About
Who writes this, and what it will not tell you
The person behind it
My name is Taro Schenker. I build measurement tools for the web. This is not the first face-analysis tool I have shipped — the same landmark engine runs on a sibling site, The Face Report, which has put it through tens of thousands of real analyses. The relevant experience is not a cosmetology qualification: I do not have one, and I say so on every page that gives styling advice. It is that I have watched this geometry meet real faces at volume, and seen where it goes wrong.
A concrete example, because “we test rigorously” is worth nothing without one. The eye-shape tool classifies the tilt of your outer corners. Its neutral band used to be centred on 0° — a level eye — which sounds obviously right. It is not: human canthal tilt peaks at about +4°, so a band around zero called 74.7% of people “upturned” and handed anyone with genuinely downturned eyes the advice for the opposite problem. I found the direction of the error on eight test faces here, then measured it against 27,343 real analyses on the sibling site running the same code before changing anything. Recentred on the real peak, the split is 16% upturned, 58% level, 25% downturned.
That is the standard I hold this to: when the tool and the world disagree, the tool is wrong, and the fix waits for data rather than intuition. The methodology page sets out exactly what is measured and where it stops.
Why this site exists
Most tools that read a face from a photo are built to score it. They return a number out of ten, a percentile, a verdict on whether you are attractive. That product has an audience and I am not pretending otherwise — I run one. But it answers a question that has no useful answer, and it makes people feel worse.
Face shape is different. It is a proportion, not a rank. A round face is not a worse face; it is a face that a blunt jaw-length bob will widen and a long soft layer will not. That is a question with a real answer, and the answer changes what you do on Saturday morning at the salon. So this site measures proportions and stops there. You will not find a score, a rating, a percentile or a “how do I compare” anywhere on it, and that is a deliberate constraint enforced in the code, not a promise.
How the recommendations are sourced
The measurement is geometry and is reproducible. The styling advice is not — it is craft knowledge, and it comes from the settled consensus of working hairdressers and opticians: balance width against length, soften a strong angle rather than repeat it, keep volume away from the widest point. Where sources broadly agree, we say it plainly. Where they do not, we say that too rather than picking the tidier answer.
Every cut in the catalogue carries two things most listicles omit: the reason it suits a shape, tied to that shape’s styling goal, and an honest note on upkeep. Upkeep is the half people regret, and a guide that lists a blunt micro-fringe without mentioning the six-week trim is not being helpful.
Nothing here is medical, dermatological or psychological advice. If you are considering a cosmetic procedure, this site is not the input you want — see a clinician.
Money, plainly
- Nothing is paid for. No brand, salon or retailer has paid to appear on this site, and none has been given a review before publication.
- There are currently no affiliate links. If that changes — the glasses guides are the likely place — the page carrying them will say so above the recommendation, not in a footnote, and a commission will never move a frame up a list.
- The tools are free and stay free. There is no account, no email wall and no upsell on any measurement.
Images, and what is real
The hairstyle thumbnails are AI-generated illustrations. They are labelled as such on every card and in the footer. They exist to show the silhouette of a cut, not to imply that a real client walked out of a real salon looking like that. No before-and-after on this site is a real person, because there are none.
The overlay drawn on your own photo after a measurement is the opposite: it is real, it is your face, and it draws the same landmark points the verdict was computed from. A test enforces that the picture and the maths cannot drift apart, because a diagram that stops matching the calculation is a diagram that lies convincingly.
Corrections and contact
If the tool gets your face wrong, or a guide states something you know to be untrue, tell me and I will fix it or explain why I disagree. That is a real address read by the person who wrote the page:
Errors that changed an answer get noted on the page they affected, with the date — the canthal-tilt correction above is the first of them.
Publisher
Face Shape Report is published by Known by One, LLC, a Delaware limited liability company.131 Continental Dr, Suite 305, Newark, DE 19713, United States.
Founder and sole author: Taro Schenker.
The same company publishes The Face Report, which shares this site’s measurement engine and none of its content. Disclosed because a sibling site is the kind of thing you should hear from us rather than find out.