A while ago, a founder sent me the watch he wanted to launch. The images had been made with AI and, at first glance, they looked finished.
Looking closer, I could see the problems. The dial had inconsistencies. The strap was attached to the case in a way no lug or spring bar could follow. When I explained that all of this had to be reworked, he answered: “But the concept is done. You just need to make the technical file, so it shouldn’t cost much.”
I hear this sentence more and more often. It comes from founders who are motivated and who have already spent hours on their idea, so it deserves a real answer.
What an AI image gives you
AI visuals are a good thing for a project. Before, people described their watch with words and a few photos of other models, and I had to guess what they had in mind. Now they show me the mood, the shapes, the colours and the details they care about. We start the conversation from something concrete, and that saves time.
So if you have visuals, bring them.
What it can’t give you
An image generator draws what looks right. It doesn’t know which movement goes inside, or how the case parts fit together. Nobody asked it whether a factory could make them, or whether the strap would hold on a wrist for years.
The same founder had a second concept, with light inside the watch: luminous, electronic elements in the case. On screen it was striking. In reality it raised technical questions (where the light comes from, how it is powered, how the watch stays water resistant) that would call for a real research phase, with studies and probably several prototypes. It might have been worth exploring, but it was far from simple. For him it was, because the light was already in the picture.
Why “the concept is done” is rarely true
The image captures an intention, and that is valuable. The design itself still has to be done, from the proportions and the construction to the materials. A technical file cannot describe a watch that has not been designed yet. If I wrote one straight from the image, the factory would come back with questions, or make its own choices, and the watch would no longer really be yours.
Another founder told me his whole concept on a call, in detail, and concluded the same way: since everything was already thought through, it should be quick and cheap. I understand the reasoning, but the time goes precisely into turning those ideas into parts that fit together, and that work hadn’t started yet.
How I work from your visuals
I start by asking what you wanted to say. Who is the watch for? Which part matters most to you: the case shape, the dial, a detail on the crown? What should people feel when they wear it? There is always an intention behind the visual, even when the image shows it poorly.
Then I redraw the watch, front and side, with real proportions. We decide together what becomes the signature of the watch and what can be simplified. I work the volumes in 3D, choose materials and finishes, and keep comfort in mind, because a watch is worn all day.
Before anything goes to a factory, the studio prints the watch in resin at full scale. You hold it, you try it on, and we adjust. Then comes the tech pack, the document that explains the watch to the factory. I wrote a separate article on what a watch tech pack contains.
Send us your visuals
At ACTE.5, we take AI visuals for what they are: a starting point that tells us a lot about you. Send them with a few words about your brand, and we will tell you honestly what can be kept, and what needs to change before a factory can make the watch.
