How long should a YouTube hook be?

The decision is usually made inside fifteen seconds, and often inside five. Write the opening as one sentence that could stand alone, then give yourself another ten seconds to justify it before any setup begins.

Should I still use an intro animation?

Not at the start. A branded sting in the first five seconds spends your most valuable attention on something the viewer did not come for. If you want one, place it after the hook has landed, or drop it entirely.

What makes a hook fail?

Three things, almost always: it restates the title without adding anything, it begins with housekeeping like a greeting or a subscribe request, or it promises something the video never delivers. The third one is worse than a weak hook because it damages retention across the whole video.

Does the first shot matter as much as the first line?

Yes, and it is the half people neglect. The line and the image arrive at the same time, and an interesting sentence over a stock establishing shot reads as generic before the sentence has finished.

Can I fix a video with a bad opening after publishing?

Not really. You can change the title and thumbnail, which affects who arrives, but the opening is baked into the file. This is why it is worth writing the hook before the script rather than after.

The first fifteen seconds decide the video

Everything about a video that you can control converges on one moment. The title and thumbnail decide whether someone arrives. The first fifteen seconds decide whether they stay — and if they do not stay, nothing else in the video exists.

That is not a motivational framing. It is a mechanical one. Average view duration is the number that drives whether a video is shown to more people, and the largest single drop in almost every retention graph is in the first thirty seconds. Everything you do to improve the middle of a video is competing for a smaller effect than fixing the opening.

What a hook is actually doing

A hook is not a promise of value. Every video promises value. A hook is doing something narrower: it is creating a specific gap between what the viewer knows and what they want to know, and making that gap uncomfortable enough to stay open.

That is why “in this video I’ll show you five ways to grow your channel” fails. There is no gap. The title already said that, so the sentence carries no new information — the viewer knows exactly as much after it as before it, minus eight seconds.

Compare it with a sentence that opens a gap: “Four of these five made no difference at all, and I ran them for six months before I worked out which one was doing the work.” Now the viewer has a question they did not have, and it can only be answered by staying.

That is the whole mechanism. Everything below is a way to get there.

The three ways openings fail

Restating the title. The most common failure by a wide margin. The viewer clicked because of the title; repeating it back to them is the one piece of information they definitely already have. Assume the title was read and start from the sentence after it.

Housekeeping first. Greetings, channel introductions, “before we start, hit subscribe”, a five-second animated sting. All of these spend the highest-value attention you will ever have on something the viewer did not come for. Housekeeping is not forbidden — it just cannot be first. Move it to ninety seconds in, after the viewer has decided to stay, where it also performs better because the person hearing it is now interested.

Promising what the video does not deliver. This one is worse than a weak opening, because the damage arrives later. An overstated hook gets the first thirty seconds and loses the next three minutes, and the retention graph shows a cliff exactly where the viewer realised. It also trains the audience not to trust your titles, which is expensive across a channel.

The four openings that generally work

There is no formula, but there is a short list of shapes that reliably open a gap.

The contradiction. State the thing everyone believes, then say it is wrong. Works because it creates the gap instantly and the viewer cannot resolve it themselves. Requires that you actually have the counter-argument — this is the shape most often used dishonestly.

The result first. Start at the outcome and work backwards. “This channel makes more from four videos a month than it did from twenty.” The gap is how, and it is unavoidable.

The specific number. Not “a lot” — the figure, immediately, with enough oddness that it sounds real. Specificity is what separates a hook from a claim.

The stakes. Name what goes wrong if the viewer is on the wrong side of this. Works best for anything with a cost — money, wasted months, a policy that demonetizes you. Fails when the stakes are invented.

What they have in common: each one is a complete sentence that could stand alone, and each one is about the subject rather than about the video. Sentences that are about the video (“I’m going to break down…”, “we’ll be covering…”) are always weaker than sentences that are about the thing.

The first shot is half of it

The line and the image arrive together, and most people write the line and let the editor pick the image. That is backwards.

An excellent opening sentence over a generic establishing shot — a city timelapse, a stock person at a laptop, a slow zoom on an unrelated graph — reads as generic before the sentence has finished. The viewer’s judgement about production effort is visual and it is faster than listening.

The first shot needs to do one of two things: show the thing being talked about, or contradict the line in a way that creates its own small gap. What it must not do is decorate.

This is where writing the opening as a scene rather than a sentence helps. When you split the script into shots — the scene splitter does this for a finished script — the opening becomes a shot with a required visual rather than a line with an optional one, and the weakness becomes visible immediately.

Write the hook first

The practical change that improves openings more than any technique: write the hook before the script, not after.

Written afterwards, the hook is a summary. It describes a video that already exists, which is why it so often just restates the title — that is what summaries do.

Written first, the hook is a constraint. It states a claim, and now the script has to earn it. This tends to improve the whole video, because a script written to justify a specific claim has a spine, and a script written to cover a topic usually does not.

A workable sequence:

Write the hook as one sentence.
Ask: does this contain information the title did not? If not, rewrite.
Ask: can the video actually deliver this? If not, rewrite — do not soften it later.
Write the first shot next to it. Not “b-roll” — the actual image.
Write the script to justify the claim.
Check the opening survives the edit. Openings get padded during editing more than any other part.

The hook generator produces openings in these shapes with a first shot attached, which is useful mostly as a way of seeing several angles on the same material quickly. Treat what it gives you as drafts to react against — the one you write after rejecting five is usually the good one.

Where the fifteen seconds actually go

A rough allocation that holds up across most long-form explainers:

Seconds	What is happening
0–5	The claim. One sentence, on the subject, with a real first shot.
5–12	The justification — the single strongest piece of evidence that the claim is true.
12–15	The turn into the video. Not a table of contents; the first real step.

What is not in there: greetings, channel branding, an outline of what is coming, a subscribe request, or an animated logo. All of those can exist later.

And one more constraint worth checking before any of it: the title. The hook and the title have to be different sentences doing different jobs, and if the title gets truncated in the feed the hook may end up being the only complete claim the viewer sees. The title counter shows where yours cuts on each surface — it is worth knowing which half of your title people actually read before you write the sentence that follows it.