A voiceover script format that most video teams use is the two-column AV script: what the viewer sees goes in the left column, and what they hear (narration, music, sound effects) goes in the right, scene by scene with a time stamp. Copy the template below, edit the 60-second example, and use the live word count to check the narration fits your running time.
Key Takeaways
- Two columns: one column notes visuals and one notes what you hear. The format is simpler than a narrative screenplay, which is why it suits first-time script writers.
- Write the narration to be spoken. Contractions, short sentences and numbers as digits read aloud best.
- Budget words by time. At about 150 words a minute, 60 seconds holds about 150 words. Leave room for pauses and music.
- Mark everything that is not narration. Music, effects and on-screen text each get their own label so nobody reads them aloud by mistake.
- Give the narrator the clean text. Copy only the voiceover lines into a text-to-speech tool or send them to a voice actor.
What does a two-column voiceover script look like?
Here is a 60-second product explainer for a made-up app, TaskPilot. The left column is the picture, the right column is the sound. The time stamps let an editor cut to the voice.
| Time | Visual (what we see) | Audio (what we hear) |
|---|---|---|
| 0:00–0:08 | Close-up of hands moving sticky notes on a monitor. Cut to a phone buzzing with alerts. | VO: Your to-do list lives in six places. Notes, email, chat, and a few sticky notes on your monitor. |
| 0:08–0:18 | Screen recording: scattered items slide into one clean list. On-screen text: “One list”. | SFX: soft whoosh as items land. VO: TaskPilot pulls them into one list. Forward an email, tag a chat message, or type a note. It lands in the same place. |
| 0:18–0:32 | Calendar view. A task is dragged onto Thursday. | MUSIC: light and steady, under the voice. VO: Then drag a task onto your calendar to give it time. If Thursday fills up, TaskPilot shows what to move, so nothing slips past its deadline. |
| 0:32–0:46 | Team view: three avatars tick off tasks while a progress bar fills. | VO: Working with a team? Share a list, assign owners, and see progress at a glance. No status meeting needed. |
| 0:46–0:56 | Phone screen showing “12 of 12 done”. | VO: Set up your first list in a few minutes. Then get on with the work. |
| 0:56–1:00 | Logo and web address on a plain background. | MUSIC: resolves. VO: TaskPilot. Everything in one place. |
What goes in each part?
Every part of the table has one job. Use this as a checklist when you write your own.
| Part | What to write | Why it helps |
|---|---|---|
| Time | Start and end of the scene, in minutes and seconds | The editor can cut the picture to the voice, and you can see pacing at a glance |
| Visual | The shot, the action, any on-screen text in quotes | The narrator and the designer see the same picture |
| Audio label | VO, MUSIC, SFX at the start of each line | Nobody reads a sound cue aloud |
| Narration | Full sentences, exactly as spoken | What you write is what gets recorded or generated |
| Pauses | A new line or “(pause)” where the voice should breathe | The word count stays honest and the read stays natural |
Copy the narration and check the word count
This block holds only the voiceover lines from the example, with the labels and cues stripped out. That is the text you would paste into a voice tool. The count under the title updates if you edit it in your own copy, and turns red if it runs past 60 seconds at 150 words a minute.
TaskPilot narration only 0 words
Count the words in your own narration column
Paste your right-hand column below to see its length and how many words fit a target. Nothing is sent or saved.
Copy a blank two-column template
Pick the layout that fits where the script will live. Each block copies as plain text you can paste into a document, a spreadsheet or an email.
Table template for Word, Google Docs or Notion
Insert a table with 3 columns and as many rows as you have scenes. Paste the header row, then fill one row per scene. Keep the time column narrow and give the audio column the most room, because it holds the most words.
Header and first row Template
0:00-0:10 | [Shot, action, on-screen text] | MUSIC: [mood, level]
VO: [Narration, exactly as spoken]
Plain text template for email or chat
When a table will not survive the trip, use stacked blocks. One block per scene, with the same three labels.
Scene block Template
VISUAL: [Shot, action, on-screen text]
MUSIC / SFX: [Cue and level, or none]
VO: [Narration, exactly as spoken]
NOTES: [Tone, emphasis, pronunciation]
Spreadsheet template
A spreadsheet suits long scripts, because you can sort by scene and add a word count column. Paste the line below into the first row and add one row per scene.
Column headings Template
Narration ready? Paste the right-hand column into Kveeky, pick a voice, set the speed, and hear the timing before you edit the picture.
Voice your script freeHow do you write the narration column?
Follow these steps in order. Writing the audio first, then the picture, keeps the story clear.
- Set the running time and word budget. Multiply the minutes by 150. A 2-minute video holds about 300 words. The words to minutes calculator shows other paces.
- Write the narration first, as one block, in short spoken sentences. Read it aloud and cut anything you stumble on.
- Split it into scenes. One idea per scene, each about 5 to 15 seconds.
- Add the visual for each scene, then the music and sound effect cues, with a label on every cue.
- Add the time stamps from your word count, then adjust once you hear the voice.
- Export the clean narration. Copy only the VO lines for the voice tool, and keep the full table for the editor.
For radio and audio-only scripts, there is no picture column, so see our radio ad script templates. To plan where the file will play, check the audio ad specs.
Common mistakes in a voiceover script
- Writing for the eye. Long clauses and brackets are hard to speak. Split the sentence.
- Describing what the picture already shows. Let the visual carry the detail and the voice carry the meaning.
- Mixing cues into the narration. “(music swells)” inside the VO line gets read aloud by a text-to-speech tool. Put it in its own labelled line.
- No time stamps. Without them, the editor guesses at the pace.
- Unspoken numbers and names. Write “2026” as “twenty twenty-six” if the voice might misread it, and add a pronunciation note for names.
How Kveeky helps with voiceover scripts
Kveeky turns the narration column into a voice track, so you can hear the timing before the edit and change a line without a re-record.
Hear your narration column free
Sign up in seconds. No credit card needed.
Frequently asked questions
What is a two-column AV script?
It is a script split into two columns. One column notes the visuals and the other notes what you hear, such as narration, music and sound effects. It is simpler than a narrative screenplay format.
Which column goes on the left, video or audio?
Most templates put the visuals on the left and the audio on the right, but we found no official rule. Pick one layout and use it for the whole script, or follow your client’s template.
How many words should a 60-second voiceover script have?
About 150 words at an average pace of 150 words per minute. Plan for 100 to 130 if the video has music, pauses or on-screen text that needs time to read.
Should I write the voiceover or the visuals first?
Write the narration first, then add the visuals. The story is easier to keep clear when the words come first, and the narration sets the length of each scene.
How do I send a script to a voice tool or a voice actor?
Send only the voiceover lines, with no cues or labels inside them. Keep the full two-column table for the editor, and add notes on tone and pronunciation in a separate line.
Sources
- Two-column AV script structure (one column for visuals, one for what you hear; simpler than a narrative script): Ohio State University, “Pre-Production 101: Choosing a Script Format”, retrieved 16 October 2026. The page does not say which column goes on which side.
- Average speaking rate of about 150 words per minute: see the sources on our words to minutes calculator, which cites the National Center for Voice and Speech.
- Kveeky formats, plans and rights: Kveeky pricing, retrieved 16 October 2026.
- The TaskPilot example, column labels and templates are our own work. TaskPilot is a made-up product.