You are running an experiment with no measurement
A shelf of products, applied together, judged by how your face looks in whatever bathroom light you happen to be standing in. No control, no baseline, no record.
DermProof measures your skin against your own baseline for fourteen days, then tells you whether each product did what the bottle promised — and says so plainly when it didn't. No lamp to buy. No account. Nothing leaves your phone.
One email when TestFlight opens. Nothing else, and no sharing with anyone.
Hydrating
Niacinamide serum · used 26 of 28 routines
Above your personal noise floor of ±2.1. Trend holds after weighting each photo by capture quality.
How well your own data supports the claim. Not a probability that the product works.
The problem
Skin moves slowly and your memory of it is worthless three weeks later. That gap is where the whole industry lives: every bottle is "clinically proven" on somebody else, and nothing you own tells you whether it did anything to you.
A shelf of products, applied together, judged by how your face looks in whatever bathroom light you happen to be standing in. No control, no baseline, no record.
A trial average across strangers says nothing about your skin, your climate, or the four other things you put on your face that morning.
Before-and-after pairs are the oldest trick in beauty marketing, because the answer can be changed entirely by the light, the angle, or which day you chose.
How it measures
Anyone can score a selfie. The reason most skin apps can't be trusted is that the measurement drifts more than the skin does. DermProof spends its effort on the drift.
The guided camera locks exposure, white balance and focus before the shutter, and records the settings alongside the photo. A brighter bathroom no longer looks like brighter skin.
Apple Vision does face detection and region mapping on-device.
Baseline capture takes several frames seconds apart. Your skin cannot change in that time, so the spread between those frames is pure measurement noise — and it becomes the bar every later change has to clear.
A change smaller than your noise floor is reported as noise.
The verdict comes from a trend fitted across all of your readings and weighted by capture quality, not from baseline versus latest. One bad photo can no longer flip the answer.
Results carry an uncertainty interval, not a made-up percentage.
Humidity, UV and season are recorded with each reading using coarse location. When the environment explains the change as well as the product does, the verdict is withdrawn rather than awarded.
A dry week doesn't get blamed on your serum, or credited to it.
Honest results
Every app in this category is built to show you progress. An app that can only find progress isn't measuring anything. Here is the full set of answers DermProof can return.
The metric behind the claim moved in the right direction, by more than your noise floor, and the trend held across your check-ins.
Fourteen days of your own data, and the claim didn't show up. Worth knowing before you rebuy it.
The movement is inside your noise floor, or there are too few good captures. Silence beats a confident guess.
A result that looked positive but tracks the humidity or UV as closely as the product gets pulled back rather than shown to you as a win.
What the first fortnight looks like
Guided capture with locked camera settings, several frames, and your personal noise floor computed from them. Then you add the products you already own and what each one claims.
A tap to record the routine you actually did, and one 90-second photo check-in a week. Logging is what lets a change be attributed to a product rather than to luck.
Each claim gets an answer, an uncertainty interval and an evidence strength — or an honest "needs more time". This is the earliest point at which skin data can support a conclusion.
Every week sharpens the trend and tightens the interval. Swap a product and you have a measured before to compare it against, instead of a memory.
Where we sit
Hardware scanners solve the lighting problem with a lamp and a cable. Photo diary apps skip it. We solved it in software, on the camera you already carry.
| DermProof | Clip-on skin scanners | Photo diary apps | |
|---|---|---|---|
| What you buy first | Nothing. The iPhone you own is the instrument. | A device, typically a few hundred dollars, plus a subscription. | Nothing. |
| Keeping captures comparable | Locked exposure, white balance and focus, recorded per capture, plus a personal noise floor from replicate frames. | Controlled lighting built into the hardware. | Usually a face-alignment guide only. Room light still moves the result. |
| Can it tell you a product failed? | Yes. "Not supported" is a first-class result, and so is "needs more time". | Hard to, when the same product also recommends what to buy next. | Rarely. The product is progress. |
| Product recommendations | None. DermProof sells no products and takes no affiliate revenue. | Core to the pitch. | Common, often ad-funded. |
| Where your face is analysed | On your device. No account, no upload, no cloud model. | Images generally go to the vendor's cloud for analysis. | Varies. Usually an account and a server. |
| When you can start | The day you download it. | When the device ships. | Now. |
Categories, not specific vendors: hardware and pricing details change, so check any competitor's current claims yourself before comparing.
Privacy
There is no DermProof login because there is no DermProof copy of your face. Backup is your own iCloud and a file export you control.
Derm Duo
Pair anonymously with one other person and run the same product side by side. It is the fastest way to learn that a product everyone swears by does nothing for your skin — without either of you sharing a photo or a name.
Free, and staying free.
Supported for you, inside the noise floor for them. Same bottle, different skin.
Pricing
The annual trial runs 14 days on purpose: it is exactly how long the first verdict takes, so you see the thing you are paying for before you pay for it.
Free
$0
Pro
$69.99 / year
Straight answers
On its own it isn't, and that is the honest answer. What makes the readings comparable is that capture conditions are locked and recorded, that your own replicate frames define how much the measurement wobbles, and that no change smaller than that wobble is ever called a result. DermProof measures change in your own readings over time; it does not claim an absolute clinical value for your skin.
No. DermProof measures visible change in tracked metrics and tells you whether a product's claim held up in your own data. It does not diagnose, treat or screen for any condition. See a dermatologist for anything medical.
Because skin changes slowly and shorter windows mostly measure noise. Fourteen days is the minimum at which a hydration or texture trend can separate from your own variability, which is also why the annual free trial is exactly that long.
A summary of how well your own data supports a claim: which direction the metric moved, how big the move is relative to your noise floor, how many good captures back it, and how tight the trend is. It is not a probability that the product works, and it is not clinical certainty.
No, and we don't plan to. DermProof sells nothing, takes no affiliate revenue and has no opinion about what you should buy. It only reports what happened to your skin with what you already own. That is what makes a "not supported" result believable.
Yes. Photos and all pixel analysis stay on the device, using Apple Vision. There is no DermProof account. The only things that can leave are anonymous, non-image rows you opt into: a claim result for pooled stats, a Duo pairing code, and coarse weather context. One tap deletes them.
No hardware at all. Consistent conditions help — same room, same time of day is ideal — but the app locks and records the camera settings so ordinary variation is measured rather than mistaken for your skin changing.
DermProof opens on TestFlight this autumn. Leave an email and you get the invite before the App Store does.
Free to start · No account · Nothing uploaded