A hook is a purchase. It buys the next three seconds

The hook is not a length. It is a job. Whatever it takes to convert a scrolling viewer into a watching one is the hook. Everything after that is the ad.

On vertical video that is the first two to three seconds. The number describes viewer behaviour rather than a rule. The decision to stay gets made before the subject has been understood.

The word is borrowed twice over. In songwriting the hook is the phrase that lodges in your head. In journalism the lede is the sentence that commits you to the article.

Advertising took the term for the same reason both of those did. The opening is a separate craft from the thing it opens.

So a hook is not a shorter version of the ad. A summary of your offer in eight words is still a summary. A scrolling viewer has not agreed to hear one.

The opening earns the right to make the pitch. It is not the pitch made quickly. That confusion is where most weak openings come from.

Reducing the hook to a line of dialogue is the mistake that costs the most performance. Four channels fire at once. They can contradict each other.

Two of the four are silent. Those two carry a sound-off viewer on their own.

The standard measure is hook rate. It is the share of impressions that reach a short early threshold, commonly three seconds. It is blunt. It is also the best instrument available, because it isolates the opening from everything else.

The trap is treating it as the score for the ad. A hook rate is easy to raise with a shock, a false promise or a loud first frame.

Those techniques reliably lift the top of the funnel and lower everything below it. Attention bought under false pretences gets spent on annoyance.

So read hook rate next to a hold measure. A high hook rate collapsing straight afterwards means the opening wrote a cheque the ad did not honour.

A modest hook rate with strong hold is the opposite. The opening is under-selling something people actually want, and improving it is the cheapest gain available.

One caution on the number. Thresholds and impression counting differ between platforms. Both change over time. Build your reference from your own recent ads in the same placement.

Two practices make it usable. Record the hook rate of every ad in one sheet at the moment you turn it off. Dashboards get restructured and history quietly goes missing.

Separate the four channels when you vary an opening. Changing the first frame and the first line at once produces a number you cannot attribute to either.

Ten ads is enough to have a median. Twenty is enough to trust it. That is the difference between measuring an opening and rating it.

That the hook must state the offer. It must not. Nobody scrolling has agreed to hear a pitch. The hook exists to buy the right to make one.

That the hook is a formula. Templates circulate, work for a season, then burn out precisely because they circulated. The underlying moves are durable. A contradiction. A named audience. A specific number. A demonstration. A confession. A question the viewer already asked.

That the hook belongs to the copywriter. On a muted feed the first frame does as much work as the first sentence. Teams where the writer and the person choosing the opening image never speak produce a predictable failure.

That a hook can be fixed with edits. A generic opening sentence is not rescued by any caption style. Rewriting the sentence is the whole job.

And the belief nobody argues with. That one hook is enough. The hook has the most variance in outcome. It also has the lowest cost to vary. That combination makes it the highest-return thing in the account to test.

Record six openings against the same body in one sitting and you have six creatives for the price of one recording. Six different shapes, not six rewordings of one. Six versions of the same contradiction produce six similar numbers and a wasted week.

Cutroom is built for exactly that loop. Each opening is a batch and an export rather than a second edit, so the sixth costs what the first did.