The first three seconds: hooks that survive the scroll

This is the least glamorous advice in short-form and the most reliably true: a good video with a weak opening performs worse than a mediocre video with a strong one. Not because the platform is unfair, but because the platform is only ever measuring what people did, and almost nobody did anything.

A hook is not a promise of value. It is a reason to delay the scroll for one more second, and then another. That is a lower bar and a more useful one, because it stops you writing openings that describe the video instead of starting it.

"In this video I am going to show you three ways to..." is a description. Nobody has ever stopped scrolling for a table of contents.

1. State the conclusion first. "You are exporting your videos at the wrong resolution." The whole payoff, up front. This works because it is genuinely informative in three seconds, and people stay to find out if it applies to them.
2. Name the audience precisely. "If you post to more than two platforms, this will save you an hour a week." A narrow opening outperforms a broad one — the people it excludes were never going to watch.
3. Show the result before the method. Open on the finished thing. This is why before-and-afters endure.
4. Contradict something people believe. "Scheduling does not hurt your reach." Only works if you can actually back it up — see the note below.
5. Start mid-motion. No greeting, no intro, no logo. Begin on the second sentence of what you were going to say.
6. Ask a question the viewer has actually asked. Not rhetorical. A question they typed into a search bar last week.

"Wait for it." "You will not believe what happened." "The secret nobody tells you." The pure curiosity gap — withholding the subject to force a watch — has been used so heavily that audiences read it as a signal of *low* value rather than high.

The version that still works is a specific curiosity gap: tell people the subject, withhold the answer. "There is one setting that decides whether your Reel gets recommended" is honest about what it is about and still worth staying for.

1. Watch your own retention graph. The drop in the first few seconds is the hook. Everything after is the video.
2. Change one thing. If you rewrite the hook and the thumbnail and the caption, you have learned nothing about any of them.
3. Compare across at least five posts. Short-form performance is spiky; one video proves nothing in either direction.
4. Watch it on a phone, with sound off, mid-scroll. That is the actual viewing condition. Judging a hook on a desktop preview flatters it enormously.

Worth saying, because hook advice tends to imply otherwise. A strong opening on an empty video produces a high three-second retention and a terrible completion rate, and completion is the thing being scored. The hook buys attention; the video has to be worth the attention it bought.

The clearest sign you got both right is not views — it is sends. Someone forwarding your video to a specific person is the strongest signal that the payoff landed, which we go into in sends per reach.