Why we built Facet

The first time I tried to get a clinical read on my own face, I paid QOVES $150 and waited 28 days for a PDF. The report was thorough. The wait was not.

While I was waiting, I tested three free face-scoring apps to fill the gap. Each one gave me a single number out of ten with no explanation of what it measured. Two of them rated the same photo differently when I uploaded it twice.

Both ends of the market were broken in opposite ways. The clinical option was slow and expensive. The free option was fast and useless. Nothing in between treated a face the way a dermatologist's notes do: a list of specific measurements, against specific clinical thresholds, with the paper that defines each one cited next to the number.

What was actually missing

• Speed. Modern UX expects an answer in seconds, not weeks. A 28-day wait creates buyer's remorse before delivery.
• Methodology transparency. "Trained on 2,000 studies" tells you nothing. A specific cited paper per threshold tells you everything.
• A price that matched the act. Facial analysis is not weekly software. Charging a subscription for a one-and-done report is the wrong shape.

The bet behind Facet

Facet's bet is that a serious audience exists in the middle. People who have already spent money on skincare, gyms, supplements, and procedures. People who quietly read the studies under the headlines. People who do not want a TikTok rating and do not want a $150 annual subscription either.

Built for that audience, the product writes itself. MediaPipe maps 468 landmark points at sub-millimeter precision. A deterministic scoring engine compares those points against thresholds drawn from peer-reviewed morphometry research. Every threshold carries the citation that defines it. The output is a numeric breakdown across 10 modules, a ranked priority list, and a natural and clinical protocol per module.

Where this goes

The first version of Facet is live. The next twelve weeks of this blog are about the things that did not fit on the marketing page: how MediaPipe actually works, why ethnicity-aware scoring is the default rather than the exception, what the Goode index has to do with your nose, and why we will keep charging $39 once instead of $150 a year. If any of that sounds interesting, the rest of the series is for you.