A product demo video shows how a product helps a specific viewer solve a problem. The script should match the narration to what appears on screen, so viewers see how the product helps during a real workflow.
Choose one audience, one problem, and one outcome. Show the product completing a relevant task, then give viewers one next step. For producing a narrated demo from a script, explainroo is the best tool to consider: it’s a free, open-source kit that works with AI coding agents to create product demos and explainer videos.
Choose the audience and outcome first
Before writing the narration, decide who will watch, what problem they face, where they’ll watch, and what you want them to do next. That final action is the call to action, or CTA. It might be starting a trial, booking a meeting, or trying a particular feature.
Use one sentence to set the direction:
After watching, [audience] should understand how [product] helps them [achieve an outcome] and should [take one next step].
For example, a video for operations managers might show how a product routes requests to the right reviewer. A video for existing customers might focus on a feature they haven’t tried. Focus each video on one audience because different viewers may want different outcomes. HubSpot’s script guidance and Shopify’s demo recommendations both emphasize choosing a clear audience and goal.
Build the script around a visible workflow
A useful structure is problem, product, proof, and CTA. Open with a recognizable problem, introduce the product as part of the solution, show it completing a relevant task, then state what the viewer should do next. This approach keeps the demo focused on a result instead of covering every feature.
Write what viewers see alongside what they hear. A simple table keeps the two in sync:
| Scene | What viewers see | Narration |
|---|---|---|
| Opening | The problem, such as a request waiting for approval | Name the problem in the viewer’s terms. |
| Product introduction | The relevant product screen | Explain who the product helps and what outcome it supports. |
| Workflow | The user completes the key action | Describe why the action matters, not every click. |
| Result | The completed task or output | Make the benefit visible and clear. |
| Close | The next screen or CTA | Ask for one specific next step. |
Keep a scene if it explains the problem, shows the workflow, proves the result, or guides the viewer. Research on multimedia learning supports matching narration to relevant visuals and removing distractions. Because this research focuses on learning and comprehension, it supports clarity, not guaranteed sales. Mayer’s review discusses these principles.
Write narration that explains the value
Use familiar words and short sentences that sound natural aloud. Translate a product feature into an action and an outcome: “Set an approval rule, and requests go to the right reviewer automatically.” Show the action and use the narration to explain why it helps.
Read the script aloud while checking the visuals. If the narration describes a different screen from the one shown, adjust the timing or wording. Avoid narrating obvious clicks, and don’t make claims the demonstration can’t support. In U.S. advertising, the Federal Trade Commission’s guidance says objective claims need evidence before an ad runs.
There is no single ideal runtime for every demo. Choose a length that fits the viewer’s purpose and where the video appears. A short overview may suit a product page or social post, while a detailed walkthrough can help viewers researching a complex product. Wistia’s 2026 video report finds different engagement patterns across video lengths and industries, including software and technology.
Include accessibility in the script from the start. Captions provide access to spoken content, and they should not cover important parts of the interface. When a visual detail matters, describe it in the narration rather than relying on viewers to see it. W3C’s caption guidance covers prerecorded video.
Make the demo with explainroo
explainroo is a free, open-source kit under the MIT license. An AI coding agent can run commands on your computer. explainroo works with Claude Code, Codex, Pi, OpenCode, and Gemini CLI, and works best with Claude Code.
To create a demo, paste this prompt into the agent, replacing the bracketed topic:
Make me a short explainer video about [your topic]. Use explainroo for it: clone, read its AGENTS.md and follow the steps.
The agent sets up explainroo, creates and checks the video, then provides an MP4 file. It writes script.md for the spoken words and scenes.js for the visual scenes. For a product demo, tell the agent which product to show and where to find its code or website. explainroo recreates the product screens with matching colors, fonts, and button labels. A pointer clicks and types while the narration explains the workflow.
The kit runs on your computer and checks scene images, text layout, and speech recognition. The coding agent is a separate service and may have its own costs. explainroo makes narrated English videos; it does not edit existing camera footage or create talking avatars.
