A skincare companion app that matches products to your skin, catches risky ingredient pairings, and keeps a routine simple enough to actually stick to.
Try the prototype ↗Skincare shelves keep growing, but most people still can't say what their skin actually needs, or which two products in their bathroom shouldn't be worn on the same day. Advice is scattered across videos, forums and packaging that assumes a dermatology degree.
The result is a familiar pattern: an impulse buy after a good review, a shelf of half-used bottles, and a routine that never quite settles into a habit. This project asks what a phone in your bathroom could do about that.
BLOOM starts with a short skin-profile quiz, then builds a simple AM/PM routine around it, checks every new product for ingredient conflicts in plain language, and tracks how skin actually changes over time.
The goal was never to replace a dermatologist. It's to make the everyday decisions, what to use, in what order, whether two products fight each other, easy enough that the routine actually survives past week one.
In early interviews, almost no one struggled to buy skincare. They struggled to use it consistently, correctly, and without accidentally cancelling one product out with another. Three problems came up again and again.
Before sketching a single screen, I talked to people who already own more skincare than they use, read through survey responses on how people track (or don't track) their routines, and reviewed the skincare and shopping apps already on their phones.
Before sketching screens, I grouped the app around the four things people actually do with their skincare: follow a routine, add a new product, check two products against each other, and see whether any of it is working.
Rather than a full ingredient encyclopedia, BLOOM narrows to five things: know the skin, add what's already on the shelf, build the routine, catch the conflicts, and show the progress.
Six short questions replace a guessed skin type, so every recommendation that follows starts from something real.
Point the camera at a label or search the catalogue, and BLOOM checks a product before it ever touches your skin.
AM and PM steps are split automatically, in the right order, and a skin score at the top shows how well the routine is actually being kept.
Add a new product and BLOOM checks it against the current routine, flagging conflicts in plain language instead of a chemical name no one recognizes.
A weekly selfie prompt and a simple skin-score trend, so a slow-moving routine still has something to show for itself.
BLOOM was designed as a personal exploration of how far a phone can go in replacing the guesswork of a skincare routine, without pretending to replace a dermatologist. Tap through a working prototype covering the routine, adding a product, the conflict check and progress.
The conflict checker was the feature testers cared about most, more than personalization itself. People didn't want another recommendation engine; they wanted reassurance that they weren't accidentally undoing their own routine. Designing the warning to explain itself in one sentence, instead of a chemistry lecture, mattered more than getting the algorithm exactly right.
The next round means testing the conflict rules against a dermatologist's review rather than public ingredient lists alone, widening the quiz for more skin tones and conditions, and running the photo-progress flow with real testers over a full routine cycle rather than a single sitting.