Finding your personalised New Year Song
Solo Project from Concept to Deployment
Duration
4 Days (2025)
My Role
Solo Project (Planning, Designing, Development)
Overview
What is New Year Song?
In Korea, there’s a fun belief that the first song you listen to on New Year’s Day sets the tone for the entire year. Inspired by this, I built this project to help you find your own "destiny song."
I took full ownership of the end to end process from planning and design to final deployment. I proactively utilized AI to transform my vision into a functional reality.
Problem Space
What is the issue?
"I want a song that matches my New Year's mood."
Every year, YouTube is full of general "New Year’s First Song" playlists. But for many users, these don't feel personal enough.
1. Search for my vibe: Users spend too much time clicking through many songs just to find the one that fits their personal New Year's wish.
2. Checking the lyrics: Users have to listen to each song and check the lyrics manually to see if the message truly connects with their wish.
Research
Analyzing SNS trends
Popular music curation posts on social media show that viral content prioritises lyric analysis and the specific message of the music over simple popularity. This indicates that users are searching for a story that fits their personal situation rather than just a pleasant melody.
User Voice
Clear patterns identified in user comments where personal experiences are frequently shared in relation to song lyrics.
Setting the Hypothesis
Recommending music matched to user wishes instead of manual searching will provide a more personal and meaningful start to the new year.
Database categorisatio
Building a curated database of 200+ songs frequently featured in popular YouTube "New Year’s First Song" playlists. Every track is pre-sorted into one of six core categories to create a structured foundation for a precise and logical matching process.
AI Personalisation Logic
Utilising AI to analyse the specific intent of a user’s wish and matching it with the most relevant song within the selected category. The system generates a personalised recommendation reason by linking the emotional context of the wish directly to the song’s lyrics.
Impact
Launched as a seasonal web service, this project recorded 298 total visits during its 2week launch period. It might not be a very high number, but more than half of the visitors finished the whole journey, showing a 54.7% task completion rate. When I looked at the user data, 'Career' was the most selected keyword. Also, reading the wishes from users showed that their New Year goals were very specific. The service successfully recommended the right songs for each person.
Total visits
Task completion rate
Top user choice
What I Learned as a Product Designer
Built and deployed the entire web service alone using AI without any development background, improving my AI skills.
Managed the project from end to end, gaining a better understanding of PM, engineering, and operations perspectives for future teamwork.
Felt a deep emotional connection by reading user wishes, growing my curiosity for products that truly connect with people instead of just making services I want to build, and inspiring me to do more of this in the future.