· By Angie Mok
How a Yellow Dude video gets made in 2026
The TL;DR version
Roughly five weeks per video
Topics are Wayne's call, from his own training and what the numbers say the audience needs
Research is checked for attribution before writing, and expert coaches sign off the technical claims
Concept, storyboard, illustration and animation are drawn by people
A real voice actor performs every script
AI is a productivity tool in the process, not a replacement for the people (details at the bottom)
Everyone who worked on a video is named in the description
A Yellow Dude video takes about five weeks. Almost all of that is drawing: concept, storyboard, illustration, animation, frame by frame, by people.
One day of it is writing. The other twenty-odd days are drawn in five-day stages, and no stage moves until someone has signed off the one before it.
Here's the whole thing.
Where an idea comes from
Every video starts on one board. Wayne, who founded the channel and has trained calisthenics for most of his adult life, decides what's worth making. That call is his, and it comes out of his own training experience.
We also watch the niche. If a topic is clearly landing for people, that tells us there's an audience for a better explanation of it.
We also read our own numbers. Twenty-six months of them say something blunt about this channel: training and body topics work, general lifestyle and wellness topics don't. So an idea gets tagged green, yellow or red before anything else happens, and the rule is simple: does this make you better at training, or is it just another habits video.
Most ideas die here. That's the point of the board.
Anything that survives the board gets built from scratch: our research, our framing, our drawings.
Coaches see it twice
This is the step people don't expect.
Once the topic is finalized, we go to a network of expert coaches (people who train and teach calisthenics and movement for a living) for tips on what belongs in it. What's worth including, what we've framed wrong, what a coach would actually tell someone standing in front of them. Their input feeds the research and the outline.
The topic itself is Wayne's call. Which routine, which progression, which exercise, and what a viewer at that level can actually do.
So a script begins with tips from people who coach this professionally, not just with what we could find written down. Only then does anyone start writing.
Those same coaches come back later, at draft stage, to sign off on the technical claims: the ones flagged in the script and carried through storyboard and animation. Two passes: what goes in, then whether what we said about it holds up.
That second pass exists for a simple reason. The person writing the script is not the person training at the level the script describes. Reading everything published on a movement is not the same as having coached someone through it, and no amount of research closes that gap. So the people who do this for a living go first, and they go last.
Coaches don't decide what we make. They decide whether what we say about it survives someone who does this for a living.
Research
Expert tips, primary sources, and studies read rather than cited secondhand.
Everything we draw on gets checked for attribution before writing starts: journals, papers, published articles. If a claim came from someone's work, it carries their citation.
Two rules we don't bend. Every study gets verified against the actual paper, and anything we can't verify stays out. And we check whether a famous finding still holds, because the most-quoted study in fitness is very often the one that failed to replicate.
Citations appear on screen, on the exact sentence they support, and are never read aloud.
The script — one day
The outline usually arrives with the idea itself, from Wayne. Where it doesn't, it's built from our own standard structure: hook, intro, value, endscreen. Either way, nothing moves until he has read the finished script, signed off the structure and picked the hook.
The script is drafted in six layers, worked in order, each one gating the next:
- Research and attribution review: the gate above
- Outline: beats, not sentences, and one sentence stating what the viewer will understand by the end that they didn't before
- Full draft: visual direction written inline for the illustrators, and every technical claim flagged for checking
- Hooks: written last, once the whole script exists, so the body never leans on any one of them. Three to five options; Wayne picks
- Humanize: one pass hunting the tells (filler transitions, repeated crutch phrases, triplet structures, anything that reads like a blog post instead of a person talking)
- Sponsor integration: if there's a sponsor
The drafting happens layer by layer, and each layer is reviewed and approved by a person before the next one starts.
Three checks it has to clear before Wayne sees it. Does it explain why, not just what. Is there at least one thing in it the viewer hasn't heard before. And would someone who already trains calisthenics still find it worth watching. Any no, and it goes back.
Every technical claim carries a flag that stays with it through drawing and animation, so whoever has to draw that movement can see exactly what's being claimed and query it if it looks wrong. An expert coach signs the claims off at draft stage.
One day of writing. Two hours of review. Then it leaves the keyboard and doesn't come back.
Everything after this is drawn by people
Concept and rough sketch: 5 days. The script becomes a shot-by-shot plan. Every frame planned, nothing left for the editor to invent. It's slower than handing an editor the audio, and it's why the videos don't drift.
Gate 1: animatic review. A rough timed pass with draft voiceover, watched before anyone spends real hours on it. No music, no effects, no polish. Three questions: does the story hold, does the pacing work, does the order match the plan.
Storyboard and illustration: 5 days. Every frame drawn. Yellow Dude is hand-illustrated, frame by frame, in the same palette and the same two typefaces every time.
Animation: 5 days. Rigged figures posed and keyframed in After Effects. No stock footage, no live action, ever. 4K minimum.
Gate 2: draft 1 review. The cut goes into Frame.io and gets commented on frame by frame.
Revision and draft 2 — 5 days. Sketched placeholders get replaced with finished assets and every note from gate 2 gets worked through. Then gate 3, the draft 2 review. The last look before the file goes final.
Final 4K file.
Thumbnails
Thumbnails take a full day of their own (concept, draft, review, final), absorbed into the concept and drawing stages rather than bolted on at the end.
Three named gates after the script, on top of the six layers inside it. Most of them catch something. This is where the five weeks actually go.
The voice
A voice actor performs every script. The recording is then processed so the Yellow Dude character sounds the same across the whole catalogue.
Music is licensed from Epidemic Sound. Every asset, every track, cleared.
Everyone gets named
Every person who worked on a video is in the description. Writer, illustrator, storyboard, animator, editor, localization.
That isn't standard in this niche. We decided at the start that if someone drew a frame of it, their name is on it, and that includes creators outside the team whose work an idea came from.
What this looks like on a collaboration
When we make a video with another creator, we invert the process.
They pick from a shortlist of their own proven videos. They write the script themselves, in their own voice, and they record it themselves. The words are theirs and the performance is theirs. Every exercise demonstration is animated on our side, so the ask on them stays small.
Then our whole pipeline runs on top of it: storyboard, illustration, animation, review gates, the lot. Their idea, their words, their name, and our five weeks of drawing.
We've done this with Hampton from Hybrid Calisthenics and Matt from Red Delta Project, with more lined up. Every expert we work with is on the experts page.
Where AI fits in all this
AI shows up across this process the way it shows up in most modern studios: a tool for doing the work better and faster, not a replacement for the people doing it. Where it helps:
- Writing: brainstorming angles, tightening grammar, and finding a sharper way to say something so a line lands with the effect we want. Drafting is AI-assisted, layer by layer, with a person directing and approving each layer, and nothing ships until Wayne signs off the finished script.
- Story: pressure-testing pacing and the emotional arc of a video before storyboarding starts.
- Research: helping find and cross-check sources faster. The verification itself is done by a person, against the actual paper.
- Illustration: anatomy and perspective references when planning a difficult pose. Every frame that ships is drawn, rigged and animated by our illustrators and animators.
- Localization: helping bring our videos to the translated channels in other languages.
- Voice consistency: the processing described above, applied to a real actor's recording.
-
What it never does: pick the topics, replace the coaches, draw the frames, or publish anything a person hasn't reviewed and put their name to.
The five weeks are the difficult part. The credits are the name. That's the whole standard, and it's the same standard whether a video is sponsored or not.