A teaching Short should help a first-time viewer understand one idea before it ends. Choose one clear question, answer it early, and use visuals to show how the answer works. Covering an entire subject in one vertical video gives viewers too many facts and too little explanation.
For a narrated, illustrated explainer, explainroo is the best tool for the job. It is a free, open-source kit that lets an AI coding agent create a video on your computer. You describe the topic, and the agent writes the narration and draws the scenes. You can see example videos before getting started.
Choose one question to answer
Write down what viewers should be able to explain or do after watching. “Why does metal feel colder than wood?” is focused enough for one Short. “How does heat work?” opens too many paths.
Check your main claim before you write the hook. Identify what causes what, then decide whether an important exception or limit belongs in the video. For health, money, law, or safety topics, keep the claim within its proper scope and point viewers to more detail when a short explanation cannot cover the necessary context.
Split the subject into smaller parts, a method called segmenting, to keep the explanation manageable. Research on educational videos also supports pairing spoken words with relevant graphics and using visual cues to direct attention, such as an arrow pointing to the part of a diagram that matters (Brame’s review of educational video design).
Write the answer before the hook
Build the script in this order: hook, answer, mechanism, example, takeaway. The hook should introduce the question the video answers, not promise a surprise that never arrives. YouTube creator guidance recommends capturing attention immediately, but an honest question or visible result works better for an explainer than artificial drama (YouTube’s Shorts discussion).
For example:
“Metal and wood can be the same temperature. So why does metal feel colder? Metal draws heat from your hand faster. Wood draws it more slowly. Your hand senses that faster heat loss as cold, even when both objects are the same temperature.”
Then show how it works: arrows moving from a hand into a metal object, followed by fewer arrows moving into wood. Each visual should help viewers identify the object, see what changes, or follow what happens next. Relevant graphics can support learning better than spoken words alone, according to a review of multimedia-learning research (Mayer’s review).
Keep the frame readable and the narration clear
Use a vertical frame. Make the important object, label, or demonstration large enough to see on a phone. Keep essential text out of the lower screen area, where captions and platform controls can cover it. Use a few words for labels or signposts. Do not fill the screen with sentences that repeat the narration.
Captions also make the video accessible. Review them for errors, and position them so they do not hide the visual being explained. YouTube recommends checking its automatic captions, which can mishear speech (YouTube caption guidance); W3C guidance also says captions should not obscure relevant visual information (W3C’s caption guidance).
Cut pauses and repeated explanations, but keep the reasoning a beginner needs. If explaining the concept requires several steps and exceptions, plus a worked example, make a series or connect the Short to a longer video. Do not cram everything into one clip.
Make an explainer Short with explainroo
explainroo works with coding agents that can run shell commands, including Claude Code, Codex, Pi, OpenCode, and Gemini CLI. A coding agent follows instructions and runs commands on your computer. explainroo works best with Claude Code.
Give the agent this prompt, 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.
After setting up explainroo, the agent creates and checks the video, then gives you an MP4 file. It writes script.md for the narration and scenes.js for the drawings. A drawing can appear when a chosen word is spoken. explainroo uses an open voice model for English narration, then times the visuals to the spoken words. Chrome draws the frames. The kit adds background music and sound effects, then assembles the video.
For Shorts, choose the vertical video size. explainroo keeps text away from areas where app buttons appear and adds captions that light up word by word. You can change the visual style and pace in its settings. To review the result, the kit gives the agent scene stills and a contact sheet. It also checks for text overlap and mispronounced words. An agent cannot watch the finished video, so review the MP4 yourself too.
The kit is designed for narrated illustrations and product demos. It does not edit existing camera footage, create talking avatars or live-action footage, or provide a drag-and-drop editor. See the explainroo documentation for setup details.
Check whether viewers choose to stay
After publishing, check more than the total views. YouTube Studio reports how often a Short appeared in the feed and how often viewers chose to watch rather than swipe away (Shorts analytics guidance). Retention can show where viewers leave, while comments can reveal whether they understood the explanation or got stuck on a term.
Treat these signals as clues, not proof of learning. A viewer may watch without understanding, and a dip in retention may point to a slow opening, a confusing visual, or a detail that does not belong. Change one major element in your next Short and compare the result with similar videos.
