Spotify Classical: UX Research

Understanding Spotify’s classical music listeners and how to improve their experience

Research question: How can we improve Spotify’s classical music experience?

Timeline: March to June 2020

Role: UX researcher, UX designer

Classical music listeners have long complained about their experience on Spotify — specifically, that there is no good way to browse or compare different recordings of the same classical work from the music streaming app’s enormous library of classical music.

In this case study, I share how I conducted primary and secondary user research and a variety of analyses to answer the question:

How can we improve Spotify’s classical music experience?

✅ Final Recommendation — Champion Listeners

To answer the research question, I conducted several months of primary and secondary research including:

• interviewing and surveying listeners and musicians and surveys

• pouring through app feedback

• creating two personas

• analyzing technical and business issues

• investigating task flows for a half dozen apps

• and running prioritization matrices to help decide what to build

. . . and after all this, the best overall business solution suggested that:

❌ Spotify should do nothing. Classical problems are inherent and complex in Spotify’s metadata structure. Any changes to improve classical discovery and searching would be extremely cost-prohibitive given how small the classical music market is.

However, my first job was to be a champion for users, so the final design recommendation addressed the best solution for listeners:

✅ Spotify should acquire, re-brand, and adapt an existing classical music service to give classical listeners a vastly improved experience and to incentivize premium subscriptions.

Real World Results Validate Research

Eerily enough, even though this was a passion project and Spotify Classical exists only as a prototype, both of the above results actually happened.

• ❌ Spotify did nothing for classical listeners — since 2020 and at the time of writing, Spotify has not significantly improved its classical music experience, and it remains frustrating to use. Instead, Spotify focused on podcasts. So my initial solution proved right, unfortunately.

• ✅ A Spotify competitor acquired a classical music app, re-branded it, and launched a new service. In 2021, Apple Music, which had the same classical problems as Spotify, acquired an existing classical musical service — Primephonic, rather than IDAGIO. It adapted and rebranded it, launching its own classical-only service, Apple Classical, in 2023 to boost subscriptions.

Real-world events eventually validated my recommendations — just for different companies. Not bad for my first UX project ever.

Here’s How I Did It:

1. Context, Goals, Constraints

2. Understanding classical listeners

3. Metadata problems

4. Competitive analysis,  Affinity Diagram, Task Analyses

5. Personas, User Journey

6. Deciding what to build

7. 🔥 Design recommendations
   Answering the research question

8. Results & Next Steps

9. Key Takeaways

10. Toolstack

Context

Improving Spotify for classical music listeners

Classical music listeners have long complained there is no easy way on Spotify to find, browse or compare different recordings of the same classical work from its enormous library of classical music.

A classical music listener and musician myself, I spent the spring and summer of 2020 as the sole UX researcher and product designer exploring solutions to how Spotify might improve their experience for classical music listeners.

Goals

What at first appeared to be a deceptively simple research question — "How might we improve Spotify’s browsing and comparison for its classical music listeners?” — proved to be far more complex than I had ever imagined.

Trying to get a handle on the problem, I ended up grappling with a host of thorny issues:

• The unique needs of classical listeners

• The business challenges of the classical music market

• The technical problems with classical metadata

• How Spotify handles classical music; and

• Business issues driving Spotify’s subscription race with competitors, especially Apple Music

Constraints

Business constraints — I had no access to Spotify’s stakeholders or business objectives during the project, nor any real budget. So I conducted guerrilla research and made reasoned decisions about product design based on that research as well as Spotify’s past actions and recent initiatives — for example, in 2020 Spotify had just invested a huge amount into podcasts.

Technical constraints — Similarly, with no access to Spotify’s engineering or developers, I had to reach out to software engineers I happened to know, to get some technical advice on issues of metadata, databases, and backend servers, that would drive design decisions and recommendations.

Classical music listeners struggle with Spotify

Talking to real listeners first

To better understand the needs and frustrations of Spotify’s classical listeners, I started by talking to real listeners:

• Interviewing more than a dozen musicians and classical music aficionados

• Conducting surveys of music listeners, and

• Posing questions to classical fans on forums like Talk Classical and Reddit.

A life-long musician myself — I play 📯 French horn and 🎸bass guitar — I leveraged my connections with a major Florida symphony general manager to recruit and interview professional and semi-pro musicians and dedicated classical music listeners, many of whom use Spotify. (I would also later employ this same group in user testing.)

Combing App Store reviews for real user feedback

In addition, I combed through hundreds of app reviews left by actual users on Apple’s App Store and Google Play for a number of music streaming apps, including Spotify, Apple Music, Amazon Music, YouTube Music, Tidal, IDAGIO, and Primephonic.

While not a substitute for a real research team and budget, having access to opinions posted by thousands of real users proved a gold mine, in terms of uncovering insights.

Key insights

Difficulty browsing and comparing works

• 😣 Classical listeners struggle to find music on Spotify

• 🔍 Search can be hard if you don’t specifically know what you want

• 🍎 🍊 Comparing recordings is especially difficult

Secondary insights

• Classical listeners hold ambivalent attitudes toward Spotify

  • ➕ positive: it has a huge library of classical music

  • ➖ negative: its experience for classical listeners is poor

• Familiarity with and use of Spotify is high, but much less with classical streaming services Idagio, Primephonic

• Most listeners had extensive analog collections — 💿 CDs or vinyl — but a plurality also listened to rock/pop/jazz

• Most classical listeners prefer listening at home

• Many listeners turn to classical for "relaxation”

• Many listeners use Spotify primarily as “research” — they search Spotify to find new releases and then buy them on CD or vinyl to listen at home

Spotify’s structural problem — metadata

Artist, Song, Album

Spotify and most other major music streaming services were built for popular music — music is listed by artist, song, and album. This information about each track is called metadata. However, classical music is not like popular music — it typically has triple the metadata.

Nonperforming composers and multiple movements

Here’s one classical metadata example: a recording of the first movement of Bach’s Brandenburg Concerto No. 1 (one of more than 115 recordings on IDAGIO) :

• Composer name: Johann Sebastian Bach

• Work: Brandenburg Concerto No. 1 in F major BWV 1046

• Movement: I. (Allegro)

• Album: Johann Sebastian Bach Brandenburg Concertos

• Conductor: Karl Richter

• Performing ensemble: Münchener Bach-Orchester

• Soloists: Hermann Baumann (Horn), Karl-Heinz Schneeberger (Violin)…

• Period: Baroque

• Year: 1999

Because Spotify uses the artist-song-album format, all this classical metadata must be shoe-horned into 3 slots. The result for classical music listeners when searching looks like this graphic — a confusing jumble of results all with no easy way to discern different works, movements, performers, conductors, ensembles, or albums. This problem is unique to classical music.

Classical markets and apps

Key insights from secondary research

While conducting primary research interviewing classical listeners, I spent a couple of weeks digging into secondary research about streaming music services, especially then-new classical-only apps like Idagio and Primephonic. I read dozens of reports, news articles, and studies. Key insights included:

• 😔 Spotify has a long history of poor classical experience

• 😖 But none of the other major music streaming services handle classical well, either

• Metadata is the main culprit ^^

• Several startups — Primephonic, IDAGIO — have stepped in to take advantage of this small, but fiercely loyal audience

• Classical listeners who’ve tried new services like IDAGIO love them

• Classical libraries at new classical-only services tend to be small, limited

White paper supplies key personas

⚠️ Easily the most important secondary research I encountered was a white paper commissioned by then-new classical streaming service Idagio called The Classical Music Market. It detailed a gap in the music market no one had yet filled [at the time]. Without UXR resources of my own, I would rely on this paper to create primary and secondary personas for Spotify Classical.

Competitive Analysis — Spotify needs more and more subscribers

One of the key business facts gleaned from some competitive analyses I conducted during my research is that classical music is just a tiny fish in a huge streaming ocean. Although classical accounts for 2.5 percent of album sales in the United States, it accounts for less than 1 percent of total streams.

So for most major music streaming services, classical is just not a priority. . . bad news for classical listeners.

However, a competitive analysis of the total music streaming market showed Spotify in a race to maintain dominance — especially against Apple and Amazon, which have been gaining in subscribers, and possess their own hardware (Apple) and/or video streaming (Apple, Amazon).

🔥 My analysis of the market showed that, in the on-going fight for premium subscriptions, that 1% of streaming classical listeners could make a real difference from a business standpoint to Spotify

Putting it all together

Affinity diagram focuses and uncovers insights and issues

From the huge amount of information collected through primary and secondary research, I created an extensive affinity diagram to better focus on the issues most important to listeners and how I could answer the research question.

The diagram was most helpful in organizing comments and opinions by subject, allowing me to discover, for example, the incredible UI design frustrations of opera lovers trying to use Spotify:

However, it also allowed one to see just how many (and which) positive aspects diagram Spotify had to offer classical listeners:

Buried among the facts and quotes were two comments about a classical music streaming app’s significant accessibility issue that would drive design decisions later in the Spotify Classical project.

Comparing what works & what doesn’t — task analysis

To best understand the frustrations and needs of classical users, I tested out a series of identical tasks using six (6) different music streaming apps — Spotify, Apple Music, Amazon Music, IDAGIO, Primephonic, and Naxos.

Results aligned closely with user research.

• Some tasks — like comparing different recordings of classical music — were difficult and frustrating to perform on Spotify.

• However, on other tasks — like specific searches — Spotify performed as well or better.

Task analysis would prove invaluable not only for creating a realistic and actionable customer journey but also for later interaction and user interface design.

Frustrations on a musical journey

For the final parts of the UXR project I synthesized data and insights gained from primary research and — borrowing many aspects from the classical music personas created for Idagio — created primary and secondary personas for Spotify’s classical listeners.

Needs, goals, and pain points

Marie, the primary persona, is a more casual classical listener with a family, all of whom use Spotify for different. Her goal is to be able to access both classical and other genres. But she feels overwhelmed by search results.

Andreas is a “superuser” and opera aficionado who regularly attends concerts. He needs to be able to compare different recordings of works, but feels that Spotify's search is “totally unacceptable” — he can’t find anything.

Customer Journey

Using actual user quotes and feedback from interviews and app reviews, the maps for Marie and Andrea’s musical journeys demonstrated how frustrating the current search, browse and compare processes are for classical listeners.

I used direct quotes from the affinity diagram helped make the point. For example:

“The only way to find what you’re looking for is to start playing a track and skip around until you find the right one.”

In the later design stage, focusing on the prime areas of frustration — search, browse and compare — would prove critical to the success of the design.

Deciding what to build, if anything

It was now clear there were significant issues with Spotify’s classical experience and that some new apps had successfully addressed them. What solutions could benefit both classical listeners and Spotify?

I had no stakeholders to consult or to decide for me.

Fix Spotify or Create New App

The solutions broadly suggested by the research were:

1. Fix Spotify itself — this would require expanding metadata on the backend, and altering the UI (and UX) significantly on the front end; or

2. Create a classical-specific version of Spotify (like Idagio or Primephonic)

Neither solution was without significant costs or risk:

• 🔥 The first option would affect the vast majority of Spotify’s non-classical listeners without providing them any additional benefit while helping a small minority of classical listeners.

• But creating a new classical app from scratch for a relatively small user base seemed a risky undertaking with low ROI

What’s worth doing?

Was either method worth it? If I had access to stakeholders, I could have gotten expert input on the business and technical issues. But I didn’t. Rather than just arbitrarily pick some direction, I tried to determine a decision Spotify itself might choose, balancing business, technical, and user needs.

I compared eleven different options including:

• Fix search on Spotify

• Fix search + fix all metadata on Spotify

• Create new Spotify app from scratch
  (like the short-lived Spotify Stations)

• Acquire and re-brand a classical music streaming service”
  (e.g. Idagio, Primephonic)

• Do nothing

It became clear that the options that would require overhauling Spotify’s metadata would likely have the lowest ROI, since they’d only benefit classical listeners (a small minority of users). But I was wary of my biases for action — I needed a way to better compare options.

Comparing solutions

Prioritization matrices winnow solutions

To fairly compare all the options, I created a series of prioritization matrices to focus on better choices based on

• 💰 Cost

• 🛠️ Effort

• 💎 User value

• 🍎🍊 Market differentiation

• 📈 ROI

With the help of some developer and business analyst friends, I made a series of broad assumptions and guesses, especially regarding cost and effort.

• For example, I assumed it would be easier to buy Idagio and adapt it than build a new app from scratch — but it’s hard to know if there would be a huge number of legal or technical issues.

Points were assigned to options according to which quadrant they fell on the matrix and then added to provide numerical comparisons.

And the winner is . . .

Design recommendations

Champion the listeners

All the options that involved Spotify making changes to its metadata and user interface scored very poorly and were deemed not worth the effort/cost.

So after all this time and effort, it came down to the two most plausible options:

Option 1 — ❌ Do Nothing

In the end, the disheartening but maybe predictable best choice for Spotify from a business — though not UX — standpoint was keep Spotify as it was. Do nothing.

However, as I considered myself, being a UX designer, a champion for Spotify’s classical listeners, doing nothing felt unacceptable to me — and a lost business opportunity.

Instead, my final design recommendation was for the second-most likely scenario:

Option 2 — ✅ Adapt an existing app

⚠️ Acquire and re-brand an existing classical music streaming app — like IDAGIO or Primephonic — to help drive premium subscriptions

Strategy moving forward

Under this adapt strategy, Spotify could use the new classical-only app to encourage new subscriptions, plus incentivize current “freemium” or basic plan members to upgrade to a premium level.

As part of this strategy, I envisioned Spotify adding features that the acquired app (IDAGIO) didn’t have, like 🎧 podcasts, and adapting innovations it already had like 🎟️ events.

Spotify’s listeners could then have their huge classical music library and discover it, too.

Results

Both design recommendations play out

Well, it turns out I was right. Unfortunately.

😕 Now, updating this case study three years later in 2023, my original analysis proved accurate — Spotify has recently expanded into audiobooks but has still done nothing for classical listeners to improve search, browsing, and comparison.

• Do nothing (about classical) has proved an effective and profitable “solution” for Spotify; their focus is elsewhere

🔥 On the other hand, it turns out my final design recommendation actually happened.

• However, itwasn’t Spotify that bought IDAGIO and created Spotify Classical.

• But Apple Music bought Primephonic in 2021 and launched a new standalone app, Apple Music Classical, in March 2023

So the results of my research have been borne out.

📱 In terms of the Spotify Classical project, this research proved invaluable to the design phase, allowing me to focus on the key aspects of the music streaming app that was most important to classical listeners.

Next Steps — applying research

• For the next steps in the Spotify Classical project, I switched hats from 🖋️ UX researcher to 📱Product designer.

• The goal was to explore how Spotify might adapt a newly acquired IDAGIO app into Spotify Classical with the business goal of driving premium subscriptions

• I would rely on and return to the insights gained in the UX research done in this case study to drive design decisions during design, testing, and prototyping

Key Takeaways

• 🔑 Stakeholder involvement is key — I needed stakeholders to provide important insights, guidelines and data to inform the research (and, later, the design). Without access to business leaders, developers, designers, researchers, and marketing experts, I was forced to make a lot of assumptions and guesses.

• 🧶 Mind the scope — A UX newb at the time, I tried to tackle a huge problem and it’s no wonder the project became unwieldy and huge. In the future, I would learn to chunk problems better.

• 📊 Synthesizing large amounts of data — I spent a great deal of time organizing user feedback into the affinity diagram. And while I uncovered valuable insights, next time I would likely not spend so much time on it.

Toolstack

• Google Forms (surveys)

• Google Sheets (analysis)

• LucidChart (user research, analysis)

Other case studies from the Spotify Classical project:

📱 Spotify Classical: Product Design

🕶️ Spotify Classical: Accessibility

🎧 Spotify Classical: Podcasts

🎟️ Spotify Classical: Events

^ Back to top

> More projects