method

What the research establishes and what it refutes

106 criteria drawn from 24 papers, each with its source and its strength of evidence. The golden ratio and symmetry are refuted; what actually predicts, and what the engine refuses to score, is set out here.

Analyse my photo
guidance

A coach for men and women alike

automated analysis
menwomenseparate criteria

The scale is not the same

Feminine facial shape predicts attractiveness in women. Masculinity in men predicts nothing at all: a null effect across 400 raters. And a wide nose, which the classical canons penalise, goes with higher ratings in one of the populations studied.

A single scale therefore gets half its users wrong, sometimes in the opposite direction.

What this coach refuses

A product that says what it won't sell you is the more credible one.

  • Blood type — 1,415 references screened, nothing conclusive
  • Mewing — the palatal suture is fused in adults
  • Facial yoga — measures muscle tone, never appearance
  • Perfect symmetry — predicts nothing once averageness is controlled
lifestyle−4 years perceived age

Pigment spots

The one lever that holds across all origins. Causal evidence, odds ratio 6.07.

lifestyle7 facial cues

Sleep

Eyelids, dark circles, complexion, redness, puffiness. Free, reversible in a few nights.

lifestyle−3.7 years perceived age

Sagging and wrinkles

Sun protection, quitting smoking, sleep. Retinoids have the strongest evidence.

groomingimmediate effect

Facial hair

Shifts perceived dominance and age. No universally optimal length exists.

photo qualitycorrected on the spot

Framing and light

A 12° tilt manufactures an asymmetry that isn't there. An unusable photo is refused.

Startfree, no sign-up

Analyses and recommendations are generated automatically from a database of scientific papers, each piece of advice carrying its source. They are not a substitute for medical advice.

example

What it looks like what you get

illustration data
Illustration of a facial mesh overlaid on a portrait
28 measurementspose validated
estimate
4,8/ 10
plausible interval
3,4 — 6,2
0510

The interval is wide on purpose. Two cultural groups of humans agree only at r ≈ 0.62 on the attractiveness of the same face: announcing a score to the nearest tenth would be invented precision.

computed sub-scores

General proportions4.6
Mouth and lips5.1
Jaw and chin4.2
Eye region4.9

skin · analysed on the device

Skin-tone evenness2,41middling
Dark circles6,2 ptsimprovable
Surface regularity18,3favourable

These are the best-evidenced dimensions in the base — and the only genuinely modifiable ones.

not scored, and why

  • · Symmetry — measured but never scored: it predicts nothing once averageness is controlled for.
  • · Nose — 5 criteria in the base, all refuted on the populations measured.
  • · Averageness — needs at least 10 reference exemplars to mean anything.
and what you do with it

The plan that comes out 3 levers

ranked by documented effect, sources included
lifestyle3 to 6 months
Pigment spots
−4 yearsof perceived age

Daily sun protection. Causal evidence, odds ratio 6.07.

Fink 2012 · Vierkötter 2009

lifestylea few nights
Sleep
7 cuesfacial cues affected

Eyelids, dark circles, tone, redness, puffiness. Free and reversible.

Axelsson 2010 · Sundelin 2013

shottoday
Framing and light
immediatecorrected on the spot

12° of deviation manufacture an asymmetry that is not on your face.

built-in quality control

and what nothing changesBizygomatic widthNasofrontal angleMidface height— named to spare you time and money spent on them.
what the research shows

What can actually change and what it doesn't

What moves
Skin, sleep, sun
Fewer pigment spots means four years off your perceived age. The one lever that holds for every face.
What doesn't move
The geometry of the face
Proportions, widths, angles: none of it shifts without surgery. 64 criteria out of 106.
Criteria database
106 criteria
24 papers · every piece of advice carries its source · 830 calibration exemplars
the glow-up

What actually changes, and how long it takes

documented effects, sources in the report

Of the 106 criteria in the base, 64 cover traits that nothing can change. The glow-up is everything else — so it is the whole of what is up to you.

  1. today
    Framing and light
    immediate

    12° off-axis manufactures an asymmetry that is not on your face.

  2. a few nights
    Sleep
    7 cues

    Eyelids, dark circles, complexion, redness, puffiness. Free and reversible.

  3. 4 to 8 weeks
    Skin
    best evidence

    Retinoids carry the strongest evidence in the entire corpus.

  4. 3 to 6 months
    Pigment spots
    −4 years

    Of perceived age. Causal evidence, odds ratio 6.07. Holds across every ancestry.

  5. 6 months and beyond
    Sagging and wrinkles
    −3.7 years

    Of perceived age. Sun protection, quitting smoking, regular sleep.

These effects add up

No single lever transforms a face on its own, and nobody has measured their sum. They do all bear on perceived age and health, and they are the only ones in the base whose effect depends on neither ancestry nor sex.

A glow-up is measured, not noticed

A mirror does not compare: it shows a face today, and memory of the one from three months ago is poor. That is why the product records follow-up points instead of handing out a score and stopping there.

method

Why progress is worth more than a score

An absolute score is fragile. Two cultures agree only to r ≈ 0.62 on the same face, and the best published engine tops out at 0.78 — against an average of raters, not against any real person.

Comparing a face to itself changes everything: those error sources become constant and largely cancel out in the difference. What does not cancel out is the shot itself — hence the refusal to compare two photos whose angle or lighting differ too much.

The classical canons, meanwhile, are refuted. The golden ratio and the Marquardt mask do not hold: of 77 “ideal” parameters tested, only 16 survive. Symmetry, this industry's number-one selling point, predicts nothing once averageness is controlled — three independent studies, one spanning ten cultures.

For adults only, on your own portrait. No photo is transmitted or stored. Analyses are automated and are not a substitute for medical advice.

Analyse my photofree, no sign-up