How to Make Explainer Videos (2026)
The short answer
To make an explainer video: scope to exactly one concept (one mechanism, one question, one 'how X actually works'), hook with the question plus a promise ('why planes don't fly straight - 45 seconds'), explain through one engineered analogy and one visual diagram rather than definitions, build the visuals as animated illustrations that show the mechanism moving, narrate conversationally at 140-155 wpm, and end by closing the loop the hook opened. 45-75 seconds, 9:16. Reelry generates the full illustrated explainer - script, diagrams, voiceover, assembly - from a topic prompt.
The explainer is short-form's knowledge format: how mortgages actually work, why the sky is blue, what an API does, how casinos engineer the floor. Done right it is the most save-and-share dense content there is - people send explainers to whoever they just argued with. The vertical version is its own discipline, distinct from the corporate 2-minute animation: one concept, a question-shaped hook, an analogy doing the heavy lifting, and diagrams built for a phone screen. This guide covers concept scoping, hook construction, analogy engineering, the illustrated visual system, narration, and the niches where explainer channels compound.
Specs at a glance
| Ideal length | 45-75 seconds; one concept fully landed beats two concepts gestured at |
|---|---|
| Hook window | First 2-3 seconds: the question + the promise ('why X happens - and it's not what you think') |
| Concept scope | One mechanism per video; if the explanation needs 'and also,' it needs another video |
| Aspect ratio | 9:16 vertical, 1080x1920; diagrams designed for phone-width, not shrunk from slides |
| Platform limits | TikTok up to 10 minutes, Reels and Shorts up to 3 minutes; explainers earn 90 s only with exceptional material |
| Narration | Conversational, 140-155 wpm, second person where possible ('your bank does this with your money') |
| Posting cadence | 1 daily is achievable with an AI pipeline; series ('how money actually works, part 6') compound |
Free tool for this format: Facts Video Generator
Generates verified did-you-know facts with kicker lines and hooks across eight categories - mine it for the surprising facts that become explainer topics, then build the mechanism video around the best one.
Why this format works
- Curiosity gaps are the strongest cognitive hook there is: a well-formed question the viewer cannot answer creates an itch only the explanation scratches.
- Explainers are send-content: 'this explains what I was trying to tell you' is one of the most common share motivations on every platform.
- Understanding feels like a gift, and audiences follow accounts that reliably deliver it - explainer channels have among the best follow-through-rates from a single viral video.
- The format is fully faceless and fully AI-producible: script, illustrated diagrams, voiceover, and assembly are exactly the pipeline AI video tools automate.
Step-by-step guide
1.Scope to one mechanism
An explainer explains one thing: not 'economics' but 'why your rent went up when the building didn't change.' The scope test: can you state the question in one sentence a 12-year-old would understand, and the answer in two? If the answer requires 'and also' or 'but first you need to know,' split it - the prerequisite is its own video, and series ('how money actually works, part 3') outperform compressed single videos anyway. Concept selection is most of the format's craft: surprising mechanisms behind familiar things is the reliable well.
2.Build the question-promise hook
The hook has two parts: the question that opens the curiosity gap ('why do casinos have no windows?') and the promise that scopes the payoff ('the answer is a billion-dollar design decision - 60 seconds'). The question must be one the viewer realizes they cannot answer about something they thought they understood - familiarity plus ignorance is the trigger combination. Counterintuitive framings ('everything you know about X is backwards') work but must be honored by the content; over-promised hooks are the format's spam signature.
3.Engineer the analogy before writing the script
The analogy is the actual explanation - definitions are what people forget, analogies are what they retell. Engineer it deliberately: find the structure of the mechanism (flows, feedback, bottlenecks), then map it onto a domain the viewer's hands know (plumbing, traffic, kitchens, queues). 'Inflation is too much water in the soup' beats two paragraphs of monetary policy. One analogy per video, extended through the whole explanation - switching analogies mid-video resets the viewer's model and loses them.
4.Design diagrams that move
Explainer visuals are animated diagrams, not decoration: the mechanism should visibly operate - arrows flow, queues build, the bottleneck clogs. Build for phone-width: one visual idea per frame, labels in large type, color used to track the moving part. Illustrated style keeps abstract content warm; Reelry generates diagram-style illustrated frames and animates them, which is precisely the explainer's production need. The test: could a viewer follow the mechanism with the sound off? The diagram should almost manage it alone.
5.Narrate like explaining to a smart friend
Conversational register, 140-155 words per minute, second person wherever the mechanism touches the viewer ('when YOU tap your card, here's the four-step race that happens'). No jargon undefined, no hedging preambles, and one deliberate beat of pause after the core insight lands - the pause marks 'that was the thing.' End by closing the loop: restate the hook's question and the now-obvious answer in one line ('no windows, because time is the product').
6.Serialize and let the comments commission topics
Explainer channels compound through series ('how your money actually works,' 'design secrets of everyday things') because each video recruits for the others. The comment section is the topic engine: every 'wait, but why does X...' reply is a pre-validated next video - answer it with a video reply and the format becomes a flywheel. Post daily if production allows; the AI pipeline (script to diagrams to voiceover to assembly) makes daily achievable for a single creator, with verification as the human step that cannot be skipped.
Examples by niche
Money/finance niche
'Why your bank wants you to tap, not insert.' Hook: the question everyone's hands know. Analogy: payment rails as toll roads - each method takes a different road with different tolls, and the bank owns more of the tap road. Diagram: the four-step race animated between card networks. Loop close: 'every tap is a fraction of a cent choosing sides.' Finance explainers convert at the highest follow rates in the format - money mechanisms touch everyone and are explained to no one.
Everyday-design niche
'Why casino carpets are ugly on purpose.' The mechanism: visual fatigue management - your eyes rest on the tables and machines because the floor refuses to host them. Diagram: sightline heatmap animated over a casino floor illustration. Series frame: 'design secrets of everyday things, #12.' The everyday-design well (airport layouts, supermarket milk placement, traffic light timing) is effectively bottomless and entirely faceless-producible.
Tech-for-normal-people niche
'What actually happens in the 2 seconds after you hit send.' Analogy: the message as a postcard cut into pieces, mailed separately, reassembled at the destination - packets without ever saying packets. Diagram: the postcard pieces racing different routes. The translation discipline - explaining infrastructure through objects - is the entire niche, and 'explain it like I'm not in tech' audiences are huge, loyal, and underserved by content that always either dumbs down too far or jargons up too fast.
Common mistakes
Two concepts in one video
The video about inflation AND interest rates explains neither - each concept's prerequisites crowd out the other's payoff. One mechanism, fully landed, with the second concept teased as the next video. Scope discipline is the format.
Definitions instead of analogies
Reading a definition over a diagram is a lecture slide, not an explainer. The analogy carries the understanding; the definition, if needed at all, arrives after the viewer already gets it ('that race is what's called payment routing').
Over-promised hooks
'Scientists HATE this' framing on a mundane mechanism torches the channel's credibility - the explainer audience is precisely the audience that punishes clickbait debt. The hook's promise must be paid in full by the content, every video.
Templates
One-concept explainer template (60 seconds)
0-3s: question + promise hook. 3-10s: the familiar thing, reframed as a puzzle ('you do this every day - here's what you never see'). 10-40s: the mechanism via one extended analogy, diagram animating in sync, the moving part color-tracked. 40-50s: the core insight + a beat of pause. 50-60s: loop close (question restated, answer now obvious) + series tag ('part 7: why the opposite happens in Japan').
Related resources
For hook formulas you can apply across all these formats, read the TikTok hook formulas that convert guide on the Reelry blog.
Generate your first reel with Reelry
Reelry produces complete illustrated TikTok reels from a text prompt - script, frames, voiceover, animation, and assembly - in under 5 minutes.
Starter plan from $19/month · 7-day money-back guarantee · Free plan available, no credit card required
Create your first reel - freeReelry for specific creators
Related guides
How to Make History Videos (2026)
The narrative sibling: mini-documentaries with verification discipline.
How to Make Educational TikTok Videos Without Being Boring
The broader educational content playbook.
How to Make Product Videos (2026)
Problem-demo-proof structure when the explainer sells something.
Frequently asked questions
How long should an explainer video be?
45-75 seconds for one concept: hook (3s), puzzle setup (7s), mechanism via analogy and diagram (30s), insight and loop close (15s). The corporate 2-3 minute explainer is a different product for a different context - in vertical feeds, one fully-landed concept under 75 seconds wins.
How do I make explainer videos with AI?
The explainer is the most AI-native format there is: generate the script structure from a topic prompt, illustrated diagram frames per beat, voiceover, and assembly - Reelry's pipeline does all four from a single prompt, in 9:16 with captions. The human steps that remain: verifying the mechanism is actually correct, and engineering the analogy, which is where explainers live or die.
What makes an explainer hook work?
Familiarity plus ignorance: a question about something the viewer encounters daily but realizes they cannot explain ('why do casinos have no windows?'), paired with a scoped promise ('the answer is a design decision - 60 seconds'). The gap between 'I know this thing' and 'I can't answer this question about it' is the strongest curiosity trigger available.
Where do I find good explainer topics?
Three wells: mechanisms behind daily actions (tapping a card, the milk at the back of the store), 'why is it like that' design questions (casino carpets, airport layouts), and the comment section - every 'wait, but why...' reply under your videos is a pre-validated topic. Reelry's facts generator surfaces surprising verified facts; the ones that make you ask 'wait, how?' are explainer topics.
Do explainer channels need an expert on camera?
No - the format is natively faceless: animated diagrams, illustrated scenes, and voiceover carry it completely, and some of the largest explainer brands have never shown a face. What replaces the on-camera credential is correctness over time: verified mechanisms, sources for contestable claims, and visible corrections when wrong. The trust is in the track record, not the face.