Most lists of Meta ads hooks are lists of sentences. You copy the sentence, it does not work for your app, and you conclude that hooks are a lottery. They are not. What travels between accounts is the pattern underneath the sentence, and there are only about a dozen that matter.
A pattern is a repeatable psychological move. The words change completely for every product, every audience and every angle. The move does not. Learn the twelve and you stop staring at a blank document, because the question changes from "what should this ad say" to "which move have we not made yet".
Below are the twelve patterns we write against, what each one is actually doing to the viewer, an illustrative example for a mobile app, and the creative zone each one tends to occupy. The examples are written to show the shape. They are not taken from client accounts and they carry no result claims.
Key takeaways
- A hook is a psychological move, not a sentence. Copying wording between accounts fails, copying the pattern works.
- Loss framed hooks tend to outperform gain framed ones because losses weigh roughly twice as heavily as equivalent gains (Kahneman and Tversky, 1979).
- Twelve patterns is not twelve zones. A pattern is how you say it, a zone is the position you say it from, and it is the zone that determines whether the algorithm reads two ads as different.
- The permission hook and the negative admission are the two most underused patterns in mobile app advertising, and both work by removing pressure rather than adding it.
- Judge hooks on hook rate, which moves earlier than cost per action, and judge the concept underneath on what happens after the first three seconds.
What Meta ads hooks actually have to do
A Meta ads hook is the first three seconds of the ad: the line, image or moment that decides whether the rest of the ad is watched at all. That is its entire job. It does not have to explain the product, establish the brand or make the offer, and hooks that try to do those things are the reason most app ads lose the viewer before the pitch arrives.
The mechanism is worth being precise about. The viewer is mid scroll with a thumb already moving and sound frequently off. They are not deciding whether to buy, they are deciding whether to keep watching, and they make that decision on recognition rather than on argument. A hook works when it names something the viewer already carries around, which is why the strongest Meta ads hooks are usually about the viewer and not about the app.
Everything after the hook has a structure too, and it matters, but it only gets to matter if the first three seconds earn it. We cover the four beats that follow in our guide to writing UGC ad scripts that convert.
The 12 hook patterns
Each pattern below is one psychological move. The illustrative examples use a handful of imaginary apps, a sleep app, a budgeting app, a fitness app and a language app, so you can see the same patterns applied across categories.
1. The loss frame
Names something the viewer is quietly losing rather than something they could gain. This is the most reliable single upgrade available to most app advertisers, because losses weigh approximately twice as much as equivalent gains (Kahneman and Tversky, Prospect Theory, 1979) and yet the default hook in almost every category promises a gain the viewer has already heard.
Budgeting app: "You are not overspending. You are losing about forty pounds a month to things you forgot you subscribed to."
Zone: urgent valence, higher language intensity, speaks to the actual self.
2. The pattern interrupt
Breaks the visual or verbal prediction the viewer has already made about what an ad looks like. Opening mid sentence, an odd camera angle, an unexpected setting, a first frame that does not look produced. It buys about a second, which is enough for a real hook to land behind it.
Fitness app: opening mid thought, phone propped against a kettle, "so I have got about four minutes before this boils, but".
Zone: this one cuts across all zones. It is a delivery choice rather than a psychological position, and in our experience it also decays fastest, because the audience learns the trick.
3. The negative admission
Concedes something that costs the advertiser to say. It works because a claim that is against the speaker's interest reads as true, and that credibility then transfers to whatever is said next. Most app ads are relentlessly positive, which is precisely why an honest limitation stands out.
Language app: "This will not make you fluent. It made me able to order dinner without switching back to English, which is all I wanted."
Zone: warm valence, conversational intensity, actual self.
4. The specificity hook
Opens on a precise, slightly awkward detail rather than a category statement. Precision reads as lived experience. Round numbers read as marketing and odd ones read as memory, so "eleven times" persuades where "constantly" does not.
Sleep app: "I checked my phone eleven times between one and four in the morning last Tuesday."
Zone: neutral valence, conversational intensity, actual self.
5. The question hook
Asks something the viewer answers internally before they have decided to engage, which turns watching into participating. The failure mode is asking a question that is really a pitch, such as "want to save money on your bills". A question only works if the honest answer makes the viewer slightly uncomfortable.
Budgeting app: "Do you know what you spent last month? Not roughly. The number."
Zone: mid intensity, neutral to urgent valence, actual self.
6. The contrarian claim
Disputes something the viewer believes. Disagreement holds attention longer than agreement does, because the viewer stays to see whether you can justify it. The rule is that you must actually be able to, within the ad. A contrarian claim you cannot defend is a bounce with extra steps.
Fitness app: "Counting your steps is close to the least useful thing you can do for your fitness."
Zone: urgent valence, high language intensity, works on solution aware audiences who already have a habit to defend.
7. The social proof hook
Signals that people like the viewer already do this. The word doing the work is "like the viewer", not "people". Download counts and star ratings are weak proof because they describe a crowd nobody identifies with, while a specific similar person is strong proof.
Sleep app: "Every nurse on my ward has this on their phone" does more than "two million downloads" ever will.
Zone: warm valence, mid intensity, ideal self.
8. The curiosity gap
Opens a small, specific gap between what the viewer knows and what they want to know. Curiosity is the discomfort of an information gap (Loewenstein, 1994), and the important word is small. A narrow, concrete gap pulls. A vague one reads as clickbait and the viewer has been trained to ignore it.
Sleep app: "The reason you wake up at three in the morning has nothing to do with what time you went to bed."
Zone: neutral valence, conversational to mid intensity. Close the gap inside the ad, or retention collapses at the ten second mark.
9. The identity call-out
Names a group the viewer belongs to, so the ad arrives addressed rather than broadcast. It also does useful qualifying work, filtering out people who would install and never convert. The more specific the group, the better it performs, which feels wrong to most advertisers and is consistently true.
Language app: "If you have kept a streak going for over a year and still cannot hold a conversation, this is for you."
Zone: mid to high intensity, ideal self, product aware.
10. Before and during, not after
Shows the middle of the process rather than the finished transformation. The polished after state is unrelatable and reads as advertising, while the messy middle is where the viewer actually is right now. This is one of the easiest patterns to produce and one of the least used.
Fitness app: "Week three. Still hate it. Still going."
Zone: warm valence, conversational intensity, actual self.
11. The permission hook
Releases an obligation instead of adding one. A large part of most audiences is not undermotivated, they are carrying guilt about a standard they have not met, and the entire category is busy adding more pressure to it. Self Discrepancy Theory (Higgins, 1987) separates who you are from who you feel you ought to be, and this pattern works on that gap by shrinking it rather than pressing on it.
Fitness app: "You do not need a five in the morning routine. You need twenty minutes and nobody watching."
Zone: warm reassurance, low language intensity, ought self. The most underused pattern in the list.
12. The cost of inaction
Loss aversion extended across time. Not what the viewer is losing today but what another year of exactly this costs them. It is distinct from the loss frame because it is cumulative and future tense, and it lands hardest in categories where nothing dramatic happens if you do nothing, which is most of health, finance and learning.
Budgeting app: "Another year of this and the gap between what you earn and what you keep will be exactly where it is now."
Zone: urgent valence, high intensity, ought self.
Patterns are not zones, and the difference matters
Here is where most hook libraries go wrong. A team learns twelve patterns, writes twelve hooks, and every one of them is urgent, high intensity and aimed at the ideal self. Twelve different sentences, one psychological position. The audience experiences that as repetition and so does the auction.
A pattern is how you say something. A zone is the position you say it from, and we define a zone with three dimensions: emotional valence from warm reassurance through to urgent warning, the self concept the ad speaks to, which is who you are now, who you want to be or who you feel you should be, and language intensity from conversational through to hard direct response. The full system, and why Meta's retrieval stage treats zone difference as genuine signal difference, is in our guide to Meta ads creative fatigue for mobile apps.
The practical test is a grid. Put the twelve patterns down one side and your zones across the top, then mark what you have actually run in the last ninety days. Most accounts we audit have four or five cells filled and the rest of the grid untouched, and the untouched cells are where the cheap performance is sitting.
Note also that the same pattern moves zone depending on how you write it. A loss frame delivered warmly and conversationally to the actual self is a completely different ad from a loss frame delivered as an urgent warning to the ought self, even though both are pattern one. That is the leverage: twelve patterns crossed with the zones gives you a working library, not a list of twelve ads. The mechanics of loss framing specifically, including where it backfires, are in loss aversion in Meta ads.
How to tell whether a hook is working
Hook rate, three second video views divided by impressions, is the right metric and it moves earlier than cost per action does. Read it at ad level, compare hooks within the same concept rather than across the whole account, and give each one enough spend to mean something before you judge it. We set out the full method in our guide to hook rate as the earliest fatigue signal.
Two diagnostics worth knowing. A high hook rate with weak conversion means the hook is writing a cheque the rest of the ad cannot cash, usually a curiosity gap that never closes or a contrarian claim with no argument behind it. A low hook rate with strong conversion on the people who do watch means the pattern is wrong but the underlying angle is right, which is the easiest fix in performance creative: keep the concept, change the first three seconds.
One more thing on cadence. Hooks wear out at different speeds by category, and in our experience mobile gaming creative fatigues in one to two weeks at scale, health and fitness in three to four, and finance or SaaS in five to eight. Pattern interrupts wear out fastest of all. If you want a refresh rate to plan against at your own spend level, our creative refresh calculator will give you a number.
Frequently asked questions
What is a hook in Meta ads?
A hook is the first three seconds of an ad: the line, image or moment that decides whether anybody watches the rest. It is not the product, the brand or the offer. Its only job is to interrupt a scroll and give one specific person a reason to stay, which is why we treat it as roughly 80 percent of the work in any performance creative.
How long should a Meta ads hook be?
Three seconds, and usually fewer than twelve spoken words. Hook rate is measured on three second video views, so anything that has not landed by that point has effectively not been seen at all. A hook that needs setup is not a hook, it is context that has been put in the wrong place.
What is the best hook for mobile app ads?
There is no single best hook, only the best hook for the psychological position you are trying to occupy. Loss framed hooks tend to outperform gain framed ones because losses weigh about twice as heavily as equivalent gains, but a library built entirely from loss frames fatigues as one thing. The best hook is the one filling a gap in your coverage.
How many hooks should I test on a mobile app?
Test hook patterns rather than hook wordings. Five to eight genuinely different patterns per concept will tell you more than fifty rewrites of the same line, because the algorithm and the audience both read reworded versions of one idea as a single piece of creative. Once a pattern proves itself, then it is worth producing variations inside it.
How do I know if a hook is working?
Hook rate, which is three second video views divided by impressions, is the cleanest read available and it moves before cost per action does. Judge a hook on hook rate and the concept underneath it on what happens after the three second mark. A high hook rate with poor conversion means the hook is writing a cheque the rest of the ad cannot cash.
Want this run for you?
Knowing twelve hook patterns is the easy half. Running them deliberately across every psychological zone, at a production cadence that matches how fast your category fatigues, week after week without quietly collapsing back into the two patterns your team finds most comfortable, is the hard half.
That is what we do as a performance creative agency for mobile apps: map the zones an account is not covering, then write and produce against the gaps. If you would rather have it run for you, apply to work with us. We take a small number of mobile app clients per quarter.