for an unaware audience because the cost is money, not health.
The reverse warning
The line sounds like a warning and turns out to be a boast. It borrows the attention of a warning without the risk of fear content.
Do not buy Black Girl Vitamins Collagen Powder if you don't want these side effects.
- Est. spend
- $1.3M
- Days live
- 110
- Format
- Video, 66s
- Variants
- 2
The "side effects" are the benefits the creator lists next. The joke only lands because the opener sounds like a real warning for one second. Both "Do not buy" openers tagged as warnings reach the top tier, but two ads is not a sample.
The witnessed threat
Someone else got hurt, and the speaker was there. Travel Insider opens "She watched TSA empty her entire bag." (est. $354K, 70 days), and the story that follows is the warning. Across the board, cautionary-tale ads, built on what went wrong for someone, reach the top tier 14% of the time (129 ads), well above average. More in cautionary tale ads.
The insider's warning
A professional tells you what they see every day. Uproot Clean's opener ("There is something living inside your washing machine right now, and after 19 years fixing them, I can tell you no cleaner on that shelf will touch it.") is the clearest example on the board, and we break it down in the hooks pillar. The credential sits in the same sentence as the threat, which is what makes it believable.
How far is too far
Two of Meta's Advertising Standards set the outer limits:
- Shocking or sensational content. Meta's violent and graphic content policy says ads must not contain shocking, sensational or excessively violent content. Gross-out close-ups and disaster imagery belong here.
- Personal attributes. No asserting or implying the viewer's health, finances or other attributes. "Your kidneys are failing" is out. "Most people store bread wrong" is in.
Within those limits, the data points to a third limit that is about performance, not policy: proof.
| Proof level (warning ads only) | Ads | Reach top tier | Median est. spend |
|---|---|---|---|
| Strong | 12 | 25.0% | $489K |
| Medium | 90 | 12.2% | $298K |
| Unproven | 47 | 12.8% | $247K |
A warning backed by strong proof (a demonstration, a test, a credential) reaches the top tier a quarter of the time in this small sample, with a median estimated spend more than double the board's. A warning is a claim, and the viewer discounts it until the ad shows something.
How to write a warning hook that holds up
- Aim at a habit, an object or a category. "Your washer", "your bread bag", "used cars". Not "your body" or "your condition".
- Prefer solution-aware buyers. Telling people their current fix is wrong is the strongest use of a warning on our board (16.3% top tier).
- Put the credential next to the threat. "After 19 years fixing them" in the same sentence.
- Show proof within seconds. A test, a close-up, a side-by-side. Warnings with strong proof do best.
- Make the cost money or time where you can. Financial warnings ("will cost you thousands") avoid most of the policy risk of health warnings.
- Try the reverse warning. "Do not buy X if you don't want..." gets the stop of a warning with a positive payoff.
- Keep claims checkable. If a warning states a fact about toxins, bacteria or risk, you need to be able to source it. That responsibility sits with the advertiser.
For openers that flip a belief rather than raise a threat, see contrarian hooks. To find every warning opener running in your niche with its estimated spend, you can filter the Ad Radar board by hook pattern.
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.