AI skincare & wellness app

Problem

Skincare is made of small daily actions, and without them nothing happens, yet those are exactly what slips.

Hypothesis

If the routine is the app's entry point, the daily actions get ticked off instead of remembered.

Solution

Daily routine tracking: laser sessions and supplements. The app's entry point.

Problem

Skin progress was judged by eye, so changes were noticed too late or not at all.

Hypothesis

If a scan scores eight metrics and Trends plots them over time, progress becomes visible.

Solution

An AI photo scan scores 8 skin metrics; Trends shows how they move over time.

Problem

The device, the subscription and the profile lived in different places, and support got the same questions over and over.

Hypothesis

If the service layer sits in one section, people sort it out themselves.

Solution

Profile, connected device, supplement subscription: the app's service layer.

Problem

Health access is asked for at onboarding, before anyone understands why, and gets declined.

Hypothesis

If it is asked when the data becomes relevant, with what and why spelled out, consent is given knowingly.

Solution

A three-step connection: entry card → what data is used and why → granular data selection. Surfaced when it becomes relevant, not forced at onboarding.

Problem

Skin depends on sleep, stress, activity and weather, and none of that connection is visible to a person.

Hypothesis

If today's context is explained from live signals, the advice stops feeling arbitrary.

Solution

Cards on the home screen explain today's skin context from live signals: sleep, stress, activity and recovery, weather and UV.

Problem

A below-baseline score left a person alone with 'why', and sceptical about the score.

Hypothesis

If the likely reason sits next to the result, a score reads as an explanation rather than a verdict.

Solution

For below-baseline scan results an inline “contributing factors” hint explains the likely why: dark circles ← short sleep, redness ← cold, dry weather. One primary factor at a time, never for good results.

Problem

Wellbeing is wider than skin, but asking about it daily burns people out on questionnaires.

Hypothesis

If a short check-in across four areas runs every two weeks, the trend is visible without survey fatigue.

Solution

A biweekly quiz across four areas (physical, energy, emotional, social) maps answers to named states and comparison charts with AI summaries. I restructured the entire question system.

Problem

Routines, scans and insights lived as separate lists and never added up to a picture over time.

Hypothesis

If they come together in one calendar, a person sees the link between what they did and what changed.

Solution

The bigger picture: routines, scans and insights across time.

Problem

Some features were simply never found, because nothing pointed at them on first open.

Hypothesis

If the key places are highlighted with animated hints, the features get found without a manual.

Solution

First-open highlighting of key features; hint animations built with Lottie and Figma Make improve discoverability.

Problem

The app is hard to describe, because its value only lands after your own scan.

Hypothesis

If you can run a scan for someone else, the product sells itself.

Solution

Run a skin scan for someone else to show them the app: a growth loop designed into the product.