Open your retention curve on almost anything you've made. There's a cliff in the first few seconds.
That cliff is normal. It's people deciding whether this is for them, and no amount of craft removes it entirely. What you control is how steep it is, and that's the hook's entire job.
Not to be clever. Not to be a slogan. To make staying feel more interesting than leaving, for a few seconds, until the piece can earn the rest on its own merits.
What a hook actually is
The opening move of the piece. Commonly the first three to five seconds on short-form, the first fifteen to thirty on long-form. But the boundary isn't a duration, it's a decision point. The hook ends when the viewer has decided to stay.
Two things it isn't:
It isn't the title. The title and thumbnail got them to click. The hook stops them leaving. Different job, different moment, and a hook that just restates the title has wasted the most expensive seconds you own.
It isn't a tagline. A tagline is a compression of what you are. A hook is a structural move that opens something.
The five that work
Almost every hook that works is doing one of these. They're not a formula. They're a description of what's happening underneath.
1 · Open a question the viewer wants closed
Not a rhetorical question like "ever wondered why…?" (nobody has). A genuine gap between what they know and what they want to know.
"This costs eleven dollars to make. It sells for four hundred."
You now want to know how, and the piece has permission to take its time.
2 · Show the end state first
Start at the result, then explain how you got there. The finished thing, the transformation, the number.
This works because it converts the whole piece into an explanation of something the viewer has already decided is interesting. It's also the most reliably transferable structure across formats: it works in a fifteen-second Reel and a forty-minute documentary.
3 · Contradict something they believe
"Almost everything about how you're supposed to price this is wrong, and it's costing you about a third of your income."
Genuine tension does the work. This one has the highest ceiling and the highest risk: if the piece doesn't actually deliver a contradiction, you've trained the viewer not to believe your openings.
4 · Start in the middle of the action
The cold open. No logo, no "hey everyone", no explanation. Something is already happening.
It works because the first seconds are your most valuable real estate and titles spend them on information nobody asked for. Moving the introduction thirty seconds later costs almost nothing and buys the attention to deliver it.
The failure mode is being merely confusing rather than intriguing. The test is whether the first shot raises a question, not whether it's loud.
5 · Present an image that demands explanation
Purely visual. Something strange enough that the viewer needs to know what they're looking at.
Highest skill ceiling, least reliable to plan, and the one most dependent on actually having the shot. If you're going to use it, it belongs on the shot list as a named shot, not hoped for.
What almost never works
Introducing yourself. "Hi, I'm X from Y, and today we're going to talk about…" is three seconds of information the viewer did not ask for, delivered in the window where they're deciding whether to leave. Move it. If it has to exist at all, it belongs after the hook has done its job.
Explaining what the video will cover. A table of contents is not a reason to stay.
Slow logo animations. A brand asset placed in the exact spot where you have the least attention and the most competition.
Starting with context. "Before we get into it, some background." The background is why they'll leave.
The promise problem
This is the part most hook advice skips, and it's the part that matters most over time.
A hook makes a promise, and the rest of the piece has to keep it.
An oversold hook doesn't prevent the drop-off. It moves it. Instead of leaving at three seconds, they leave at twenty-five, having felt slightly cheated. You've converted a clean bounce into a bad experience, and on any platform that weighs watch-through, that's worse than not having hooked them at all.
There's a longer-run cost too. An audience learns what your openings are worth. If your hooks routinely overpromise, your regulars start discounting them, and the same words stop working on the people most likely to convert.
Reading whether yours work
Don't take anyone's word for which hook style suits your audience. It's an empirical question and you have the data.
Long-form: open the retention curve and look at the first thirty seconds. Compare your best five and worst five pieces. The shape of that opening drop is your hook working or not.
Short-form: watch-through is your proxy, and it's coarse. What you're looking for is a consistent gap between the top and bottom of your own catalogue.
The step everyone skips: check whether the pattern in your winners is absent from your losers. If your top five all open with a question and your bottom five do too, you've found a habit, not a signal. That check is the whole difference between analysis and confirmation. It's the same discipline as outlier reading.
Hooks in a brief
If you're commissioning, the hook is a decision, and leaving it to the edit means leaving it to whatever the footage allows.
Put in the brief: which of the five structures, what the promise is and, critically, where the payoff lands. Then make sure the shot list contains what the hook needs. A "show the end state first" hook requires the end state to have been shot, in a frame that works cold, and that's a shot-list line.
The honest caveat
There's a lot of confident advice about hooks online, most of it derived from a handful of very large accounts. That's a badly biased sample. Those accounts get enormous distribution for reasons largely unrelated to their openings, and their hook style is at least as much a consequence of their scale as a cause of it.
What holds up: the first seconds carry disproportionate weight, and an opening that raises a question outperforms one that supplies context. Beyond that, the specifics belong to your audience, and the only way to know is to look at your own curves.
Where Acumin fits
Content Analysis reads your hook and format mix against your cohort, and Outlier Studio surfaces the posts that beat your own baseline: the raw material for the comparison above. A Concept Read gives an idea a confidence band before you shoot it, and it is explicit about being a band rather than a prediction.
None of it can tell you a hook will work before it exists. Nothing can. What it can do is stop you concluding something from five posts.
How to use this tomorrow
Take your three best and three worst performing pieces. Write down, in one line each, what structural move the first five seconds makes.
If the six lines look the same, your hook isn't your variable and you should be looking somewhere else. If they don't, you have your next test.
Related: Thumbnails that read at feed size is the other half of the same moment. Running a content experiment is how to settle which hook structure actually works for you.