How to Choose an AI Skin Analysis App
Eight questions that separate a measurement from a product finder. Two of the answers below are ones where Skinzy is not your best option, and they are marked as such.
Work Out What Kind of Scanner You Are Using
Nearly every skin scanner falls into one of three groups, and the group tells you more about its recommendations than its feature list does. None of these is disreputable — but knowing which one you are holding tells you what its output is optimised for.
Run by a skincare company, free, usually inside their site or app.
The scan ends at that brand's catalogue. It is built as a product finder, and it works well as one — but a concern it has nothing to sell for is a concern it has little reason to surface.
Run by a shop that carries many brands, free, inside their app.
Wider catalogue than a single brand, so the recommendations can be more varied. The output is still a basket: it resolves to products the retailer stocks, and stocking is a commercial decision.
Sold to brands and retailers, who put their own name on it. You meet it as someone else's scanner.
The customer is the brand licensing it, not you. It is typically the most technically capable option in the category, and the one whose results you are least able to trace to a published method.
The Eight Questions
Does it measure your skin, or ask you about it?
A questionnaire returns your own self-assessment in a nicer format. A measurement can disagree with you, which is the only way it can tell you something you did not already know.
Measures. Five regions are cut from facial landmarks and scored independently across five traits, from the photo.
Does it tell you when it is unsure?
Every photo-based measurement has conditions it handles badly. A tool that returns the same confident-looking number for a bright window shot and a grainy midnight selfie is hiding the difference from you.
Yes. Every scan carries a confidence value between 0.40 and 0.95, and a low enough one drops the plan back to maintenance rather than introducing an active.
Does it publish what it cannot do?
Capability limits are the first thing marketing removes. A published limit is a small piece of evidence that the rest of the claims were written by someone willing to be held to them.
Yes. It measures oiliness, dryness, acne, pigmentation and sensitivity, and states plainly that it does not measure aging, wrinkles, fine lines or firmness. It also names pigmentation as its own weakest axis.
Who profits from the recommendation?
This is the question that decides what the other answers are worth. A scan attached to a catalogue has a reason to find concerns that catalogue addresses.
Nobody. Skinzy sells no skincare, takes no sponsorship and carries no affiliate tag. It names ingredients rather than products, so the recommendation cannot resolve to a SKU anyone is paid to move.
What happens to your photo?
You are uploading your face to a health tool. The answer should be one sentence and easy to find.
It is scored and discarded inside the request. Only the resulting scores and recommendations are stored, and the web app loads no analytics, cookies or third-party scripts at all.
Does it recur?
A skincare plan runs months. Access that expires mid-plan is a different product from access that covers it.
Unlimited scans, all five zone scores and a routine preview are free permanently. Plus is sold once — yearly or lifetime — and never auto-renews.
Can it see everything you care about?
Scope is not the same as quality. The narrower tool may be better at what it does and still be the wrong tool for you.
Only if your concerns are the five it measures. Anything aging-related — wrinkles, fine lines, firmness, elasticity — is outside its scope entirely, and a scanner built for those will serve you better.
How much has it been validated?
Worth asking of everyone, and rarely answered by anyone. Be suspicious of accuracy claims in this category generally; almost none are backed by a published benchmark.
Partially, and we would rather say so. Score stability under changing light is measured and published. Accuracy against clinical ground truth is not, the calibration corpus is small, and nothing here has been through peer review. No scanner in this category should be treated as a diagnosis.
What We Claim, and What Backs It
Anyone can say they are trustworthy. Below is every claim Skinzy makes about itself, each one next to the specific thing in the product that makes it true — so you can check rather than take it on faith. There is deliberately no “most accurate” claim in this list: no scanner in this category publishes a benchmark against its competitors, this one included, so it is not a claim we can stand behind.
Five zones are rasterised from MediaPipe FaceMesh landmarks and scored independently across five traits.
Every scan ships a confidence value between 0.40 and 0.95, derived from image quality, blur, cross-zone variance and how many axes could actually be measured.
Ranking ties resolve on a fixed, explicit key rather than on whatever order the data happened to arrive in, so two separate runs of the same scan return byte-identical recommendations.
The photo is decoded, scored and discarded inside the request. Only scores and recommendations are persisted.
Recommendations name ingredients, not SKUs we are paid to move. No retailer link on the site carries an affiliate tag.
Nothing loads from a third-party origin, so there is nothing to opt out of.
Skinzy scores oiliness, dryness, acne, pigmentation and sensitivity. It does not measure aging, wrinkles, fine lines or firmness, and says so in its own machine-readable index.
Pigmentation is the least reliable of the five, and the Home FAQ says so rather than burying it.
Questions People Ask
It depends on what you need measured and what you are willing to trade. A brand-owned scanner is free and polished but ends at that brand's catalogue. A white-label scanner licensed to a retailer is usually the most technically capable and the hardest to trace to a published method. Skinzy measures five traits across five facial zones, sells no skincare, and publishes what it cannot do — but it does not measure anything aging-related, so it is the wrong tool if wrinkles or firmness are your concern. The eight questions on this page are more useful than any ranking.
Free is normal in this category, so it is not the useful question. The useful question is what the free version is for. Most free scanners are product finders paid for by the catalogue they recommend from, which is a real service but a different one from a measurement. Ask who profits from the recommendation before you ask what it costs.
Almost no one in this category publishes a benchmark, so treat every accuracy claim — including any you read here — with suspicion unless it comes with a method you can check. Skinzy publishes measured score stability under changing light rather than an accuracy figure, because stability is what it can actually demonstrate. None of these tools is a diagnosis, and none can assess a mole or a lesion.
No. It sells no skincare, takes no sponsorship and carries no affiliate tag, and it recommends ingredients rather than specific products — so a recommendation cannot resolve to a SKU anyone is paid to move.
It cannot diagnose anything. It cannot assess a mole, a rash or a lesion, and no result from one should delay seeing a clinician about a changing spot. Skinzy specifically does not measure aging, wrinkles, fine lines or firmness at all.
Check Our Answers
Every claim above is documented somewhere you can read rather than taken on trust.