A hook is the first line — spoken or written on screen — that earns the viewer's next second of attention. A verbal hook is what you say in the first breath of the video; a text hook is the words layered on top of the footage. Most creators only ever write one of each, and most never test whether a different version would have worked better.
Verbal hooks vs text hooks — and when you need both
A verbal hook is the opening sentence you say out loud. It's the "every architect and interior designer knows this moment" line that tells the right person, in the first three seconds, that this video is for them.
A text hook is different. It's the words on screen — and it exists for two specific reasons. First, if your verbal hook is weak or missing (you jumped straight into the story without a clean opening line), a text hook can carry that job instead. Second, it adds a secondary layer of context the verbal hook doesn't have room for — a specific promise, a number, a name.
Most videos need both, doing two different jobs at once. The verbal hook earns the ear. The text hook earns the eye, especially for anyone scrolling with the sound off.
Why most creators only test one hook per video
Here's the part almost nobody does: writing one text hook, publishing it, and never touching it again. That's leaving most of the upside on the table. A text hook is designed to grab attention, but the biggest miss isn't the wording of any single hook — it's only ever testing one.
Changing a few words in a text hook takes seconds to edit and re-upload. But most creators never bother, because they assume the video itself is what performs or doesn't. In our experience, the hook does more of that work than people expect.
Trial reels: A/B testing hooks on the same footage
The fix is what we call a trial reel: the exact same video, re-uploaded with a different text hook, shown to a fresh cold audience. Same footage, same edit, same everything — except the words on screen for the first second.
Trial reels exist because a cold audience has never seen the video before, so each version gets a fair, unbiased read on which hook earns the scroll-stop. We've seen a change of a couple of words move a video's views by thousands. That's not a rewrite — that's a few words, tested honestly instead of guessed once.
If you're only publishing one hook per video, you're not finding out which version of your idea actually works. You're finding out whether your first guess was good enough, which is a very different question. For more on running this test properly, see trial reels and multi-hook testing.
Examples of hooks that stopped the scroll
Not every hook is a sentence. Some of the strongest hooks we've shot are pure motion — a pattern interrupt that raises a question before a single word is spoken. In one video, the subject wasn't on screen at the start; she moved into frame and slammed something down on the table. That single beat of motion created enough curiosity to stop a scroll before anyone knew what the video was even about.
The same video used a second visual hook seconds later: holding the microphone up to a cheese grater. It doesn't explain anything — it raises a question. Why is this person talking to a cheese grater? That confusion is the hook. It bought enough time for the actual content — a nutritionist talking about something her audience genuinely cared about — to land. That video reached over a million accounts and was shared more than 60,000 times, without the subject spending any extra time in the studio to get it.
Color grading did some of the same work. A distinctly vintage, slightly old-school grade on that video made it visually different from everything else in the feed at first glance — before the viewer had even processed the hook. Visual distinctiveness is its own kind of hook.
The common thread across verbal, text, and visual hooks is the same: each one exists to buy a second or two of attention before the actual value of the video has to prove itself. Get that first second right, and the rest of the video gets a fair chance.