2026
SoundScout
Completed for HCI 530: Mobile Design @ DePaul University
End-to-end design of a social music rating and discovery app. Built around the idea that the best recommendations come from friends and not algorithms.
Role
Timeline
8 weeks (Jun 2026 - Mar 2026)
Skills
Product Thinking
User Research
Prototyping
Tools
Figma
React
Claude Code
Problem
Algorithms don't get you like your friends do
Many music fans like to keep track of their listening activity using spreadsheets, playlists, or other 3rd party tools that don't go beyond surface-level analytics or generic recs. When looking for new music, existing apps just point users to stranger's ratings or opaque algorithmic suggestions that feel impersonal. No current app surfaces suggestions from friends in a meaningful, transparent way.
Competitive landscape
Last.fm
✓
x
~
x
Musicboard
✓
✓
✓
x
RateYourMusic.com
✓
✓
~
x
AirBuds
~
x
✓
x
~ = Partial Support
process
Research
Interviews with 4 music listeners revealed that people were already cataloguing music and discovering music from trusted friends, but struggled with generic streaming recommendations and lacked an organized logging method. A competitive landscape analysis of similar platforms revealed that no single product meaningfully combined streaming history, a structured rating system, and friend-weighted discovery.
Define
Synthesized findings into user needs: catalogue what you've listened to, share ratings and reviews with friends, and discover new music primarily through friend's ratings rather than impersonal suggestions. Interview findings directly informed the app's architecture.
Design
5 main tabs each serve a specific purpose. Two full iterations were completed covering the Home, Discover, Activity, Profile, and album pages along with rating, reviewing, and search flows. Claude was used for ideation, visual design exploration, and design system help (ex: developing color and type specifications). All design decisions and designs shown are my own.
Test
Tested 4 users on finding an album suggestion, rating and reviewing it, and adding a friend to determine taste compatibility. Feedback and results were overwhelmingly positive, but some users struggled with the rating system input and locating the search/add function.
Key Screens
Home
An overview of recent activity from friends and the highest rated album in your circle that week, with recently released music from your saved artists and quick access to recently logged music for rating and reviewing.
Discover
Suggested music page led by social context and an AI-powered recommendation system. All suggestions are surfaced with distinct reasons, coming from friends that have similar ratings of shared albums and preferred genres. What your friends think is displayed before the ratings of the whole user base.
Activity
A pure reverse-chronological feed, with all logged music, ratings, and reviews from everyone you follow. An unfiltered display of the raw social data that powers discovery across the rest of the app.
Profile
Your taste fingerprint: number of logged and reviewed albums, with your selected favorite genres and albums displayed. The data displayed is the foundation of the taste compatibility feature with friends.
Recommendation Settings
Three primary adjustable factors that influence how your recommendations are generated, so you can take control of your music discovery. This the use of AI transparent and user-controlled instead of a blackbox.
Key decisions
Social-First Discovery
Purely algorithmic recommendations are opaque, while pure social feeds can get noisy. Neither matched how people valued their friend's opinions when it comes to a mutual understanding of each other's taste.
Resolution: The Discover page leads with friend context (their scores and words) before surfacing community data on individual album pages.
10-Point Rating Scale
Apps like Letterboxd or GoodReads use a 5-star rating system, where nuance and specificity is lost. A rating of 4/5 is vastly different from a 4.5/5, but more specific ratings aren't available.
Resolution: A 10-point rating scale, used by most major music publications, provides granularity in ratings that reflect exactly how you feel about a piece of music. This rating renders as a color-coded score badge.
What changed
Global Search
Two users struggled to identify the location of the search bar while tasked with adding a friend. They expected the search to be visible on the Home and Discover pages as well. The search bar was added to the Home and Discover pages, with a distinct purpose of browsing rather than adding an album or user (what the Add tab does).
Rating Modal
Users showed difficulty in using the button-based keypad system, and the small review text field indicated that they had limited space (left). I opted for dual slider and keyboard input for the rating instead and increased the size of the review space (right), leaving more space for thoughts.
Score Badges
Users found that the gold score badge (given to albums with a rating of 9.0 and above) wasn't distinct enough from other badges (left). I added a gradient in order for it to visually resemble gold (right), making it stand out as a gold-medal album should. Additionally, I included another badge for specificity and optimized them for light and dark modes.
what i learned
Clarity of Job Beats Richness of Features
The hardest part wasn't deciding what to include, but ensuring that every feature had an unambiguous purpose. The difference between the Home, Discovery, and Activity pages seemed unclear at first, until I could articulate that one is a personal catch-up, one curates suggestions, and one shows unfiltered data, respectively.
Current state
Using AI Tools to build a Functional Prototype
I have continued to use Claude Code to see how far I can take this idea based on the designs I created in Figma. Using React Native, Expo Go for previewing, and my Figma file and design system as the blueprint, I created an interactive prototype that I plan to test further once all features are implemented.
© Jack Thomson 2026