How FaceIQ Labs Scores Work: The 0–10 Scale, Four Pillars and Overall Score

Key facts

• FaceIQ Labs uses one 0–10 scale for everything from individual ratios to the overall score; 5 is the population average.
• The scale is non-linear: 7 is roughly the top 6% (about 1 in 18, around the 94th percentile) and 8 is about 1 in 250 (around the 99.6th percentile).
• Rarity is symmetrical, so a 3 is as uncommon as a 7.
• Four pillars are scored independently: Harmony, Angularity, Dimorphism and Features.
• Harmony uses 70 measurements (37 from the front photo, 33 from the side photo) compared with ideal ranges adjusted for gender and ethnicity, and is weighted roughly 60% front photo and 40% side photo.
• Angularity has 5 assessments, Dimorphism has 10, and Features has around 30.
• The overall score is not an average: it is a power mean with a very small exponent, close to a geometric mean, that heavily penalises the weakest pillar.
• Four 7s give 7.0, but three 8s and a 4 give roughly 6.7.
• The overall score is only calculated once all four pillars are complete.
• On Basic and Pro, Progress Tracking plots saved analysis scores over time.

What does a 0–10 score mean on FaceIQ Labs?

FaceIQ Labs scores everything on one 0–10 scale, from individual ratios up to the overall score. A 5 is the population average. The scale is non-linear, so each step up is rarer: a 7 is roughly the top 6% (about 1 in 18) and an 8 is about 1 in 250. Moving from 5 to 6 is common, while the top of the scale is essentially theoretical. Rarity is symmetrical, so a 3 is as uncommon as a 7. Percentiles shown in the app come from the same curve: a 7.0 reads as roughly the 94th percentile and an 8.0 as the 99.6th.

What do the FaceIQ Labs score labels and rarity levels mean?

FaceIQ Labs attaches a label and an approximate rarity to each score band so a number can be read without memorising the curve. For example, 7 is labelled Great at about 1 in 18, and 8 is labelled Exceptional at about 1 in 250. The full bands are: 9.5–10, Near / idealised perfection, 1 in 100,000+. 9, Approaching perfection, 1 in 20,000. 8.5, Outstanding, 1 in 2,500. 8, Exceptional, 1 in 250. 7.5, Excellent, 1 in 60. 7, Great, 1 in 18. 6–6.5, Good / very good, 1 in 3.5 to 1 in 6.5. 5–5.5, Average / slightly above, 1 in 2. 4–4.5, Below average, 1 in 2.2 to 1 in 3.5. 3–3.5, Poor / below average, 1 in 6.5 to 1 in 18. Under 3, Poor to extreme disharmony, 1 in 60 and rarer. The label wording is written for Harmony (for example Good harmony or Exceptional harmony) and is swapped for the pillar being viewed, so Angularity shows Good angularity. Features uses its own shorter set: Poor, Weak, Fair, Average, Good, Strong, Exceptional, Elite.

What are the four pillars FaceIQ Labs scores?

FaceIQ Labs scores a face on four independent pillars, each on the 0–10 scale: Harmony (proportion), Angularity (definition), Dimorphism (how masculine or feminine each region reads) and Features (quality rather than shape). They measure different things and are scored differently, so one face can rate differently on each. A face can be highly harmonious and only moderately dimorphic, or the reverse, which is why the pillars are scored separately.

How is the Harmony score measured?

FaceIQ Labs measures Harmony, the proportion pillar, from 70 landmark-based measurements: 37 on the front photo and 33 on the side photo. Each is compared with an ideal range from published aesthetic research, adjusted for gender and ethnicity, and scored out of 10. The pillar score is a weighted combination. Ratios carry different weights depending on how much they influence perceived attractiveness. Harmony is weighted roughly 60% from the front photo and 40% from the side photo, and side photos are standardised to the Frankfort horizontal plane during landmarking. It is the only pillar that is pure geometry, and the one most sensitive to landmark placement.

How is the Angularity score measured?

FaceIQ Labs measures Angularity, the definition pillar, through five assessments: cheekbone prominence, jaw definition, cheek leanness, chin definition and submental (under-chin) definition. Each is rated on a continuous morph from soft to sharply defined and scored against the ideal level for the user's gender. Angularity is mostly about soft tissue: body fat, hydration, and how much bone shows through.

How is the Dimorphism score measured?

FaceIQ Labs measures Dimorphism by scoring how masculine or feminine each of ten regions reads: jaw, brow ridge, eyebrow thickness, eyes, face shape, facial hair, hair length, hairline, lips and nose. Each is placed on a masculine–feminine morph and scored on how well it matches the ideal for the selected gender. Because a face can be highly harmonious and only moderately dimorphic, or the reverse, this pillar is scored separately from Harmony.

How is the Features score measured?

FaceIQ Labs measures Features, the quality pillar, with around thirty assessments covering skin (acne, scarring, pigmentation, rosacea), under-eyes, facial creasing, sclera, eyelids, nose shape, hair health, brows and lashes, plus symmetry. They are rated by the vision model rather than measured from landmarks, and are the most changeable with skincare, sleep, grooming and treatment.

How is the overall FaceIQ Labs score calculated?

FaceIQ Labs combines the four pillar scores with a power mean using a very small exponent, which behaves close to a geometric mean and heavily penalises the weakest pillar. It is not a simple average, so a lagging pillar drags the overall score down. Four 7s give 7.0, but three 8s and a 4 give roughly 6.7, not the 7.0 a simple average would produce. Improving the weakest pillar therefore moves the overall score far more than improving the strongest, which is why the dashboard shows an impact-if-perfect figure for each pillar.

When does FaceIQ Labs show an overall score?

FaceIQ Labs calculates the overall score only once all four pillars (Harmony, Angularity, Dimorphism and Features) are complete. Until then, the dashboard shows the individual pillar scores with a prompt to finish the remaining assessments, so a missing overall score means at least one pillar is unfinished.

Why does FaceIQ Labs weight the weakest pillar so heavily?

FaceIQ Labs weights the weakest pillar heavily because, as the company describes attractiveness research, a single conspicuous weakness is noticed more than several modest strengths. A face with no glaring issues therefore outscores a face with one brilliant feature and one obvious flaw. This is why the overall score uses a power mean rather than a plain average.

How are scores combined within each pillar?

Inside each FaceIQ Labs pillar, individual scores are combined with the same weakest-weighted idea as the overall score. Harmony ratio scores use a gentler power mean that is still weighted toward weaker items, and Angularity, Dimorphism and Features combine their individual assessments the same way. A single very poor ratio or assessment therefore costs more than its raw weight suggests, and it will usually appear near the top of the flaws list.

What are strengths, flaws and impact in a FaceIQ Labs report?

In a FaceIQ Labs report, Harmony results are summarised as strengths, meaning ratios comfortably inside their ideal range, and flaws, meaning ratios outside it. Each flaw shows an impact: how many points the score would rise if that measurement alone were brought into range, accounting for overlaps between related ratios. Sorting by impact is the fastest way to see what matters. Items marked possible are borderline: a small landmark shift would change the verdict, so FaceIQ Labs advises reading its accuracy guidance before acting on them. The Free plan shows 10 harmony ratios and 2 strengths/flaws.

Can I track how my FaceIQ Labs scores change over time?

Yes. FaceIQ Labs keeps the scores of every saved analysis, and on the Basic and Pro tiers Progress Tracking plots them over time, so users can see whether changes in diet, skincare or grooming show up in the numbers. Photo setup must stay identical between scans for the trend to mean anything. FaceIQ Labs recommends photos taken from six feet away with the back camera at 2x zoom, at eye level, with even lighting and a neutral expression.

Frequently asked questions

Is a FaceIQ Labs score of 7 good?

Yes. On the 0–10 scale, 5 is the population average and 7 is roughly the top 6% of people (about 1 in 18, around the 94th percentile). It is labelled Great for Harmony, with equivalent wording for the other pillars.

Why isn't my overall score just the average of my four pillar scores?

FaceIQ Labs combines Harmony, Angularity, Dimorphism and Features with a power mean close to a geometric mean, which heavily penalises the weakest pillar. For example, three 8s and a 4 give roughly 6.7 overall, not the 7.0 a simple average would give.

Why don't I have an overall score yet?

FaceIQ Labs only calculates the overall score once all four pillars are complete. Until then the dashboard shows the individual pillar scores and a prompt to finish the remaining assessments.

What does impact mean next to a flaw in my Harmony results?

Impact is how many points your score would rise if that single measurement were brought into its ideal range, accounting for overlaps between related ratios. Sorting flaws by impact shows what matters most.

How is Harmony different from Features?

Harmony is pure geometry: 70 measurements from photo landmarks compared with ideal ranges. Features rates quality, such as skin, under-eyes, hair and symmetry, using the vision model rather than landmarks, and is the pillar most changeable with skincare, sleep, grooming and treatment.

What labels does the Features pillar use?

Features uses its own shorter label set instead of Harmony-style wording: Poor, Weak, Fair, Average, Good, Strong, Exceptional and Elite.

Can I see how my scores change over time?

Yes, on the Basic and Pro tiers, where Progress Tracking plots saved analysis scores over time. Keep your photo setup identical between scans so the trend is meaningful.