git:20261002.1fcdb26 to git:20261002.d5668a9

- description: Use when writing, rewriting or reviewing the opening seconds of a short-form video (Reel, TikTok, YouTube Short, LinkedIn video), a video script opening, or a storyboard's first frames. Builds an aligned visual, text, spoken (and audio) hook on one idea, picks a proven hook format, and checks the payoff into the body.

- A hook has one job: make the right viewer opt in. It needs topic clarity and on-target curiosity, and curiosity comes from contrast [7][1]. Every hook has layers: what you show (visual), what you write on screen (text), what you say (spoken), plus the audio/sound effects [1][2]. The difference he names between 500 and 500,000 views is maximum alignment between those layers on ONE idea [1].

-

- ## Core model (internalise before writing)

- 1. **Curiosity loop through contrast.** The viewer believes A; the hook shows B. The distance between A and B is the contrast; bigger contrast, deeper hook. You have about 5 seconds to make it stick [1].

- 2. **Viewers read the screen before they hear you.** Order of intake is visual, then audio, then back to the visual to confirm what they heard ("visual, audio, visual" sandwich) [1]. Eyes take in far more per second than ears [1][3]. So the spoken hook is the least important of the three layers [3]; text and visual carry the stop. (He ranks text above visual in one post [2] and visual above all in others [1][3]; both agree spoken comes last. Treat text + visual as one unit that must work with sound off.)

- 3. **Misalignment = comprehension loss = bounce.** If the spoken line says something the visual does not show, the viewer gets confused and leaves [2][1]. Even a slightly different visual weakens the hook [1].

- 4. **Spend disproportionate effort on the opening.** Hooks are the "80 of the 80/20" [1][7]; he recommends ~80% of effort on the opening [11].

- ## Method (run in this order)

- 1. **Find the key visual first.** Before writing words, list the visuals you can put in the first 3 to 5 seconds. The better the visual, the more you lean on it. If there is no good visual, either manufacture one (motion graphics, stock) or seriously consider not making the video [1].

- 2. **Mine the biggest contrast.** List the most interesting facts or angles; score how surprising each is (he uses a 1 to 100 shock score) and keep the ones most viewers do not know [8]. The fact with the biggest contrast decides the format [1].

- 3. **Pick one hook format that both maximises contrast and matches the key visual** [1]. Any topic can be framed in any format, so choose by fit to the visual, not habit [1]. His six archetypes [1]:

- - **Fortune Teller**: present vs future. Establish current reality, show how the future changes, tease it ("This is the future of home design"). Good for launches, news, innovation [1][4].

- - **Experimenter**: peer-to-peer demo. Base pain point, show you tried a new method/tool, show why it is better. Good for product demos and B2B tools; best when the product itself is on screen in the hook [1].

- - **Teacher**: teacher-to-student lesson. Pain point, method/learnings, why it works. Builds authority [1].

- - **Magician**: a visual or verbal stun gun ("check this out", snap, wig reveal, rapid match cut) or a "visual pacifier" (passive activity on screen). Stack it in front of any other format [1].

- - **Investigator**: a secret or finding the viewer does not know; frame against their current not-knowing [1].

- - **Contrarian**: state outright what you believe that most people do not ("you're doing X wrong") [1][4].

- Short alternate set from his IG: Fortune Teller, Contrarian, Insider, Proven Winner (lead with a result that validates the claim), Eye Magnet, Connector (link two things that do not normally go together) [4]. A longer set of nine formats (Secret Reveal, Case Study, Comparison, Question, Education, List, Contrarian, Personal Experience, Problem) appears in his scripting video [8]. To choose, check which one or two formats win in your niche via outlier videos [8].

- 4. **Write the spoken hook: 2 to 4 lines** [1] using the three beats [6][9]:

- - **Context lean** (1 to 2 sentences): state the topic plainly so wrong viewers exit and right viewers lean in, via common ground, a benefit/pain point, a simplifying metaphor, or a mind-blowing fact [6][9]. When showing the key visual, literally describe what is on screen ("These are life-size floor plans...") for pure alignment [1].

- - **Scroll-stop interjection** (1 sentence): a contrast connector (but, however, yet) that stuns [6][9].

- - **Contrarian snapback** (1 sentence): snap the viewer the opposite way of the lean, still on topic [6][9]. Skip the explicit interjection if the visual is already contrast enough [1].

- 5. **Check the words carry the power-word slots**: Subject clarity word, Action word, Objective/state-change word, Contrast word (the most important, compares outcome vs base state). Two more optional slots raise intrigue [5]; per a secondary write-up they are Proof and Time [10].

- 6. **Write the on-screen text hook** to support the key visual and to re-confirm it after the viewer hears the speech [1]. It is larger overlay text, not captions [11]. Prefer a promise/category over an unfamiliar term: the 15M-view floor-plans video used "future of home design" plus an arrow pointing to the key visual [1][9].

- 7. **Visual motion**: just enough motion. Too much overwhelms, too little bores [6]. Visual hook families: graphic and text overlays, pattern interrupts, subject motion, visual effects, unique visual selection [3].

- 8. **Watch it back with sound off, then on.** Ask: going visual, speech, visual, is the viewer clear what this is about, and does it match where the rest of the video goes? If not, stop and rework the hook; if no clear aligned hook exists, can the idea [1].

- 9. **Pay it off immediately.** Deliver the first hit of value right after the hook, not buried later [6]. Never open a loop with nothing at the end [9].

- ## Self-review checks (run before returning output)

- - [ ] One idea: visual, text, spoken (and audio) all say the same thing; nothing in speech the picture does not show [1][2].

- - [ ] Sound-off test: the frame + on-screen text alone tell the viewer the topic [1][2].

- - [ ] Topic is clear within the first 1 to 2 seconds; no throat-clearing ("guys, this is crazy...") [7].

- - [ ] Clarity: fewer words, simpler words, active voice, about 6th-grade reading level; read it in isolation and confirm only one interpretation [7][9].

- - [ ] Relevance: framed as "you/your" unless the "I" result is itself the hook; framed as a need-to-have pain point, not a nice-to-have [7][9].

- - [ ] Contrast present: A (common belief/status quo) vs B (faster, better, cheaper). Use stated contrast unless the audience's baseline is certain [7][9].

- - [ ] Format named and it is the one that best fits the key visual [1].

- - [ ] Spoken hook is 2 to 4 short lines; short, staccato sentences at the start [1][6].

- - [ ] Subject, Action, Objective, Contrast words all present [5].

- - [ ] Hook checks: Alignment [1], Speed to value [6], Clarity [8], Curiosity through contrast [1].

- - [ ] Next beat pays off the promise directly [6][9].

- ## Failure modes to avoid

- - **Abstract key phrase with no showable visual.** His own sub-100K example: "future of storytelling" over floating rooms plus jargon ("generative world model") caused comprehension loss; the fix was something concrete and showable, or cutting the idea [1].

- - **Spoken hook contradicts or drifts from the visual** [2].

- - **Polishing the spoken line while ignoring the visual stun** [3].

- - **Delay**: topic arrives late, vague suspense [7].

- - **Confusion**: convoluted sentence, passive voice, jargon [7].

- - **Irrelevance**: "I" framing that makes viewers decide if they identify with you [7].

- - **Disinterest**: clear and relevant but no contrast [7].

- - **Trick hooks** that pull the wrong audience, and **empty snapbacks** with no payoff [9].

- - **Unfamiliar idea with no anchor**: "cult hop" by wrapping it in a known reference (brand, celebrity) [6].

- - **Using AI to invent hooks from a blank prompt**; he uses AI to rewrite for clarity at a 6th-grade level, keeping the meaning [7][9].

- ## Worked examples (illustrative, written for this skill, not Kallaway's)

- **1. SaaS feature launch (Fortune Teller + Magician)**

- - Before: Spoken "Hey everyone, we're super excited to share our new AI update that leverages agents to streamline workflows." Text: "BIG NEWS". Visual: logo animation.

- - Problems: delay, jargon, logo shows nothing, text says nothing, no contrast.

- - After: Visual: screen recording, a 40-row spreadsheet filling itself in under 3 s. Text: "Spreadsheets that fill themselves" + arrow at the filling column. Spoken: "Check this out. This spreadsheet is filling itself in. You paste a list of companies and it researches every row. Most teams still do this by hand for hours."

- **2. Coaching tip (Contrarian)**

- - Before: Spoken "I've been thinking a lot about productivity lately." Visual: talking head. Text: none.

- - After: Visual: hand dragging a to-do list into the trash. Text: "Stop writing to-do lists". Spoken: "If you write a to-do list every morning, you're probably getting less done. Most people think the list is the plan. But the list is the problem."

- **3. Product demo (Experimenter, stated contrast)**

- - Before: Text "Our editor is fast". Spoken "Our video editor has many features."

- - After: Visual: split screen, left timeline being hand-cut, right one cut automatically, timer on both. Text: "4 hours vs 4 minutes". Spoken: "You're editing podcast clips by hand. I gave the same episode to this tool. It cut 10 clips before I finished my first."

- ## Output format: hook storyboard

- ```

- FORMAT: <archetype> (+ Magician if stacked) CONTRAST: A = <belief> -> B = <new reality>

- KEY VISUAL: <the one shot the hook is built around>

- 0.0-1.0s VISUAL: <shot, motion, framing> TEXT: <overlay, position, arrow/target>

- SPOKEN: <line 1, context lean> AUDIO: <sfx/music cue>

- 1.0-2.0s VISUAL: ... TEXT: ... SPOKEN: <interjection or continuation>

- 2.0-3.0s VISUAL: ... TEXT: ... SPOKEN: <snapback / promise>

- SOUND-OFF READ: <what a muted viewer understands from frame + text>

- NEXT BEATS:

- Beat 1 (3-6s): <first payoff of the promise, what is shown>

- Beat 2 (6-10s): <next contrast / re-hook>

- CHECKS: alignment Y/N | sound-off Y/N | topic by 2s Y/N | contrast Y/N | S-A-O-C words Y/N | payoff Y/N

- ```

- ## Sources

- 1. Kallaway, "I Studied 100 Viral Hooks, These 6 Will Make You Go Viral" (YouTube xnOe8aA9Pmw), transcript: https://you-tldr.com/transcript/xnOe8aA9Pmw (video: https://www.youtube.com/watch?v=xnOe8aA9Pmw)

- 2. Kallaway IG, "Hook Alignment": https://www.instagram.com/reel/DbyRmzAu81r/

- 3. Kallaway IG, "Visual Stun Gun": https://www.instagram.com/reel/DbA5ScOOYDl/

- 4. Kallaway IG, "these 6 hook formats always outperform": https://www.instagram.com/reel/Das2r1lOB0B/

- 5. Kallaway IG, "6 Hook Power Words": https://www.instagram.com/reel/DbQgNloOigA/

- 6. Summary + description of "How to Create Irresistible Hooks" (LmXpbP7dD48): https://videohighlight.com/v/LmXpbP7dD48

- 7. Summary of "Give me 15 mins, and I'll make your hooks impossible to skip" (2byPP_9F0-Q): https://videohighlight.com/v/2byPP_9F0-Q

- 8. Summary of "How to Script Viral Videos 10x Faster" (_Z11mjFh2zY): https://videohighlight.com/v/_Z11mjFh2zY

- 9. Secondary analysis of [6] and [7] with quotes: https://build-os.com/blogs/source-analyses/kallaway-irresistible-hooks

- 10. Secondary analysis of "The ONLY 6 Words You Need to Hook ANY Viewer": https://build-os.com/blogs/source-analyses/kallaway-6-words-hook

- 11. Creator Spotlight profile (six ingredients, hook layers, 80% effort): https://www.creatorspotlight.com/p/kallaway

- 12. Kallaway's 7-day course outline (Day 3 = hook framework): https://shortformsystem.co