The Viral Hooks Masterclass

You have about three seconds before a viewer decides whether your video deserves the next thirty. That decision is not made by the quality of your edit, your lighting, or your punchline. It is made by whether the opening gives them a reason to stay. A hook is not a magic phrase you bolt onto the front of a finished video; it is a promise, made in the first frame, that the rest of the video has to keep.

This guide is about building that promise deliberately: the frames that work, the mechanics that make them land, how to write ten of them in twenty minutes, and how to tell whether the one you picked is actually doing its job.

What a hook has to accomplish

A working opening does three things at once, usually in under eight words.

1. It establishes stakes. Something is at risk, wrong, expensive, or about to be missed.
2. It promises a specific payoff. The viewer should be able to describe, out loud, what they will know or see by the end.
3. It tells the truth about the video that follows. A hook that oversells is worse than a boring one, because the drop-off happens at second four instead of second one, and it teaches the platform that your video holds attention badly.

The third point is the one most creators skip. A hook is not a separate creative act from the video; it is a contract. If the opening says "three mistakes," the body should name three mistakes in the first fifteen seconds, not spend forty seconds on a story first.

Four frames that keep working

Frames are reusable structures, not scripts. Each one works because it exploits a different reason people keep watching.

1. The mistake frame

You are doing something wrong and it is costing you. This works because it reframes the viewer as the subject of the video rather than an observer.

• Weak: "Here are three ways to grow on TikTok."
• Strong: "You are making these three mistakes on TikTok, and the third one caps your views."

The specificity in the second version is doing most of the work. "Three mistakes" is a promise you can keep in a thirty-second video; "ways to grow" is a promise that cannot be kept at all.

2. The insider frame

Knowledge that feels gated. It works because curiosity is stronger when the viewer believes other people already know the answer.

• Weak: "How to edit videos faster."
• Strong: "Editors charge four figures for this export setting. It takes ten seconds."

The risk here is overclaiming. If the setting turns out to be ordinary, the viewer feels cheated and the comment section says so.

3. The specific-number frame

Numbers make abstract promises concrete, and they make a video easy to evaluation: the viewer knows exactly how long the payoff will take.

• Weak: "Some tips for better captions."
• Strong: "Your captions are unreadable on a phone. Here is the one measurement that fixes it."

A number does not have to be large. It has to be checkable.

4. In medias res

Start mid-action, at the moment of highest tension, and explain afterwards. This is the only frame that works with no spoken words at all, which makes it the strongest option for sound-off feeds.

• Weak: opening on you sitting down to talk.
• Strong: opening on the half-second where the clip is already moving, then cutting back to how it got there.

The three channels have to agree

A hook lands through three channels at once, and most weak openings have them fighting each other.

• The first frame: what is on screen before anything moves. A face mid-expression beats a face at rest; motion beats a static card; a frame that looks like an accident beats a frame that looks like a title slide.
• The on-screen text: the line a viewer can read without sound. It should not repeat the spoken line word for word, it should compress it.
• The first spoken line: the sentence that has to survive being heard at half attention, in a noisy room, at 1.2x speed.

When all three carry the same promise, the opening feels inevitable. When the text says "budget editing" and the spoken line says "let me tell you a story," the viewer's attention splits and the scroll wins.

Writing ten hooks in twenty minutes

The fastest way to a good hook is to write a bad ten and pick the least bad one. Working alone, that beats waiting for inspiration.

1. Write the video's payoff as a single flat sentence, with no style: "the export bitrate that stops Shorts looking soft."
2. Write it as a mistake the viewer is making.
3. Write it as a number.
4. Write it as a warning.
5. Write it as the moment of payoff, with no setup.
6. Write it as a question the viewer has actually typed into search.
7. Write it addressed to one specific person: "if you film on a phone, this is for you."
8. Write it in the negative: "stop doing this before you export."
9. Write it with the cost attached: "this mistake costs you a re-upload."
10. Write the boring version, on purpose, so you have a baseline to compare against.

Then read all ten out loud, in the order you would actually say them. Two things usually happen: one reads naturally and three read like advertising. The natural one is your hook. If none of them read naturally, the problem is usually that the video's payoff is not specific enough yet, and no amount of wording will fix that.

The failure modes

Bait. The hook promises something the video does not deliver. It buys three seconds and loses the next thirty, which is a worse trade than never being clicked at all.

Borrowed hooks. Copying a phrase that worked for someone else, with a different audience, in a different niche. The words transfer; the trust does not.

Hook tuned to the wrong audience. A hook that maximises curiosity from people who will never care about your subject pulls in traffic that does not convert, follow, or return. Short-term numbers go up, and the account gets harder to grow.

Over-editing the opening. Six cuts, three sound effects and a zoom in the first second tells the viewer you are compensating. One clear visual beat beats six busy ones.

Assuming the hook is the problem. Often it is not. If the first three seconds hold but the drop-off happens at second nine, the opening worked and the first section failed. That is a retention-graph problem, not a hook problem, and the fix is in the edit, not the copy.

Measuring whether it worked

The number to watch is the share of impressions that survive to three seconds. Compare it against your own median, not against someone else's screenshot. A useful routine:

• Post the video, wait 48 hours, and record views at three seconds and views at the end.
• If three-second retention is above your median but completion is low, the hook is doing its job and the body is not.
• If three-second retention is below your median, the opening is the weak link; the body never got a chance.
• If both are high and views are low, the problem is distribution and the thumbnail or caption, not the video.

For a more careful version of this, including how to compare two hooks without splitting an audience you do not have yet, see testing hooks with a small audience. For reading the shape of the drop-off itself, how to read a retention graph goes deeper.

A worked rewrite

| Flat version | Hook | | --- | --- | | "Tips for better lighting" | "Your key light is on the wrong side. Here is how to tell in two seconds." | | "How I edit my videos" | "I cut this 40-minute interview into nine Shorts in one sitting. Here is the order I do it in." | | "Caption styles explained" | "This caption style is unreadable at arm's length. Most editors ship it anyway." | | "Why your videos flopped" | "Your video did not flop. Your first frame did." | | "My favourite tools" | "Three tools I pay for and one I would cancel tomorrow." |

Notice that every strong version commits to something the video can be checked against. That is the whole trick: specificity is what makes an opening feel like it is worth three seconds.

The checklist

Before you export, read your opening and answer these honestly.

• Can a viewer say what they will get, in one sentence, after hearing only the first line?
• Does the first frame contain a face, motion, or an unfinished action?
• Does the on-screen text compress the spoken line instead of repeating it?
• Does the video deliver the promise inside the first fifteen seconds?
• Would the hook still make sense to someone who has never seen your other videos?
• Is there anything in the first second that is not doing work: a logo, a title card, a slow zoom?

If all six pass, post it. If not, fix the weakest one before you export, because the opening is the only part of the video every viewer is guaranteed to see.