Guide
Early views over impressions. That is the whole formula
Hook rate is the share of impressions that reach a short early threshold, commonly three seconds. It measures the opening alone rather than the ad. That is what makes it useful, and also what makes it trivially easy to inflate. This page covers the formula, why cross-platform comparisons are meaningless, and how to build a baseline from your own ads. By the end you will know what number your next opening has to beat.
- Division, and two numbers you already have
- 1
- Recent ads that give you a real baseline
- 10
- Metrics it must always be read beside
- 2
Divide early views by impressions, and know which threshold you divided by
Take the number of views that reached the early threshold and divide by impressions. That is the whole calculation.
The threshold is where the trouble starts. Three seconds is the common one. Two seconds exists. So does a percentage-watched threshold.
An impression is also not one thing. Platforms differ on when a served video counts, particularly for autoplay in a scrolling feed.
So a hook rate is a ratio between two numbers that both mean something slightly different depending on where you read them.
That is not a reason to ignore the metric. It is the reason a hook rate from one placement cannot be compared to one from another, including any published benchmark.
What the number is made of
Two inputs, one division. Both inputs are defined by the platform, which is why the output does not travel between them.
Impressions
Counted the platform's way
Views past the threshold
Often three seconds
Divide
That is the hook rate
Compare to your own
Never to a benchmark
It isolates the opening, which is the only reason to look at it
Almost every other creative metric measures the whole ad. A change anywhere shows up everywhere, so nothing is attributable.
Hook rate reads the first frame, the first sound, the opening line and any text on screen. Nothing after that.
That isolation is what makes it worth having. Vary the opening and hold everything else steady. The difference in hook rate is the difference the opening made.
It also tells you which problem you do not have. A healthy hook rate with poor results means people are arriving and leaving. That is a body problem or an offer problem.
Two of the four channels it measures are silent. The first frame and the on-screen text carry a muted viewer on their own. An opening line that exists only as audio does not exist for them.
It travels badly even inside one account. A hook rate on a static image placement and one on video measure different behaviours. Keep every comparison inside the same format.
This is the whole editor
Highlight a phrase and a clip lands on those exact words. No timeline, no keyframes, no layers.
It is easy to inflate, and inflating it moves money in the wrong direction
A shock, a false promise or a loud first frame raises hook rate reliably. Every one of those also lowers everything below it.
Attention bought under false pretences gets spent on annoyance. The funnel widens at the top and narrows harder underneath.
That is why the number is only readable next to a hold measure. A high hook rate that collapses immediately means the opening wrote a cheque the ad did not honour.
A modest hook rate with strong hold is the better problem. The opening is under-selling something people actually want, and improving it is the cheapest gain available.
Read cost per landing page view alongside both if you want a third opinion. It reflects intent rather than curiosity, and curiosity is what a shock buys.
Reading hook rate beside hold rate
The metric alone cannot tell you which of these you have, which is why it is never read on its own.
| What the pair says | What to do | |
|---|---|---|
| Hook high, hold collapses | The opening overpromised | Fix the opening, not the body |
| Hook low, hold strong | Under-selling a good ad | The cheapest gain you have |
| Hook low, hold low | Nothing is landing | New concept, not a new edit |
| Hook high, hold strong, no sales | The creative is fine | The offer or the page is not |
Build your own baseline from ten recent ads, then ignore every published number
Take your last ten ads in the same placement. Calculate hook rate for each. Take the median. That is your bar, and it is the only one that means anything for your account.
It costs about ten minutes. After that, every argument about an opening is settled with a number instead of a preference.
Recalculate it quarterly, because thresholds, counting and audience all move. A baseline from eighteen months ago is a historical document.
Then use it the way it was meant to be used. As a gate on new creative before it earns wider budget. And as the metric you vary openings against.
Record six openings against the same body in one sitting and read them against the median. Six shapes, not six rewordings. Near-identical variants produce near-identical rates.
In Cutroom each of those six is a batch and an export rather than a second edit, so the number you are chasing costs 110 credits to chase.
What three seconds gives an opening to work with
At roughly two and a half words a second. The metric is measuring a decision made inside eight words.
- Two seconds5 words
- Three seconds8 words
- The whole thirty second ad75 words
It cannot tell you why, and chasing it alone is how accounts get worse
Hook rate is a symptom reading. It says whether the opening is stopping people. It says nothing about which of the four channels caused it.
Diagnosing that needs variation. Change the first frame alone. Change the opening line alone. Change the on-screen text alone. The metric becomes an explanation only when you feed it a controlled comparison.
It is also useless below the volume where the ratio is stable. On a small budget, a day of data is noise wearing a decimal point.
An ad optimised purely for hook rate is an ad optimised for curiosity. That is not the same as an ad optimised for revenue. Keep the hold measure next to it every single time.
Treat any hook rate reported against a different threshold as a different metric wearing the same name. That is exactly what it is.
Write the threshold beside the number every time you record one. That single habit prevents most of the arguments this metric causes.
Questions people ask
- What is a good hook rate?
- Whatever beats the median of your own last ten ads in the same placement. Published benchmarks are calculated on different thresholds and different impression counting, so a number from somebody else's account is not a target, it is trivia.
- Is hook rate the same as three-second view rate?
- Usually, and only if the threshold matches. Some reporting uses two seconds, some uses a percentage watched. Check which definition your platform is using before comparing two numbers, including two of your own from different placements.
- Can a bad hook rate be fixed in the edit?
- Sometimes, if the problem is the first frame or missing on-screen text. If the opening sentence is generic, no caption style rescues it and rewriting the sentence is the whole job.
- Is hook rate the right thing to optimise for?
- It is, until your ads already stop people and still do not sell. That is a body or an offer problem, and raising hook rate makes it more expensive rather than less. Read hold rate and cost per landing page view alongside it.
One division, one threshold, one baseline built from your own account. Then vary the opening against it, which in Cutroom is a 110 credit ad rather than a second edit.