Role: Product designer
Team: Independent
Tools: Figma, Font Ninja, Miro, ColorZilla
Duration: 4 weeks
The Hook
Creating playlists should be fun and effortless, yet many Spotify users feel overwhelmed by the process. This project challenged me to explore how AI could simplify playlist creation while keeping it personal, engaging, and aligned with users’ moods and contexts.
The Context
Spotify, with over 600 million users worldwide, is the leader in music streaming, but the market is increasingly saturated, with competitors like Apple Music, Amazon Music, and Deezer constantly innovating.
Over a few months, I independently explored how AI could enhance the playlist experience, handling research, ideation, prototyping, and testing. My goal was to design a feature that could fit seamlessly into Spotify’s existing mobile experience while addressing real user frustrations.
The Challenge
The core challenge was understanding how to make playlist creation easier without sacrificing the personal touch that users value. Many Spotify users struggle to create playlists that match their mood, activity, or a specific context, often abandoning the process because it feels time-consuming or cumbersome.
Adding to this complexity, the feature had to integrate naturally into Spotify’s familiar interface, respect technical constraints, and strike a delicate balance between AI-driven automation and user customization. It wasn’t just about designing a cool feature; it was about solving a genuine problem in a way that felt effortless and emotionally resonant.
The Process
I began by conducting a competitive analysis to understand the landscape of AI features in music streaming. I analyzed Apple Music, Amazon Music, and Deezer., only Amazon Music and Deezer offered AI-powered features, albeit with notable limitations. For example, less than 5% of Deezer's premium users in the US had access to its AI features, and Amazon Music wasn't positioned as a premium platform.
Spotify, on the other hand, was already successfully generating playlists based on listening habits, making the introduction of a conversational AI feature a logical extension of its current strategy.
To validate the real need for this feature, I conducted a study with 10 Spotify users via a quick surveys, using a 1-10 scale used the "spark effect" to simplify the process and avoid user fatigue. The results highlighted a common frustration: users often felt overwhelmed when creating playlists and wanted more personalized search options. To explore this further, I interviewed five participants and found that users wanted playlists tailored to their current mood or context, but many abandoned the process when it became too long or complex.
By synthesizing this information using the JTBD framework and sorting insights on Miro, I identified a clear need: active users aged 25 to 35 want AI-generated playlists that save them time and personalize them to their preferences.
Using this information, I designed a strategic user journey that would ensure high visibility while seamlessly integrating with Spotify's existing interface. The AI functionality was integrated into the same place that traditional playlist creation process , i added a onboarding with skippable step, and existing Spotify design patterns were reused to optimize user time(creation playlist screen and playlist screen .
The journey begins with a pop-up window where user can be choose emotion to reinforce the vibe of playlist and they can prompt they idea, followed by a loading screen required for AI generation with a wording to say 'Creating your mood, please wait' and a spinner, After that ends with the user arriving at the standard playlist interface with the generated content, ensuring consistency with the overall Spotify experience.
During the ideation phase, I created a pop-up window with a robot icon and the tagline "Create your own vibe in two clicks," emphasizing the time-saving benefits. New users are guided through the onboarding process, with a progress bar and the ability to return to a previous step.
Prompt help them naturally express their mood or preferences, with a recommendation such as "I want a city pop vibe for cooking."
I initially considered adding a duration selector but removed it due to technical complexity and the difficulty of ensuring accurate playlist length. I also replaced the large text area with a single input field featuring an integrated validation button, and removed emojis from mood selection to create a more premium look. This iteration also optimized the popup’s screen space.
To test the concept, I built a high-fidelity prototype and i using screenshots of the Spotify interface and layering interactive elements on top for the home page . Tools like ColorZilla and Font Ninja ensured visual fidelity by capturing Spotify's exact colors and fonts, and Material Design 3 components were integrated to ensure consistency. This approach enabled rapid iterations and user testing, while reducing development time
The Solution
The final design made the AI playlist feature easily discoverable alongside standard playlist creation. Users could tap “Create with AI,” describe their mood or preferences, and instantly receive a generated playlist, landing directly in the familiar Spotify interface.
Prompts were clear and context-driven, encouraging users to create playlists that reflected their personality and current mood. While the feature automated playlist creation, it preserved user control, allowing for edits and refinements. This seamless integration maintained consistency with Spotify’s existing mobile experience while delivering an innovative AI-powered solution.
The Impact
Spotify eventually launched a similar AI playlist feature two months later, validating the user need I had identified. Comparing my design with Spotify’s implementation, my UX hypotheses were largely aligned
with theirs.
To explore improvements, I rewrote the header as “Ready to create your own” to create an emotional connection and updated carousel prompts to better reflect user activities.
In an A/B test with four users, I asked which interface felt more engaging. My version was preferred for clarity, engagement, and mood-based playlist creation, and all participants correctly identified Spotify’s original interface. This confirmed that thoughtful UX and design can make a feature feel more personal and emotionally resonant, even if Spotify’s version is more direct.
The Refection
This project reinforced that having an idea is only the beginning; execution, iteration, and alignment with user needs are what create value.
I strengthened skills in research, UX writing, prototyping, and product strategy, and I learned how to balance automation with customization. Looking ahead, I would explore expanding this feature to desktop platforms, testing richer AI interactions, and continuing to refine user prompts to make the experience even more engaging and intuitive.