Table of Content
3 sec The window that decides most feed video | 47 sec Average time on one screen today | 5x Readers who see the headline vs the body |
The first time I watched a retention graph fall off a cliff at the two-second mark, I understood the job differently. I had spent an afternoon polishing a script, and the audience left before the second sentence. Nothing after the opening line mattered, because almost nobody reached it.
That is the uncomfortable truth about hooks. A hook is not a clever sentence. It is a promise the reader believes and wants kept. It works when it opens a small, specific question in the reader's head and makes the answer feel close.
This piece is the version of that lesson I wish someone had handed me early. It is built on how attention actually behaves, the psychology researchers have documented, and the formulas I still reach for when a draft is dying in the first line. I have kept the theory short and the examples usable. Steal the templates, test them against your own audience, and keep the ones that work.
What a hook is really up against
Before the tactics, it helps to see the room you are writing into. Most advice starts from a number that was never true.
The goldfish line is a myth. You have heard that humans now hold attention for eight seconds, one less than a goldfish. That figure came from a 2015 Microsoft Canada slide that cited a company called Statistic Brain, and when the BBC went looking for the underlying study in 2017, no one could produce it. The claim was invented. Do not build your strategy on it.
What is real is more useful. Gloria Mark, who has measured attention in real settings since 2004, found that the average time people spend on a single screen fell from about two and a half minutes to roughly 47 seconds. People are not incapable of focus. They have gotten faster at deciding whether something is worth theirs. That decision is where the hook lives.
47 sec Average time on a single screen today, down from about 2.5 minutes in 2004. Gloria Mark, attention researcher | 3 sec When Instagram and Facebook first count a video as viewed. The opening is a separate battle. Platform view definitions |
65% / 45% Share of viewers who reach ten and thirty seconds once they clear the first three. Reported Meta benchmarks | 5x More people read the headline than the body copy, so the opening spends most of your budget. David Ogilvy |
Ogilvy said it about print sixty years ago: five times as many people read the headline as the body, which means once you have written the headline you have already spent eighty cents of your dollar. The medium changed. The math did not.
Why a good hook works
A scroll-stopping hook is rarely loud. It is incomplete in a way the brain wants to fix. Two findings explain most of what works.
1994 The information gapThe economist George Loewenstein described curiosity as what happens when your attention lands on a gap in your knowledge. It feels closer to an itch than to interest. He named three conditions: you have to know enough to sense something is missing, the missing piece has to feel specific, and the answer has to feel reachable. Too vague fails the second test. Too impossible fails the third. | 1927 The open loopBluma Zeigarnik found that people remember interrupted tasks far better than finished ones, roughly twice as well in her experiments. An unfinished thought stays active in working memory. Every cliffhanger, every list you have to keep reading, runs on this. Open a loop in the first line and the reader carries it forward for you. |
Both come down to the same move. Say enough to create a specific question, then make the reader travel to get the answer.
The hook types, rated
Not every hook does the same job. Here is how the main types compare, rated on how reliably they stop a cold scroll, where they fit, and what tends to go wrong. Filled squares mark stronger scroll-stopping power.
| Hook type | How it stops the scroll | Best for | Power | Main risk |
|---|---|---|---|---|
| Curiosity gap | Opens a specific, unanswered question | Short video, subject lines, thumbnails | ■■■■■ | Vagueness, or a promise the content never keeps |
| Callout / targeting | Names the exact person or moment | Niche audiences, communities | ■■■■■ | Narrow reach by design |
| Bold claim | States a strong, testable promise | Faceless video, ads, landing headlines | ■■■■□ | Sounds hollow without proof |
| Direct promise | Names the exact payoff up front | Tutorials, B2B, search-led content | ■■■■□ | Forgettable if the payoff is generic |
| Contrarian | Contradicts a belief the reader holds | Opinion, thought leadership, ads | ■■■■□ | Needs a real argument behind it |
| Story cold-open | Drops the reader mid-scene | Long video, newsletters, case studies | ■■■■□ | Slow to pay off; can lose the impatient |
| Stat or shock | Leads with a surprising number | Data content, PR, health and finance | ■■■□□ | Wears off fast; easy to overclaim |
| Question | Asks something the reader silently answers | Casual social, community posts | ■■■□□ | Weak if the answer is an obvious yes or no |
Nine formulas I keep reaching for
Templates are not a substitute for a real idea. They are a faster way to shape one. Fill each with your specific payoff, then read it out loud.
THE SPECIFIC GAP The [surprising thing] almost no one [does] about [topic]. Example The email mistake that quietly kills your reply rate. | RESULT FIRST How I [got a result] in [timeframe] without [the expected cost]. Example How I booked 12 sales calls in a week without a single cold call. |
THE NAMED ENEMY Stop [common advice]. Here is what to do instead. Example Stop posting daily. Here is what actually grew my account. | THE CALLOUT If you [specific situation], this is for you. Example If your videos die at three seconds, read this. |
COST OF INACTION You are losing [thing] every [period] and you cannot see it. Example You are losing readers in the first line and your analytics are hiding it. | MID-SCENE OPEN It was [time] when [tense event], and [unresolved detail]. Example It was 2 a.m. when the campaign went live, and the numbers made no sense. |
THE CONFESSION I [did the thing people avoid admitting], and here is what happened. Example I deleted 40 percent of my posts, and my reach went up. | THE REFRAME NUMBER [Surprising number] of [group] [do this]. Here is why it matters. Example Most social video is watched on mute. Your hook has to work without sound. |
THE ANSWERED QUESTION What is the one [thing] you would [action] if [condition]? Example What is the first line you would cut if you read your last post out loud? | RULE OF THUMB Fill every bracket with a number, a name, or a concrete image. Example Vague words are where hooks go to die. Specific is the whole game. |
What kills a hook
Most weak openings fail for the same handful of reasons. Here is what to cut and what to write in its place.
| Cut this | Write this instead |
|---|---|
| Warming up with context nobody asked for | Start on the most interesting second of the story |
| A vague promise, such as “some tips to improve” | A checkable payoff, such as “the three edits that doubled my open rate” |
| Burying the hook under a logo or intro animation | Put the brand at the end, after the viewer decides to stay |
| Writing for everyone | Name the one person who needs this |
| A yes-or-no question | A question whose answer requires reading on |
| Wordplay that hides the point | Plain language that telegraphs the payoff |
| A hook the content does not deliver | A promise you keep in the very next line |
What the people who studied this say
The habit of stopping the scroll is old. The names change, the finding holds.
“The headline is the ticket on the meat.” Ogilvy meant the opening exists to flag the readers who are actual prospects, and to spend your budget where nearly everyone still looks. David Ogilvy Confessions of an Advertising Man | “Curiosity is a gap in one's knowledge.” In his 1994 work, Loewenstein framed curiosity not as interest but as the discomfort of a missing piece the brain is driven to close. George Loewenstein The Psychology of Curiosity, 1994 |
“The eight-second stat was never real.” When the BBC traced the famous goldfish comparison in 2017, the trail ended at a source no one could verify. Stop quoting it. BBC Busting the attention span myth, 2017 | “Attention did not shrink. It got faster to spend.” Mark's long-running measurements show time on a single screen falling to about 47 seconds, which reframes the whole problem. Gloria Mark Attention research, University of California, Irvine |
How I write and test a hook
This is the loop I run for anything that has to earn its first line, whether it is a Reel, a subject line, or the top of an article.
1. Write the ending first. Know the payoff before the opener, so the hook can promise it honestly.
2. Draft ten openers, not one. Volume beats polish here. Most will be flat. You are hunting for the two that sting.
3. Read each one out loud. If it needs a run-up to make sense, cut the run-up.
4. Delete the first sentence. The real hook is usually line two. Try removing line one and see if it gets stronger.
5. Make the promise specific. Swap every vague word for a number, a name, or a concrete image.
6. Check the loop closes. If you open a question, the content has to answer it. A broken promise costs you the next post, not just this one.
7. Ship two versions. Send both to comparable audiences and let the early retention curve pick the winner.
8. Keep a swipe file. Save the openers that worked and a note on why. That file becomes your fastest tool.
The VerdictSpecific beats clever, every time After years of writing openings and reading the graphs they produce, the part that still surprises me is how little of it is about being clever. The hooks that stop the scroll are almost always the specific ones. They name a real person, promise a real payoff, and open a question the reader cannot leave alone. The polish comes later, in the body, where the audience you earned is finally paying attention. If you take one habit from this, make it writing ten openers and throwing away nine. The winning line is rarely the first one you love. It is the one that survives being read out loud, that a stranger would understand in a second, and that your content can actually deliver on. The information gap, the open loop, the formulas, the rated table: all of it is just a faster path to that line. Write the promise. Keep it. Then watch the curve stop falling. |