Visual Hook
Follow shared release-shell rules in:
Use this skill when the task is about making a short-form visual opening stronger.
This skill applies to:
- TikTok, Reels, Shorts, and other scroll-feed videos
- slideshows / photo carousels
- AI-generated image or video prompt planning
- storyboard first frames
- visual hook critique and rerun notes
It is not a generic beauty or art-direction skill. A good visual hook is not merely pretty. It creates a clear reason to keep watching.
Core Rule
The first image or first 1-3 seconds should make the viewer ask a concrete question before they consciously decide whether to keep watching.
Good questions:
- "What happened?"
- "Why does it look like that?"
- "How is this going to change?"
- "What is that action doing?"
- "Is this about to get worse or better?"
Weak questions:
- "What brand is this?"
- "Why is this product shot here?"
- "Is this an ad?"
Baseline Checklist
A strong visual hook should usually have:
- one obvious subject
- one visible conflict or tension
- one action already in progress
- one piece of evidence the viewer can understand without explanation
- one unfinished outcome that invites the next shot
If the hook requires the caption, voiceover, or product claim to make sense, the visual is probably too weak.
Hook Families
1. Pain Evidence
Open on the problem itself.
Use when the product or story has a visible pain point.
Examples:
- a person sees flat, oily hair in the mirror and immediately touches it
- a close-up of scalp oil, flakes, thinning appearance, or a bad styling result
- a brush, sink, shirt collar, skin patch, messy drawer, or visible failure cue
Avoid:
- clean product still life before the problem is visible
- vague "sad person" acting with no concrete evidence
2. Contrast / Transformation
Open with a visible before/after tension.
Use when the content promise is change, progress, or repair.
Examples:
- same person, same mirror, bad state implied first, improved state teased next
- a split-second cut from social failure to routine reset
- a slideshow cover that places the worst visual state against the best credible state
Avoid:
- impossible miracle framing
- changing too many variables at once, such as different person, room, lighting, and angle
3. Mechanism Curiosity
Open on a specific action that looks like it has a reason.
Use when the product has a process, device, ingredient, or unusual application step.
Examples:
- parting hair before spraying along the scalp
- adding essence or oil into a device
- mist coverage moving along a visible path
- red light or another physical mechanism interacting with the target area
Avoid:
- abstract ingredient animation with no real-world action
- product close-ups that do not show how the product is used
4. Status Loss / Social Stakes
Open with a visible consequence of the problem.
Use when the story angle is dating, work, confidence, presentation, or identity.
Examples:
- someone dressed well for a date notices their hair looks bad in the mirror
- a message, glance, or reaction implies social rejection while the visual problem is visible
- a polished outfit clashes with one obvious flaw
Avoid:
- melodrama that overwhelms the product problem
- social stakes that feel unrelated to the visual evidence
5. Interrupted Moment
Open in the middle of a lived-in action instead of at the setup.
Use when the scene needs to feel TikTok-native and less like an ad.
Examples:
- the hand is already lifting hair
- the sprayer is already close to the scalp
- the phone is already in hand after a disappointing notification
- the person is already halfway through getting ready
Avoid:
- walking into frame
- blank room, empty sink, product on table
- slow cinematic open before anything happens
6. Proof Close-Up
Open with a tactile, specific detail that looks hard to fake.
Use when trust matters more than aesthetics.
Examples:
- nozzle-to-scalp distance
- visible spray/mist path
- part line exposure
- device loading step
- texture change, volume lift, or shine/oil contrast
Avoid:
- beauty-commercial macro shots that look disconnected from use
- clean renders with no hands, scale, or context
7. Aspirational Leisure / Soft Status
Open on an easy, desirable life moment that the viewer wants to enter.
Use when the hook is driven by lifestyle pull, beauty, social proof, romance, friendship, outdoors, travel, or "I want this kind of day" energy.
Examples:
- an attractive couple relaxing on paddleboards in clear water
- a wide overhead view where people look small inside a beautiful leisure setting
- a low-effort selfie angle with sun, water, food, skin, and relaxed body language
- a casually fit person on a board, boat, beach, patio, picnic, or vacation activity
Why it works:
- it sells a state, not a feature
- it makes the viewer borrow the identity of the scene
- the hook question becomes "Where is this?", "How do I get that day?", or "Who are they with?"
Keep it TikTok-native:
- use relaxed handheld, selfie, drone, or action-camera framing
- let the frame feel discovered, not staged
- include a small lived-in detail such as snacks, paddle, wet hair, sun glare, or imperfect pose
- keep people doing something, not just posing
Useful subtypes:
- : first-person body/board/boat/beach view that makes the viewer feel physically placed in the scene
- : show a real or simulated viewer request, then answer it through the visual sequence
- : open on a leisure setup with visible objects, then reveal "everything I brought / use / pack / own"
- : posture, tan lines, wet hair, snacks, drink cans, and gear make the moment feel lived-in instead of staged
For the comment-request subtype, the hook has two layers:
- visual layer: "I want to be lying there"
- content layer: "What exactly is on that board / in that setup?"
This works well for product sets, bundles, accessories, travel gear, outdoor gear, beauty prep, and lifestyle kits because the viewer already wants the scene before the item list begins.
Avoid:
- generic travel-influencer beauty shots with no human action
- model-catalog posing
- luxury polish that feels unreachable or ad-like
- scenery-only openings unless the environment itself is surprising
- pretty people with no relational or activity context
Slideshow Guidance
For slideshows, the first slide has to do the work of a video opening frame.
If the user is making a professional-share or method-breakdown deck, keep the copy dense enough to signal expertise. Each page should ideally contain one concrete judgment plus one operational detail, not just atmosphere or broad inspiration.
Prefer:
- a single strong contradiction
- direct evidence in the image
- large, readable visual forms
- a second-slide promise that naturally answers the first-slide question
Avoid:
- collage overload
- small before/after details that cannot be read on a phone
- cover images that rely on long text to explain the premise
Video Guidance
For video, do not spend the first second establishing the room.
Prefer:
- action already underway
- handheld or mirror-native framing when appropriate
- a clear cut from problem to process
- a reason for the next shot to exist
Avoid:
- cinematic product hero open
- slow motion packaging shots
- empty B-roll before the human problem is clear
- AI-perfect rooms that feel like stock advertising
Generated Media Prompt Rule
When writing image or video prompts, specify the hook job before the style.
Prompt order:
- visual conflict
- subject and action
- evidence detail
- camera/framing
- realism constraints
- style and quality notes
Do not start from "beautiful cinematic shot" unless beauty itself is the tension.
Critique Questions
When reviewing a proposed visual hook, answer:
- Can a viewer understand the situation without audio or caption?
- What exact question does the first frame create?
- What is the visible conflict?
- What action is already happening?
- What evidence makes it believable?
- Does it feel like a real TikTok moment or an ad setup?
- What should the next shot answer?
Common Failure Modes
- product appears before the problem
- the frame is attractive but conflict-free
- the scene is too clean, studio-like, or AI-polished
- the key detail is too small for mobile
- the hook depends entirely on text
- the opening starts before the interesting moment
- the transformation is overclaimed or visually impossible
Output Shape
When asked to create visual hooks, prefer a compact structure:
text
Hook family:
First frame / first slide:
Action in progress:
Evidence detail:
Viewer question:
Next shot:
Avoid:
When asked to critique hooks, prefer:
text
Verdict:
What works:
What weakens the stop-scroll:
Specific fix:
Rerun prompt note: