How to write a video script (with a template and an example)

A video script is the document a video is made from: what the viewer sees on the left, what they hear on the right, one row per moment, each with a start time. To write one, settle who it is for and the one thing they should take away, outline it in timed beats from the hook to the call to action, write the narration for the ear, then give every line a picture. Narration runs at about 130 to 150 words a minute, so a one-minute video holds roughly 120 words once you leave room for shots with no talking.

The two-column script format

Most nonfiction video (explainers, tutorials, ads, training, documentaries, YouTube) is written as an A/V script: two columns, for audio and video, with one row per moment. A row is a shot or a short run of shots and the words and sounds that play over it. Video usually goes on the left and audio on the right. Some tools flip them, which matters less than keeping each row's picture and words side by side. A narrow time column at the far left shows the pacing at a glance.

The format earns its keep before the shoot. In a single column it is easy to write three paragraphs of narration with nothing to show; in two, the empty cell gives it away. The narrator reads down the right, the editor down the left, and everyone shares the row numbers. The A/V script template has the columns set up.

Shorthand you will see in each column

When to use screenplay format instead

A short film with characters, scenes and dialogue and no narrator is written as a screenplay: scene headings such as INT. KITCHEN – NIGHT, action lines, and character names above their dialogue. Commercials and social ads usually stay in two columns even when actors speak, so the client can approve the picture and the words together.

How long a video script should be

Narration guides put a natural pace at 130 to 150 words a minute. Newsreaders go faster: one study of radio news measured BBC English bulletins at an average of 168. A tutorial, where the viewer is also watching your hands, should go slower. Plan at about 140, then time your narrator, because theirs is the only pace that counts.

Then write under it. A video needs moments where nobody talks: a reveal, a result, a beat of music between sections. Wall-to-wall narration makes a video feel like a podcast with pictures.

Pick the length from where the video will play, not from how much you have to say: as a starting point, a social ad at 15 to 30 seconds, an explainer at 60 to 90, a tutorial as long as the steps need and no longer.

Structure: hook to call to action

Short videos that hold attention tend to use the same five beats. The time budgets below are starting points. Adjust them to the content, not the content to them.

The first five seconds

The hook is where viewers decide whether to stay, so it gets the most rewriting. On YouTube you can watch this happen: the retention report in YouTube Studio includes an Intro figure, the share of viewers still watching after the first 30 seconds. A greeting, a logo sting or a reminder to subscribe spends those seconds on nothing the viewer came for. Open with something they can see.

Starting mid-task is the screen version of in medias res. Whichever hook you use, the promise beat after it should say plainly what the video will deliver. For channel videos, the guide to writing a YouTube script goes further into retention.

The middle and the end

In the body, open each beat with a line that says where the viewer is (“Step two: the anchors.”), keep to one idea per beat, and change the picture often; even a cut to a closer frame of a talking head resets attention. In longer videos, raise a question before you answer it (“There's one tool here you won't expect.”) so there is a reason to stay through the next beat. End on the payoff, not a summary: show the result or return to the opening image, give one call to action, and stop. The logo and the thank-you, if any, go last.

How to write a video script

1. Write a three-line brief

   Answer three lines before anything else: who is watching, the one thing they should remember, and what they should do next. Add where it will play, since a muted feed and a training portal want different openings. If the takeaway won't fit in one sentence, you have two videos. A logline does the same job for a film.

2. Outline in timed beats

   Give each beat from the structure table a time budget that adds up to your target length, and one line on what it has to do. In a tutorial the body beats are the steps; in an explainer, the problem, how the solution works, and the proof.

3. Write the audio column first, for the ear

   Draft what will be said before you plan a shot. It is easier to find pictures for a clear argument than an argument in a pile of footage. A fast first draft: record yourself explaining it to a friend, then clean up the transcript.

   Listeners can't reread, so the rules of readability tighten: one idea per sentence, contractions, numbers said the way people say them (“about a third,” “eighty dollars”), and signposts (“Two things. First…”). Spell hard names phonetically in brackets for the narrator.

4. Give every line a picture

   Fill the video column row by row: the shot (WS, CU, screen recording, B-roll) and any text on screen. An empty cell means footage you have not planned. Then apply show, don't tell to the narration. If the picture already shows it, cut the words and spend them on what the picture can't show: why, how much, what happens next.

5. Plan for the sound being off

   Some viewers will watch muted: in a feed, on a train, at a desk without headphones. Put the facts that matter in SUPERs (the name, the date, the price, the action). Captions do the rest, and the script is most of that work: YouTube Studio's auto-sync takes a pasted transcript and sets the timings for you.

6. Read it aloud with a timer and cut

   Read the audio column at the narrator's pace, time each row, and write the real start times in. Anything you stumble over twice gets rewritten. If it runs long, cut whole sentences rather than a word from each; a short script read well beats a long one read fast.

7. Turn the script into production lists

   Sort the video column by location and it becomes the shot list; the B-roll you named becomes a list to gather. On YouTube the start times also give you chapters: put timestamps in the description, the first at 0:00, at least three in all, and each chapter at least ten seconds long.

Video script example: a 75-second explainer

Here is a complete script for a short explainer: a public library announcing that it will lend tools and gadgets. The library, the people and the numbers are invented. It has about 130 words of narration and dialogue, fewer than the table allows for 75 seconds, because seven seconds go to a montage with music only and five to the end card. In the audio column, / marks a short pause and // a longer one.

What the script does on purpose:

• It opens on a picture and a confession, not a greeting. The library is not named until 0:12, when the viewer already knows why it matters.
• No line describes the shot. The narration adds what the shelf can't show: the date, the count, the card.
• The name, price, loan period and date are on screen as SUPERs, so it works muted.
• Seven seconds have no words. Then the ending returns to the drill from the opening and asks for one thing.

A video script template to copy

Put the header at the top of the document, then build the rows from the table below, adding or removing body rows to fit your beats. The video script template has the same structure as a file you can download in Markdown or Word.

Title: [Working title]
Length: [0:00] · Plays on: [YouTube, a feed, a landing page, a training portal]
Viewer: [Who is watching and what they already know]
Takeaway: [The one sentence they should remember]
Action: [The one thing they should do next]
Narrator: [Name] · Draft: [1] · Date: [Date]

For narration recorded on its own, the voiceover script template drops the video column. For a talk delivered live with slides, see how to write a presentation script.

Mistakes that sink video scripts

The most common is writing for the page: clauses, parentheses and precise figures that a listener loses on the first pass.

• A greeting before the hook. “Hi everyone, welcome back” and a logo sting spend the seconds viewers use to decide. Move both to the end or cut them.
• Narration that describes the picture. “Here you can see the shelf” wastes a line the picture already delivered.
• Two takeaways. A video that teaches two things is remembered for neither. Make the second video.
• Three calls to action. Like, subscribe, visit, download. Pick the one that matters and ask once.
• Never timing it. Scripts run long. Find out at your desk, not in the edit.

Writing video scripts in Strokes

A two-column script is a table, so any editor with tables will hold one, Word and Google Docs included. In Strokes, write it as a table in a document and keep each video's brief, script and shot list in one project folder. Version history saves as you work, so the 90-second cut is still there to compare and restore after you make the 60-second one. Select a line of narration and press ⌘K to tighten it, or ask the AI Assistant to point out sentences that will be hard to say aloud. Then export to PDF or Word for the narrator and the client.