Most videos that feel long or unfocused don’t have an editing problem — they have a script problem. Writing the video before recording it saves takes, cuts rambling, and makes the edit faster because you know exactly what you said and why.
This guide is practical: a structure that works for almost any talking-head video, a template to copy, rules for writing text that will be spoken aloud, and the errors to avoid. It pairs well with a teleprompter, since a script you can read smoothly is half of a natural delivery.
The four parts of a video script
Almost every successful short-to-medium video follows the same skeleton: hook, promise, body, close. You can write it in any order — many creators write the body first and the hook last — but the four parts should be there.
1. Hook (first 5–10 seconds). One line that earns the next thirty. Name the problem, ask the question, or show the result upfront. Write it last, when you know what the video’s best moment is.
2. Promise (next 10–15 seconds). Tell the viewer what they’ll get and roughly when. “By the end of this video you’ll be able to X.” A promise gives structure to the viewer’s attention and to your script.
3. Body (the content itself). Three to five points, one idea per point. Introduce each point in a single sentence before explaining it, so the viewer always knows where they are.
4. Close (last 15 seconds). Sum up in one sentence, tell the viewer exactly what to do next (one single call to action), and stop. Two calls to action halve the clicks.
A template you can copy
This template fits a 3–5 minute video. Paste it into the editor and replace the bracketed text:
[HOOK — one line, the problem or the result]
In this video I'll show you [PROMISE — what they'll learn].
Takes about [X] minutes.
[POINT 1 — one-sentence intro]
[Explain point 1: 2–4 short sentences]
[Example or mini-demo]
[POINT 2 — one-sentence intro]
[Explain point 2: 2–4 short sentences]
[Example or mini-demo]
[POINT 3 — one-sentence intro]
[Explain point 3: 2–4 short sentences]
[Example or mini-demo]
So, quick recap: [ONE SENTENCE recap].
[CLOSING LINE — invite one action]Timing rule of thumb: conversational English is around 130–150 words per minute. A 3-minute video needs roughly 400–450 words. If your script is 900 words, it’s a 6-minute video — decide on purpose, not by accident.
Five rules for text that will be spoken
Write short sentences. One idea per sentence. When a sentence needs a second comma, split it. Spoken language tolerates fragments better than subordinates.
Use the words you actually use. “Buy” beats “purchase”, “use” beats “utilize”, “help” beats “facilitate”. Simple words are easier to say and easier to understand.
Write for one person. Picture a specific viewer and talk to them. Scripts addressed to “everyone who might be interested” sound like announcements; scripts addressed to one person sound like a conversation. And when the script needs to reach the person who’ll record it, a password-protected link beats an email attachment.
Read it aloud and fix the stumbles. Every sentence that makes you trip in a normal voice will trip you harder on camera. Rewrite those on the spot — shorter, simpler, closer to how you’d say it.
Cut the first 10%. First paragraphs warm up; viewers don’t wait. Delete the intro you wrote to start writing, and start at the point where the video actually begins.
Formatting the script for a teleprompter
A script that will be read from a screen benefits from light formatting:
- Split paragraphs into lines of one or two sentences.
- Insert blank lines where you want to pause — they are visible breathing marks.
- Bold the words you tend to rush past, and highlight numbers or names you always forget.
- Spell out symbols you’ll read aloud: “50 percent” instead of “50%”, “dot com” instead of “.com”.
The instructions page shows where each of these options lives in the editor and the teleprompter view.
Common mistakes
- Writing for the eye, not the ear. If it reads well but sounds stiff, it’s a newspaper sentence. Rewrite it as something you’d say.
- No hook. Starting with “Hi, welcome back to my channel” spends your best seconds on the least informative line. Greet later, or let the hook do it.
- Three ideas per paragraph. The viewer can’t reread. One idea per paragraph, one paragraph per block on screen.
- Numbers nobody can follow. Round them, slow down before them, and repeat the important ones. “Around 150 words per minute” is better on camera than “147”.
- A close with no action. “Thanks for watching” is fine, but add one specific thing to do next: watch a related video, try the tool, download the template. Just one.
- Never rehearsing. A single read-aloud pass catches 80% of the problems in two minutes.
From script to recording
Once the script is done, the workflow is short: paste it into the editor, format the pauses, choose the font size, and do one dry run to pick the scroll speed. If it’s your first time reading on camera, our first-time teleprompter guide covers the reading side in detail — the script is only half of a natural delivery.
Open the editor, paste your template, and write the video before you record it. Your future self, editing half the footage, will thank you.