How We Made the CRINIÈRE Episode with AI
We did not make another beautiful hair commercial — we made an episode that shows a creative idea being found in front of you. The brief was simple: make an ad for a hair product. The obvious direction was a woman using the product, looking in the mirror and feeling beautiful — exactly the ad everyone has already seen hundreds of times. The crack came from a different question: why would anyone care about her hair?
1. The brief, and the direction we rejected
"We need an ad for a hair product." The first direction was obvious immediately: a woman uses the product, her hair looks better, she looks in the mirror and feels beautiful.
And that is precisely the problem. The first idea everyone reaches for is usually the idea everyone has already made. So we stopped and asked the question that actually matters — not "how do we show the product", but "why would anyone care".
2. The crack: look for the behaviour, not the hair
Instead of showing the hair over and over, we looked for the behaviour that happens when hair is not cooperating. A woman who does not like her hair:
- checks it in the mirror;
- checks it again;
- changes her plans;
- ties it up;
- hides it;
- only lets herself be photographed from one angle.
So the problem is not only that the hair does not look good. The problem is that her whole day starts to revolve around it. And from there came the idea:
We are not selling beautiful hair. We are selling a whole day where she no longer thinks about it.
3. Why this became an episode instead of staying an ad
We could have made a standalone commercial and been done. But then it would have been one more video that looks good. Instead we placed the commercial inside an episode of our series.
The classroom
The instructor sits facing her students and rejects their generic ideas.
The crack
She leads them to the real question, and from it to the idea.
The screen lights up
She presses the remote and the camera moves into the screen.
The commercial
The viewer watches the campaign that was born in that conversation.
The ad never feels like an interruption — it becomes the outcome of the story. The viewer is not enduring a commercial on the way to content; they are watching the commercial as proof that the idea worked.
4. How we built the episode with AI — in layers, not one prompt
All of it was made with AI, but not all at once. Four layers, each with its own job:
01 · A lead character
The instructor is not someone reading lines. She has a defined character: sharp, unimpressed by generic ideas, sparing with words, and quick to spot when people are describing a picture instead of an idea. Once the character is clear, her reactions stay consistent.
02 · A fixed world
The classroom is not a backdrop; it creates the feeling of a real creative workshop. So we held the position of the screen, the arrangement of the students, the camera angles, where the instructor stands and the lighting — so the whole episode happens in one world.
03 · The product inside the plot
The product does not appear only at the end. The students discuss it, the instructor asks why anyone would care about it, the screen shows the campaign, and the commercial proves the crack. The product is not stuck onto the story — it drives it.
04 · An ad built on behaviour
Not aesthetics talking about themselves, but a sequence of actions that delivers the message without stating it.
5. The commercial itself: what she does, not what she says
The ad is built from two halves, and the message sits exactly in the gap between them.
The message is never said to the viewer. It is built through her behaviour — and that is what makes it memorable.
6. The real challenge was not beauty
The challenge was not getting AI to produce a beautiful woman or a good-looking product. The challenge was making all the parts hold together:
- the instructor staying the same character throughout the episode;
- the classroom staying in the same place;
- the product not changing between shots;
- the spray still looking like a spray;
- the hair cover staying the same cover;
- the move from classroom to screen feeling real;
- the dubbing belonging to the right character;
- and the commercial feeling like the result of the crack, not a random clip dropped into the episode.
This is exactly why an AI video can look beautiful and still not work. Beauty is only the execution layer.
7. Three questions before you open a video tool
What is the person in the video supposed to feel?
Not what they are supposed to see. What they are supposed to feel.
What real pain or behaviour am I showing?
If there is no real behaviour, what you have is probably just aesthetics.
What will the viewer remember once it ends?
Here the viewer is not meant to remember the product, but the feeling of a day where she finally does not have to think about her hair.
Want the working system, not just the idea? Open the creative guide in OMER METHOD →
What is not in this article
What you have here is the thinking behind the episode. What is not here is the working system:
- the structure for cracking an idea;
- the prompt templates;
- the method for building characters;
- reference mapping;
- how scenes are divided;
- controlling the dubbing;
- the failure-fixing process;
- and how an episode, an ad, a character and a call to action are connected.
Knowing what the idea was is not the same as knowing how to produce one yourself.
The same principles sit behind everything I build — from Monica Andrew to the Better production file.
From "I have a cool idea" to "I know how to produce it"
The guide library holds the whole process: how to crack an idea, how to build content people care about, how to plan characters and a world, how to work with references, how to write prompts you can actually produce, how to break a video into scenes, how to control dubbing and sound, how to diagnose why a result is not working, and how to add the moment of WOW without wrecking the story.
Do not settle for knowing that the idea was "don't show the hair, show the life after it works". Learn how to reach ideas like that again and again.