How a hook works
The hook is whatever a person meets first: the opening frame of a video, the words burned across an image, the first line of the caption before the See more cut. A feed moves under someone’s thumb at speed, and the decision to stop or keep going is made almost before they have read anything. The hook is the part of the ad that decision is made on.
It does two jobs at once. It interrupts — through movement, a human face, an unexpected image, or a line that names a problem the viewer actually has. And it qualifies, signalling who the ad is for so the wrong people scroll past and the right ones stay. Delivery then feeds on that behaviour: creative that holds attention earns more of the impressions the auction has to give.
Why the hook matters
Everything after the hook is only seen by the people the hook kept. A carefully argued offer, a strong guarantee and a clean landing page all sit behind it, and none of them get a chance if the first moment is skipped. That is why two ads with identical bodies can perform very differently — the difference is upstream of everything you were measuring.
It matters more on mobile-heavy markets like Nepal, where most impressions are served to a phone, often with the sound off and on a connection the viewer is conscious of. If the opening only works with audio, or only after the video has buffered past its first seconds, it is not really a hook. The hook rate metric is what tells you how well it is doing.
Common mistakes with hooks
The most common one is the warm-up: a logo sting, a slow establishing shot, or a sentence of throat-clearing before the point arrives. The second is writing the hook for the marketing team rather than the buyer, opening with the company name and how long it has been trading rather than with something the viewer recognises as their own situation.
The third is treating the hook as the whole job. A hook that overpromises collects cheap attention and expensive disappointment: people stop, then leave at the offer or the landing page, and cost per lead rises even though the attention metrics look healthy. Comparing hook rate against hold rate is how you catch that.
How to act on it
Write several hooks for every concept and change nothing else. Keep the offer, the body and the call to action constant so the test tells you something clean about the opening rather than about the whole ad. Hooks are cheap to produce and quick to judge, which makes them the most efficient thing to iterate on in Meta ad creative.
Then review them the way your audience will: on a phone, sound off, at scrolling speed. If you cannot tell what the ad is about before your thumb wants to move, neither can they. Keep the hooks that stopped people and held them, and retire the ones that only did the first.