Video Ad Hooks: 12 Openers That Stop the Scroll (With Examples)

A video ad is won or lost in the first two seconds, and the hook is the line that buys the next eight. The best hooks do one of three things fast: name a problem the viewer has, promise a specific payoff, or open a curiosity gap they need closed. Below are twelve hook patterns that stop the scroll, the exact opening lines they use, and why each works. The single most important rule underneath all of them: test many hooks against the same body, because you cannot predict which one your audience responds to, and the winner is often the one you liked least.

Most ads do not fail because the product is wrong or the offer is weak. They fail because nobody watched past the first frame. On a feed that autoplays and scrolls in under a second, the hook is not the intro to your ad, it is the entire audition. Get it right and the rest of the ad gets a chance. Get it wrong and your production budget, offer, and targeting never matter. These twelve patterns are the ones that reliably earn that chance.

1. The direct problem call-out

Name the exact problem your best customer has, out loud, in the first line. It works because the right viewer feels seen and the wrong viewer scrolls, which is exactly the filtering you want. "If your bank statements come as locked PDFs, this is for you" stops the people who have that problem and nobody else.

Opening line pattern: "If you [specific painful situation], watch this." The more specific the situation, the sharper the stop.

2. The bold claim

Lead with the payoff, stated as a confident, specific promise. This works when the claim is concrete enough to be interesting and true enough to defend. "I cut my ad production time from a week to an afternoon" beats "save time on ads" because the number makes it real.

Opening line pattern: "This [did specific result] in [specific timeframe]." Only make a claim you can support, since an inflated one costs you trust the moment the viewer feels it.

3. The curiosity gap

Open a loop the viewer needs closed. State something surprising or incomplete that only makes sense if they keep watching. "The reason your ads stop working after two weeks has nothing to do with your budget" forces the question and buys the next few seconds.

Opening line pattern: "The real reason [common belief] is wrong is..." Then actually pay it off. A curiosity hook that does not deliver trains people to scroll past you next time.

4. The negative or contrarian take

Go against the obvious advice. Contrarian openers stop the scroll because they violate an expectation. "Stop testing new audiences, your problem is the creative" lands harder than agreeing with what everyone already believes.

Opening line pattern: "Stop [common advice]. Do this instead." Use it when you genuinely have the better take, not as a gimmick.

5. The before-and-after tease

Show the transformation in the first frame, before you explain anything. Visual change is the fastest hook there is because it needs no words. The messy desk cutting to the clean one, the locked file becoming a spreadsheet, the tired parent to the sleeping baby.

Opening line pattern: no line needed, the cut is the hook. If you add words, make them "watch what happens when..." and let the visual carry it.

6. The question hook

Ask the exact question your customer is already asking themselves. A well-aimed question feels like the ad is reading their mind. "How much are you actually losing to manual data entry every month?" pulls in anyone who has wondered.

Opening line pattern: "How much [cost the viewer feels] are you [paying / losing / wasting]?" Keep it pointed at a real, felt cost, not a rhetorical throwaway.

7. The social proof open

Lead with the fact that other people already chose this. Numbers and specificity make it credible. "Fourteen thousand DTC brands switched to this last year" works when the number is real and relevant to the viewer.

Opening line pattern: "[Specific group] already [made this choice], here is why." Never invent the number. A fabricated proof point is the fastest way to lose the exact trust you were trying to borrow.

8. The demonstration open

Just start doing the interesting thing. No intro, no setup, the product mid-action from frame one. This works for anything with a satisfying or surprising mechanism, because showing beats telling and it needs no verbal hook to land.

Opening line pattern: "Watch this." Then the demo. The restraint is the point. Every second of preamble is a second the viewer uses to leave.

9. The relatable confession

Open with a small, honest admission your audience shares. Vulnerability reads as human, which is disarming on a feed full of polished ads. "I ignored this for way too long because it seemed like a hassle" invites the viewer who has done the same.

Opening line pattern: "I put off [thing] for [too long] because [honest reason]." The honesty has to be genuine, because a fake confession reads as a script.

10. The price or number hook

Lead with a striking number, cost, saving, or comparison. Numbers cut through because they are concrete and fast to process. "This costs less than one freelance creator and makes forty ads" front-loads the value.

Opening line pattern: "[Specific number] versus [the alternative]." Make the comparison honest and current, especially any price you name against a competitor.

11. The urgency or timing hook

Tie the message to a moment that makes it relevant right now: a season, a deadline, a change the viewer is facing. "If you are planning Q4 ads, do this before you spend a dollar" earns attention from anyone in that window.

Opening line pattern: "Before you [imminent action], watch this." Only use real urgency. Manufactured countdowns train people to ignore you.

12. The pattern interrupt

Open with something visually or verbally unexpected that breaks the scroll rhythm. An odd first frame, an abrupt statement, a motion the eye did not predict. It works because the brain notices anything that does not match the pattern of the feed it was skimming.

Opening line pattern: whatever the viewer did not expect, kept relevant to the product. Interrupt for a reason, not just for shock, or you get the attention and waste it.

How do you write a good video ad hook?

Write a good video ad hook by making the first line do exactly one job: name a problem the viewer has, promise a specific payoff, or open a curiosity gap. Keep it to one clear idea, put it in the first two seconds, and make it visual wherever you can, since a hook that shows beats one that only tells. Then write several different hooks for the same ad, because you cannot reliably guess which one your audience will respond to.

The mistake most advertisers make is polishing one hook instead of testing ten. The winning opener is frequently the one the team liked least, which is why testing beats taste. If you are short on angles, running the product through a quick idea-generation pass to produce dozens of hook variants gives you more to test than you would write by hand.

How many hooks should I test?

Test at least three to five distinct hooks per ad concept, and more if your budget allows, because hooks are the highest-leverage variable in a video ad and the cheapest to change. Keep the body of the ad constant and swap only the opening, so you learn which hook holds attention rather than confounding it with other changes. Hooks also fatigue faster than bodies, so plan to refresh the winning openers well before you rebuild the whole ad.

Producing that many variants by hand is the real constraint. This is where AI creator video helps: you can generate multiple hooks against the same product from a URL, which makes hook testing a volume game instead of a bottleneck. Start a batch with the AI UGC ad generator, or if you want the format handled end to end, the AI video ad maker renders each hook as a finished, captioned ad ready to upload.

Where to start

Start with three hook patterns that fit your product: usually the direct problem call-out, the bold claim, and the demonstration open, since they cover the three reasons someone stays, they have the problem, they want the payoff, or the visual is interesting. Write two or three versions of each, keep the ad body identical, and put all of them in market at once.

Read the results by hold rate and cost per result, not by which one you would have picked. Kill the losers fast, scale the winner, and when it starts to fatigue, write new hooks before you touch anything else. The hook is the cheapest lever you have and the one that moves performance the most.

On some placements the hook does more than hold attention, it decides whether you pay at all. YouTube skippable in-stream ads only charge once a viewer watches 30 seconds or interacts, so a hook that survives the first few seconds changes your effective cost directly, which is worth keeping in mind alongside the way each YouTube ad format bills.