Ad hooks: hook ads for Facebook and Meta, the best ad hooks written for your product

Three seconds, and it is three things at once rather than one sentence. Meta's help center article on video ad best practices says to feature the brand and key message within the first 3 seconds, to use motion or a compelling visual in the first frame, and to design for sound off. Read together, those three instructions describe a hook with three layers, and each layer has its own budget.

The on-screen text row is the one most teams skip. If most people watch muted, as Meta says, then for most of your audience the caption is the hook and the spoken line is decoration. A useful test: mute the ad and watch the first three seconds. If you cannot tell what the product does or who it is for, the hook is not finished.

The spoken budget comes from the same arithmetic as a full video ad script written to a word budget: about 2.5 words per second for a natural ad read, so three seconds holds seven or eight words. Anything longer is a hook that has not ended by the time the viewer decides.

There are about eight shapes that account for most hooks worth testing, and the reason to know them is so the five hooks you test are five different ideas rather than five wordings of one. Below, each shape is written for the same illustrative product, a cordless car vacuum, so you can see how different the openers are when only the angle changes.

Count the spoken lines and every one sits at eight words or fewer, which is the point. Notice too that the first frame is doing as much work as the line: the result-first hook is carried almost entirely by the cut, and the pattern interrupt only works because glitter is visibly absurd.

The mistake the table is meant to prevent is testing five problem callouts and calling it a hook test. That tells you which wording of one idea wins, not which idea wins. Pick three or four different rows, write one strong version of each, and you learn something from every result. The same shapes applied to organic TikTok, where the opener has to survive without paid distribution behind it, are on TikTok hooks.

Yes, but not as a logo. This is where Meta's guidance and most creator advice seem to disagree and actually do not. Meta's help center tells advertisers to feature the brand and key message within the first 3 seconds. Creator-side advice says never open on the logo. Both are right, because the product in someone's hand in the first frame is the brand, and it is the version a viewer does not scroll past.

The failure both warnings point at is the opening brand card: a logo animation or a packshot with a tagline, the thing a viewer has learned to recognize as an ad and skip. Meta's same article says ads that blend in with organic content perform better and to avoid overly ad-like creative, which is the case against the brand card in Meta's own words.

So the practical rule for a hook is simple. The product is visible, in use, by second three. The brand name is on the packaging, in the caption or in the creator's line, not in a separate title sequence. And the logo, if it appears at all, belongs at the end with the call to action, where the people who are still watching are the ones who want to know who made it.

Three to five hooks on one body is the practical range, and the reason is arithmetic rather than taste. Every hook you add splits the same budget further, so each variant takes longer to collect enough impressions to judge. Two hooks rarely tell you much because a close result is noise. Ten hooks on a modest budget means none of them gets enough delivery to be read. Three to five different angles, each written as a different row of the table above, is where most accounts get a readable answer.

The production side is where the saving is. Five hooks on one body is five ads from one shoot, as long as the brief asks the creator to film every opener in the same session, with the same lighting and wardrobe, so the edit can swap the first three seconds without a visible join. Commissioning five separate videos to test five openings costs five times as much and changes the body at the same time, which means you cannot tell which change caused the result.

That is the brief this page writes: one body, the call to action, and the numbered hook list the creator films back to back. How many finished videos a month an account actually needs, worked out from spend and frequency, is in how many UGC videos do I need, and the full field list for the creator document is on the UGC script template page.

Better than your own account's median, measured on the same body. Hook rate is not a metric Meta reports under that name. It is a calculation media buyers make from Ads Manager columns: 3-second video plays divided by impressions. It answers one question, how often the first three seconds earned a pause, and nothing else.

You will find agency blogs quoting 25 to 30 percent as a floor and 40 percent as excellent. Those are not Meta's numbers, Meta publishes no hook rate benchmark, and the figure moves with placement, audience and category enough that someone else's average is a weak yardstick. The comparison that holds up is internal: run several hooks on one body, and the winner is the one with the best hook rate that also holds its click-through rate.

That last clause matters. A shocking opener can double the hook rate and attract people with no interest in the product, which shows up as a strong hook rate with a falling hold rate and click-through rate. A hook that stops fewer people but the right ones is the better hook. If you want the brief behind the winner turned into more creator videos for Meta placements specifically, UGC for Meta ads covers the formats and ratios.

Because the same people have seen them. Creative fatigue on Meta usually shows up in the first three seconds first: the hook rate slides while the hold rate for the people who do stop stays roughly where it was. That pattern is useful, because it tells you the body is still fine and only the opener needs replacing.

The cheap response is to refresh the hook before touching anything else. Keep the demonstration, keep the offer, write three new openers from rows of the hook table you have not used yet, and film them against the existing body. A new first frame and line is a far smaller production job than a new ad, and it resets the part of the ad the audience has learned to ignore.

The expensive mistake is reading fatigue as a failed product or a failed audience and rebuilding the campaign. If hook rate is falling while hold rate is flat, the product has not stopped being interesting, the opening has stopped being new. For the full sequence of which layer to vary after the hook, the structural guide in short-form video ads walks through briefing creators for hook-based ads, and if you are weighing a writer against a tool for the words, what video ad scriptwriters charge puts the freelance rates side by side.