Voice-over for videos and presentations: how to prepare the text

3 min read

A video with a clear voice-over is easier to follow than one with text on the screen: a product demo, an explainer for a website, a lesson in an online course, a presentation that has to run without you. The voice can only be as good as the text it reads, though. A script written for the eye often sounds stiff when spoken. Here is how to prepare a text that sounds natural when read aloud.

Voice-over for a video and a presentation

Measure the script by time

A calm voice reads about 130 to 150 words a minute, which is roughly 800 to 900 characters. Count before you write the rest:

  • A 30-second ad: about 70 words.
  • A 1 to 2 minute intro or product video: 150 to 280 words.
  • A 5-minute explainer or lesson: about 700 words.

If the video is already edited, time each scene and write the text to fit it, not the other way round.

Write for the ear

Listeners cannot go back and reread a sentence, so:

  • Keep sentences short. One idea per sentence, 15 to 20 words at most.
  • Put the main word at the end. "The order arrives in two days" is easier to catch than a long sentence with the key fact buried in the middle.
  • Avoid lists in brackets and long chains of numbers. Say "about a third" instead of "32.7 percent" unless the exact figure matters.
  • Read it aloud yourself. If you stumble, the voice will sound unnatural there too.

Mark pauses and emphasis

A plain text gives no hints about rhythm. Add simple notes:

  • A new paragraph for a longer pause, for example when the picture changes.
  • Bold or capitals on a word that should be stressed.
  • A note on pace: slower for instructions, livelier for an ad.

Spell out names, numbers and terms

Write unusual words the way they are pronounced, in brackets or in a separate list: "Crumb (rhymes with drum)", "SaaS (say: sass)". Decide how numbers should sound: "2026" as "twenty twenty-six", "$19.90" as "nineteen ninety". Do the same for abbreviations, web addresses and product codes.

Choose the voice and the mood

Think about who is listening and where:

  • Calm and clear for a tutorial or a course.
  • Energetic for an ad or a social media video.
  • Warm for a story, a greeting or a brand video.

Name the language and, if it matters, the accent. Describe the voice by its character, not by a famous person: voices of real people are not imitated.

Mixing with music and the slides

A clean voice track is the most flexible: you add music in the video editor and keep it quiet under speech. For a presentation, split the script by slide, so each audio piece starts with its slide. Leave a second of silence at the start and end of each piece for easy editing.

Before you publish

  • Listen with the picture. Check that each sentence matches what is on the screen.
  • Check names and numbers once more by ear.
  • Say that the voice is synthesized in the description or credits, if a synthesized voice is used.

In short

A good voice-over script is measured by time, written in short spoken sentences and marked with pauses, stress and pronunciation. If you need the text read for you, order a voice-over: you paste a script of up to 1,500, 5,000 or 15,000 characters and describe the mood and pace, and get an MP3 file of the reading in a natural synthesized voice, plus WAV in the larger packages. Payment is released only after you approve the work.