How to make an explainer video
Tell Vizard Agent what to explain and who it is for, and it writes the narration, generates the visuals to match, and returns a finished video with voiceover and captions. You supply no footage. Naming the audience matters more than naming the topic — it decides the vocabulary, the pace and what gets assumed.
What is the short version?
Three moves, and none of them involves finding footage. An explainer normally stalls on visuals: the script is written in an hour and then somebody spends two days looking for stock that matches it. Vizard Agent generates what the narration needs, so the script and the pictures arrive together.
- Go to Vizard Agent and say what needs explaining.
- Say who it is for — customers, beginners, social.
- Read the narration, correct the facts, and publish.
The rest of this page is what goes into each of those, and where the approach stops working.
What do you need before you start?
A topic and an audience. That is genuinely all — this is the one format on this list that needs no source material at all. What decides the result is how narrow the topic is: "explain our pricing" produces something forgettable, "explain why our pricing is per-seat and not per-usage" produces something worth watching.
- One question, not a subject. An explainer answers a question. If you cannot phrase it as one, the video will wander.
- The audience, named. "For customers" and "for beginners" produce different vocabulary and different assumptions about what needs defining.
- Any terms that are yours. If your company uses a word in a specific way, say so — otherwise the general meaning is used.
- A length. Sixty seconds is a common default because it fits most places, but say what you need — thirty for social, three minutes for a walkthrough.
What do you type into Vizard Agent?
Vizard Agent ships this as a template, and both brackets earn their place: the topic decides the content, the audience decides everything else. Filling in only the first is the most common way to get a video that is technically correct and completely flat.
Prompt
Three variants worth keeping:
When the topic is yours, not general
When it has to hold on social
When it is going in front of a customer
What are the steps inside Vizard Agent?
Three, and the third is not optional. Vizard Agent writes the narration, generates the visuals for it, records the voiceover and cuts the video as one request — which means the first time anyone reads the script closely is when it is already a finished video. Read it anyway.
- Type the topic and audience into the Vizard Agent chat.
- Pick the tier that matches the job. For anything customer-facing the product recommends Max; the free Flash tier is built for quick, simple edits.
- Read the narration before publishing. An explainer that is wrong is worse than no explainer, and this is the format where a confident sentence is easiest to miss.
What does the result look like?
A finished video with narration, generated visuals cut to the script, captions and music. Vizard Agent keeps the assets in the project, so a second version — shorter, different audience, different platform — costs a sentence rather than a fresh build.
It is not instant, and this category was not separately measured. Comparable work runs a median of 28 to 38 minutes end to end depending on how much source material is involved, and across all projects the median cost by tier is Flash 47, Pro 55, Max 242, Ultra 263 credits. Explainers use no source footage, which puts them at the lighter end of that range.
When does this not work well?
An explainer is a claim about how something works, delivered confidently, with pictures. That is exactly what makes a wrong one dangerous: nothing in what Vizard Agent returns signals which sentences it was sure about and which it inferred, because the narration reads the same either way. Six cases are worth checking before you publish.
- Facts specific to your organisation. Your pricing, your process, your SLA. Vizard Agent knows the general shape of these things, not yours. State them in the prompt or expect the general version.
- Anything contested or moving. A topic where the answer changed this year, or where experts disagree, gets one confident version of it. Check it.
- Regulated explanations. Medical, legal and financial explainers carry rules about what may be said and how. Treat the script as a draft for whoever signs those off.
- Highly visual technical detail. Some things need a real diagram with real numbers on it. Generated visuals illustrate; they do not document.
- Very broad topics. "Explain AI" produces a video about nothing. Narrow it until it is a question.
- Precise terminology. Where a word means something specific in your field, say what you mean — the general sense is otherwise the safe assumption.
If the explainer is the authoritative source for something — a compliance process, a safety procedure — write the script yourself and have Vizard Agent build the video around it.
How do you fix a result that came back wrong?
Reply in the same Vizard Agent conversation instead of starting over. The script and the generated visuals are still there, so you are not billed again for work already done — only for what the correction costs. Fixing a fact is a script change; fixing the whole angle is closer to a rebuild.
| What you see | What to say next |
|---|---|
| A fact is wrong | Correct it directly: "seats are named users, not concurrent logins" |
| Too much jargon | "Assume they have never heard of any of this — define every term" |
| Too slow to the point | "Get to the answer in the first ten seconds, explain after" |
| The visuals do not match | "The middle section should show a warehouse, not an office" |
| Too long | "Cut it to 45 seconds — drop the history, keep the mechanism" |
How does Vizard Agent compare to the usual explainer route?
The usual route is script, storyboard, source or animate the visuals, record a voiceover, edit. The visuals are the bottleneck — an animator or a stock search stands between a finished script and a finished video, and both cost days.
| Vizard Agent | Script → storyboard → animate | |
|---|---|---|
| Source material needed | None | Stock licences or an animator |
| Time to a first cut | Tens of minutes | One to three weeks |
| Cost of a second audience version | Another sentence | Most of another project |
| Factual accuracy | Yours to check | Yours to check |
| Bespoke visual style | Described in words | Exactly as designed |
Neither replaces the other. Vizard Agent wins when the explanation matters more than the art direction, and when you need four versions for four audiences; a designed animation still wins when the video is a brand asset that will run for two years.
Common questions
Do I need any footage at all?
No. This is the one format here that starts from nothing but a sentence — Vizard Agent generates the visuals to match the narration it writes. Attaching your own footage is possible and will be used, but it is not required.
Can I write the script myself?
Yes, and for anything authoritative you should. Paste the script and ask Vizard Agent to build the video around it. You keep the wording exactly; Vizard Agent handles voiceover, visuals, captions and timing.
How accurate is the narration?
Accurate enough to be convincing, which is the problem — nothing in what Vizard Agent returns signals a shaky sentence. For general topics it is usually right; for anything specific to your organisation, assume it is the general version until you check.
Can I use my own voice?
Attach a recording and say to use it. Voice cloning needs a clean sample of one person, roughly 10 to 90 seconds, with no music or second voice over it. For an explainer that will represent your company, it is worth the extra step.
What length works best?
Whatever the explanation needs. Sixty seconds suits most placements, thirty works on social, and a complex process may want three minutes — a rushed minute explains less than an unhurried ninety seconds.