Spotify
Designing an Emotional AI Plug-In for the World's Leading Music Streaming App
Created Spotify Spectrum, an AI plug-in designed to help users select music based on their mood, addressing emotional health needs by offering a more personalized, comforting experience
Project Summary
Research that led to Spotify Spectrum, our AI plug-in designed to help users discover the perfect song for any moment.
Deliverables
See the full project breakdown below. ⬇️
My Role
Researcher and Ad Voice
Ella Gormanlove, Art Director
Smera Dhal, Art Director
My Team
My Research Tools
Google Scholar, Midjourney and Notebook LM
Spotify is the most popular music streaming app in the world. With over 210 million subscribers, the brand welcomed 15 million subscribers in the first quarter of 2020 - with the COVID-19 pandemic being a key contributor. The Q1 findings released by Spotify Technology S.A. stated that two in five consumers they surveyed in the US said they were listening to music to manage stress more than they typically do.
The Context
Spotify Users by the Millions
How can Spotify spread positivity by using technology to bring listeners closer to the creators and communities they love?
The Ask
Insight One ➡️ Music Triggers Distinct Neural and Emotional Responses
Most Spotify listeners view music as a way to boost their mood, yet music also evokes a wide range of emotions. Elements like BPM and tempo can activate different neural responses. For example, rap music tends to enhance memory engagement, while rock music elicits neural reactions linked to emotional intensity.
How Spotify Listeners Use Music
Insight Two ➡️ Music’s Emotional Impact Varies by Age
The ways that music listeners use music a emotional health aid varies based on generational differences.
Gen Z (18-27) and younger Millennials (up to 40) mostly use music to ease loneliness rather than for emotional management. Older Millennials (40+) and Gen X (50+) mainly use music to manage their emotions rather than to soothe loneliness.
Insight Three ➡️ Spotify’s AI Model is Presumptive
Spotify’s current artificial intelligence model, which involves using MIR (also known as music information retrieval), uses hard contextual information such as location, time, time of data collection and acoustic characteristics of songs to give users music suggestions how it assumes the user is feeling.
-
Sonic Science study conducted by Spotify & Neuro-Insight, surveying 624 participants over a decade. Sonic Science Video Link
Spotify Technology S.A. (2020, April 29). Spotify Technology S.A. announces financial results for first quarter 2020. Business Wire.
Assuncao, W. G., Piccolo, L. S., G., & Zaina Luciana, A. M. (2022). Considering emotions and contextual factors in music recommendation: A systematic literature review. Multimedia Tools and Applications, 81(6), 8367-8407. doi:https://doi.org/10.1007/s11042-022-12110-z
Hennessy, S., Sachs, M., Kaplan, J., & Habibi, A. (2021). Music and mood regulation during the early stages of the COVID-19 pandemic. PLoS One, 16(10), e0258027.
Herrero, E. M., Singer, N., Ferreri, L., McPhee, M., Zatorre, R., & Ripolles, P. (2020). Rock’n’roll but not sex or drugs: music is negatively correlated to depressive symptoms during the COVID-19 pandemic via reward-related mechanisms.
Krogh, M. (2023). Rampant Abstraction as a Strategy of Singularization: Genre on Spotify. Cultural Sociology, 0(0). https://doi-org.proxy.library.vcu.edu/10.1177/17499755231172828
Till, C. (2023). Spotify as a technology for integrating health, exercise and wellness practices into financialised capitalism. Big Data & Society, 10(2). https://doi.org/10.1177/20539517231210278
Ziv, N., & Hollander-Shabtai, R. (2022). Music and COVID-19: Changes in uses and emotional reaction to music under stay-at-home restrictions. Psychology of Music, 50(2), 475-491.
The research uncovered key issues for us to address related to the current user experience for Spotify listeners.
UX & Product Considerations
Spotify’s AI should give users more control over AI-generated music suggestions.
Spotify’s AI should be improved to reflect a wider range of emotions.
Our final product should positivity across all demographics.
Spotify Spectrum is an AI-powered tool that gives users greater control over song recommendations during emotionally intense moments.
This feature redefines Spotify’s approach to music suggestions by recognizing that musical preferences exist on a spectrum and ensuring AI-generated recommendations better reflect a wider range of emotions. By enhancing user control and emotional accuracy, Spotify Spectrum creates a more personalized and uplifting listening experience for all demographics.
Solution
Refer to the slide show, ad showcase video and key features section below for more details. ⬇️
Spotify Spectrum Intro Screen
Spotify Spectrum Instructions
Spotify Spectrum Send-Off
Spotify Spectrum - Invisible Grid
Spotify Spectrum
Spotify Spectrum Product Ad Showcase Video
Video by Smera Dhal
Spotify Spectrum: Core Features – Axes, Icons, and the Invisible Grid
Based on research on emotions and music preferences, Spotify Spectrum lets users navigate an interactive grid, instantly connecting each point to a song.
Artwork designed by Ella Gormanlove
The Turtle icon on the X-Axis represents slower songs.
Users can move their finger toward it to find music with a slower tempo.
The Sun icon on the Y-Axis represents happier songs.
Users can move their finger toward it to find upbeat music.
The Cloud icon on the Y-Axis represents sadder songs.
Users can move their finger toward it to find music with a sad or angry tone.
The Hare icon on the X-Axis represents faster songs.
Users can move their finger toward it to find music with a quicker tempo.
When users are experiencing strong emotions, especially distress, they don’t want to waste time choosing the right song to improve their mood. Spotify Spectrum leverages the platform’s existing AI model to suggest music while empowering users to maintain control over their music preferences in real time.
The project received an average peer score of 4.32 out of 5 on strategy alone, and a 4.35 out of 5 overall.
Impact 💥
I'm proud of the secondary research I conducted for this project.
Utilizing AI tools like Google's Notebook LM to analyze research papers helped uncover valuable consumer insights that were pivotal in shaping the final solution.
Project Reflection
Check out my other projects below. ⬇️
