Most people choose skincare the same way: a recommendation from a friend, a product that shows up in a feed, or an ingredient list they only half understand. The routine that results is built around an average face rather than yours. AI skin analysis is one of the newer attempts to change that, using a photo taken on your phone to describe what your skin looks like today and to suggest products accordingly. It is worth understanding what these scans actually do, because the technology is often described in vague or inflated terms.
What a skin scan looks at
A selfie scan reads visible characteristics of the skin in a photograph. The most common signals are hydration, which appears as how smooth and plump the skin looks; texture, meaning unevenness, visible pores and fine roughness; redness, which shows up as areas of increased color; sensitivity indicators, such as patterns associated with irritation; and tone, which covers evenness and the presence of spots or dullness. The scan behind 2250 reads more than ten of these signals in about 60 seconds, from any phone, with no appointment.
It is important to be clear about the word “reads.” The scan does not look beneath the surface, and it does not measure the skin the way clinical instruments do. It interprets what is visible in the image and compares those patterns with what it has learned from a large set of examples. That is a real and useful capability, and it is also a bounded one.
How a photo becomes a recommendation
The process has three stages. First, the image is checked for quality, since lighting, angle and focus all affect what can be read. Second, a model trained to recognize skin characteristics scores the photo across the signals described above. Third, those scores are matched to a set of formulation options. In the 2250 system, the result is one base product, a cleanser, serum or moisturizer, combined with two pods that each carry a specific active ingredient. With three bases and fifteen pods, that produces more than a thousand possible combinations, so two people with different scan results are unlikely to receive the same routine.
The advantage of this approach is consistency. A trained model applies the same criteria every time, and it can be repeated whenever the skin changes. You can read more about the reasoning behind this on our page about the science behind 2250.
How it compares with a quiz and a consultation
A questionnaire depends on what you know and what you notice about your own skin. People often describe their skin type from habit, and skin type is not fixed: it changes with seasons, stress, travel and age. A quiz also cannot see anything, so the recommendation reflects your description rather than your skin.
An in-person consultation with an esthetician or dermatologist brings professional judgment, physical examination and the ability to ask follow-up questions. It is the right choice for persistent concerns, and a scan does not replace it. What a consultation does not offer is convenience and repetition. It is difficult to book one every time the weather changes.
A selfie scan sits between the two. It is more objective than a quiz because it starts from an image, and it is far easier to repeat than a consultation. It is less thorough than a professional exam, and that trade-off should be stated plainly.
What a scan cannot do
A scan is not a medical diagnosis. It cannot identify skin conditions, and it should not be used to decide whether a mole, rash or persistent irritation needs attention. The FAQ on our site says this directly: the output is a cosmetic recommendation tailored to the skin’s current needs. If you have concerns beyond routine care, a dermatologist is the right person to see.
A scan is also a snapshot. Skin that looks calm on a humid afternoon can look different after a week of dry indoor air, which is why the ability to rescan matters as much as the first result. The goal is a routine that follows your skin as it changes, and treating a change in results as normal rather than as a failure of the system.
Finally, results depend on the photo. Soft, even daylight facing a window gives the model the clearest information. Strong overhead light, heavy makeup, filters and shadows all reduce accuracy. Taking the picture without makeup, with the camera at eye level and the hair pulled back, is a simple way to improve what the scan can read.
Why this matters for how people shop
The larger point is not the technology itself but what it changes about buying skincare. When a routine starts from your skin rather than from a marketing claim, it becomes easier to judge whether something is working, because you know what it was meant to address. It also reduces the habit of adding products one at a time in the hope that one of them solves everything.
If you want to see how the process works in practice, you can scan your skin in about a minute and see how the routine is assembled from the result.
