user research spotify

We stand with Ukraine and our team members from Ukraine. Here are ways you can help

A Community Of Over 800,000

A UX/UI Case Study on Spotify

• September 29, 2020

For this case study, I will be focusing on the modern music streaming platform — Spotify.

This case study was completed as an independent project during a four day sprint while I was a student at Ironhack. I was challenged to create a new feature to an already existing and highly adopted app, so I chose to work on one of my favorite apps for streaming music, Spotify!

Music streaming has changed  so much  over the past few years in response to the digital revolution. What seems to feel like a lifetime ago, people were using MP3, Napster, iPods and even resorting to not so legal ways to download and listen to music. And before that, there were CDs, cassettes and vinyls. Although the idea of having a record player at home and lining up outside the store to buy the latest music drops seems charming and entertaining, I couldn’t imagine not having music at finger tips — available whenever and wherever I want it! The evolution of music streaming in itself is a topic I could talk about for hours, but for this case study, I will be focusing on the modern music streaming platform —  Spotify .

This week’s design challenge at  Ironhack  was to  incorporate a new feature  for an existing and highly adopted app. I chose Spotify because it completely changed the streaming industry in 2008 by setting the trend for subscription based music streaming services. It has also been my music streaming platform of choice for years now.

During this  four day sprint,  we focused more on  visual design  and  Atomic Design Principals  and developed our ideas to hi-fidelity prototypes. I was both excited and nervous about this weeks project since it was my first solo design project, but I am  big  on challenging myself, so I felt that it was the perfect time to tackle my own project.

So, let’s get started!

Just to give you a bit of background, Spotify is a music streaming platform — the biggest in the world by number of subscriptions. It was founded in Sweden by Daniel Ek and Martin Lorentzon, and eventually launched in 2008. As of March 2020, they have  248 million monthly active users , compared to Apple Music that has about 68 million users. Spotify uses a freemium model that offers users a free tier that includes ads, and a premium subscription for a monthly payment.

“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.” — Spotify Mission Statement

Lean UX Canvas

Using the Lean UX Canvas, I outlined the current state of Spotify and wrote down my assumptions that I began within boxes 1–4. Since this is a living document, I revisited it to make updates as I went through the Design Thinking Process.

Competitive Feature Analysis

I started off my research with an analysis of Spotify’s direct competitors — Apple Music, Youtube, Soundcloud, and Google Play. I compared their features in the snapshot below.

Competitive Positioning Chart

I mapped out the competitors’ positions in relation to the axes to better visualize areas of opportunity and the competition. I decided to compare them based on the level of social interactions that they facilitate and the level of personalization that each platform allows for their users.

Spotify’s app can be personalized for premium users but is more so for private use. The “Blue Ocean” is an opening in the current market without many competitors which symbolizes an area of opportunity that I believe Spotify can enter with an additional feature. Although I had originally wanted Spotify to move towards a social adaptation, I decided to pivot later on in the process to create a more advanced personalization feature after getting some user insight.

User Research

I started off my user research with a survey of 12 questions. Surveys have been a really valuable tool for gathering a large amount of quantitative data, so I took some time to craft my questions to reflect the information I really needed at the beginning of my process.

I posted my survey in a Slack channel and a few Reddit pages related to music and received about  65 responses in 24 hours.

Here are some important statistics relevant to my project:

• About 60% of respondents said they use Spotify
• 90% said they listen to music for both pleasure and when doing other things (driving, working)
• Only about 1/3 of people like the personalized playlists that their platform creates for them

The last bullet point validated the need for a better-personalized playlist feature , so I decided to run with it and dive into some more user research to see where I can add value!

I conducted 5 interviews with people who use Spotify and/or Apple Music. I got A LOT of valuable insights from my interviews. One of the most surprising pieces of information that I found was the fluid nature of my interviewee’s streaming service choices depending on their needs. Their preferences constantly shifted based on the music experience they were looking for, attributing each platform its own function and purpose.

For example, users brought up the variety and diversity of music that Youtube and Soundcloud offered. When asked about finding new music, 4/5 people said they find more new music either through Youtube or Soundcloud.

“If I was able to find new music easily on Spotify, I wouldn’t be using Soundcloud or Youtube.”

However, I had to keep in mind that these platforms have an advantage for music discovery because some music is exclusively on Youtube or Soundcloud.

I converted all of my recorded interviews and survey results into digital sticky notes using  Miro . This week I discovered a new tool in Miro that made data synthesis so much faster! In the sticky notes feature on the sidebar, there’s a “ Bulk Mode ” button that allows you to type several sticky notes at once instead of doing them one by one. This cut my synthesis time down by so much! (Credit to our TA, Kathryn)

Affinity Map

After dropping all of my data onto sticky notes, I organized them onto my Affinity Map with themes and color-coded them by streaming service . I found that several of the insights were very much interrelated with one another, so connections between ideas are indicated with arrows .

The most prominent themes are:

• Finding new music

Many of the smaller headers are sub-themes that branch off of the two bigger themes.

Some key findings:

• Most of the data points were related to finding new music and playlists
• People find a lot of new music on Youtube
• Playlists are defined by their “vibe”
• The word “vibe” came up a lot in the interviews — this will be an important term that I will revisit later in my case study
• Songs on a playlist should have a similar sound and consistent genre variation
• People value finding new music/artists that are aligned with their preferences

Customer Profile

Stratagyzer’s Value Proposition Canvas  is another great tool for identifying customers’ jobs (things they need to get done), pains, and gains. This tool helps us better understand the customer we are creating value for. The customer profile connects directly to the second part of the canvas — the value map, which describes the elements of the value proposition.

Customer jobs can be  functional, social and emotional . Most of the customer jobs above are either emotional or social, which indicates that several gains and pains also fall into that category.

Task Analysis

The task analysis is a visual used to map out the path that users take to create a playlist. Although there are several different ways to create a playlist on Spotify, I used my interview data to choose the steps that users would usually take.

User Journey Map

A user journey map is useful for getting a more complete view of the user’s experience for a specific task. The goal is to find the significant low points or problems that the user is facing. It’s also a great time to highlight the thoughts and emotional aspects of the user’s experience.

The user’s journey is centered around creating a new playlist for an upcoming activity (like a road trip). First, the user may think about a future activity, which will motivate them to create a playlist to make it more enjoyable. The user experiences low points when:

• Thinking about suitable songs to start the playlist
• Looking through their library to find songs to add to the playlist
• Listening to the songs recommended by the Spotify Algorithm
• Trying to find similar songs to the ones they already have on their playlist

After identifying these four pain points, I quickly wrote down some opportunities and ideas to use during ideation.

I synthesized the pain points into a  problem statement.

Problem Statement

Then, I converted the problem statement into four  “How Might We”  statements:

• HMW make sorting through the personal library easier and more specific to users preferences
• HMW make newly recommended songs align more with users taste and requirements for the  vibe
• HMW help users curate songs with a similar sound to make creating a playlist easier

This is just another way to frame the problem by turning them into questions to better prepare myself for ideation!

Ideation is one of my favorite parts of the process!

Collaborating with my classmates  Sebastian Benitez ,  Christina Rice  and  Dave Ostergren  for a brainstorming session was awesome! We came up with about  80+ ideas in 15 minutes  using the  time-boxing  method to really push us forward.

Here’s what our brainstorming session looked like:

 After brainstorming some ideas for each HMW statement, I reshaped the best ideas into stars and diamonds then made  three umbrella headers  that the ideas are categorized into —  A sorting/filter feature, a “vibe” feature and a playlist creating helper.

Organizing all of the the contents and ideas from the brainstorming session was a vital step before moving on to prioritizing the ideas since there were so many to consider and analyze. I really enjoyed this part of the process because it allowed me to arrange the detailed ideas under three overarching ideas and set myself up for prioritization.

MoSCow Method

I further categorized the ideas from brainstorming into the MoSCoW Method chart pictured below to  prioritize  the features into Must Haves, Should Haves, Could, Haves, and Won’t Haves.

The must haves are the  mandatory features  that   are key to solving the users problems. These include the  three umbrella solutions and their corresponding ideas.

After prioritizing the features, I was able to define the products necessary to make my idea work, in addition to the gain creators and pain relievers on my Value Proposition Canvas.

The products and services on the left of the canvas are color coded to correspond with the gain creators and pain relievers on the right.

Just to recap, the pain points are :

Although the sprint was short and my workload for the week was high, I still wanted to incorporate a solution that solved both problems related to creating a playlist AND finding new music.

My Minimum Viable Product includes a solution to address both pain points (creating a playlist and finding new music).

Job-To-Be-Done

When Spotify users want to make a playlist for a future activity (I used an upcoming road trip for this scenario), they will “hire” Spotify to create a desired “vibe,” curate songs in their library (already saved songs), and be exposed to similar songs — the perfect opportunity to suggest new songs for the playlist based on user specifications like specific genres, artists and tags .

At the end of the day, the goal is to make whatever experience that the user is creating the playlist for to be more enjoyable. So my job is to help them fulfill their vision so that they are excited about hearing it in the future, and hopefully listening to it several more times.

Next, I created the user flow based on this job-to-be-done to create a visual representation of the path that the user will take to create a playlist with the new feature. The user flow also includes instances of human interaction (taps) and places where onboarding is necessary.

 PROTOTYPING

I created a low-fi prototype for the process of creating a playlist, incorporating the new filter feature.

Low-Fidelity Prototype

After conducting usability testing using  Maze , the heat maps indicated that many miss-clicks occurred where there were no clear guidelines for each step and also when people followed their natural tendencies. Seeing where people tapped enabled me to  design the feature so that it would fit into their mental models  instead of trying to go against it.

For example, many people clicked  “liked songs”  on this screen because that’s what they are used to doing when they create playlists. So I used this insight to make the next iteration more intuitive by guiding the user through the liked songs page to pick a few songs to start the playlist.

Another indication that people go to their “liked songs” or “library” to create a playlist using songs they already have saved.

Mid-Fidelity Prototype

For my mid-fi, I used the atomic design principles to build out my “atoms” and “molecules” that would later become icons, buttons, and footers that I used repeatedly throughout the entire design.

 After testing my mid-fi, I added a few more changes to the layout of the screens and updated the prototype with colors and pictures for my hi-fi.

Hi-Fidelity Prototype

Here is a high level overview of my hi-fi prototype. I tried my best to make it look as close as possible to Spotify’s interface!

Here’s a step-by-step micro-level view:

The process begins by going to the library and pressing “create playlist”and giving it a name. Let’s say we are going on a road trip!

After pressing “add songs” users would go directly to the liked songs page and would be prompted to choose as many songs as they would like to start the playlist as a basis for the “vibe.” Once a few songs are selected, users go to “Find your vibe” which is where they can specify preferences for suggestions.

The tag can be defined as a label or set or labels attached to each song. Songs can have several tags that describe their “vibe” or mood. Tags are reported by users themselves, and the most prominent tag is displayed on these screens. However, if the user wants to find songs with the same tag, they can click each tag to be taken to the “tag dashboard” where songs with the same tag are listed. — This is yet another way that users can add songs of a certain “vibe” and find new music!

Then, users go through the filter feature of drop down menus to select their specific criteria for suggestions. Here, the user can be as  specific  or as  vague  as they would like, to accommodate very specific preferences and those who are more open to a variety of suggestions.

Spotify will first recommend songs within the library of saved music and users will have the option to add all or a add a few to their playlist.

Then, there will be an option to add *new* songs that are similar and also fall under the user’s preferences that were designated earlier.

The user can also share their playlist with a friend for collaboration!

Here’s a link  to my presentation if you would like to see a live walk through of the hi-fi!  (Go to slide 30)

Success and Failure Metrics

We will know if the new feature is a success through these metrics:

• Less time creating playlist
• Higher Satisfaction
• More People Use Feature
• Save newly recommended music

And we will know if the feature is a failure through these metrics:

• More time creating playlist
• Lower Satisfaction
• Users don’t Use Feature
• Disproval of recommended music

Knowledge Gaps

If I had more time, I would’ve loved to do some in-person usability testing with some of my friends to get some more insights on the way they use Spotify. I would’ve also loved to dive deeper with a second round of interviews focusing on the other music streaming platforms that they mentioned, like Youtube, to see which features they find most valuable and exactly how they use them. I believe that there are features that Spotify could potentially adopt from the other music streaming platforms that would give users additional value.

• Finish atomic design inventory
• Test the hi-fi prototype and make another iteration
• Develop the tag feature for multiple functions

I am not the user even if I use the app! Although I am a loyal Spotify user , I had to keep reminding myself not to accept my assumptions as facts. It’s a universal UX principal to always keep the users at the center of the design.  

Adopt the users’ language and make it a part of the product! Like I mentioned earlier, the word “ vibe ” kept coming up in my interviews, so I knew that this is something that may be universally understood across the music listening community. I incorporated it into my design to further relate to the users and use a language that they can appreciate.

Overall, this project was extremely rewarding! Spotify is an already amazing music streaming platform with features that go above and beyond. So the challenge to create an additional feature that adds value was definitely challenging. I enjoyed going every step of the Design Thinking Process and being able to go all the way through a hi-fi prototype!

Let me know if you have any thoughts or comments! Feel free to connect with me on  LinkedIn  as well. And if you’re in the mood to listen to some of my *very much incomplete* playlists, here’s my  Spotify .

• Consumer products , Content and Copy , Design , Interface and Navigation Design , Product Releases and Redesigns , Prototypes , Visual Design

Delawit Assefa is a UX/UI Designer in the Washington D.C. Area, a recent first-generation college graduate looking for new opportunities. She is passionate about user-centered design, problem-solving, and storytelling through different mediums and is looking forward to creating meaningful experiences and bridging the gap between technology and human experiences through design in her career.

Related Articles

The next big ai-ux trend—it’s not conversational ui.

• Artificial Intelligence , Design , New and Emerging Technologies , UX

Discover the future of user interfaces with aiOS, an AI-powered operating system that promises seamless, intuitive experiences by integrating dynamic interfaces, interoperable apps, and context-aware functionality. Could this be the next big thing in tech?

• The article explores the concept of an AI-powered operating system (aiOS), emphasizing dynamic interfaces, interoperable apps, context-aware functionality, and the idea that all interactions can serve as inputs and outputs.
• It envisions a future where AI simplifies user experiences by seamlessly integrating apps and data, making interactions more intuitive and efficient.
• The article suggests that aiOS could revolutionize how we interact with technology, bringing a more cohesive and intelligent user experience.

Share this link

• August 8, 2024

Tell us about you. Enroll in the course.

First name *

Last name *

Role/title *

Company email address *

Linkedin profile *

Interest in this course: What do you hope to get out of it

If applicable, what AI projects are you working on? describe the initiatives and their objectives and your role in the initiatives

If applicable, what challenges are you faced with in these initiatives?

Describe a future state: If successful, this course will have helped you to ___________

Anything else you'd like us to know about your work or your interest in this course?

Please leave this field empty.

Join 740,000 designers, thinkers, and doers.

This website uses cookies to ensure you get the best experience on our website. check our privacy policy and.

How We Improved Data Discovery for Data Scientists at Spotify

At Spotify, we believe strongly in data-informed decision making. Whether we’re considering a big shift in our product strategy or we’re making a relatively quick decision about which track to add to one of our editorially-programmed playlists, data provides a foundation for sound decision making. An insight is a conclusion drawn from data that can help influence decisions and drive change. To enable Spotifiers to make faster, smarter decisions, we’ve developed a suite of internal products to accelerate the production and consumption of insights. One of these products is Lexikon, a library of data and insights that help employees find and understand the data and knowledge generated by members of our insights community.

We’ve learned a lot since we first launched this product. In this blog post, we want to share the story of how we iterated on Lexikon to better support data discovery.

Improving the data discovery experience

Diagnosing the problem.

In 2016, as we  started migrating to the Google Cloud Platform , we saw an explosion of dataset creation in BigQuery. At this time, we also drastically increased our hiring of insights specialists (data scientists, analysts, user researchers, etc.) at Spotify, resulting in more research and insights being produced across the company. However, research would often only have a localized impact in certain parts of the business, going unseen by others that might find it useful to influence their decision making. Datasets lacked clear ownership or documentation making it difficult for data scientists to find them. We believed that the crux of the problem was that we lacked a centralized catalog of these data and insights resources.

In early 2017, we released Lexikon, a library for data and insights, as the solution to this problem. The first release allowed users to search and browse available BigQuery tables (i.e. datasets)— as well as discover knowledge generated through past research and analysis. The insights community at Spotify was quite excited to have this new tool and it quickly became one of the most widely used tools amongst data scientists, with ~75% of data scientists using it regularly, and ~550 monthly active users.

However, months after the initial launch, we surveyed the insights community and learned that data scientists still reported data discovery as a major pain point, reporting significant time spent on finding the right dataset. The typical data scientist at Spotify works with ~25-30 different datasets in a month. If data discovery is time-consuming, it significantly increases the time it takes to produce insights, which means either it might take longer to make a decision informed by those insights, or worse, we won’t have enough data and insights to inform a decision.

Our team decided to focus on this specific issue by iterating on Lexikon, with the goal to improve the data discovery experience for data scientists and ultimately accelerate insights production. We were able to significantly improve the data discovery experience by (1) gaining a better understanding of our users intent, (2) enabling knowledge exchange through people, and (3) helping users get started with a dataset they’ve discovered.

Understanding Intent

To kick things off, we spent time conducting user research to learn more about our users, their needs, and their specific pain points regarding data discovery. In doing so, we were able to gain a better understanding of our users intent within the context of data discovery, and use this understanding to drive product development.

Low-intent data discovery

Let’s say you’re having a rough day and you want to listen to some music to lift your spirit. So, you open up Spotify, browse some of the mood playlists, and put on the  Mood Booster  playlist. You’ve just had a low-intent discovery experience! You had some broad goal to lift your mood and you didn’t have extremely strict requirements on what you wanted to listen to.

Within the context of data discovery, a data scientist with low-intent has a broad set of goals and might not be able to identify exactly what it is they’re looking for. This mode of discovery is particularly important for new employees or for people who are starting on a new project or team. For example, as a data scientist, I may want to:

• find popular datasets used widely across the company,
• find datasets that are relevant to the work my team is doing, and/or
• find datasets that I might not be using, but I should know about.

In order to satisfy the needs of low-intent data discovery, we revamped the homepage of Lexikon to serve personalized dataset recommendations to users. The homepage provides users with a number of potentially relevant, algorithmically generated suggestions for datasets including:

• popular datasets used widely across the company,
• datasets you’ve recently used,
• datasets used widely by the teams to which you belong, and
• recommendations for datasets you haven’t used, but might find useful.

While we did experiment with more advanced methods for serving recommendations, including using natural language processing and topic modeling on the dataset metadata to provide content-based recommendations, we determined through user feedback that relatively simple heuristics leveraging data consumption statistics worked quite well. In the first version of Lexikon, most traffic to BigQuery table pages was driven by search. After making these changes, we now see that 20% of monthly active users navigate to BigQuery tables through personalized recommendations on the homepage.

High-intent data discovery

You’re walking down the street and hear a passing car blasting a great song you haven’t heard in a while. You can’t get the song out of your head and need to listen to it immediately. So you pull up Spotify on your phone, search for the track, and play it (on repeat). You’ve just had a high-intent discovery!

A data scientist with high-intent has a specific set of goals and can likely articulate exactly what they’re looking for. This mode of discovery is often more important to more tenured data scientists who may be familiar with some datasets, but may be looking for something they haven’t used before that meets a certain set of criteria. For example, as a data scientist with high-intent, I may want to:

• find a dataset by its name,
• find a dataset that contains a specific schema field,
• find a dataset related to a particular topic,
• find a relevant dataset located in a particular BigQuery project,
• find a dataset that my colleague has used of which I can’t remember the name, and/or
• find the top datasets that a team has used because I’m collaborating on a new project with them.

To better serve the use case of high-intent data discovery, we iterated on the search experience. First, we focused on the search ranking algorithm. We learned through data analysis that although we have tens of thousands of datasets on BigQuery, the majority of consumption occurred on a relatively small share of top datasets. Data scientists in a high-intent mode of discovery were often looking for one of these top used datasets that met their needs. So, we adjusted our search algorithm to weight search results more heavily based on popularity. Following this change, in user feedback sessions, data scientists reported that the search results not only seemed more relevant, but they were also more confident in the datasets they discovered because they were able to see the dataset they found was used widely by others across the company.

In addition to improving the search rank, we also introduced new types of entities (e.g. schema field, BigQuery project, person, team, etc.) to Lexikon to better represent the landscape of insights production. Our belief was that by making these types of entities more explorable, we would open up new pathways for data discovery. For example, a data scientist might be looking for the best dataset to use that contains a  track’s URI  track_uri. In this case, a user can search for “track uri”, navigate to the “track_uri” schema field page, see the top BigQuery tables that contain the schema field, and navigate to the dataset page. In addition to the schema field page, we’ve added BigQuery Project, people, and team pages, which can serve as a similar stepping stone on the pathway to data discovery. Since launching these new entity pages, we’ve seen that they’ve proven to be a critical pathway for discovery, with 44% of Lexikon’s monthly active users visiting these types of pages.

Enabling knowledge exchange through people

Imagine you’re starting to explore the genre of jazz. You happen to notice that your coworker has a jazz album on Spotify pulled up on her desktop screen. You strike up a conversation and learn that she is a jazz aficionado. She has become your new genre guide. We’ve found that there are similar opportunities for people-to-person knowledge exchange with data discovery.

In addition to using learnings from user surveys, feedback sessions, and exploratory analysis to drive product development, we also conducted research on knowledge management theory to better understand how we might adjust our approach (recommended reading:  Knowledge Management in Organizations: a critical introduction  by Hislop, Bosua, and Helms).

With the first iteration of Lexikon, we used the knowledge management strategy of codification, which is based on the objectivist perspective of knowledge. This perspective assumes that knowledge can take the form of a discrete entity and can be separated from the people who understand and use it. In the case of Lexikon, we initially believed that if data producers did a great job describing their datasets there would be little-to-no need for person-to-person knowledge exchange. However, in reality, while the first iteration of Lexikon reduced the need for person-to-person knowledge exchange in discovery contexts, there were still instances in which people found it useful to connect with others to find the right data. Rather than fight this, we decided to embrace the idea by (1) mapping expertise within the insights community and (2) providing supplemental information in collaboration tools.

Mapping Expertise

Through user research, we learned that data scientists who failed to discover the data they were looking for would often fall back to finding an expert in the insights community on a given topic and connecting with them in person or online. However, in some cases, data scientists found it difficult to find the right person to talk to about a particular topic. This was especially true for new employees who hadn’t yet built personal connections with members of the insights community. So, we introduced a feature in Lexikon that allows you to search for people working in the data and insights space related to a given keyword (i.e. “experts”). These results are powered by summarizing an employee’s insight production and consumption activity related to the given keyword. For example, an employee who queries/owns datasets, views/owns dashboards, authors research reports, and/or runs A/B test experiments related to the given keyword will be returned in the list of results. More weight is given to actions related to insights production (e.g. owning a dashboard) rather than insights consumption (e.g. viewing a dashboard).

Providing supplemental information in collaboration tools

Following the release of the first version of Lexikon, we found that data scientists continued to talk with each other about datasets in Slack. Rather than discourage this discussion, we felt like we could help improve the person-to-person knowledge exchange by providing supplemental information. So, we built a Lexikon Slack Bot to improve discussions about datasets. When a user shares a link to a dataset in Lexikon, the Slack bot provides a brief summary of the dataset including:

• description,
• usage stats, 
• data lifecycle information, 
• access tier, 
• an overview of the most used schema fields in the table, and 
• links to view more information in Lexikon, request access, or open directly in BigQuery. 

Not only does this provide useful information to users in the moment, but it has also helped raise awareness and increase the adoption of Lexikon. Since launching the Lexikon Slack Bot, we’ve seen a sustained 25% increase in the number of Lexikon links shared on Slack per week.

Helping people get started with a dataset they’ve discovered

You just listened to a track by a new artist on your Discover Weekly and you’re hooked. You want to hear more and learn about the artist. So, you go to the artist page on Spotify where you can check out the most popular tracks across different albums, read an artist bio, check out playlists where people tend to discover the artist, and explore similar artists. Using these features on the artist page after your first listen allows you to truly discover and build a connection with the artist. Similar to artist discovery, one of the most critical steps in data discovery is the final step—starting to use the dataset you’ve discovered.

Through user research, we learned that data scientists would often have a lot of questions about how to start using a dataset, which slowed down their ability to start using the dataset they just discovered. So, we developed the features Schema-field consumption statistics, Queries, and Tables commonly joined to address this last mile of discovery.

Schema-field consumption statistics

Datasets often contain dozens or even hundreds of schema fields. Once you’ve determined that you’ve found the right dataset, it can be quite daunting to try to understand all of the available fields and determine which ones are actually relevant. In addition to basic metadata about the schema fields, we included consumption statistics at the schema field level. This shows the number of queries referencing the schema field and the number of unique people who have queried the schema field. This feature gives Lexikon users a way to sort the list of available fields by usage to easily find the ones that are likely to be the most relevant.

Data scientists are often curious to see how a dataset is actually used in practice. In the first version of Lexikon, we introduced example queries that allowed data producers to submit example queries to give data scientists an idea of how they might use the available dataset. We found there were a few issues with this approach. First, we ran into challenges encouraging data producers to share example queries for all datasets. Second, of the example queries that were submitted, they often became outdated quickly given the ever-changing landscape of data. For example, an example query might be out-dated because it included a join to a deprecated table. So, we abandoned the curated example query and instead allow users to search through all recent queries made on the given dataset. This gives users the opportunity to see a variety of up-to-date queries that use the dataset, and the ability to search for specific queries on the dataset (e.g. “show me queries on this table that reference this specific field”). Since launching this feature, we’ve seen that 25% of users who visit a dataset page use the queries feature.

Tables commonly joined

It’s rare that a single dataset will contain all of the information for which a data scientist is looking. It’s likely the case that they’ll need to join a dataset with others in order to answer the question they have. So, we built a feature on a BigQuery table page that allows the user to see tables that are most commonly joined with the given dataset. While this isn’t the most widely used feature, we’ve seen that it is consistently used by 15% of users who visit a dataset page.

Final Thoughts

By understanding the user’s intent, enabling knowledge exchange through people, and by helping people get started with a dataset they’ve discovered, we’ve been able to significantly improve the data discovery experience for data scientists at Spotify.

Since making these improvements to the data discovery experience in Lexikon we see that adoption of Lexikon amongst data scientists has increased from 75% to 95%, putting it in the top 5 tools used by data scientists. For comparison, more people report using Lexikon than BigQuery UI, Python, or Tableau at Spotify. Lexikon’s user base has organically grown from ~550 to ~870 monthly active users as it has proven to be useful to data consumers in non-insights specialist roles (e.g. engineers, data-savvy product managers, etc.). We’ve also seen a significant increase in engagement with the average number of sessions per MAU increasing from ~3 to ~9 since our initial launch. In addition to these encouraging adoption and engagement metrics, we’ve learned from surveying data scientists that after making these improvements data discovery is no longer identified as a primary pain point in insights production.

Join the band

If you’re interested in helping us tackle similar problems or you’re a data scientist that’s looking to work at a company where producing impactful insights is becoming easier every day, visit the  Join the Band  page to view open roles.

Acknowledgments

We’ve had a number of folks help get this product to where it is today. Shout out to our current team (Ambrish Misra, Bastian Kuberek, Beverly Mah, David Lau, Erik Fox, and Nithya Muralidharan) and others who have contributed to Lexikon (Adam Bly, Aliza Aufrichtig, Colleen McClowry, David Riordan, Edward Lee, Luca Masud, Mark Koh, Molly Simon, Mindy Yuan, Niko Stahl, and Tianyu Wu).

Andrew Maher  is a Product Manager for Spotify’s Insights Platform Product Area.

Sign up for engineering updates

The Story of Spotify Personas

Spotify personas are the topic of much discussion by those in the product, design, and user research communities. Here, Olga Hörding, Mady Torres de Souza, and Sohit Karol explain how we developed our personas tool, how we use it today, and why it’s so useful for an autonomous, cross-functional organisation like Spotify. Why not listen to our companion playlist for this article?

Here at Spotify, we often ask ourselves whom we’re designing for. And since listening to music is so universally popular, it might seem at first that the answer is ‘everyone’. After all, Spotify is available as a free and paid product. It can be used by anyone with a phone, computer, car, set of smart speakers, or many other devices. It’s present in over 79 markets and it offers experiences – like Daily Mixes – that are personalised to every single listener.

Yet designing for a mass, generalised audience isn’t likely to end up pleasing ‘everyone’. So in 2017, our team was challenged to create a better understanding of existing and potential listeners. We wanted to agree on how to differentiate the needs of these listeners and the problems our products could solve for them. We needed a solution that was durable and flexible enough to work for autonomous teams, working out of different offices, in different countries and on different parts of our products. And we were determined to put a face to our listeners – an identity that everyone at Spotify could recognise and talk about with ease.

We responded to this challenge by designing personas.

How did we craft the personas?

User-centred design has several schools of thought on how best to create and use personas. The general idea is that capturing and clustering the needs, goals, habits, and attitudes of existing and potential users helps to build a solid understanding of the problem space. For us, our personas tool is an example of a boundary object – a durable and reliable artefact that’s flexible enough to inspire discussions, share information, and adapt to the needs of the product development process. And we developed it in two phases:

Phase 1 (2017)

In Phase 1, we scoped our analysis to US listeners. We picked this market due to its size and the variety of listening behaviours that emerge from the way of life there – for instance, long commutes, suburban lifestyles, and so on. At the start, we discussed the idea of clustering behaviours gathered from our current data. But we moved away from this approach because it revealed only superficial knowledge about our listeners and concealed the reasons behind their behaviour. It also failed to help us understand why potential customers listen to music. So instead, we decided to study listeners of different ages, incomes, family types, lifestyles, music cultures, and more. We used a combination of diary studies and contextual inquiries to collect this data.

Early in the analysis, we noticed that people’s needs or reasons for listening to music were consistent, even in different clusters — that is, to kill boredom, to feel productive, to entertain themselves, etc. But what was different was their attitude towards music consumption, the value they saw in paying for music and their behaviours around devices in different contexts.

As a result, we ruled out the idea of clustering based on needs alone and used a combination of Alan Cooper's method and the Grounded Theory approach to build our personas instead. We transcribed our interviews minute-by-minute. Then, we coded and clustered them into needs, attitudes, device habits, contexts, and other dimensions in order to identify the best cluster combinations. Two tools — Mural and Airtable — were particularly useful during this phase.

Phase 2 (2018)

In our next phase, we built on a key Phase 1 insight – that when it comes to music listening, context matters. Sure, there’s value in creating abstract dimensions, such as needs and motivations. But ultimately, people use Spotify in the real world. Their device ecosystems, physical and mental abilities, and other contextual factors shape their listening choices. And so, combining the learnings from Phase 1 with a literature review of theories from sociotechnical systems and adaptive computing, we decided to focus Phase 2 on how people listen to music together.

In this phase, we sought to unpack the nuances and complexities that arise when people listen together at home, in the car, with kids, and more. And since this work built on our previous research, we once again kept our sampling within the US. We included roommates, empty nesters, partners with and without kids, households with toddlers, teenagers, and others. Our goal was to ensure we had an extensive variety of situations where people came together to listen to music.

Unlike in Phase 1, we followed up our diary studies and contextual inquiries with a bottom-up analysis using the Grounded Theory Approach. Qualitative coding revealed insights that we would have otherwise missed and resulted in the Listening Together Framework™, our tool to communicate the outcomes to a broader audience.

While people might have the same problems or needs, the existing habits determine the existing methods they use to address those problems. Attitudes determine how different people will adopt products designed to meet their needs.

Next, how should we represent our listeners?

Representing personas poses a tricky challenge: we want them to be relatable, but they’re not 1:1 matches with real people. Believable human traits and flaws help create empathy with problems and needs. But we don't want groups to be wrongly excluded based on the characteristics we've picked. So finding a balance is a crucial step if we’re to create useful and believable archetypes.

For that reason, we arbitrarily picked genders, names and appearances that matched the range of people we interviewed. While personas exist independently from these traits, they were fundamental to make them memorable as people. And deciding which human characteristics to include in each of the personas was especially challenging. To do so, we reduced the representation of personas to keywords, colours, symbols, and energy levels reflecting their enthusiasm for music. This exercise helped us navigate through the variations of poses, facial features, clothing, and visual styles we created.

To balance out these specific traits, we used flat illustrations with our brand colours, giving them a more abstract look. Avoiding a too-realistic representation made the material easy to refresh with evolving illustrative styles. It was also much easier to reproduce in high or low fidelity, since sketching a specific pose or picking a colour palette would be enough to refer to a persona.

How did we share our work?

We didn’t wait until our personas were complete before sharing them – we actually started thinking about communication as soon as we began our research. We spent a lot of time testing our asset ideas in pilot workshops. The goal was to integrate with our existing practices seamlessly. And by following our team needs, we crafted a communication strategy for Personas that includes digital assets, physical assets, and workshops.

Digital assets

Traditionally at Spotify, we create Google presentations when reporting back research – and sometimes, these get lost amongst all the many other presentations produced! But this time around, we envisioned our personas work to be relevant for at least a couple of years. So we created an interactive website, shared across Spotify offices through announcements and posters. Having a digital source of truth for the research was especially handy whenever we needed to update the study or add new learnings.

Physical assets

Raising awareness about the personas was useful, but we didn't want to stop there. We wanted to create fun, playful ways for the teams to incorporate them into their workflows. So we created assets that teams could use on their own, whether they were running one-hour mini-workshops or design sprints over several days. These assets were made available through our personas website.

Our team hanging out with the personas cardboard cutouts and the card game we've created to share the insights.

One of the most powerful modalities for learning that emerged during our pilot workshops was ‘learning by doing’. So the user research team hosted workshops with product teams and helped them to use personas in a way that was relevant to their specific areas.

What was the impact?

Since our teams are so autonomous, we realised right from the start that the personas would be relevant to all of them in different ways and at different stages of their work. For that reason, no one was mandated to use personas. Yet, as a reliable, durable, and carefully designed information artefact, we’ve seen many teams beyond the product organisation adopt them into their work and vocabulary over time – including those across Marketing, Content, and Brand.

For instance, teams that want to create features from scratch can now choose their personas, map out the existing opportunities, pick a direction, and start ideating from there. Although personas don’t replace user research, they can help us create educated hypotheses and save us time – meaning we don’t need to run foundational research every time we want to explore a new topic within the music listening experience. Our teams can now focus their resources on diving deeper into problems from the level set by the personas.

Equally, when teams are more focused on maintaining features, they can now map out their work and see how different personas would use it. They can create mental model diagrams for different personas and discover how they experience their journeys. And in doing so, they can refine the features to better fit certain ways of listening to music, while making sure they don’t alienate others.

Crucially, the personas are slowly becoming a part of our internal vocabulary – a means of helping teams to select and identify which ways of listening are being affected. We can’t optimise a feature for every single one of our listeners. So today, it’s common to see teams having their product roadmaps centred around specific personas instead.

A long process, with long-lasting results

Sometimes, in order to move fast, you have to move slow. Foundational research initiatives, like the development of personas, take time and are resource-intensive. Yet the learnings benefit us long into the future. And here are just a few of them:

When in doubt, over-communicate. We need a regular cadence to share details and progress around the organisation – this might add overhead, but it ensures alignment and transparency. We used Facebook Workplace, Slack, and emails to keep the stakeholders updated throughout the process.

Keep your disciplines close. Our process had to move quickly from behavioural analysis to fieldwork, then straight onto asset creation and scoping needs, attitudes and habits, through the use of surveys. The speed we moved was only possible by having design, user research, and data science integrated throughout the process.

Know your audience. Adopting new frameworks may be a significant change for some product teams. So we spent lots of time getting to know their workflows, running pilot workshops, and inviting them to fieldwork sessions in order to build trust and reduce any potential resistance to change.

As Spotify continues to grow, we expect to expand and adapt our personas for markets outside the US, as well as broadening out our area of study to also include podcasts. There are exciting times ahead and plenty more work to be done – we’re looking forward to the next chapter in the story of Spotify personas. :)

Mady Torres de Souza

Mady designs with the Home Consumer Electronics team. She converts oxygen into activities like obsessive cooking and taking electronics apart.

Olga Hörding

Sohit Karol

User Researcher at Spotify working on personalized listening experiences.

Introducing Spotify’s New Design Principles

Design principals talk design principles. Here’s how we aim to make Spotify feel Relevant, Human, and Unified.

• Design Systems

A Design System Isn't One-Size-Fits-All

At Spotify, we’re constantly investing in how we support and deliver a design system to the business. Here's what we learned along the way.

Designing Data Science Tools at Spotify: Part 1

A product designer explains how she built usable, well-designed tools to help data scientists transform a mass of raw data into valuable insights.

• Editorial Design

Designing for Belonging: Why Image Localization Matters

A closer look at the ways in which we try to ensure that Spotify looks just right, no matter where it appears.

Want Spotify Design updates sent straight to your  inbox ?

By clicking send you'll receive occasional emails from Spotify Design. You always have the choice to unsubscribe within every email you receive.

• Humans of User Research
• Getting Started
• AI in UX Research
• 📚 ALL Resources
• 🗓 UserCalendar

🗞 How UX Research at Spotify is getting ahead of shifting trends

Spotlight > Articles > Video > Audio

😌 Humans of User Research w/ Shweta Singh

Hi Shweta, tell us a little about yourself. I’m Shweta Singh, a UX Researcher at ADP, working primarily in an effort to reduce the amount of time and money spent on time-consuming and costly human resource management compliance tasks. Owing to the work I did at TCS for 3.5 years, I built strong analytical, leadership, project management, and collaboration skills that helped me greatly when I finally found my passion in user experience research in the digital product industry. I believe my true potential resides in directing the strategy of digital product design through human insights. I'm an active volunteer for social causes- education, diversity and inclusion. I'm a part of organizations like ENGin and GlobalTech Network. This enables me to connect with people from across the globe which I love.

What originally got you interested in UX Research? Let's start with what got me interested in Design first. One of the first books I read on Design was Emotional Design by Don Norman. I know people usually start with Design of Everyday Things but I’m glad I picked this one up first. I understood how good design can influence how a person feels, and it struck me as incredibly powerful. Moreover, I read more and more about User Experience and started practicing UX Research within the team I was working on at the time. What appealed to me about UX Research was the search for the right problem to solve and it stayed with me.

What does your perfect evening look like after a day full of user studies? After a long day of user studies, I would curl up in my bed with a fun show on my laptop. On certain days I like to doodle. These are my favorite ways to unwind after a long day.

What can people reach out to you about and how can they find you? One of the questions that I’d like to always bring up with other UX Researchers is how they break into the field and their thoughts on mentorship in UX. Some other topics of my interests are mixed-method research and how to deliver quality insights in the new remote world. Let's connect! Drop me an email at [email protected] or connect on Linkedin .

🧐 Highlights.

How user research at Spotify is getting ahead of shifting trends in global culture As we leap into spring, user researchers at Spotify are thinking carefully about how we can adapt to stay ahead of shifting global patterns and ensure an engaging, relevant experience for our users everywhere. Alexandra Hornsby 03/10

When should I report a one-off insight? Though one participant’s opinion isn’t statistically significant, there are times where a piece of feedback may point to a larger issue. Nikki Anderson 03/12

The state of user research 2022 report: highlights, themes, and takeaways What did we learn from this year's State of User Research Report? Here are the key trends to note about buy-in, budget, ops, and more. Katryna Balboni 03/14

🛬 Approaches.

Transfer your academic skill set to the world of UX While switching industries may feel like a daunting change, it helps to remember that you’re not going alone—you’re bringing years of expertise with you. Below, we explore how to best incorporate prior knowledge into your UX practice. Janelle Ward 03/10

The state of ResearchOps: untapped yet The field of research operations (ResearchOps) has garnered attention in recent years, but dedicated roles in organizations are uncommon and resources are scarce today. Kara Pernice 03/13

Applying psychology to UX research Four principles from Psychology that are particularly relevant to UXR: 1) Users don't think like us. 2) Users don't have good insight into the reasons for their behaviour. 3) The best predictor of the user's future behavior is their past behavior. Luísa Quinn 03/14

Translating user research into design Ensuring your hard-earned research insights turn into compelling design solutions. Taylor Palmer 03/13

Overview: setting the right metrics for product launches Product launches are fraught with challenges—from budgeting to timelines to collaboration breakdowns, there are many places and ways a launch can go wrong. But one area where headaches are preventable is metrics: i.e., how you measure the success of your launch and subsequent product performance. Mark Simborg 03/13

Prototyping in data science Introduction of the term “discovery-driven prototyping” to describe how prototyping in data science can lead to the exploration of possible paths for decision-making. Fabien Girardin 03/14

Research teams of 1 (Sponsored) Being a solo researcher presents a unique set of opportunities and challenges. Great Question is compiling an eBook of best practices to support these folks, and they'd love your contributions. If you are a research team of 1, participate in their study here: Participate

Enter your email to receive the latest updates in your inbox.

🎥 Video of the Week.

Bridging UX and team science Stephanie presents an introduction to user experience design (UXD) & user experience research (UXR) and to key topics in these fields. She discusses how team science practitioners and researchers can integrate UXR and UXD both into their work and into growing their own web presence, using an interactive exercise and opportunities for group discussion. YouTube 03/10

🔉 Audio of the Week.

Racine Brown on Anthro to UX with Matt Artz In this episode of the Anthro to UX podcast, Racine Brown speaks with Matt Artz about his UX journey, medical anthropology, and leading a UX team. Racine earned a PhD in Applied Anthropology from the University of South Florida. He has worked for Radiant Digital and is starting a new UX Research role at AnswerLab. Spotify Apple Google 03/08

Happy Researching, Jan

• Share on Twitter
• Share on Facebook
• Share on LinkedIn
• Share on Pinterest
• Share via Email

User Weekly Newsletter

Join the newsletter to receive the latest updates in your inbox.

This website uses cookies to ensure you get the best experience on our website Got it!

7 jobs in 4 locations

Data, Research & Insights

We believe in mixed methods and diverse teams, investing a lot in creativity and experimentation to achieve broad narratives we can act on as a business.

We’re a driving force behind business decisions - big and small. It’s our job to understand the ecosystem of listeners, fans, creators, content, and advertisers, and use our learnings to help define where we go and what we do next.

Advising from afar isn’t what we do. Instead, we like to make a direct impact and become valuable, long-term partners to the teams we serve. All so we can deliver what we’re here for - hands-on user research and analytics, sophisticated modeling and forecasting, and cutting-edge technical research.

Data Science

Data analysis is how we get to know our business better, understand what people like, connect them with the right content, and communicate effectively. It’s also how we further our key product and business goals too.

User Research

We deliver a deeper understanding of our listeners, creators, and partners, their behavior, and the challenges they face. We then turn this into insights for the business, the kind that can drive truly user-centered innovation and product development.

Product Insights

Not an individual discipline, but the integration of many from areas like Data Science, User Research, Analytics Engineering, and more. It’s a novel approach, but one that’s really working for us. We work as one with the Product teams, helping to guide decisions by backing them up with a blend of insights and analysis.

Research Science

We create cutting-edge experiences for our listeners and creators. With our pioneering research, we’re helping to redefine what state of the art means in our industry and beyond, regularly contributing to the wider research community by publishing papers. AI heavily informs our research and helps us achieve our goals, from matching listeners with the content they want - or didn’t even know they wanted - to AI-assisted music creation tools.

Market Research

We look at the world around us and extract what it can tell us about consumers in any given market. Economics, demographics, attitudes, perceptions, and experiences, we gather and analyze insights in a broad range of qualitative and quantitative ways to better inform business decisions.

Data & Insights Leadership

We ensure Data, Research & Insights as a whole are working on the most important and most impactful areas. We work closely with leadership in Product, Design, Engineering, and Business, keeping our ambitions aligned and helping to successfully navigate any ambiguous problem space.

Fresh opportunities

Hang on, just getting the latest stats...

0 jobs in all locations in all job types

Fresh content

Title: NerdOut@Spotify, Join us for Wizards, Portals, and Spotify for Backstage

Title: NerdOut@Spotify, 27: Measuring Developer Productivity

Title: A Product Story, 10: In pursuit of the perfect session - a DJ just for you

Title: NerdOut@Spotify, 26: A Trillion Events

Title: NerdOut@Spotify, 25: Voice Translation *Release Notes*

• Who we are Who we are
• Our culture Our culture
• Equity & impact Equity & impact

One band, no solo artists

It’s our culture. It’s our values. It’s who we are and what we’re not. It’s why we do things the way we do and why that matters. It’s all here in our band manifesto.

Quick clicks

Explore our global locations

Students , make an impact not the coffees

Read about diversity, equity & impact

Diversity, equity & impact

Spotify Community

• Get Premium
• App & Features

User Research for Spotify

Help wizard.

Type in your question below and we'll check to see what answers we can find...

Loading article...

Submitting...

If you couldn't find any answers in the previous step then we need to post your question in the community and wait for someone to respond. You'll be notified when that happens.

Simply add some detail to your question and refine the title if needed, choose the relevant category, then post.

Just quickly...

Before we can post your question we need you to quickly make an account (or sign in if you already have one).

Don't worry - it's quick and painless! Just click below, and once you're logged in we'll bring you right back here and post your question. We'll remember what you've already typed in so you won't have to do it again.

Please see below the most popular frequently asked questions.

Loading faqs...

• Ongoing Issues

Please see below the current ongoing issues which are under investigation.

Loading issue...

Loading ongoing issues...

• Help Categories

Help categories

Account & payment.

• Subscriptions
• Premium Family
• Premium Student

Using Spotify

• Your Library
• Content Questions
• Spotify for Developers
• Other (Podcasts, Partners, etc.)

Listen Everywhere

• iOS (iPhone, iPad)
• Desktop (Windows)
• Desktop (Mac)
• Desktop (Linux)
• Subscribe to RSS Feed
• Mark Topic as New
• Mark Topic as Read
• Float this Topic for Current User
• Printer Friendly Page
• All forum topics
• Previous Topic

• Mark as New
• Report Inappropriate Content
• App & Features

Suggested posts

Let's introduce ourselves.

Hey there you,   Yeah, you!    Welcome - we're glad you joined the Spotify Community!   While you here, let's have a fun game and get to know…

• The Star ePaper
• Subscriptions
• Manage Profile
• Change Password
• Manage Logins
• Manage Subscription
• Transaction History
• Manage Billing Info
• Manage For You
• Manage Bookmarks
• Package & Pricing

Spotify expands AI Playlist feature to new markets including US, Canada

Tuesday, 24 Sep 2024

Related News

King visits Comac Beijing Aircraft Technology Research Institute

Uzbekistan holds week of information and communication technologies, trading ideas: globetronics, pgf capital, pansar, nextgreen, heitech padu, cosmos technology.

FILE PHOTO: A screen displays the logo of Spotify on the floor at the New York Stock Exchange (NYSE) in New York City, U.S., December 4, 2023. REUTERS/Brendan McDermid/File Photo

(Reuters) - Spotify is expanding its tool that helps create playlists using generative artificial intelligence for premium users to four new markets, including the United States and Canada, the Swedish streaming giant said on Tuesday.

AI Playlist, which is currently offered in beta, helps subscribers personalize their selection of songs that can be refined via additional text prompts.

Spotify is expanding AI Playlist to fresh markets, including Ireland and New Zealand, in a bid to attract new subscribers by sprucing up its app with AI amid growing competition from rival services by Apple and Amazon.com.

Spotify said the feature, which was launched in the United Kingdom and Australia in April, will not produce results for non-music-related prompts such as current events or specific brands, as it is still in beta.

The company also offers tools such as 'daylist', a personalized playlist that updates daily with new music recommendations, and 'AI DJ', which creates music recommendations based on users' listening habits.

The company's paying subscribers rose about 12% from a year earlier to 246 million during the second quarter.

(Reporting by Jaspreet Singh in Bengaluru; Editing by Shreya Biswas)

Found a mistake in this article?

Report it to us.

Thank you for your report!

Stronger ties, growing partnerships at 21st China-Asean Expo

Next in tech news.

Trending in Tech

Air pollutant index, highest api readings, select state and location to view the latest api reading.

• Select Location

Source: Department of Environment, Malaysia

Others Also Read

Best viewed on Chrome browsers.

We would love to keep you posted on the latest promotion. Kindly fill the form below

Thank you for downloading.

We hope you enjoy this feature!

Complex problems, exceptional colleagues, and a direct impact on the joy of 615 million people. Our beliefs drive our success.

Browse our jobs site to see the latest openings, and learn about our mission, our beliefs, our culture, our values, and everything else in between.

Spotify’s 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.

The race is on to find others who share our passion for building the future of audio.