A face rating is only useful if you can see the working.

Why most face ratings are meaningless

A rating with no visible method has three problems at once. You can’t verify it, you can’t reproduce it, and you can’t act on it. It’s a vibe rendered as a decimal.

The tell is simple: if an app gives you a score but can’t show you which measurements produced it, there’s nothing underneath to inspect.

What a defensible score requires

• Named metrics. Specific, measurable quantities — not adjectives.
• Published reference ranges. Each metric compared against a documented ideal range from anthropometric research, with the citation attached.
• Reproducibility. Two scans under the same conditions should land in the same place.
• Actionability. The result has to separate what you can change from what you can’t.

How GlowAI builds its harmony score

75 measurements are taken from a front photo and a side profile, each compared against ideal ranges drawn from published work by Farkas, Powell, Naini, Sarver, Goode and Mommaerts. Every metric in the app carries its citation. Those individual results then roll up into a single composite:

The composite at the bottom is the score people screenshot. The five rows above it are the reason it means anything.

How to read your score without wrecking your week

A harmony score describes how your proportions compare to population averages. That’s all it does. It is not a measure of how attractive you are, and it has nothing to say about how you come across to another person — expression, warmth and grooming carry enormous weight that no static ratio captures.

The useful move is to ignore the headline number after the first look and go straight to the ranked list of fixable metrics. That list is where the score stops being a rating and starts being a plan.

Next: how the scan itself works, or how those metrics become a daily routine.

Common questions