Spotify Case Study: Quick Filters Feature

In the realm of digital music streaming, Spotify has long stood as a bastion of variety and personalization, a place where every user’s taste finds a melody. Beneath this landscape's surface lay an uncharted challenge: the quest for a more intuitive and personalized music discovery experience.

“Our mission is to unlock the potential of human creativity by giving a million creative artists the opportunity to live off their art and billions of fans the opportunity to enjoy and be inspired by it.”

This is the Spotify mission statement. As concise and accurate as this mission statement is, a platform with the express goal of providing users with opportunity and inspiration fails to offer significant access to these opportunities within their app.

Identifying The Challenge: The Call to Adventure

The Initial Problem

When users listen to music, they want to discover songs and artists that adhere to their musical tastes so they can grow their music library.

In recent years, Spotify has been experimenting with different features to enhance song discoverability for their users. Despite these attempts, they seem to place their focus on the song and not the user, neglecting their user’s preferences.

User Research

I first refined this problem by conducting interviews with Spotify users to identify their pain points within the app and determine if they possessed the same discoverability grievances as I did. I wanted to identify common practices users employ to achieve their desired listening experience on Spotify.

Users utilize playlists to create song lists for specific events or situations.
Users discover new songs primarily through the Spotify recommendation algorithm.
Users had mixed opinions about the algorithm’s accuracy in assessing song cohesiveness.
Users expressed a key desire to exercise some control over recommendations.
Most users don’t venture past the Home page or Library page

These were essential insights from users that provided me with a roadmap and allowed me to pinpoint where to focus my efforts.

The People Problem

After synthesizing the user research findings, namely the trends and insights, I refined my problem:

When users are looking to play new songs, they want to quickly discover music without explicitly searching for it, so they can listen to new content that fits the criteria they are searching for, but they cannot do that because...

Users are unable to apply their preferences to discoverability effectively
Searching for specifics is time-consuming and inconvenient

Brainstorming: The Quest for a Solution

After user research, I had my map, but I still needed to figure out how to use the map and arrive at my destination. How would I identify the landmarks on the map in the real world?

Market Research

To answer this question, I first conducted market research with the goal of identifying how similar platforms to Spotify attempt to solve the problem I had identified.

I examined four popular platforms and competitors to Spotify:

Youtube Music
Pandora
Apple Music
SoundCloud

Though they are all competitors to Spotify, these platforms accomplish similar goals in very different ways. I identified that there’s limited ability for users to set explicit preferences or filters for their listening experience across all of these platforms. This finding was precisely what I was searching for and facilitated my discovery of the main opportunity for improvement among these applications: enhanced user preference input.

Allowing users to set more detailed and explicit content preferences for better-tailored recommendations would significantly enhance the user experience on a listening platform.
After realizing the success of Spotify-generated playlists, perhaps leveraging the power of artificial intelligence to achieve a level of customization not seen before in personalized playlists would give users the freedom they desire in playlist creation

The Storm Begins

I had the map, I had the landmarks, and all that was left was to figure out exactly where the treasure was buried. I collaborated with two team members to identify questions within “opportunity areas” where we could focus on specific answers for a solution.

With the opportunity areas as guides, we brainstormed solutions that would satisfy the people problem in the areas most pertinent to the problem.

We had many, many solutions. Our goal was to record as many ideas as possible and then group the solutions into solution spaces, where we voted on the solutions that best solved the people problem.

Through this brainstorming, we identified the specific area for the solution I was looking for:

How might we better implement AI in music discovery?
How might we incorporate user input into the Spotify algorithm?
How might we create a more interactive discovery experience?

Low-Fidelity Mock-ups: Building the Skeleton

Low-fidelity mock-ups were a vital pathway to fleshing out my ideas. They provided a visualization of how these features may work, which aided in determining which feature to move forward with.

Iteration: The Treasure

With a more clear vision of the final solution, I began iterating on how it would be implemented into the existing Spotify app.

The Treasure

I concluded that implementing a “quick filters” feature would best solve the people problem by permitting users to play recommendations filtered by user-defined preferences. Through this implementation, Spotify would provide users with a concise pathway to quickly play recommendations that fit their situation (event, mood, etc.) while reducing the time users spend searching through menus.

Content Requirements

To properly render my ideated solution, I developed a content hierarchy tailored specifically for the feature I aimed to implement.

This content hierarchy outlined the content required to successfully implement the feature and served as a guideline for developing medium-fidelity mock-ups.

Medium-Fidelity Mock-ups: Putting on Some Muscle

The following are three medium-fidelity flow prototypes from my iterations that best meet the goals of the Quick Filters feature. These flows aim to implement the feature at different entry points seamlessly.

My goal with Flow A was to integrate the quick filter feature into the homepage of the Spotify application. My intention was to maximize the visibility of the feature through its own widget without drawing too much attention away from other elements of the page.

With Flow B, I experimented with an entry point on the Library page through the plus button. This button is used to create a new playlist or blend, so it seemed fitting to integrate a new playlist creation feature within this menu.

Flow C brings the entirety of the quick filters feature to a widget, eliminating the need for a new menu. My reasoning was that implementing the new feature without a separate page would entice users.

I experimented with different user flows and entry points at this stage, searching for a flow that balanced the content required for the feature with a concise and intuitive user experience. My goal was to create medium-fidelity explorations using the existing Spotify user interface and app design as a base, carefully balancing the new feature with Spotify’s present layout.

Design System: The Diet

I developed a design system to ensure the Quick Filters feature’s integration into Spotify’s user interface. The system is a framework that encompasses a spectrum of components, from typography and color palettes to interactive elements and layouts. This design system was the blueprint for the feature’s design and enhanced the overall user experience with its intuitive and user-friendly design.

High-Fidelity Prototyping: The Conditioning

Based on my medium-fidelity iterations and the design kit that I generated based on analysis of the Spotify app, I fleshed out the flows into high-fidelity explorations of the feature.

Based on user feedback and analysis of the three final high-fidelity flow designs, I decided to move forward with Flow C. Although I was initially interested in confining the feature to a widget without introducing a new page to the app, there comes a point where filling the widget with too much information impacts the user’s experience negatively. Reducing the widget to an entry point displaying recently created playlists accessible without entering the new page tastefully separates content and improves usability. Reducing the number of menus that the user must access increases intuitiveness and elevates the feature's accessibility.

The Final Product

click here to view the final product

My Takeaway

Product Insights: A memorable lesson from this project was the indispensable role of user research and collaboration in narrowing down the scope of my idea and identifying solutions. Taking the time to assess the user properly provided me with new perspectives, ideas, or ways to approach a problem. Through comprehensive user research, I ensured that the final product resonated with the users’ needs.

Design Process: A key learning experience from this project is the significance of iteration in the design process. At the beginning of the design process, it became evident to me that the key to uncovering the best idea lies in initially prioritizing the quantity of ideas and explorations. Once a promising idea emerges, the focus should shift to quality, meticulously refining the concept into a viable, user-centric design. I realized that the iterative process of exploration and refinement is the cornerstone of developing a successful solution.