Who hasn’t looked in their closet before, whether there are 10 items on their hangers or it’s bursting at the seams, and thought “I have nothing to wear”? Especially as someone interested in saving money and shopping secondhand, endless scrolling and ever-evolving micro-trends can make the task of searching for your perfect wardrobe addition feel like a fashionable David facing down the Goliath of overconsumption.
Phia, an AI-powered e-commerce solution, has started to tackle this problem for the 1.3M+ shoppers currently using their site. Through a built-in browsing extension and standalone app, users are directed to lower-priced options or comparable versions to items they seek out. However, there is still a gap - Phia is great for answering ‘Should I buy this’ version of xyz item, but it doesn’t quite help users figure out what to buy in the first place.
Most shoppers — myself and my friends included — don't know what colors flatter them, what's missing from their wardrobe, or what would actually complete their closet. This leads to aimless browsing, decision fatigue, abandoned carts, and another season of frustration.
First of all, I would be remiss to omit the fact that I love the Phia app and what they are trying to accomplish. That being said, I decided to go page-by-page and dissect the user experience. The main issues I zeroed in on were: I felt overwhelmed by the number of places I was being driven to look, and I didn’t quite feel like the app “got” me or what I needed. There are a ton of great items and savings out there, but that doesn’t help me — or the friends I spoke to — with figuring out how that maps to what is missing from their closet. What option(s) could truly round out my wardrobe and be the glue that brings my closet together?
The Phia Profile tab should focus on how to personalize the experience.
The Phia Home tab feels duplicative at times and doesn’t feel like it’s driving me to make decisions.
Initial concept - consider a Profile that that is truly that - a profile of my needs and how to fulfill those (e.g., what colors bring me to life versus wash me out, what are all my preferences, what is my closet missing?)
This exercise helped me identify what seems to be working well on the Phia app experience, and what could benefit from some enhancements. Following this initial ideation flow, I then leveraged Claude as a thought partner throughout this sprint — pressure-testing ideas, mapping user flows, refining copy, and generating wireframe references that I then translated into Figma mockups. Iteration examples with Claude are below:
One of the first mockup iterations with Claude based on my concept sketches.
An example of potential “For You” page content based on the color analyzer feature.
Final Claude version of the “For You” page - I ended up combining components from this version with a previous iteration in the Figma wireframes.
Phia already has powerful AI for price intelligence — comparing costs across 40,000+ sites in real time. The opportunity here is to extend that same AI to personalized style intelligence: helping users understand what suits them and what's missing from their own wardrobe.
The question shifts from 'Should I buy this?' to 'How do I decide what to buy next?'
This reframe matters for Phia too. Users who know what they need convert at higher rates than users who are just browsing. More confident shoppers make more purchases and generate more revenue.
<aside> 💡
As a secondhand shopper, I want to know which colors and styles suit me and what's missing from my closet, so that I can shop with confidence instead of browsing endlessly and second-guessing every purchase, or never making a purchase at all.
</aside>
StyleDNA is designed to serve both ends of the Phia user spectrum — the explorer who needs to discover their style and the intentional shopper who needs to optimize it.
<aside> 🧍♀️
Annie, 20
<aside> 🧍♀️
Dee, 34