Artificial intelligence in the hair salon: where it helps, where it doesn’t

Artificial intelligence arrived in hairdressing through the marketing door: filters, colour simulators, image generators. That’s the visible part, and it’s the least useful. What actually changes a salon’s day-to-day happens elsewhere.

Where it truly helps

The formula. A client arrives with an uneven base, an uncertain history and a specific request. The professional has to decide on a base, a highlight, a developer and a processing time. A well-built technical assistant doesn’t guess: it applies colour reasoning to a specific catalogue, and proposes a starting point that the hairdresser confirms or corrects.

The timing. Technical questions don’t arrive during office hours. They arrive on a Saturday late morning, when no one can be reached. A tool available round the clock answers at the moment when the answer has value.

Repetition. Stock tracking, restocking, follow-ups, answering administrative questions: these are tasks that require no judgement, yet eat up hours. This is automation’s natural territory.

Where it doesn’t help

Diagnosis in place of the professional. An assistant doesn’t touch the hair, doesn’t see the salon’s lighting, doesn’t know the client’s real history. Any answer given remotely remains a suggestion, never a decision.

The promise of a result. An on-screen colour simulation says nothing about what the hair will actually do. It sells a dream, and it’s the hairdresser who then bears the responsibility for it.

Empty content. Generating three posts a day doesn’t build a brand. A profession recognises itself in what is real, shown, dated, signed.

Three criteria for judging a tool

Which catalogue does it reason from? A colour recommendation only has value if it concerns the products actually sitting on the shelf. A generic assistant will give a plausible but unworkable answer.

What does it do with your data? A salon’s conversations are about its clients. They should be used to answer the question, not to feed a marketing database.

What do you actually gain, in minutes? If the tool doesn’t save time or prevent a measurable mistake, it serves no purpose, whatever the technology behind it.

What we’re building

The group develops its own technical assistants, each tied to a brand and its catalogue: Edgar on the KUST side, Lucy on the YDOL side. They are not interchangeable, and that’s deliberate: a formula isn’t given in the abstract, it’s given with the products the hairdresser has to hand.

Our rule of conduct is simple. A tool must be judged by what it saves behind the chair. If it changes nothing in the salon, it has no reason to exist.