What it measures and what it doesn't
A face analysis system locates landmarks on the image — outline, brows, eyes, nose, mouth, jawline — and calculates distances and ratios between them. From there come proportions, relative symmetries and the position of the main lines. They're geometric data: precise as measurements, silent on meaning.
What the system doesn't do is decide. It doesn't know how much time the client spends styling, doesn't know the hair's technical history, doesn't know there's a wedding in two months. The jump from data to proposal remains entirely professional, and that's exactly where the salon adds value rather than being replaced by a tool.
There's also the question of conditions. Framing, lighting and head tilt change the measurements more than you'd think: a photo taken from below lengthens the lower third, side lighting alters the outline. That's why guided capture matters as much as the algorithm, and a reading taken from a rough photo is worse than no reading at all.
Where automated analysis adds real value
The chart is qualitative and compares the observable effects in the salon when the reading of the face becomes visible to the client instead of staying implicit.
Reading by eye and assisted reading
Neither replaces the other. The difference is in what remains for the team and the client after the consultation.
| Aspect | Experienced eye alone | Experienced eye with analysis |
|---|---|---|
| Speed | Immediate for the experienced | A few minutes more, with shared output |
| Transferability across the team | Hard to teach verbally | Explicit criteria, the same for everyone |
| Client understanding | Depends on the ability to explain | She sees the proportions being discussed |
| Trace for the next visit | One professional's memory | A record with measurements and reasoned choices |
Fitting the analysis into the flow
Ask for the goal first
The analysis answers a question already asked.
Take care with capture
Front-facing, even light, head straight.
Show two measurements
Translated into consequences, not numbers.
Close with a spoken proposal
Not with a screen.
How to fit it into the flow
- Don't open the consultation with the analysis: the client's goal first, then the reading.
- Show two or three measurements, not the whole output.
- Always close with a spoken proposal, not with a screen.
Order matters more than the tool. When the analysis comes before the question «what would you like to change?», the client feels assessed by a machine. When it comes after, it becomes the visual answer to a question she has just asked, and the tool's role is obvious: it's there to explain, not to decide.
The second point concerns language. Data should be translated into concrete consequences — where the eye moves, how much work a style takes — and never read out as numbers. A consultation in which percentages are said out loud is a consultation that lost the client several minutes ago.
An AI consultation designed for the salon's real work
Saloria guides the journey on a tablet next to the client: welcome, profile, face analysis, guidance look plan, look plan, technical protocol and closing. The salon doesn't sell magic; it sells a clearer method to decide cut, colour and maintenance journey.
Frequently asked questions
Does the AI decide the cut?
No. It returns measurements and proportions. The proposal comes from combining that data with the hair's technical history and what the client wants to achieve.
Do you need a professional photo?
You need a front-facing photo with even light and a straight head. Saloria's guided capture checks those conditions first, because a crooked photo produces wrong measurements.
Does the salon keep the photo?
No. It's used during the session and deleted according to the configured retention policy, with consent collected before the shot.