How to Write Ad Hooks That Stop the Scroll

I've reviewed thousands of ad accounts over the past eight years. The single most common reason a creative underperforms is not the offer, not the targeting, not the landing page. It's the first three seconds.

If you don't stop the scroll, nothing else you built matters. The body copy, the product benefits, the testimonials, the CTA — none of it gets seen. The viewer is already on the next video.

The hook is the most leveraged variable in performance creative, and it's consistently the least systematically developed. Most brands treat it as the last thing they write rather than the first thing they engineer. That backwards approach is costing them real money every day their ads run.

Here's how I think about hook development at Impremis and the framework I use to build hooks that earn the watch and then earn the sale.

Image brief: Five horizontally-arranged labeled cards (Pattern Interrupt, Direct Problem, Credential, Curiosity Gap, Outcome Lead) with one-line examples. alt: "Five ad hook frameworks shown as cards." caption: "The five frameworks that show up in winning creative across accounts."

Why the hook matters more than anything else

The attention economics of social media have compressed to the point where you're competing not just against other advertisers but against everything else in a person's feed simultaneously — friends, family, entertainment, news, creators they follow. Your ad is asking someone to pause all of that for you. The hook is the argument for why they should.

On TikTok, the average viewer's decision about whether to keep watching happens within the first one to two seconds. On Meta Reels and Instagram Stories, similar. Even on Facebook Feed, where video autoplay has slightly more momentum, you have three seconds at most to give someone a reason to stay.

The implications are significant. A hook that lifts thumb-stop rate from 20% to 35% means 75% more people are watching your message. If downstream conversion rate holds, that's 75% more conversions from the same spend without changing anything else in the creative. No new offer. No landing page rebuild. No targeting adjustment. Just a better hook.

That's why hook testing comes first in our creative testing system. It's the highest-leverage investment you can make in a creative program.

The five hook frameworks that consistently perform

There's no single hook formula that works across all products, all audiences, and all platforms. There are five structural frameworks that show up in winning creative across the accounts I run consistently. Understanding the mechanism behind each one tells you when to deploy it and why it works.

Framework 1: The pattern interrupt

A pattern interrupt stops the scroll by presenting something unexpected in the first frame — a visual surprise, an unusual statement, a counterintuitive claim, or simply an opening that doesn't look or sound like an ad.

The mechanism is attention capture through novelty. The brain is wired to notice things that break expected patterns. When your ad opens with something the viewer didn't anticipate, their brain pauses the scroll while it processes what it's seeing.

The strongest pattern interrupts on TikTok tend to be visual or audio — an unusual setting, an unexpected sound, a reaction shot conveying strong emotion before context is established. On Meta Feed, where video competes with static content, a bold graphic or striking opening frame can function as the pattern interrupt before the audio registers.

The risk: the interrupt captures attention without connecting to the product. Hooks that are surprising but irrelevant produce high thumb-stop rates with poor downstream conversion. The interrupt has to flow into the message without a jarring transition.

Framework 2: The direct problem statement

"If you're struggling with [specific problem], keep watching."

This works through immediate identification. The viewer who has the problem recognizes themselves in the first sentence and continues watching because the content feels personally relevant. The viewer who doesn't have the problem scrolls — which is the correct outcome. You want to stop the right scrolls, not all of them.

Specificity is what determines performance. "If you have trouble sleeping" is weak — too broad to feel personal. "If you're waking up at 3 a.m. and can't get back to sleep" is strong because it describes an experience precisely enough that the person who has it feels seen.

Vague problem statements create vague identification. Precise ones create strong identification with the subset of your audience who actually have the problem — and those are the people most likely to convert.

Framework 3: The credential or social proof open

"I tried 14 different [category] before I found this one." "We tested $200K of ad spend to find the supplement that actually works." "47,000 customers switched in the last 90 days."

This framework leads with a signal of credibility before making any product claim. The mechanism is trust establishment at the earliest point in the attention window. Instead of asking someone to trust a brand claim, you offer evidence that other people have already done the evaluation work.

The specific number matters. "Thousands of customers" is weak. "47,000 customers" is strong. "We tested a lot of supplements" is weak. "We tested $200K of ad spend" is strong. Precision signals the claim is real rather than approximate.

This framework performs particularly well in high-skepticism categories where trust is the primary conversion barrier — supplements, skincare, health products, financial services.

Framework 4: The curiosity gap

"The reason your [category] isn't working has nothing to do with [expected cause]." "Most people don't know this about [familiar product or behavior]." "We weren't supposed to share this, but here it is."

The curiosity gap creates tension between what the viewer knows and what they suspect they're about to learn. The mechanism is information-driven. Once a question is implied, the brain wants the answer. Stopping the video before getting it creates cognitive discomfort that keeps people watching.

The key is establishing that the gap is relevant to something the viewer already cares about. A curiosity gap about something unfamiliar produces mild interest. A curiosity gap about something the viewer uses, does, or experiences daily produces strong engagement because the stakes of the unknown answer feel personal.

This framework has the highest ceiling of all five when it works and the lowest floor when it doesn't. If the gap isn't compelling or the payoff doesn't deliver, the viewer feels manipulated and you've lost them permanently. The promise the hook makes has to be fulfilled in the body of the ad.

Framework 5: The outcome lead

"I lost 22 pounds in 11 weeks without changing my diet." "We grew from $0 to $40K/mo in our first year using this." "My skin cleared up in 10 days after trying everything for three years."

Leading with a specific, concrete outcome before introducing the product puts the result in the viewer's mind before the skepticism machinery has time to engage. The mechanism is aspiration combined with specificity. Vague aspirational hooks ("get the body you want") produce almost nothing. Specific outcome hooks describing a result the viewer actually wants create immediate interest in how the result was achieved.

Credibility matters. An outcome that seems too good to be true triggers skepticism rather than aspiration. The best outcome hooks describe results that are meaningful but believable, specific but achievable. "I lost 8 pounds in 6 weeks" is more credible than "I lost 50 pounds in a month" even if the latter is technically true.

Hook frameworks by platform and format

Different platforms have different attention dynamics. A hook that works in Meta Feed doesn't automatically translate to TikTok or YouTube pre-roll.

| Hook framework | Meta Feed | Meta Reels | TikTok | YouTube pre-roll | Best product category | |---|:---:|:---:|:---:|:---:|---| | Pattern interrupt | Strong | Very strong | Very strong | Strong | Any, especially impulse | | Direct problem statement | Very strong | Strong | Strong | Very strong | Health, wellness, productivity | | Credential / social proof | Very strong | Strong | Strong | Medium | High-skepticism categories | | Curiosity gap | Strong | Very strong | Very strong | Medium | Information products, supplements | | Outcome lead | Very strong | Strong | Very strong | Strong | Results-driven categories |

Meta Feed tends to favor more direct hooks because viewers are often in a lower-engagement browsing mode. TikTok rewards pattern interrupts and curiosity gaps because the platform is primarily entertainment-first. YouTube pre-roll has the shortest effective window of any format — the viewer's intent is to get to the content behind the ad, not to watch your ad. Direct problem statements that feel immediately relevant are your best shot at earning the watch there.

The hook development process

Writing hooks systematically is different from writing hooks by intuition. The process I run before any major creative production cycle:

1. Map your audience's existing conversation. Before writing a single hook, collect the language your customers use when describing their problem and their desired outcome — reviews, Reddit threads, comments, customer service tickets. The vocabulary your customers use about their own experience is the most powerful hook material available. When a hook uses the exact words a viewer would use, the identification is immediate and visceral.
2. Build an angle matrix. List every distinct angle you can take to enter the conversation: problem-led, outcome-led, curiosity-led, social-proof-led, identity-led. For each angle, write three to five hook variations. You should have 15–25 hook options before you start narrowing.
3. Apply the three-second test. Read each hook out loud and stop after three seconds. Does it give a compelling reason to keep listening? Does it create a question in the viewer's mind they want answered? Does it describe something they recognize or want? If no on all three, it doesn't make the cut.
4. Prioritize for testing. Select your top six to eight across different frameworks. These become the first round of your creative test. Hold the rest in the hook library for subsequent rounds.
5. Isolate the variable. Same visual format, same body copy structure, same offer, same CTA. If you change the hook and the background music and the talent simultaneously, you can't attribute the performance difference to the hook. One variable. Clean test.

The metrics that tell you if your hook is working

Hook performance is measurable before conversion data stabilizes. The signals I monitor in the first 48–72 hours of a new creative:

• Thumb-stop rate (also called video play rate). The percentage of people who saw the ad and watched more than two to three seconds. Below 25% means the hook isn't earning the pause.
• 3-second video view rate. A strong hook should hold 30%+ of viewers past the three-second mark.
• Hook-to-hold ratio. Percentage who watched past three seconds versus percentage who watched 25%+ of the video. A large drop-off means the hook stopped the scroll but the body didn't deliver on what the hook promised. The hook and the body need to be connected by a logical thread that rewards the viewer for staying.
• Cost per landing page view (for click-based creative). If CTR is up but CPLPVs aren't improving, the hook may be attracting the wrong audience — curious but not qualified.

The CEO-level layer: hooks at scale

Hook development is one of the highest-return investments in a creative process. Not because a good hook magically fixes underperforming ads, but because hook testing surfaces learnings that are transferable across the entire creative program.

When you discover that problem-statement hooks consistently outperform outcome hooks for a specific brand's audience, that insight changes how every brief is written going forward. When you learn that a particular anxiety or desire drives the strongest response, that becomes the strategic center of the brand's messaging architecture across all channels — email, landing pages, organic social, paid.

Hooks are also the most scalable part of creative production. Producing six hook variations for a single ad concept costs a fraction of producing six entirely different ads. The creative leverage is significant.

The brands that build systematic hook libraries and test at volume produce creative programs that improve continuously rather than episodically. Each cycle feeds the next. Each winning hook becomes a model for the next generation of variants. That compounding is how creative strategy becomes a durable competitive advantage instead of an occasional win.

FAQ

How many hooks should I test per concept? Six to eight. Fewer leaves angles uncovered. More dilutes the spend per variant below significance.

Can I reuse winning hooks across products in my catalog? Sometimes. A winning framework (e.g. credential opens) almost always ports. A specific phrasing rarely does. Treat the framework as the asset, not the line.

What's a strong thumb-stop rate to target? 30%+ on Meta Feed, 35%+ on Reels, 40%+ on TikTok. Below those, the hook is the bottleneck.

Should the hook tease the offer or the outcome? Outcome usually beats offer in cold traffic. Offer tends to win in retargeting where intent is already established.

Closing

Your hook is the most leveraged element in your entire creative program. It determines whether anyone sees the rest of what you built. It sets the expectation for what follows. And it's the most testable, most iterable part of your ads.

Stop treating it as the last thing you write. Start treating it as the first thing you engineer.

Build a hook library. Test systematically. Track thumb-stop rate alongside conversion data. Feed what you learn back into the next production cycle.

The brands that do this consistently build creative engines that find winners reliably. The ones that don't are hoping a good ad shows up by accident.

Stop hoping. Start engineering.

Keep reading

Pieces I've written on related topics that pair well with this one:

• The Scroll-Stop Audit: Diagnosing Why Creative Doesn't Convert — Learn how to diagnose creative performance using the Scroll-Stop Audit framework to identify where ads fail and systematically improve hooks and conve…
• I Don't Analyze Losing Ads. Here's the System I Use Instead — Most creative analysis is just noise wearing a lab coat. The system I run on $250M+ in spend only mines winners, not individual losers.
• UGC Is Dying. Creator Communities Are Eating Its Lunch — How a 500-person creator army drove $700M+ in revenue, and how an 8-figure beauty brand we work with adapted the same model for paid social.
• Your Best Top-of-Funnel Ad Won't Look Like Your Brand — The ads that actually scale on Meta open with a problem, not a logo.
• Your Creatives Think the Buyer Is Killing Their Best Work — The fight between creative teams and media buyers is the most expensive unresolved conflict in paid media. Here is how to settle it.