# Hook anti-patterns

The openings that tend to tank first-three-seconds retention. Each one
has a documented failure mode. Catch these in critique mode, name the
pattern, then rewrite. Don't just flag them.

## Welcome / logo / channel intro

"Hey guys, welcome back to my channel."

Why it fails: every frame of throat-clearing is a frame of swipe. The
algorithm doesn't wait for the hook to warm up. The Shorts feed and
the Reels algorithm both punish delay in the hook delivery, and
viewers swipe before the actual content begins.

The fix: cut the intro entirely. Open on the payoff frame. If the
creator has a recognizable face, the face is the brand. They don't
need to name themselves at second zero.

## Logo animations and channel jingles

Visual equivalent of the welcome intro. A two-second logo before the
content is a two-second flat line on the retention graph.

The fix: branding goes at the end, not the start. The first frame
sells the watch. The end frame sells the channel.

## Vague tease

"You won't believe what happened next." "This changed everything for
me." "What I'm about to show you is wild."

Why it fails: curiosity needs a defined shape to work. A vague tease
leaves the viewer with no category for the answer, so the brain
files the hook as bait and swipes. The viewer can't predict whether
the payoff is worth the wait.

The fix: name the category of the answer without giving the answer.
"I found a hidden setting on my iPhone that doubles battery life"
keeps the answer secret but tells the viewer what kind of thing the
payoff will be.

## Bait and switch

The hook promises one thing. The video delivers another. The viewer
clicks, feels deceived, and bounces.

Why it fails: platforms detect the pattern. Satisfaction-weighted
discovery on Shorts demotes channels that consistently mismatch
hook to payoff. Reels and TikTok have similar signals. Beyond the
algorithm, the trust cost on the audience side compounds.

MrBeast's production memo treats the match between title, thumbnail,
and first sentence as non-negotiable: if viewers click expecting one
thing and see another, they feel deceived.

The fix: if you can't deliver on the hook, change the hook. The
truth is non-negotiable; the wording isn't.

## Subscribe-before-value

"Smash that subscribe button before we start." "Follow for more
content like this."

Why it fails: asking for the sub before delivering any payoff creates
psychological resistance. The viewer hasn't earned anything yet, so
the ask reads as presumptuous. Measurable retention drop in the
first three seconds. On Shorts specifically, the subscribe button
sits at the bottom of the player and accidental taps are common,
which means a "subscribe" CTA gets noise rather than signal.

The fix: deliver the value first. The right early CTA on short-form
is "watch the full video on my channel" or "save this for later,"
not "subscribe."

## Story openers on TikTok

"Let me tell you about the time I…" "So basically, a few years ago…"

Why it fails: the format pulls in long-form context that the swipe-
decision window can't carry. Storytelling can work on TikTok, but
the hook has to lead with the punchline of the story or the stakes,
not the chronology.

The fix: open on the most arresting moment of the story. Walk back
to setup after. If the story has no moment arresting enough to open
on, it might not be a TikTok.

## Face-on YouTube-style intro

"Hi everyone, today we're going to talk about [topic]."

Same root failure as the welcome intro, with extra setup baggage. The
verbal hook is "we're going to talk," which promises talking, not a
payoff.

The fix: skip the announcement of what the video is about. Just be
about it.

## Imperative-command hooks

"Stop scrolling." "Listen up." "Wait, don't swipe."

Why they're high-variance: when the payoff matches the urgency, they
hit hard. When the payoff doesn't match, they read as desperate and
the viewer swipes faster than they would have without the command.
Most uses fail because the payoff doesn't earn the urgency.

The fix: only use an imperative if the payoff is genuinely surprising
or high-stakes. Don't open with "stop scrolling" on a recipe video.

## Hook treated as ornamental

The script gets written first. The hook gets bolted on at the end
as a "catchy opener."

Why it fails: hooks that aren't part of idea-evaluation tend to
fail because they're a coat of paint on a video that didn't earn
attention in the first place. MrBeast's process and Galloway's
both treat hook-writability as part of idea-validation. If you
can't write a sharp hook for the idea, the idea probably isn't
ready.

The fix: write the hook before you write the script. If you can't
write three credible angles, kill the idea or reframe it until you
can.

## Single-layer hook

The verbal line is sharp. The first frame is a static face waiting to
talk. No on-screen text.

Why it fails: most viewers watch muted. Without text, the verbal hook
is invisible. Without a working visual frame, the viewer swipes before
the voice lands.

The fix: design all three layers (visual, verbal, text) before
shooting. See `references/three-layer-hook.md` and
`assets/three-layer-worksheet.md`.

## Brand-first opener

"Check out our new product." "We're excited to announce…"

Why it fails: leads with the seller, not the viewer. The viewer's
brain has nothing to grab onto in the first second because the hook
is about the brand, not them.

The fix: lead with the viewer's outcome or pain. The brand comes
later, in the payoff, or implied through the visual.

## Cleverness that needs explaining

A wordplay hook, an inside joke, a reference the viewer has to decode.

Why it fails: the swipe-decision window doesn't reward effort to
understand. If the viewer has to think to get the hook, they swipe.
Clarity outperforms wit at second 1.

The fix: save the cleverness for the payoff. The hook is the door,
not the punchline.

## Slow build-up

The hook is hidden inside the third sentence. The first two sentences
are setup the writer thought was necessary.

Why it fails: the algorithm reads the early seconds, not the structure
the writer intended. The "real" hook never gets reached because the
viewer swiped during the setup.

The fix: delete the first two sentences. Whatever's in sentence three
is probably the actual hook. If sentence three doesn't work as an
opener, you don't have a hook yet.

## Three-act hooks (the hook itself takes 10 seconds)

The hook has its own setup, escalation, and payoff. By the time the
viewer reaches the actual video content, half the audience has left.

Why it fails: the hook is supposed to resolve by second 3, not be a
mini-video unto itself. A hook that doesn't pay off until second 8
is a video with two hooks competing.

The fix: cut the hook down to one beat. If the hook needs setup, the
setup is the hook; start there.

## How to use this list in critique mode

When the user hands over a draft:

1. Read the hook out loud. Time it. If it runs over 2 seconds, that
   alone is the diagnosis.
2. Check it against this list. Name the specific anti-pattern by
   name. "This is a vague-tease hook" is more useful than "this
   is weak."
3. Rewrite the hook in two or three variants the user can A/B. Don't
   just flag the problem.
4. If the same anti-pattern shows up across multiple drafts from the
   same user, point out the pattern. It's usually one habit, not
   ten separate mistakes.

See `assets/hook-checklist.md` for a pass-fail check to run after
rewriting.


---
资料：Vyral Content / vyralcontent/content-skills

原始版本：https://github.com/vyralcontent/content-skills/blob/349e495d97029a3e6303a7510e3f39c73f36d8fe/skills/viral-hooks/references/hook-anti-patterns.md

许可：MIT。目录与内部链接已经整理，未收录图片已省略。

MIT License

Copyright (c) 2026 Vyral

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
