A screencast voice-over script is a script you prepare before recording a screencast. It helps you plan what you'll say and what you'll show on screen at each moment.

You don't strictly need one if you're simply talking through a presentation you already know well, but having a set of guidelines to work from is never a bad idea, and a script becomes essential once captions, translations, or reuse are involved.

The Dept of TEL has prepared a template staff can use and/or edit for their own purposes, available here.


Keep it short and single-purpose
  • Screencasts should be short and granular. Avoid long, multi-step on-screen sequences; two concise, focused screencasts beat one long one that's harder to follow and hold attention through.
  • Each recording should cover one task or one concept. If you find yourself covering more than one, split it.
  • Lead with the outcome. Open with what the viewer will be able to do by the end, not a personal introduction, viewer attention drops sharply in the first few seconds, so get to the point immediately rather than building up to it.

Pace and length
  • Aim for roughly 130–150 spoken words per minute for narration; slower than casual conversation, since a screencast has natural pauses while on-screen actions play out.
  • As a working guide: a 2-minute recording is closer to 250–280 words (about two-thirds of an A4 page), not 400. Use this to estimate how much script you need for a target length.
  • Don't pad the script to hit a length. Unnecessary detail is boring to listen to and loses the viewer's attention faster than a script that's slightly short. Keep it as lean as possible.


Write for accessibility and longevity
  • Describe on-screen actions in words as part of the narration (e.g. "click Save in the top-right corner" rather than just "click here"). This keeps the recording usable for anyone listening without watching, and makes auto-generated captions and transcripts actually make sense on their own.
  • Plan for captions/a transcript as standard, not optional; many organisations require this, and it also makes the content searchable and reusable.
  • Where possible, describe actions by function rather than exact visual position or colour (e.g. "open the Settings menu" rather than "click the grey gear icon in the corner"), this keeps the script usable even after the software's interface is updated.


Format the script for recording
  • Use a two-column layout; on-screen action next to narration, rather than a single block of prose. It's easier to record from and easier to hand off to someone else to record.
  • Write in a conversational, active, present-tense style, addressing the viewer directly ("you'll see...", "next, click...").


Structure
  • Open with the goal/outcome, not a self-introduction.
  • Walk through the task in the same order the viewer will actually perform it.
  • Close with a clear next step; a link to related documentation, the next screencast in a series, or where to get help, rather than a generic sign-off. A "thank you for watching" is fine as a line, but shouldn't be the only thing you leave them with.