Ad hooks that stop the right people: 10 patterns, the data and the rules

The first line of an ad decides whether anyone reads the second. Here is how presell ads open, what the data says, the line Meta will not let you cross, and tools to write your own.

Illustration of a copywriter selecting and refining an ad opening at a Dubai cafe.
On this page 9 sections
  1. What the data says about presell hooks
  2. Ten hook patterns, with examples
  3. The line you cannot cross
  4. Generate hooks for your angle
  5. Check a hook before you launch it
  6. From hook to page headline
  7. Hooks in video
  8. How to test hooks
  9. Misconceptions about hooks

Nobody reads an ad. People read the first line of an ad, and some of them decide to read the second. That first line, or the first two seconds of a video, is the hook, and it does more work than everything after it combined.

For a presell funnel, the hook has two jobs. It has to stop the right person (the one with the problem your page is about), and it has to set up the page: the reader taps because of a promise, and the page must pick up exactly there.

This guide starts with what 1.29 million Meta ads tell us about how presell ads open, then gives you ten patterns with examples, the rule that gets hooks rejected, and three tools to write and check your own.

What the data says about presell hooks

We compared the first lines of English-language ads that link to presell pages (18,456 ads) with those of ads that link to product pages (519,961 ads) in our research database1. Some opening patterns are much more common on one side than the other:

Interactive

How presell hooks differ

How often each opening appears in presell ads, compared with product-page ads.

More common in presell ads

  • "Stop…" / negative1.8× (4.57% vs 2.52%)
  • First-person story, POV1.6× (4.41% vs 2.73%)
  • Authority1.6× (1.63% vs 1.02%)
  • "If you…" callout1.4× (3.33% vs 2.30%)
  • Question1.3× (5.49% vs 4.36%)

More common in product ads

  • Urgency8× (1.76% vs 0.22%)
  • "New" announcement6.8× (1.78% vs 0.26%)
  • Offer1.8× (5.79% vs 3.25%)
  • "This…" curiosity1.4× (3.62% vs 2.61%)
  • Social proof1.1× (1.89% vs 1.72%)

Share of English-language ads whose first line uses the pattern. Ratios are presell ÷ product in the first tab, product ÷ presell in the second.

The split is clean. Presell ads talk about the reader and the problem: they push back on a belief, tell a story, call out a situation, ask a question. Product ads talk about the deal: what is new, what is discounted, what ends tonight. Urgency is eight times rarer in presell ads, which makes sense: a presell page asks for ten minutes of attention, not a decision before midnight.

Presell hooks are also a little longer: a median of 12 words, against 11 for product-page ads and 10 for collection ads. They can afford it, because they are selling a story rather than a price.

Ten hook patterns, with examples

Every example below is written to stay on the right side of the platform rules covered in the next section. Notice that most of them describe a situation or a belief rather than telling the viewer what is wrong with them.

1. The situation callout

Names a group by what they do or where they are, so the right people recognize themselves.

  • "Side sleepers: this one is for you."
  • "For cat owners in small apartments, the litter box is three meters from the sofa."
  • "If you sleep on your side, read this before buying another mattress."

2. The question about the world

A question the reader has asked themselves, framed as a general puzzle.

  • "Why do tight socks leave marks by 6 p.m.?"
  • "Why does the litter smell come back two days after a full change?"
  • "What angle should a winged liner actually follow?"

3. The "stop" line

Pushes back on a habit. It is the pattern presell ads over-use the most, because it sets up the "real cause" that the page will explain.

  • "Stop buying scented litter."
  • "Stop blaming your mattress."
  • "Don't buy another eyeliner stamp until you know your angle."

4. The first-person story

Starts a story that the page will finish. It works because people follow stories about other people further than claims about themselves.

  • "For six years I blamed my mattress."
  • "I changed the litter every week for two years. The hallway still smelled."
  • "My mother wore a necklace for three weeks that turned her neck green."

5. The wrong diagnosis

States the myth and the correction in one breath.

  • "It's not the mattress. It's the space between the knees."
  • "Puffiness isn't always about salt."
  • "A scent doesn't remove ammonia. It adds to it."

6. The precise moment

A time, a place and a detail. Specifics read as truth.

  • "7:10 a.m., puffy eyes. Gone by lunch. Every single day."
  • "9 p.m., the sock lines that take an hour to fade."
  • "Friday, 8 p.m., guests at the door. The litter box is by the entrance."

7. The number

Promises a finite, skimmable payoff. On presell pages, "N reasons" is the most common title pattern of all.

  • "6 mistakes that bring the litter smell back, even with a clean box."
  • "3 things a mirror tells you about your eyeliner angle."
  • "8 weeks, 5 criteria, 3 ways to get a full-body red light session."

8. The real authority

Borrows trust from a profession, but only when a real expert stands behind it.

  • "A physical therapist's first question about morning stiffness: how do you sleep?"
  • "What vets look at first when the litter box smells."

9. The discovery

For readers who are not looking for anything yet. Starts far from the product.

  • "A thousand-year-old habit for the first glass of water of the day."
  • "The bathroom plastic nobody counts: deodorant tubes."

10. The objection, admitted

Says the skeptical thought out loud before the reader does.

  • "I didn't believe a pillow between my knees would change anything."
  • "Yes, it costs more than supermarket litter. Here is the math per month."

The line you cannot cross

Many brands write their best hook, launch it and get it rejected. The usual reason on Meta is the personal attributes policy. Ads must not assert or imply personal attributes of the viewer: race or ethnicity, religion or beliefs, age, sexual orientation, gender identity, disability, physical or mental health (including medical conditions), vulnerable financial status, voting status, trade union membership, criminal record or name2.

Meta's own examples make the logic clear. "New diabetes treatment available" is allowed; "Do you have diabetes?" is not. "Meet seniors" is allowed; "Meet other seniors" and "Are you 18 years old?" are not. The product can be described; the viewer cannot be diagnosed.

For presell hooks, where callouts and questions are the bread and butter, that means writing about the situation instead of about "you":

Risky: implies something about the viewerSafer: describes the situation
"Do you wake up with back pain?""Why so many side sleepers wake up stiff"
"Over 50? Your skin needs this.""A body oil made for skin after 50"
"Your puffy eyes are telling you something.""Puffy eyes at 7 a.m., gone by noon: a common pattern"
"Struggling with anxiety at night?""What a 10-minute evening routine changed for one reader"

Generate hooks for your angle

Fill in your angle and get ten hooks, one per pattern. Keep the reader field descriptive ("side sleepers", "cat owners", "runners"), never a health condition:

Interactive

Hook generator

One angle, ten patterns. Change any field and every line updates.

  1. Callout{reader|cap}: this one is for you.
  2. QuestionWhy do so many {reader} blame {wrong}?
  3. StopStop blaming {wrong}.
  4. StoryFor years I blamed {wrong}. It was {real}.
  5. MythIt's not {wrong}. It's {real}.
  6. Moment{moment}, wide awake again. {wrong|cap} wasn't the problem.
  7. Number{number} things {reader} try before they look at {real}
  8. AuthorityWhat a {expert} checks before blaming {wrong}
  9. DiscoveryThe {product} that made {reader} stop blaming {wrong}
  10. ObjectionI didn't believe a {product} would change anything. I was wrong.

Templates are starting points. Rewrite each one in your customers' own words, and only keep the authority line if the expert is real.

Check a hook before you launch it

The scorer reads the first line of your ad the way a scrolling thumb does: is it short enough to show before "See more", does it open with a pattern that stops the scroll, is it specific, does it avoid hype and shouting?

Interactive

Hook scorer

A heuristic. It cannot see your image, and it does not know your audience: your test results always win.

The scorer rewards "you" in the first line, but it is not mandatory: a first-person story often stops the scroll better than a direct address, and it never risks implying anything about the viewer.

The headline generator turns your product and problem into page headlines for each awareness level.

From hook to page headline

The hook is half of a pair. The other half is the first line of the landing page, and the two must read as one movement: the hook opens a loop, the headline confirms it and promises to close it.

The hook (ad)The headline (presell page)
"Stop blaming your mattress.""For six years I blamed my mattress. The problem was the empty space between my knees."
"Why does the litter smell come back two days after a full change?""6 mistakes that bring the litter smell back in an apartment, even with a clean box"
"Same goal, three ways: spa LED bed, panel or mat?""Red light mat, LED bed or panel: 8 weeks, 5 criteria, one verdict"

Then check the headline on its own. A good presell headline is specific, names the reader or the situation, opens a loop and points to a concrete result:

Interactive

Headline scorer

To check the pair together, use the message match scorer.

Hooks in video

In video, the hook is the first two or three seconds: what the viewer sees, the words on screen and the first sentence spoken. The same ten patterns apply, with one addition: the visual has to make the same promise as the words. A "stop blaming your mattress" hook works better over a shot of someone sitting on the edge of a bed at dawn than over a product turning on a white background.

Video is not a requirement for presell funnels, though. In our data, 54% of presell ads use a single image, while 62% of product-page ads are videos. A strong photo with a strong first line is the standard format here.

Tag each hook so you can read results by ad: the UTM link builder keeps the names consistent.

How to test hooks

  1. Write twenty hooks per angle. Use the ten patterns, two each. The first five will be obvious; the good ones usually come after.
  2. Keep the five most different. Testing five versions of the same idea teaches you nothing.
  3. Change only the words. Same image or video, same page, same audience. Otherwise you cannot tell what made the difference.
  4. Read the funnel, not only the click. Click-through rate tells you which hook stops people; cost per purchase tells you which hook stops the right people. A provocative hook can win on clicks and lose on sales.
  5. Iterate around the winner. If the "wrong diagnosis" hook wins, write ten more wrong-diagnosis hooks for the next round.

A better hook on the same budget can change the economics of a whole campaign: our angles guide includes a calculator that shows what a higher click-through rate is worth.

Misconceptions about hooks

Myth"The more shocking the hook, the better."
Reality

Shock stops everyone, including the wrong people, and the page then has to disappoint them. The best hooks are specific rather than loud.

Myth"Hooks need emojis and capital letters."
Reality

Shouting reads as an ad, and an ad is exactly what people scroll past. Calm, specific sentences look like something a person wrote.

Myth"A question hook should ask about the viewer's problem."
Reality

That is where most rejections come from. Ask about the situation ("Why do sock lines last for an hour?"), not about the viewer's body or health.

Myth"Once a hook wins, run it forever."
Reality

Hooks wear out as the audience sees them. Keep a pipeline of new hooks on the same angle, and rotate before the cost per purchase climbs.

The best hooks rarely come from a brainstorm. They come from customers: the sentence a reviewer used to describe their mornings, the belief they held before, the moment they noticed the problem. Collect those sentences, and the hooks write themselves.

Frequently asked questions

What is an ad hook?

The hook is the part of the ad that makes someone stop scrolling: the first line of the text, the first seconds of a video, or the headline on the image. Everything else in the ad only works if the hook did its job.

How long should a hook be?

Short enough to be read before the text is cut. In most feeds only the first line or two show before "See more", which is why about 125 characters is the common guideline. Presell hooks in our data have a median of 12 words.

Can I say "you" in an ad?

Yes, as long as the ad does not assert or imply a personal attribute of the viewer, such as a health condition, age, religion or financial situation. "Why side sleepers wake up stiff" is fine; "Do you have back pain?" is the kind of line Meta rejects.

Should the landing page headline repeat the hook?

It should reuse its key words and promise, then go one step further. The visitor needs a second of confirmation that they are in the right place. See message match.

How many hooks should I test?

Write around twenty per angle, keep the five most different, and test them with the same image or video so that only the words change. Keep the winners and write the next batch around them.

Sources

  1. Meta Ad Library, Meta. Public source of the ads in our research database
  2. Privacy Violations and Personal Attributes, Meta Transparency Center
  3. Health and Wellness, Meta Transparency Center (updated July 2026)

ClickBanger Editorial We build presell pages for e-commerce brands and spend our days reading ads, landers and test results. Questions or a correction? Write to us.

Keep reading

All articles

Strategy11 min read

Message match: connect the ad promise to the landing page

Message match keeps the ad promise clear on the page it opens. Compare the claim, item, reader and offer directly. A fast exit may flag a problem, but it does not tell you its cause. Check the received page before you change the message or judge the result.

From product URL to presell page in minutes

Advertorial, listicle, quiz, comparison and 12 more formats, written from your product page and ready to edit.