Almost everything is decided in the first two seconds

The asymmetry

Attention in a short-form feed is spent in a decision that happens faster than deliberate thought. A viewer is not weighing your video against alternatives; they are reacting to whether the first moment gives them a reason not to move. Everything downstream is gated on that reaction.

This produces a strange asymmetry in effort. Most teams write the body carefully, then write the hook last, in a hurry, as a line to put on top. The proportions are exactly inverted relative to how many people will encounter each part.

A hook is a frame, not a sentence

The most common failure is treating the hook as copy. A strong sentence delivered over a shot that gives nothing away still loses the viewer, because the image arrives before the words are processed. The frame and the line are one decision, not two.

Ask what is on screen at the moment the video appears. If the answer is a logo, an establishing shot, or someone drawing breath to start talking, the line does not matter. The decision has already gone the wrong way.

Commit to one

Scripts that offer five hook options are deferring the hardest decision to whoever holds the camera — usually the person with the least context about why any of them were suggested. A committed hook is a claim that can be wrong, which is what makes the result informative.

Testing alternatives is genuinely valuable, but as separate posts with one variable changed, not as a menu inside a single script. A menu produces an outcome you cannot attribute to anything.

Why hook formulas stop working

Lists of proven hook templates circulate constantly, and they do work — for a while. The reason they decay is not that viewers get smarter about the words. It is that a hook earns attention partly by being unexpected, and a template applied at scale is by definition expected.

This means the useful thing to extract from a hook you admire is not its wording but its move. “Open on the failure state before showing the fix” is a move; “POV: you just discovered…” is a template with an expiry date. Moves survive being widely used because they can be executed a hundred different ways.

It also explains why the same hook can work brilliantly and then flatly, with nothing else changed. Novelty is part of the mechanism, and novelty is consumed by use. A hook that stopped working was not necessarily wrong; it may simply have been spent.

What to check before filming

Describe the first frame in one sentence without mentioning the brand. If that sentence is not interesting, the video is not going to be. Then read the first line aloud; if it sounds like something written rather than said, it will sound worse on camera.

Check that the frame and line are doing different work. If the text just describes what is already visible, you have spent the most valuable two seconds you have saying one thing twice — a common failure, because it feels coherent while adding nothing.

Then ask what a viewer who does not know your brand would understand from the first moment alone. Not what they would feel, which you cannot know, but what information they would have. If the answer is nothing, the opening is asking for patience that no feed grants.