36 ad hooks, seen, read and heard
Most people decide in the first second or two whether to keep watching an ad. That moment is the hook, and it has three parts: what they see, what they read on screen and what they hear. Below are 36 hooks shown the way they appear in the feed, across many product types, with formulas you can adapt and a simple way to test them.
Why the first three seconds decide
In the feed, your ad competes with everything else on the phone. Viewers don’t decide to skip it; they just keep scrolling unless something stops them. The share of people who keep watching past the first few seconds is often called the hook rate (3-second video plays divided by impressions). A weak hook means that even a great product demo in second eight is never seen.
Where viewers leave
Share of viewers still watching, second by second.
The hook has one job: make the right person curious enough to see the next sentence. The body of the ad does the selling. And a hook only works if the avatar and angle behind it are right, which is why it’s one part of our creative strategy, not the start of it.
The three layers of a hook
Seen, read, heard
What happens in the first two seconds of a video ad.
- The first frame: what they seeMovement, a surprising object or a face with a clear emotion. It has to stop the thumb before a single word is read. Here: a heavy pan hitting concrete.
- The text on screen: what they readMany people watch without sound. Five to eight words, big, high contrast, placed where the platform’s buttons don’t cover it.
- The spoken line: what they hearThe first sentence out loud. It can be longer than the text and add a detail, but it makes the same promise.
The best hooks work with the sound off and on. Write the on-screen text so it stands alone for silent viewers, and let the spoken line add a detail for those listening.
The hook library
Each hook shows the first frame (🎬), the text on screen and the spoken line. Filter by type to compare hooks that work the same way.
36 hooks, as they appear in the feed
Tap a type to filter.
12 hook formulas
Fill in the blanks with your product, your avatar and their words. The best versions sound like something a person would say, not like an ad.
| Formula | Example |
|---|---|
| [Specific person], watch this | Runners with bad knees, watch this |
| Why do you [frustrating thing]? | Why do you wake up tired after eight hours? |
| Stop [common habit] | Stop washing your face twice a day |
| [Common belief] won’t [result they want] | More cushioning won’t save your knees |
| I changed one thing and [result] | I changed one thing and she stopped pulling |
| Don’t buy [category] until you [check this] | Don’t buy a laptop backpack until you check the zipper |
| [Cheap option] vs. [your option], [time] later | €30 pan vs. €120 pan, two years later |
| [Number] signs your [thing] is [problem] | Five signs your coffee went stale |
| Day 1 vs. day [number] | Day 1 vs. day 30 |
| We made it for [group]. [Other group] bought it. | We made them for hikers. Nurses bought them. |
| POV: [the result, as a moment] | POV: your leggings stay up the whole workout |
| Things only [group] understand | Things only 5am runners understand |
Match the hook to what they already know
The right hook depends on how aware the viewer is of their problem and of you. A product-aware viewer wants proof; an unaware viewer needs a story or a moment they recognize. More on awareness in our advertising psychology guide.
| Awareness | Hooks that fit | Example |
|---|---|---|
| Unaware | Identity, story, curiosity | Things only 5am runners understand |
| Problem aware | Problem, question, callout | My back at 3pm |
| Solution aware | Myth bust, comparison, warning | More cushioning won’t save your knees |
| Product aware | Demo, result, social proof | 1 hour underwater |
| Most aware | Offer | Buy 2. The 3rd is on us. |
How to test hooks
One body, five hooks
Test the hook on its own, then build on the winner.
- Keep the body the same. Edit three to five different openings onto the same 20 to 30 seconds. Then the only difference is the hook.
- Judge the hook by hook rate, the ad by what it sells. A hook that stops people but brings the wrong buyers is not a winner. Check hold rate and cost per purchase too.
- Test different types, not different words. A callout against a myth bust tells you more than two versions of the same callout.
- Reuse the winner. A hook that works can open new bodies, new creators and new formats. It’s often the most reusable part of an ad.
We cover how to set up the tests in the account in our Meta ads account structure guide, and 43 full ad concepts to put the hooks into in the ad concepts library.
Common mistakes
- Starting with a logo, an intro or a slow pan. The first frame has to already be the hook.
- Text on screen that says something different from the voice.
- Text placed where the caption or buttons cover it.
- A hook that promises something the ad never delivers. It raises hook rate and kills trust.
- Clever hooks that the target buyer doesn’t recognize themselves in.
The short version
- The hook has one job: make the right person curious enough to see the next sentence.
- A hook has three layers: the first frame, the text on screen and the spoken line. All three make the same point.
- Write the on-screen text so it works with the sound off.
- Match the hook type to how aware the viewer is.
- Test hooks on the same body and judge them by hook rate and by what they sell.
FAQ
What is a hook in a video ad?
The first one to three seconds of the ad: the first frame the viewer sees, the text on screen and the first spoken line. Its job is to stop the right person from scrolling and make them curious enough to keep watching.
What is the difference between a visual hook and a spoken hook?
A visual hook is what the viewer sees: the first frame, the movement and the text on screen. A spoken hook is the first sentence said out loud. Many people watch without sound, so the visual hook has to work alone, and the spoken hook adds a detail for those listening.
What is a good hook rate?
Hook rate is 3-second video plays divided by impressions. It varies a lot by platform, placement and audience, so compare hooks against each other in the same account rather than against a general benchmark.
How many hooks should I test per ad?
Three to five different hook types on the same body is a good start. Keeping the body the same means any difference in results comes from the hook.
What makes a hook work?
It speaks to a specific person, shows or says something they recognize or didn’t expect, and makes a promise the rest of the ad keeps. Movement in the first frame and big, short text on screen help it work with the sound off.
Should the on-screen text match the voiceover?
It should make the same point, but it doesn’t have to be the same words. The text should be short enough to read in a second; the voice can be a full sentence that adds a detail.