Video is the most powerful content format in marketing right now. Short-form video delivers the highest ROI of any content type, with 49% of marketers naming it their top channel in 2026 (HubSpot State of Marketing). But most brands waste the format by writing video copy the same way they write blog posts. That's the mistake this lesson fixes.
Why Video Copy Is Different
Written copy lets readers set their own pace. Video is a one-way contract, the viewer either stays or leaves. Every sentence has to earn the next one.
Spoken rhythm matters more than grammar. A sentence that looks clean on a page can feel exhausting out loud, too many syllables, too much nesting. Write for the ear, not the eye.
Read your script aloud before shooting. If you run out of breath mid-sentence, it's too long. If your eyes glaze over on a paragraph, the audience's will too.
Visual cues are part of the copy. A script isn't just what someone says, it's what the viewer sees. Good video copy accounts for both tracks simultaneously.
The First 3 Seconds: Short-Form Hook Formula
On TikTok, Reels, and Shorts, viewers decide within 3 seconds whether to scroll past. That's not a metaphor, the algorithm measures it. Your opening line is not an introduction. It's a trap.
Three hooks that consistently perform:
- Pattern interrupt: Say or show something unexpected. "Stop spending money on Facebook ads until you watch this."
- Bold claim: Lead with the result. "I went from 200 to 80,000 followers in 6 weeks with one change."
- Curiosity gap: Tease without delivering. "The mistake 90% of marketers make in their first 5 seconds."
Never open with your name, your company, or 'Today I'm going to talk about.' Viewers are not waiting for your credentials, they're waiting for a reason to stay.
Test hooks in isolation. Shoot the same video with 3 different opening lines, post all 3 as separate clips, and let the retention data tell you which hook wins.
The 3-Act Structure for 60–90 Second Videos
Most short-form marketing content lives in the 60–90 second window. That's enough time for a complete story if you're ruthless about it.
Act 1, Hook (0–5 sec): The opening claim, question, or visual interrupt. One sentence maximum.
Act 2, Value or Story (5–75 sec): Deliver the thing you promised. One core idea, demonstrated or explained with specifics, not a list of five things. Specificity is credibility.
Act 3, CTA (75–90 sec): One action. 'Follow for more,' 'Link in bio,' 'Comment below.' Never stack CTAs, it creates decision paralysis and the viewer does nothing.
Hook: "Here's why your ads get clicks but no sales." → Value: Walk through one specific landing page mismatch with a screen recording. → CTA: "I break down more of these every week, follow to catch the next one."
Long-Form YouTube Scripts: Don't Save the Good Stuff
YouTube creators lose more than 55% of viewers in the first 60 seconds (2025 retention data). That means the intro is not a warm-up, it's a battleground.
The correct YouTube script structure:
- Hook (0–30 sec): State the outcome the viewer will have by the end. Not the topic, the transformation.
- Credibility moment (30–60 sec): One sentence on why you can speak to this. Then move on.
- Value delivery: Chunk the content into 3–5 sections. Plant a forward-reference at the end of each section: 'In a minute I'll show you how to fix this at the script level.'
- Mid-video CTA (around 40–50%): Ask for the subscribe or the click here, not at the end. Most viewers who will act on a CTA have already dropped off by the final 30 seconds.
- Closing CTA: Brief, low-pressure repeat. 'If that was useful, the next video on [topic] is right here.'
Pattern interrupts every 60–90 seconds, a cut, a new visual, a bold statement, reset the viewer's attention clock and push retention rates up.
Ad Video Scripts: Hook, Agitate, Solve, Prove, Close
Paid video ads follow a tighter structure because cold audiences owe you nothing. The format that converts:
- Hook: Name the problem in the viewer's language. Not your language, theirs.
- Agitate: Make the cost of inaction real. One concrete consequence.
- Solution: Introduce your product or service as the mechanism. Focus on the change it creates, not the features.
- Proof: One stat, one testimonial clip, one before/after. Pick the most credible one, not all three.
- CTA: One action, with urgency if honest.
Format dictates compression:
| Format | Word count | What survives |
|---|---|---|
| 15-sec | 35–40 words | Hook + CTA only |
| 30-sec | 70–80 words | Hook + solution + CTA |
| 60-sec | 140–160 words | Full structure |
Pacing and Sentence Length
Sentence length controls perceived pace on camera. Short sentences feel fast. Long sentences with multiple clauses feel slow, academic, and, when the speaker tries to rush through them, anxious.
A practical rule: no sentence in your script should be longer than 12 words. If it is, split it. When you speed up visually (jump cuts, B-roll, captions), shorter sentences compound the energy instead of fighting it.
Writing Visual Cues Into the Script
A talking-head script is a monologue. A shooting script tells the editor what's happening on screen during every line. Both are valid, but shooting scripts produce better videos.
Use bracketed cues inline:
[B-roll: product on desk, close-up]
"This one tool replaced three apps in my workflow."
[Cut to: screen recording of dashboard]
"Here's what that looks like in practice."
[Cut back to camera]
"The setup takes about ten minutes."
You don't need a production team for this. Even solo creators shooting on a phone can cut to a phone screen recording or a text overlay. Cues give the editor options, and options mean better retention.
Don't write visual cues that require footage you can't actually shoot. If you're a solo creator, keep B-roll cues realistic, screen recordings, product shots, and text overlays are achievable. Abstract cinematic cues ('aerial shot of city') are not.
The Watch-Time Rewrite
Before you shoot, do one final pass specifically for drop-off moments. Read the script and mark every line where a viewer might think 'I get it, I can leave now.'
Those moments are usually:
- Summaries that arrive too early ('So, to recap...')
- Transitions that feel like endings ('Now let's move on to...')
- Repeated points dressed as new ones
- Padding phrases: 'Like I mentioned,' 'As you can see,' 'Really quickly'
Cut the lines that feel like permission to leave. Add a forward-reference where you find a natural exit: 'But here's where most people go wrong, and I'll show you in the next section.'
This single rewrite habit is the difference between 30% average view duration and 60%.
The best video scripts feel like a conversation, not a presentation. If yours sounds like a slide deck being read aloud, cut the transitions and just say the next thing.







