Narration Script Template for Video and Audio

Direct answer: a narration script template is a beat-by-beat worksheet with one row per moment and four working columns: the spoken line, the visual or action it plays against, a timing target, and a production note for the editor. Write the opening, the body beats, and the close before you polish a single sentence. This guide gives you the fill-in template, a six-beat structure that fits both video and audio, and a worked example you can adapt.

A narration script is the spoken spine of anything heard rather than read: a video voiceover, an audiobook chapter, a course module, a podcast segment, a product demo. The mistake most writers make is treating it like an essay and reading it silently. Spoken language lives or dies on how it sounds at the moment it is said, one line at a time. A template fixes that by forcing you to plan the shape of the narration before you write the words inside it. If your source is a written article you plan to voice, the guide on turning blog posts into audio covers the rewrite a page needs before it can be read aloud; for a full-length book, the audiobook narration guide handles the longer form.

What a narration script template contains

Copy these columns before you write anything. The order matters, because the spoken line depends on the moment it sits against, and the timing depends on the line.

Keep the spoken line and the production note apart, because mixing them is how a script ends up reading its own stage directions out loud. “Pause here for the reveal” is an instruction; “Now watch what happens” is narration. Both belong in the document, but only the second belongs in the voice. A spreadsheet row can hold both, and so can a document with two labeled fields; the point is that they never collapse into the same line. Microsoft's help page on speaker notes makes the same split for presentations, describing the Notes pane as a box the presenter sees on their own monitor but the audience never sees.

The six beats that give narration its shape

Almost every piece of narration, from a thirty-second ad to a three-minute explainer, moves through the same six beats. The template is just a way of holding them side by side so you can see the whole thing at once.

The beats are a shape, not a script. You still have to write the words. But once the shape is on the page, a writer can see that a rambling middle is really three unsorted Steps, or that a piece has no Turn and therefore overpromises. If you are writing a demo rather than a general explainer, the product demo script template applies the same beats to a software or physical product walkthrough, where every beat has to be provable on screen.

Narration structure for video and audio

This is where a generic template stops being useful, because video narration and audio-only narration are not the same job. Video narration shares the listener's attention with a picture, so the words can point at something the viewer can see. Audio-only narration has no picture, so every line has to carry its own context.

One rule ties both columns together: redundancy. W3C's guidance on audio and video content recommends against lines that depend on a single sense, such as “attach this to the green end,” and suggests saying “attach the small ring to the green end, which is the larger end” instead. The same habit helps an ordinary explainer. If a line only makes sense while pointing at a screen, it will confuse anyone who hears the audio on its own, and a listener who plays that audio without the picture has no screen at all.

A worked example: a two-minute explainer narration

Here is the template filled in for a short explainer about backing up a photo library. The spoken column is the finished narration; the other columns are planning aids. The row order is the order the beats play.

Scene 1, Open (about 4 seconds) Spoken: Your photo library lives on one drive, and one drive can fail. On screen: a single external drive on a desk. Scene 2, Promise (about 5 seconds) Spoken: Here is a two-step habit that keeps a second copy somewhere else. On screen: two drives side by side. Scene 3, Setup (about 6 seconds) Spoken: Start by deciding which folder is the library. In this example it is the folder called Photos, inside the Pictures directory. On screen: the folder selected, with its path visible. Scene 4, Step one (about 7 seconds) Spoken: Connect the backup drive, open it in the Finder, and drag the Photos folder into the empty space. On screen: the drag from Pictures to the Backup drive. Scene 5, Step two (about 6 seconds) Spoken: Wait for the copy to finish, then check that the folder size matches on both drives. On screen: the two folders shown with their sizes. Scene 6, Turn (about 5 seconds) Spoken: This copies the files once. For a habit that runs on its own, set a repeating backup instead. On screen: the manual copy, then a backup app window. Scene 7, Close (about 4 seconds) Spoken: Make the second copy today, then schedule the next one. On screen: the backup drive, safely unmounted.

Example narration script, fictional product setup

Total the spoken column and you get the planning estimate: about 230 words, or roughly 90 seconds of speech at 150 words per minute before you add pauses and on-screen waits. The example adds around 25 seconds of breathing room across seven rows, which lands close to two minutes. The purpose of the timing column is not to hit an exact number; it is to see, before you record, which row is quietly doing too much work. Scene 4 is the longest line and the one a viewer most needs to watch, which is exactly when the narration should slow down rather than race.

Write for the ear, not the page

Reading silently hides the problems that talking exposes. Run every line through these checks before it reaches a voice.

• Read the line aloud. If you run out of breath, cut it into two lines.
• Write numbers the way you want them said, such as two hundred thirty rather than 230, or check how the voice reads digits.
• Cut lines that only work while a cursor moves, such as as you can see here.
• Name the thing instead of pointing at it: open the Share panel, not click this.
• Give each proper noun one pronunciation note, and reuse the same spelling everywhere it appears.
• Mark deliberate pauses rather than trusting the reader to find them.
• Keep one voice for the whole piece, and describe it once in the notes.

W3C's guidance on writing script content for audio also recommends avoiding jargon, acronyms, and idioms that a listener cannot pause to look up. That is a harder constraint than a written article: a reader who meets an unknown term can search for it, while a listener simply loses the thread. When the same script will also become captions or a transcript, the W3C page on transcripts notes that captions are written to sit beside the picture, so a transcript usually needs the visual information added back, such as on-screen text and speaker names. The script is a starting point, not a finished transcript.

Once the lines are short and concrete, the narration has to be performed. If you are producing a full-length read, the Mac narration workflow covers the handoff from an approved script to finished audio, and the narration voice roundup compares the voices you might pick. For training or e-learning scripts, the training video narration guide covers the wider loop from outline to export.

Turn the script into audio

When the spoken column is final, generate the narration. Murmur is a local-first macOS voice studio for Apple Silicon and Intel Macs running macOS 15 or later that handles text to speech, voice cloning, voice design, and multi-speaker projects, and it runs the models on your own machine. That means no per-character fees and nothing uploaded to a cloud service, which matters when the script contains an unpublished product name or a client's details. Export at a level and format your editor accepts, keep the script and the audio files side by side, and name each file after its scene so a later edit maps to the right place.

Choose one voice before you generate anything and keep it for the whole piece. A voice that changes between scenes reads as a mistake even when the words are identical. Murmur costs $49 one-time, has a free plan and a 3-day Pro trial, and includes a 7-day refund policy. If you are not sure which voice fits, audition two against the same two lines and pick the one you can listen to for the full runtime. The audio export guide explains the difference between an editing master and a delivery file.

When Murmur is not the right choice

Murmur is a Mac app, so if you work on Windows, need live or real-time streaming speech, or want several people editing one project in the cloud at once, a cloud service will fit better. It also does not do speech to text, so it cannot turn a recorded interview into a script. And if a line needs a specific celebrity voice or an emotion a stock generated voice cannot carry, plan on recording a human narrator. Murmur is strongest when you want private, repeatable narration you can regenerate from the same script, line by line.

Frequently asked questions