How to Write Better Hooks for Short Video Clips

A hook is a debt

Most advice treats a hook as an opening line, which makes it a writing problem and sends people looking for templates. It is more useful to treat it as a debt. A hook opens a gap – something the viewer now wants closed – and the clip has to close it. Everything that makes a hook good or bad follows from that.

Two consequences fall out immediately, and they are the two things most hooks get wrong. The first is that a hook that opens no gap is not weak, it is absent. A greeting, a name, a topic announcement: nothing has been asked, so there is nothing to stay for. The second is that a gap you cannot close is worse than none at all, because you have made a claim about the clip that the clip does not meet, and the viewer finds out at exactly the moment they were going to decide whether to trust you again.

This also explains why hooks feel harder than they are. You are not trying to be clever in the first line. You are trying to name the most interesting thing the clip contains, early, in a form that makes the rest of it necessary.

Three hooks, one second

Before a viewer has read a word, two other things have already spoken to them. Hook advice is almost entirely about the words, and the words are the slowest of the three.

The frame arrives first and costs no time at all: a face mid-sentence, a hand holding something, a screen with a number on it, a shape that is not like the last nine clips. The words arrive next, on screen before they are spoken, which is why the caption is often the real hook and not the audio. The sound arrives only if it is switched on, and a large share of feed viewing has it off.

So the practical rule is not “write a good first line”. It is: make sure the frame, the on-screen words and the audio are all promising the same thing, and make sure the clip still works if the third of those is silent. A hook that only exists in the audio has been turned off before it played.

The tall vertical frame helps here, which is one of the few genuinely good reasons to use the space – a line of text above the picture is a hook in its own channel. Converting horizontal video to vertical covers how to get that room without wasting it.

Try it: the hook is already there

Here is the part that changes how people work. You are almost never writing a hook. You are locating one. A recording of someone who knows their subject will contain a sentence that opens a gap perfectly – it just will not be at the start, because nobody starts a recording by saying the most interesting thing they know.

Below is the first thirty-five seconds of an ordinary recording. Click any sentence to start the clip there, and look at what each opening would open, owe and cost.

Three things worth noticing. The first four options are all perfectly reasonable sentences and none of them asks anything; they are the parts of a recording where the speaker is arriving. The strong one is eighteen seconds in, and it is strong because two facts sit next to each other that should not. And the best-written sentence in the passage – the explanation at twenty-six seconds – is the worst place to start, because it is the payment rather than the debt.

That is the whole method, and it is a cutting decision rather than a writing one. Choosing the best moments to clip covers finding the passage, and turning long videos into short clips covers the route around it.

Six that work, and what each owes

These are not templates to pour a clip into. Each one is a different shape of gap, and the useful part is the second half of each entry – the debt it takes on, which the clip then has to service.

• The result, stated flat“We doubled our rates and lost one client.” Two facts that should not sit together, with no adjective doing the work. Owes: the mechanism, fast.
• The contradiction“The best thing I did last year was fire my biggest customer.” It contradicts what the viewer assumes, so the gap opens by itself. Owes: the reasoning, and it had better be real.
• The named mistake“I spent two years building the wrong thing.” People will stay for a mistake they might be making. Owes: what to do instead, not just the confession.
• The number with stakes“This one setting cost me four thousand pounds.” A figure attached to a consequence, not a figure on its own. Owes: the setting, and soon.
• The question they already haveOnly if they already have it. A question the viewer has never asked is not a hook, it is homework. Owes: an answer the clip actually contains.
• The visual revealNo words at all: something changes on screen and the viewer waits to see what. Owes: a payoff you can see, which is why it suits footage rather than talking.

Notice what none of them are. None is a question the viewer has not already asked, none is an adjective (“insane”, “crazy”, “game-changing”) doing work that a fact should be doing, and none needs the phrase “wait for it”. Adjectives are the tell: if the opening line needs one to be interesting, the fact underneath it is not interesting enough, and the viewer works that out faster than you would like.

Five ways a hook fails

1. It asks nothing. Greetings, topic announcements and channel names open no gap. They are doors, and a feed is not a corridor – nobody is walking towards your clip already.
2. It promises what the clip does not have. The most expensive error, because it works once. A hook is a claim about the clip, and a claim the clip cannot meet buys one view and spends the credit for the next one.
3. It gives the answer away. The single best sentence in a recording is often the worst possible opening, because it is the payment. A payment with no debt is just a statement.
4. The three channels disagree. The frame says one thing, the caption says another, the audio arrives mid-sentence on a third. Any one of them alone would have worked.
5. It is a hook for a different clip. Borrowed templates fit whatever they are poured into and commit to nothing. If the same first line would fit half your clips, it is not describing any of them.

The second one deserves the diagram, because it is the failure people commit on purpose.

Holding the payoff back feels like suspense and behaves like a loan. The audience it was borrowed from has mostly gone by the time it falls due, and the people who stayed were going to stay anyway. If the ending really is the best part, move it forward and open a second gap out of what it implies – an answer almost always raises another question, and that is a debt you can service.

On a looping platform this is even sharper: the clip restarts, so a viewer who has already seen the payoff gets your opening line again. TikTok clips covers what the loop does to structure.

The two-question test

Before a clip goes out, two questions settle it, and they take about ten seconds.

What did the first line ask? If you cannot state the question in your own words, there is no hook, however well the sentence reads. Be strict here – “it makes you curious” is not an answer; “it asks how you can raise prices without losing customers” is.

When does the clip answer it? Find the exact moment. If it is in the first third, the clip is doing what it said. If it is at the end, you have scheduled the payment badly. If there is no such moment, the hook is a claim about a different clip than the one you made.

A third question is worth asking for anything you publish regularly: would this first line fit half my other clips? If so it is describing none of them. Generic openings are comfortable precisely because they commit to nothing, and committing to something is the entire function.

Watch your own clip once with the sound off, reading only what is on screen. That takes twenty seconds and catches the failure the other checks miss – a hook that exists only in the audio, on a clip that will mostly be seen silent.

When the footage has no hook

Sometimes the recording genuinely does not contain a good opening line. There are three honest moves, in order of preference.

Build one from the clip’s own strongest fact. Not an invention – a restatement. If the clip explains why hourly pricing caps your income, the opening can be “getting faster at your job can cost you money”. The gap is real because the clip closes it. This is writing, but it is writing constrained by what you actually have, which is the only kind that survives contact with the viewer.

Use a title card as the opening channel. Put the question on screen for a second while the speaker is still arriving. This works particularly well on podcast clips and webinar clips, where people warm up slowly and the useful sentence comes after a run-up you do not want to include.

Or accept that this recording has no clip in it. Not every hour of footage contains one, and forcing a hook onto material that cannot pay it is how channels train their audience to scroll past them. Leaving it is a real decision and it is made far too rarely.

What none of these is: promising something the clip lacks. That trade looks like it wins and it is the only mistake on this page that gets worse the better it works.

FAQ

Related: captions for video clips, YouTube videos to Shorts and what video clipping is.