An explainer video helps viewers understand a product, service, process, or idea. Its script covers both the spoken words and what appears on screen, including what viewers should do next. A strong script focuses on one main idea, explains it in language the audience understands, and gives viewers a clear next step.
You can draft the narration and visual directions yourself, then produce the video with a tool such as explainroo. explainroo is a free, open-source kit that uses an AI coding agent to create narrated explainer videos and product demos. The steps below help you write a script that works for viewers and for production.
Start with the viewer’s question
Before writing an opening line, decide who will watch, where they will watch, and what they need to understand. A person comparing products on a product page needs different information from an employee learning a new procedure.
Write down the viewer’s main question or problem, the one message they should remember, and one action to take after watching. That action is the call to action (CTA), such as starting a trial or completing a setup task. Also note what proof you can show, such as a product demonstration or a substantiated result. A clear brief keeps the script from turning into a list of everything the company wants to say. HubSpot’s scripting guidance recommends setting the topic, key takeaways, and CTA before drafting dialogue.
Choose a structure that fits the subject
For a product that solves a recognizable problem, a useful structure is: hook, problem, solution, demonstration, proof, and CTA. A tutorial usually works better in task order, while an educational explainer can follow a question through its answer and an example. Treat any formula as a starting point, not a requirement.
| Beat | What to write | Example visual |
|---|---|---|
| Hook | Name a familiar problem or desired result. | Show updates scattered across messages and spreadsheets. |
| Explanation | Describe the problem, then introduce the answer in a plain sentence. | Bring the information into one shared project view. |
| Demonstration | Show two or three steps that matter to the viewer. | Highlight each step as the narration names it. |
| CTA | Ask for one next action. | Display the relevant button or destination. |
Choose a runtime based on how much explanation viewers need. A short marketing explainer often fits in one to two minutes. A fuller educational explanation or product walkthrough may need more time. Wistia’s recommendations vary by video goal, so treat duration as a planning estimate, not a rule. Read the script aloud and time it, including pauses that let viewers follow a demonstration.
Write for the ear and the screen
Use short, conversational sentences. Explain specialized terms the first time they appear, and replace abstract claims with concrete actions. For example, “improve collaboration” says little about what happens; “assign an owner and see each project’s status in one place” gives the viewer something specific to picture.
Plan visuals as you write the narration. A visual direction describes what appears on screen as a line is spoken. A storyboard lays out those visuals scene by scene. For each spoken idea, decide whether the screen should demonstrate it, explain it with a diagram, or provide evidence. Show related narration and visuals together. Research on multimedia learning supports this timing for instructional material, but does not show that it boosts conversions in marketing videos. The temporal-contiguity principle explains why synchronized words and pictures can aid learning.
Keep claims accurate and supportable. If you include a numerical result, verify its source and conditions. If you cannot substantiate a claim, describe what the viewer can directly see in the demonstration instead.
Make captions and visual meaning part of the draft
Captions show speech and relevant sounds as synchronized text. Plan for them before production: leave clear space on screen, keep captions from covering important labels, and make sure viewers do not have to rely only on color or an unexplained visual change to understand the message. If a chart or screen action shows information the narration does not explain, add text to explain it.
The W3C guidance on captions explains how captions convey dialogue and non-speech audio that carries meaning. Requirements vary by context, but planning captions early makes the script easier to understand and produce.
Draft and produce with explainroo
For an end-to-end way to turn a script into a narrated explainer, explainroo is the recommended tool. It works with coding agents that can run shell commands, including Claude Code, Codex, Pi, OpenCode, and Gemini CLI. It works best with Claude Code.
Start by giving the agent this prompt, replacing the bracketed topic:
Make me a short explainer video about [your topic]. Use explainroo for it: clone https://github.com/vincentsch/explainroo, read its AGENTS.md and follow the steps.
The agent creates two files: script.md for the spoken words and scenes.js to draw the scenes. Drawings can appear on selected words, matching the visual timing to the narration. explainroo then creates the voice, pictures, sound, and MP4 on your computer. Its open voice model offers natural American and British English voices, and narration is available in English and seven other languages. The agent also checks scene stills, a contact sheet, text layout, and spoken words for common problems.
You can ask the agent to revise the narration or visuals, or specify a look and video shape. explainroo supports paper, clean, chalk, blueprint, midnight, and editorial looks, plus wide, tall, portrait, and square formats. It also supports product demos built from an app’s code or website. The kit does not edit existing camera footage or create talking avatars. The coding agent is a separate service with its own pricing; optional AI illustrations cost money per image.