> ## Documentation Index
> Fetch the complete documentation index at: https://docs.snowdoughnut.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Visual hooks

> The closed list of visual openings for a video ad or post: what the viewer sees in the first second, how to film each one, and when it fits. Pick from this list in a creative brief, or when a skill asks which visual hook to use.

A visual hook is what the viewer sees in the first second, before a word lands. It works alongside the spoken line and the text on screen, which come from [hook types](https://docs.snowdoughnut.com/hook-types.md) and are written with the [text / audio hook frameworks](https://docs.snowdoughnut.com/text-audio-hook-frameworks.md). A video ad opens with one of each: one visual, one spoken line, one line of text.

A face to camera already stops the scroll. A visual hook adds movement or a reveal in that first second, so the face arrives with a reason to keep watching.

## The hooks

| # | Hook | What the viewer sees | How to film it |
| :- | :- | :- | :- |
| 1 | Slide into frame | The speaker enters an empty frame from the side and starts talking | Lock the camera on an empty shot, step in from one edge on the first word |
| 2 | Quick zoom out | A tight shot snaps wide to reveal the speaker or the setting | Start on a close detail, zoom or cut wide inside half a second |
| 3 | Random object as a mic | The speaker talks into something that is not a microphone | Hold a hairbrush, banana, or the product itself like a mic, played straight |
| 4 | Laptop turn | The speaker spins a laptop toward the camera to show the screen | Start with the lid facing the speaker, turn it on the first line |
| 5 | Camera pull back | The shot backs away to show more than the viewer expected | Start close, walk the camera back or pull out on a gimbal |
| 6 | Talk in front of the subject | The speaker stands in front of the thing they are talking about | Frame the product, place, or result behind the speaker's shoulder |
| 7 | Wired earphone mic | The speaker holds a wired earphone up like a lapel mic | Hold the earphone's mic near the mouth, as in street interviews |
| 8 | Camera rotate | The frame turns as the clip opens | Rotate the phone or gimbal 90 to 180 degrees into the opening shot |
| 9 | Close-up stack montage | Three or four fast close-ups before the speaker appears | Shoot tight detail clips, cut each to well under a second |
| 10 | Object metaphor | A prop stands in for the problem | Show the prop doing the problem: a leaking bucket for wasted spend |
| 11 | Split-screen whiteboard | The speaker on one side, a whiteboard sketch on the other | Film the speaker and the board separately, place them side by side |
| 12 | Physical timer countdown | A real timer starts as the speaker begins | Set a kitchen timer or phone stopwatch in frame and start it on the first word |
| 13 | Unrelated action while talking | The speaker talks while doing something ordinary | Cook, fold laundry, or do makeup while delivering the script |
| 14 | Driving and talking | The speaker talks from a car seat | Film parked or as a passenger, never while driving |
| 15 | Walk towards camera and talk | The speaker walks into a close shot while already talking | Start far back, walk toward a fixed camera, land in frame on the key line |

Yapper ads add four small openings of their own, filmed face to camera: hands coming together without a clap, putting the hair up or down, getting into the car and shutting the door, and propping the phone up as recording starts. They sit in the [Yapper Framework](https://docs.snowdoughnut.com/skill-yapper-framework.md#the-hook).

## When each hook fits

The MCP ranks this list against the context of the video and offers the ones that fit best (four in a dialog, five in chat), using the [context facets](https://docs.snowdoughnut.com/mcp-registry.md#context-facets). A hook whose Avoid when matches is left out.

| Hook | Best for | Avoid when |
| :- | :- | :- |
| Slide into frame | length: short; channel: tiktok, meta, organic-social | |
| Quick zoom out | business: ecommerce, local; length: short | |
| Random object as a mic | goal: awareness; audience: cold; channel: tiktok, organic-social | business: b2b, sales-led |
| Laptop turn | business: saas, b2b; goal: consideration | |
| Camera pull back | business: ecommerce, local; goal: awareness | |
| Talk in front of the subject | business: ecommerce, local; proof: numbers, reviews | |
| Wired earphone mic | audience: cold; channel: tiktok, organic-social | |
| Camera rotate | length: short; channel: tiktok, meta | |
| Close-up stack montage | business: ecommerce; goal: conversion; length: short | |
| Object metaphor | goal: awareness, consideration; audience: cold | |
| Split-screen whiteboard | business: saas, b2b; goal: consideration; length: long | |
| Physical timer countdown | offer: sale, event, trial; goal: conversion | offer: evergreen |
| Unrelated action while talking | length: long; business: b2c; audience: cold | |
| Driving and talking | length: long; business: b2c | |
| Walk towards camera and talk | goal: awareness; audience: cold | |

A new visual hook is a new row in both tables. It joins the ranking the moment it is published.

## Picking one

One visual hook per video. Variants of a video are different visual hooks on the same script, or different spoken hooks on the same visual, never both at once, so a test shows which change did the work.

The visual has to make sense with the spoken line. A countdown timer opening a story about a slow, gradual change confuses the viewer; the same timer opening "you have until Friday" lands.

Two guardrails hold whatever the hook:

* Film driving shots parked or from the passenger seat.
* A countdown or other urgency cue has to match a real deadline. A timer on an evergreen offer is the kind of claim [legal review](https://docs.snowdoughnut.com/legal-review-process.md) exists to catch.

## Related resources

* [**Hook types**](https://docs.snowdoughnut.com/hook-types.md) The spoken and written openings a visual hook pairs with.
* [**Yapper Framework**](https://docs.snowdoughnut.com/skill-yapper-framework.md) Scripts that pick a visual hook for every opening.
* [**Video editing**](https://docs.snowdoughnut.com/skill-video-editing.md) Where the opening gets cut for each placement.
* [**Creative brief**](https://docs.snowdoughnut.com/creative-brief.md) Where the chosen visual hook is recorded for the creator.

> [accordion] MCP wiring
>
> **Type:** list\
> **Domains:** creative, video\
> **Provides:** `visual-hook`\
> **Uses:** `hook-type`, `legal-triggers`\
> **Best for:** per row, in When each hook fits


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.