Close
Overview
Allure is an AI-powered wardrobe app for fashion enthusiasts and busy professionals. I designed the full product — from research to UI. The goal: help people wear more of what they already own.
Problem
The average wardrobe is 60% unworn. People don't lack clothes — they lack visibility into what they own, what pairs with what and what's actually missing. Every morning starts from scratch. Every impulse purchase adds to a wardrobe that already has too much.
Key insight
Affinity mapping across three user archetypes surfaced two unexpected findings.
Users weren't quitting digital wardrobes because logging clothes was tedious — they quit because their photos looked cheap against a phone-camera background. It drove automatic background removal and light correction on every uploaded item.
The biggest morning pain wasn't choosing an outfit — it was discovering the chosen item was dirty or wrinkled at the last minute. It drove Care Status Sync: every item carries a live status and the AI only suggests outfits from what's actually available.
People spend 15–20 minutes choosing an outfit every morning.
Outfit Builder prepares looks in advance.
1
Users discover clothes are dirty or wrinkled too late.
Care Status Sync only recommends ready-to-wear items.
2
People wear only 10–20% of their wardrobe.
3
Users often buy clothes they already own.
4
Research
Competitive analysis
Three personas
Affinity mapping
Hypothesis validation
Empathy maps
Define
How Might We
Job Stories
As-Is / To-Be CJM
User scenarios
POV statements
Design & UI
Information architecture
User Flows
Empty states
Design system
High-fidelity UI
Reflection