The Slow Build

Cadran is a small company in Nantes. Its app lets people book a doctor's appointment in under a minute. Twelve developers build it, and they are good at their jobs. Nothing else about the company is fast.

A two-line change takes eleven days to reach real patients. Nobody inside the building thinks this is strange, because it has always been that way. Then Sylvie Marchand, the technical director, asks a simple question in a glass meeting room: what does that delay cost us? Nobody can answer, and the silence buys Yasmine Belkacem exactly one quarter to find out.

Yasmine is a senior developer, not a manager. She starts with a stopwatch and a grey notebook, sitting beside Hugo Lemaire — three weeks into his first job, still waiting for a database password. She learns to look at friction through three ideas: how long a developer waits for an answer, how long a developer can concentrate without being pulled away, and how much a developer must hold in their head to change one line of code.

Then she stops guessing and starts asking. Twelve interviews. A wall covered in every step between an idea and a real patient — thirty-four of them. The worst step on that wall was written four years ago by Karim Diallo, who has watched three grand company methods die and expects this to be the fourth. He walks out of the room.

What follows is not a clean victory. A survey contradicts the wall. Build times fall and bugs rise at the same time, and Sylvie sees the bug chart first. Yasmine has to argue against her own director for the boring project instead of the exciting one. Léa Nguyen fixes her own team without asking anyone. A bad month arrives: a broken pipeline, a frozen release, and a job offer in Hugo's inbox.

What finally works is not a hero and not a tool. It is a rule small enough for everyone to follow, and a habit of measuring the wait instead of arguing about it.