Video Caption Creation

Video Caption & On-Screen Hook Writer (v3)

On-screen text is the #1 visual element in short-form video. It does the PRIMARY work of stopping a scroll. A viewer sees the text and decides in <1 second whether to stop. This skill creates the text that makes them stop.

The only question that matters: What text would make a parent stop scrolling - without hearing a single word?

Not "what complements the audio." Not "what pattern fits this clip." Just: would this make someone stop?

Previous versions: SKILL_v1_archive.md (basic Triple Word Score), SKILL_v2_archive.md (Complementarity Principle era)

The Transformation Rule (Core Framework)

From Andrew Muto's caption optimization work. This is the single most important thing in this skill.

Every weak hook can be transformed using three moves:

The pattern: Weak hooks describe the content. Strong hooks provoke an emotion. The transformation is always from WHAT THE CONTENT IS ABOUT to HOW IT MAKES SOMEONE FEEL.

Before/After Swipe File

Study these transformations. They teach you more than any framework:

The Acid Test for Every Hook

Ask these three questions in order:

1. Would someone who's never heard of this episode understand this hook? If no, you're writing inside baseball.
2. Does this hook describe the content or provoke an emotion? If it describes, transform it.
3. Say the hook out loud to a stranger. Do they lean in or look confused? Confused = rewrite.

Three Functions of On-Screen Text

On-screen text isn't decoration. It does three jobs simultaneously (Andrew Muto):

1. Immediate Context - Tell viewers what they're watching from frame one. A silent scroller needs to know in <1 second: is this for me?
2. Drama Injection - Add tension when the opening audio lacks punch. Most podcast clips start conversationally. The text supplies the urgency the audio doesn't have yet.
3. Search Optimization - Help algorithms understand your content. Topic words in the text overlay get indexed alongside the auto-transcribed audio.

When writing a hook, check it against all three. A hook that nails drama but gives zero context about the topic will get views but wrong-audience views. A hook that's all context ("Homeschool math tips") has zero stopping power.

The First-3-Words Test

The first 3 words someone reads do 80% of the work. Before finalizing any hook, check:

What are the first 3 words?

• "Your kid's Minecraft" → Instant parent identification
• "Stop raising kids" → Direct challenge, emotional
• "Never give up" → Inspirational + universal
• "55% of teachers" → Specificity (tells me this is data)
• "She's a parenting" → Story setup with credential

Weak first 3 words:

• "Here's what happens" → Generic, could be anything
• "Check this out" → Zero information
• "In this clip" → Meta, breaks immersion
• "You need to" → Commanding without earning attention
• "The research shows" → Academic, not emotional

Front-load the punch. Move the most surprising, specific, or provocative word as close to position 1 as possible.

Length Variety (Critical)

Every set of 3-4 hook options MUST include different lengths. This is non-negotiable. If all your hooks are 3-5 word punchy phrases, you're in a rut.

The trap: AI defaults to short and punchy because it pattern-matches "hook = headline." Fight this. Some of the best-performing on-screen text is conversational and long.

Hook Categories That Work

1. Polarizing Statements

Take something the audience worries about and reframe it as positive, or take something accepted and challenge it.

• "Your kid's [negative thing] is actually genius"
• "Stop raising [undesirable trait] kids"
• "Homework is a scam"
• "Your ADHD kid isn't broken"
• "Your kid's C+ is fine"

2. Counter-Intuitive Reveals

Pair a negative with an unexpected positive reaction. The juxtaposition creates the hook.

• "My son's first job sucked. Perfect."
• "Screen time was never the enemy. Boredom was."
• "The worst experiences are the best teachers"

3. Direct Challenges

Command the viewer. Works when the challenge validates what they already secretly believe.

• "Never give up on the weird kid"
• "Stop raising kids who wait for permission"
• "Make your kids love learning again"

4. Curiosity Gaps

Hint at what happened without revealing it. Story setup that demands completion.

• "She's a parenting expert. Her own kid broke her."
• "His kids skipped school for 100 days. Here's what happened."
• "Finland banned homework below age 10. Now they're #1 in education."

5. Identity Hooks

Name the viewer's identity so they self-select in.

• "If you're a [identity], watch this"
• "Things I don't explain to strangers anymore: Why my kids aren't in school"
• "My kids socialize with humans of all ages, not just 24 kids born the same year"

6. Data Punches

Lead with a specific number. Specificity = credibility = stopping power.

• "55% of teachers want to leave the profession"
• "Edison was kicked out of school for being 'too curious'"
• "5 Signs Your Kid Doesn't Actually Hate Math"

Proven Hook Library (Templates)

These are templates organized by format. Use framework fitting at runtime to adapt them to specific content.

Results and Transformation

Curiosity Gap and Contrarian

Authority and "Secret"

Question and Relatability

Podcast Clip On-Screen Hook Workflow

Step-by-step process for generating hooks from podcast transcript clips. Used by the podcast-production skill at Step 4.

Step 1: Find the Emotional Core

Read the clip transcript and ask: What's the "quiet part" we can say out loud?

Look for (SURP framework):

• Surprising fact or statistic ("55% of teachers want to leave")
• Unexpected quote or bold statement ("you've lost the argument")
• Relatable problem ("kids feel lost, so they retreat to the screen")
• Provocative opinion ("our path to victory will be to trap families")

The hookable moment is usually NOT the first sentence of the clip. Find the single line that would make someone stop scrolling if they saw it as a headline.

Step 2: Transform It

Take whatever you found and run it through the Transformation Rule:

• Is it technical/educational? → Make it emotional/dramatic
• Is it a long explanation? → Compress to a punchy statement
• Is it inside baseball? → Find the universal truth underneath

The question to ask: "A parent scrolling Instagram at 10pm, sound off, has never heard of this podcast. What text makes them stop and tap?"

NOT: "What text complements the audio?" That's a useful secondary consideration, but it's not the starting point.

Step 3: Generate 3-4 Hook Options

For each clip, write 3-4 options with LENGTH VARIETY. Each set should span at least 2 different length categories (punchy, statement, narrative, full quote).

Scan the Hook Categories and Proven Hook Library for a matching pattern. Adapt the pattern to this specific content while preserving the format's proven structure.

Warning signs you're writing weak hooks:

• Uses jargon or vocabulary from the episode
• Explains rather than intrigues
• Generic phrases anyone could write about any clip
• Requires episode context to understand
• Describes the content instead of provoking an emotion

Step 4: Quality Gates

For each option, check:

• McDonald's Test: Someone at McDonald's understands instantly?
• First-3-Words Test: First 3 words carry the punch?
• Scroll Test: Would this make ME stop scrolling in my feed?
• Grandmother Test: Does this hint at the payoff without giving it away?

Step 5: Mark Recommended Pick

Select one recommended option and explain WHY in one sentence. The rationale should reference what makes it stop a scroll - not how it relates to the audio.

Audio-Text Relationship (Secondary Principle)

Once you have a hook that stops the scroll on its own, THEN consider how it works with the audio. This is the old "complementarity principle" - still valid, just not the starting point.

Good pairing: Text creates a frame, audio fills it in. The gap between what you read and what you hear creates curiosity.

The test: Does reading the text AND hearing the audio create a fuller picture than either alone? If the text just labels what the audio says, it's wasted real estate.

Identifying Hookable Moments in Transcripts

When scanning a full transcript (before clips are selected), look for:

Tier 1: Almost always hookable

• Specific statistics ("55% of teachers want to leave")
• Moments that contradict popular belief
• Emotional peaks (speaker's voice changes, pace quickens)
• Universal pain points ("every parent knows...")
• Confessions from experts ("I'm a parenting expert and I lost it on my kid")

Tier 2: Often hookable

• Personal confessions or vulnerable moments
• "I should probably not say this, but..."
• Origin stories (how they started, what went wrong)
• Direct statements that take a side

Tier 3: Rarely hookable (avoid)

• Agreement moments ("yeah, totally, I think so too")
• Background context / scene-setting
• Nuanced, heavily-caveated statements
• Jargon-heavy industry talk
• Named concepts/frameworks that only make sense with explanation

The whole transcript matters. The strongest moment might be at minute 48. Don't mine just the first half.

The Triple Word Score System

Four signals must align so algorithms AND humans immediately know "this is for me":

1. Audio Transcript (MOST IMPORTANT)

• What the speaker says out loud - algorithms auto-transcribe this
• Topic words must appear in first 10 seconds
• Core terminology repeated naturally throughout

2. On-Screen Text Hook

• Visual overlay that stops the scroll (everything above)
• NOT a repetition of the audio - adds a new frame
• Lead with topic words when possible

3. Caption Copy

• Post description with topic-relevant keywords
• Opens with a topic-relevant phrase
• Provides context the algorithm needs

4. Strategic Hashtags

• 10-12 total (optimal range)
• Broad → Mid → Specific → Niche → Audience → Platform
• Example: #Education #Parenting #Homeschool #Microschools #SchoolChoice #HomeschoolMom #Shorts

When all four align, the algorithm recognizes the topic immediately and serves it to the right people.

Caption Writing

Short-Form Platforms (Same Caption Everywhere)

Applies to: YouTube Shorts, Instagram Reels, TikTok, Facebook Reels

One caption per clip. Don't write platform-specific versions (waste of time, same audience).

Format:

[Verbatim quote or key insight from clip - 1-2 sentences]
[Context: who the speaker is and why it matters - 1 sentence]

#Hashtags (10-12)

Optional X Variant: Shorter, more conversational. Include @handles. Drop the context sentence.

Caption Quality Check

• Opens with the strongest line from the clip
• Includes guest name and title
• Hashtags span broad to niche
• Under 150 characters for the hook portion
• No emojis in body (rare exceptions for social captions)

Output Format

When this skill is invoked (standalone or from podcast-production), produce this per clip:

### [Clip Name]

**Timestamp range:** [MM:SS-MM:SS]

**On-screen hook options:**
1. "[Hook text]" *
2. "[Hook text]"
3. "[Hook text]"
4. "[Hook text]"

(* = recommended pick. List 3-4 options. Star goes on the strongest.)

**Caption (FB, TikTok, IG, LinkedIn):**
[Caption text]

**X variant:**
[Shorter caption with @handles]

Keep it clean. No category labels, no rationale paragraphs. The hooks should speak for themselves. Use the transformation rule and quality gates internally when generating, but the output is just the options with a star.

Sub-Agent Prompt Template

When podcast-production invokes this skill at Step 4, use this prompt:

You are generating on-screen text hooks for podcast clips.

Episode: [Guest Name]
Working directory: [path to prep/]

Read the following files:
- SOURCE.md (full transcript for context)
- EDITOR_HANDOFF.md (clips already selected in Sections 3 & 4)

For EACH clip in Sections 3 and 4, generate 3-4 on-screen hook options.
Mark the recommended pick with an asterisk (*).

Output format per clip:
1. "Hook text" *
2. "Hook text"
3. "Hook text"
4. "Hook text"

No category labels. No rationale paragraphs. Just the options with a star.

Internal rules (apply these but don't show them in output):
- TRANSFORM every hook: Technical→Emotional, Explanation→Statement, Inside Baseball→Universal Truth
- The hook must work for someone who has NEVER heard this episode, scrolling silently
- Apply the McDonald's Test (instantly understandable by anyone)
- Apply the First-3-Words Test to every option
- VARY THE LENGTH across options: at least one short punchy (2-5 words), one statement (6-9 words), and one narrative or full-quote (10+ words)
- If a hook uses jargon, metaphors, or concepts from the episode, REWRITE IT as a universal truth
- Scan the Proven Hook Library in the skill for matching frameworks

Write the hooks directly into the EDITOR_HANDOFF.md clip sections.

Exemplar Channels (Study These)

Podcast Clip Masters

• Diary of a CEO (@doac.clips) - Credential leads, emotional transitions. Text builds anticipation without giving away payoff.
• My First Million (clips channel) - Entrepreneurship hooks, "how to" + numbers format
• Huberman Lab (Essentials) - Dense science distilled to protocols. Direct promise hooks.

Education/Parenting Creators

• Dr. Becky Kennedy (@drbeckyatgoodinside, 3M) - Named strategy hooks ("The REPAIR Strategy"), problem → solution format
• Big Little Feelings (@biglittlefeelings, 3.6M) - "When your toddler refuses to..." pain point hooks
• Busy Toddler (@busytoddler, 2.4M) - Action-oriented hooks
• @littlefenders (132K) - "Redefining learning + parenting / Educating kids with AI / Unschooling / ADHD Mom"

Common Mistakes

1. Correlative constructions. "Your kid isn't X, they're Y" / "Kids don't need X, they need Y" / "Not X, but Y." This is the #1 AI tell. Find another way. Say one side of it and let the viewer fill the gap. "Delete this word" beats "Stop using this word and use this other word instead."
2. Writing about the episode instead of for the viewer. "Why vs What" is meaningless to someone who hasn't watched. "One word shuts kids down. The other makes them think." is universal.
3. All hooks the same length. If every option is 3-5 word punchy headlines, you haven't explored the full range.
4. Describing content instead of provoking emotion. "Parenting expert shares advice" describes. "She's a parenting expert. Her own kid broke her." provokes.
5. Using episode-specific metaphors. If the guest coined a term or used a metaphor, it won't mean anything to someone who hasn't heard the episode. Find the universal truth underneath.
6. Giving away the payoff. If the text reveals the full insight, there's no reason to watch.
7. Generic first 3 words. "Here's what happens" could be anything. "Your kid's Minecraft" is specific.
8. Fancy vocabulary. "Pedagogical paradigm shift" fails the McDonald's test. "Schools are broken" passes.
9. Starting with the audio-text relationship. Start with "would this stop a scroll?" THEN consider how text and audio work together.

Related Skills

• podcast-production - Invokes this skill at Step 4 (on-screen hook generation)
• short-form-video - Full production workflow (this skill handles text only)
• text-content - For text-only social posts (LinkedIn, X, not video)
• youtube-title-creator - Title + thumbnail (same principle, different application)
• cold-open-creator - Cold opens use [SWOOSH] transitions, not hooks

Research References

• Short-Form Video Text Hook Strategies.md - Comprehensive report on hook psychology, typography, retention
• Andrew Muto notes: Archive/social-media/Reel/Notes from Andrew 203afe52ef5980d2bb36ebef56c524a0.md - Transformation framework, McDonald's test, 3 functions, hook categories
• Podcast-production skill improvement notes: Studio/Podcast Studio/Amar-Kumar/prep/SKILL_IMPROVEMENT_NOTES.md

v3 rewritten March 6, 2026. Replaced Complementarity Principle as core framework with Andrew Muto's Transformation Rule after Laura Froyen session revealed inside-baseball hooks. Complementarity retained as secondary audio-text consideration.