Music Is Too Important To Be Left To The Marketplace

A Conversation with Liz Pelly

During the recent book tour for Mood Machine: The Rise of Spotify and the Costs of the Perfect Playlist, author Liz Pelly and I had a public conversation on stage at SPACE in Portland, Maine. A transcript follows.

photo by Naomi Yang

Liz started by reading a few selections from the book, including the following:

I’ve long been confounded by the expectation that we simply accept the dealings of the powerful as unexplainable. It’s a perspective informed by my background in DIY music spaces, where we try to live the reality that there are other ways to think about collective culture. Grassroots music communities often work directly to demystify the process of making, distributing, and sharing art. If you show up to a gig an hour before doors, someone might ask if you can help set up chairs or run to the store for some ice. At the end of the show, it’s not uncommon to discuss what each band was paid and whether the splits were fair. When it comes to recordings and labels, transparency is held as a foundational ideal worth aspiring to. With music streaming, it was jarring watching independent artists become convinced to accept a new system that no one could understand.

In my writing and reporting, I’ve been driven by a deep impulse toward demystification — toward shedding light on the inner workings of streaming companies and debunking the myths they perpetuate. Sometimes it feels more complicated and convoluted than I could have ever imagined. Other times it all just feels like music industry business as usual. The truth is somewhere in the middle: the story of streaming is as much about what’s changed as it is about what’s stayed the same.

Damon Krukowski I am so glad you read the passage where you talk about how demystification is important to you. I love this book for that reason — because you are demystifying this platform that has thrived on mystification. I mean, the amount of effort that these tech companies put into mystifying what they do almost seems proportionate to their success in some ways. Your book demystifies what is going on with Spotify and its playlists, in part because you managed to get informants from within the company, which nobody else had done. Even though we all knew what trouble this company was making for us, the book reveals why, and also how deliberate it really is.

I want to read another few sentences from the introduction which speak to this, and to so much of what we're going through right now:

"In the story of Spotify is the story of a broken music industry desperate to keep existing after the era of digitally enabled file sharing. Of ad-tech executives bringing the logic of their industry to music in new ways. Of an already highly consolidated industry growing ever more consolidated. It is the story of the 21^st-century's overeager and opportunistic tech solutionists, of billionaires and their overhyped machines, looking around for problems to solve, arrogantly disregarding the social problems left in their wake."

I feel like we're all living this exact storyline. So what I was curious about, Liz, is what do you think the problem was that Spotify thought they were solving?

Liz Pelly If I'm to really try to put myself in the headspace of someone like Daniel Ek in 2005 when he first pitched this idea to his co-founder, Martin Lorentzon — which is a hard thing to do because I don't think anyone in this room thinks the way someone like Daniel Ek does, actually — I feel maybe it's that they recognized that, from file sharing, people had gotten used to the experience of being able to open their computers and search for any song that came to mind and hear it. But the major record labels weren't going to allow that to continue happening. So it was like, okay, how can we figure out a way to capitalize on these user habits while also making the major labels happy? Which is not the same thing as figuring out: how do we use technology to make it easier for people to support artists, get money into the pockets of the people whose work they are listening to. It's just, how do we make the major labels happy while keeping this user experience?

They were from the ad-tech industry, so I think it was also sort of like, how do we take what we've learned from our experience selling banner ads in an automated way, and apply it to this online experience that's attracting really big audiences right now.

Long before Spotify, the major labels had tried to start their own streaming services. Early on in the book, I quote from this article in the Times covering these early streaming services, where a bunch of people who represented artists were talking about how it felt like they were being made victims of a huge conspiracy, because it seemed obvious even at the beginning that streaming was probably a way to capitalize for themselves on a user experience that people - customers, consumers, internet users - were clearly interested in. Which is different from figuring out how to get revenue to artists.

Damon Krukowski That’s a good explanation of what their mindset would be, which is the problem to solve is a problem that the labels were facing, specifically the major labels, but not the content providers of this entire industry. There’s no part for the musician in that equation. And then as the section you read with quotes from Anohni and from Darius Van Arman makes clear, it also didn't include what independent labels had already figured out. For smaller labels or artists who were speaking to a smaller segment of the listening audience, there was a functioning industry. That's the thing that happens, I feel, at the moment when Spotify launches, is they actually destroy a functioning industry for a lot of us, for spaces like this one we’re in — although thank God it’s finding ways to survive — and for bands like mine, and for smaller labels, even ones as big as Jagjaguwar and Secretly Canadian. We didn't necessarily have the problem they identified, do you know what I mean?

Before streaming, Anohni saw more opportunities for non-pop music. “I’m not saying it was a noble economy, but there was more of an ecology to it. There were so many ways of approaching listening to a song. There was an innate integrity to paying a person for a record, whether you listened to it five times or five thousand times. Sometimes a record only needed to be listened to five times in order for a profound transaction to have taken place.”

Shortly after speaking with Anohni, I met up with Darius Van Arman, one of the cofounders of Secretly Group, a big independent label group that includes Anohni’s label, Secretly Canadian, and the label Van Arman started in 1996, Jagjaguwar. These days, Secretly is one of the most financially successful independent label groups, releasing records by the likes of Mitski, Phoebe Bridgers, and Bon Iver. But when he started, Van Arman was releasing considerably more obscure music. One of the first albums that his label ever released was by the Brooklyn noise band Oneida. “It was good business to work with Oneida, who made challenging, beautiful records and were a great live band,” he said. The label could sell a couple thousand CDs, cover the manufacturing and promotion costs, and have $25,000 left to split with the band. No one was quitting their jobs, but it kept them all going. People would buy the CDs at shows, or because they read a good review, and even if they only listened to it once and put it on a shelf, it had value.

How about now? “Monetization is shaped differently,” Van Arman told me. “It’s based on what gets repeat listens. It didn’t take long for artists and labels to make that connection. It’s not sustainable to put out challenging records. To be sustainable, you have put out records that are going to get repeat listens in coffee shops. That people are going to want to listen to over and over again, and that are going to be playlist friendly and easier on the ears.”

Liz Pelly I tried to do some research on the relationship between the independent music world and the digital music market in the years before streaming became normalized. In 2009, a couple years before Spotify launched in the United States, there was a conference held by The Future of Music Coalition, which is this DC-based advocacy group that basically advocates on behalf of independent musicians. All of the panels from the 2009 Future of Music Coalition conference are on YouTube. I found it to be really fascinating to go look at the conversations that were happening at that time about how independent musicians specifically were finding ways to make the pre-streaming digital situation work. At that time, independent musicians were monetizing their work online by selling MP3s, (and from what I’m told, also by selling ringtones). There was this interesting chart showing that actually for independent artists, they were starting to see growth again in their sales. So it's just interesting to think like, okay, yeah, file sharing was a problem for the major labels, but actually, were independent musicians finding ways to make downloads work — until streaming came along and then convinced independent music fans to stop paying for them?

Damon Krukowski Yeah, I think that's exactly right — billionaires creating problems to solve, but not solving problems that we have, which is a big difference. Because what you're left with is their weak solution to a problem that they defined. And somehow we're all sucked into this world where we're judging how well they solved this problem that we didn't necessarily share in the first place.

There are a number of things that Spotify employees say in the book that are kind of bombshells. For example, the section that was excerpted in Harper's about "Perfect Fit Content” — to use the phrase you reveal from Spotify’s internal communications — and the way it is designed to shape your listening habits in a very deliberate, manipulative manner.

There's another remarkable statement in the book where one ex-employee says to you in an interview, "I definitely think people are afraid of silence… And Spotify has capitalized on that pretty well." This comes in turn from something that was said at a company meeting by Daniel Ek himself, which is, "Apple Music, Amazon, these aren't our competitors. Our only competitor is silence." Again, it’s like, okay, so what world were they living in when they started thinking about music in the first place? "Our only competitor is silence." Do you think they've won that competition?

Liz Pelly When I hear that quote, I think of some other marketing materials that I've seen Spotify put out over the years where they're sort of putting forth this idea of their ideal user. If you look at what’s now called Spotify Advertising — it used to be called Spotify for Brands — there are all of these reports that you can find online. And you can find the way that they talk about their company and how they're revolutionizing music, but it's aimed towards advertisers. They'll talk about this supposed average Spotify user's daily journey, someone who wakes up and puts on their wake-up soundscape and then puts on their morning coffee podcast and then they listen to some workout music and on the way to work they're listening to their morning commute podcast and they listen to focus music all day at work and then after work they have a podcast that they listen to on their commute home and then they have a playlist that they listen to every day while they're making dinner and then after dinner they have a wind-down soundscape podcast they listen to while they're doing yoga and then they fall asleep listening to a sleep playlist. They want you to have audio on all day and seemingly without any real concern for what that audio is. Just producing data. I think also part of the perspective is: if we can convince our users to have our app on all day, then they'll see it as a really ingrained part of their daily routine and then we reduce the risk of losing them as customers. A lot of the strategic moves are formulated around risk management, like reducing the risk of losing you as a subscriber because that's how they make most of their money, it’s from subscriptions.

Damon Krukowski So silence becomes a kind of nay-saying to the whole system, which would be powerful except that musicians want to make music, not silence necessarily. It's such a strange dilemma. It's like, what is the negation of this world that they've envisioned for us to live in?

Liz Pelly It's funny. As a musician, if someone was like, oh yeah, I love your music, I keep it on repeat all day, I never turn it off. You'd be like, are you OK?

Damon Krukowski You need to chill out.

Liz Pelly It kind of reflects the more general idea in the streaming era that replayability, or how many times you stream something, equates to the value of it. If this app is on in the background all day, then we can see it as valuable to the customer based on how many minutes they spent listening. In the same way, a lot of the algorithmic recommendation systems, not just with Spotify, are focused on this metric of session length. How can we extend session length? Time spent listening or time spent dwelling on something is seen as the ultimate marker of whether or not the recommendation was successful. I think the idea of a user who is just streaming from the moment they wake up to the moment they go to sleep is, by those terms, the ultimate success.

Damon Krukowski It strikes me that this is another way that they've devalued music or devalued any specific content, because once it's turned into… I think "emotional wallpaper" is the phrase in the book. Was that from one of their employees? Or is it your phrase?

Liz Pelly Somebody who worked at an independent record label used that phrase in an early interview to describe how they felt, like streaming playlists were turning all music into “emotional wallpaper.”

Damon Krukowski Even if you go out to a venue like this one, you don't really want the band to not stop playing. It would be a problem. And at some point, if the band doesn't stop playing, you're going to leave because you will decide that the value has hit zero for you. You're going to walk out and seek the silence that Spotify is so afraid you're going to find one day.

There's a lot of marketing language and promotional language from Spotify about accessibility — making music accessible, making it part of everybody's life, how people are listening to more music than they ever did before, there's now more music online than ever, all that kind of stuff. But of course, the other side of ubiquity is valuelessness.

Liz Pelly I've heard some variation of this from a few different people while having conversations around the book — like, how can you say Spotify has made music discovery less meaningful when I discovered 5,000 new artists last year with my Spotify wrap? I mean, who actually discovers 5,000 artists in a year? Can you name some of them?

Damon Krukowski There’s this loss of specificity for what people are listening to. I'm always asking people what they're listening to, if I hear something in a coffee shop or wherever. And unfortunately now the answer is usually they don't know. Or they look at their phone and they're like, oh, we started with whatever — you know, they tell me the prompt they typed in at eight in the morning when they came to work. Even at the dentist's office, which was such a classic place to encounter a different type of music than I listen to. It’s a loss of value to any given track or any given artist or album.

So in the book you take us in great detail through exactly how bad this is. And I feel like many of us in music knew it was bad, but then we read your book and we're like, oh my god, it's really bad. It's so much worse than we realized. I don't know if you feel that way yourself. Do you feel like you brought the even worse news?

Liz Pelly I feel like I kind of lost perspective at a certain point, but I will say that it was a pretty depressing book to write. It was really sad at a lot of times. Because this is a really depraved way of thinking about culture and music, and to be immersed in that way of thinking about music for so many years… People are sometimes like, oh, it must have been really frustrating to write this book. But actually it was just kind of sad, you know?

Damon Krukowski Yes, because I think it is about the mindset. I feel that's what you really revealed. We all understood what the system was doing in its effect. But you reveal the thought that went into creating this. It didn’t happen by accident. There are a lot of very smart people, very energetic people, and very well-funded people creating this system. And they're very happy with it.

So there's this great passage… I'll read another bit from your book that to me sums this up, and also reaches outside the music world to what we're experiencing in so many parts of our lives right now:

"Spotify is not a purely musical and technological endeavor. It's a business. And if there's been any lesson of the past 15 years, it is that multi-billion dollar multinational technology corporations cannot be viewed outside of that context. There are a lot of assumptions about the neutrality of technology baked into the idea that the patterns and data gleaned from such a system would provide an honest look at the world of music listening."

Or anything else that they're measuring, right?

"In the streaming environment, personalization algorithms were employed to retain customers more than anything else. And when the mode of listening or hearing is so deeply shaped by economic incentives, what can you actually get from the data? In the end, it's not data about listening to music. It's data about listening to Spotify."

So the system is self-fulfilling and self-perpetuating. And the data that they're gathering, again it's solving a problem that isn't there. We didn't need data about ourselves as listeners — but they did.

Liz Pelly Yeah, I remember a few years ago having one of these conversations about, well, what would it look like to build a streaming service that's more equitable or fair or interesting. And I remember this idea coming up of, like, what if they just didn't collect data that they don't need to collect? Or collect the data you need to pay people, and then delete it? Libraries routinely delete data that they collect in order to function. You don't have to keep it forever. But streaming services – especially Spotify, there's a whole chapter in my book about all the data that they collect on users and it's essentially every single thing you've ever clicked on, everything you've every typed, every playlist you ever made, every playlist title. Every song you've listened to, how you got to that song, how long you listen to it, when you hit skip. You know, what does this actually say? And I think in that passage, something that I'm trying to reckon with is that oftentimes, especially in the world of recommender systems and algorithmic music recognition, there's this idea that, oh, it's actually interesting to look at the patterns, like there's all this data about global music listening on an unprecedented scale. I think as music journalists, if you had the ability to interview every single person in the world about their music listening habits, and to make this incredible database of the world's music listening habits and then try to glean insights about the state of music or the state of listening. Conceptually that would be interesting. Some people like to look at all the data that Spotify has and suggest that it is just that — as if we have interviewed 600 million people about their listening habits and can make all these charts and graphs and maps. But when the data you have is so shaped by economic incentives and algorithms, when people are nudged towards material that will boost engagement, extend session times, keep you on the platform, reduce the risk of canceled subscriptions. And then also stuff like Discovery Mode, algorithmic promotion technology, commercial deals, marketing, deals with labels that determine what might be on the front page when you open the app, you know. There's all of these different ways in which commercial deals and marketing partnerships are shaping that data. It's not a pure source of information about the world's listening.

Damon Krukowski I think they actually don't want to know what we want to listen to.

Liz Pelly Yeah.

Damon Krukowski But they want to know how to manipulate us so that we will stay on the platform. It’s like if you gather data about what people want to drink, and the answer is magically Coca-Cola. That’s not helpful information if you're trying not to drink Coca-Cola. Or even if you just want to find out what people actually like to drink.

I feel like we're living more and more in a political and economic environment that takes this same presumption about data. Even polls, it's like, well, are you really asking people what they believe or what they think, or are you gathering data about data gathering, you know? Is the incentive to learn about people, or to learn how to maintain control over people by gathering data?

I want to steer the conversation to your concluding chapter, which is wonderful and has some extremely positive and hopeful things in it. Because the same way you start off talking about your life in a DIY venue, which is itself a very hopeful gesture, I feel like you don't lose that, for all the darkness that you reveal to us.

Liz Pelly Some people have said that the book left them feeling hopeful.

Damon Krukowski I think it should. So I want to spend a little time talking about what we can do, what can come next. I'll quote from your conclusion: "The problems created by venture capital are never going to be solved by more venture capital." This is I think another crucial statement. And again, it’s opposite to what we're experiencing in so much of our technological lives. So what are the solutions that we can look at? You list a few in the conclusion, and I thought maybe we could talk about those a little more. Like you mention Resonate.

Liz Pelly Resonate doesn't exist anymore, it's since gone on indefinite hiatus. But it was a cooperative, and as far as I know, it was the first of the music streaming cooperatives. It started in 2015, first in Berlin, then it became international, and it was a group of musicians and fans and industry people saying, what if we had a cooperative streaming service that was run by the members. One member, one share, one vote. The musicians who joined became a member of a cooperative, and could go to their annual meeting to be involved in conversations about the direction of the co-op. It also had this really interesting stream-to-own model. One of the reasons why I wanted to interview someone from Resonate in the book is because there are a lot of learnings and lessons from that project. And one of them is stream-to-own, this model where instead of the per stream, pro-rata model that has been so destructive, it was basically a version of user-centric where you would tap up your account, say you’d add $10, and then it was like a specific rate per stream and then once you had streamed around 99 cents worth of streams on a track, you owned it and you could download the mp3. I thought that was super interesting because I'm a fan of buying music, but some people prefer streaming, so it’s kind of a middle ground between those models. But yeah RIP Resonate, I don't think it exists anymore.

Damon Krukowski It didn't work, but there are other companies starting up sort of like that. Resonate was for profit, right?

Liz Pelly It was a cooperative. The idea being that there’s democratic governance to benefit the members. Typically if there were a surplus of revenue at the end of the year, the members might vote on what to do with it.

Damon Krukowski There are some other companies that have started up or are trying to start up — some are for profit, some are non-profit — similarly trying to allow consumers to use streaming in the way that we've become accustomed to, yet look for another way that the money could flow and connect it more to creators. Do you feel this is a positive direction? Do you think someone will get it right?

Liz Pelly I do think it's really interesting that there are a handful of different people working on cooperative alternatives to streaming right now. These are not necessarily like public endorsements of any of these projects because I'm still in the process of researching them, but a few of the ones that are cooperatives are Tone.audio, Subvert.fm, and Mirlo. A couple of them are billed more as alternatives to Bandcamp. I think it's helpful that there are so many different people who are working on emerging cooperative digital infrastructure for music.

Damon Krukowski It feels like an idea that's in the air, and people are trying to get an actual functioning version of it off the ground. But it does seem difficult to make that jump, partly because of problems of scale, which you raise in the book too. Because people expect such a huge scale.

There's also — I'll add to your list — there's a non-profit one starting up called Campfire. And you list one that is not trying to scale up, which is a collective called Catalytic Sound. That's an interesting alternative.

Liz Pelly Yes, and the reason why I spend like five or six pages in the book talking about Catalytic Sound is not to suggest that it's the perfect answer, but it's one answer and an example of artists creating their own digital infrastructure that serves their communities. And I think it's a model that could easily be repeated. It's group of 30 musicians from the more avant-garde free jazz scene who came together to make a streaming service just for recordings by their members plus some labels in their broader community. And the idea is instead of paying artists per stream or by track, you can subscribe to get access to the streaming website. And the revenue from subscriptions gets divided evenly 30 ways every month. I think that is really interesting because it points to a way of monetizing streaming that detaches from the per stream valuation of music. They all make some music every year that's exclusive to this one project to incentivize people to subscribe to it. And the technology itself is really bare bones, it's not a super technologically complicated project. It's browser-based, you subscribe to it, you could access it similarly to how you could access Bandcamp in your browser. I like how it highlights that a lot of the problems or issues aren't necessarily issues of technology. In fact it's not that hard to create a streaming service — to have a bunch of music stored on a central server and have people stream it through the internet. What's complicated, more complicated, is figuring out the needs of specific music scenes and ways of connecting with audiences and fan bases and communities. It's not a technological solution. It's a social and economic reimagining.

Damon Krukowski I wonder if the technology that runs Spotify is not so complicated, either. It does seem like a relatively simple process.

Liz Pelly I can give you another example of how simple it can be to start a streaming server. A couple of months ago I did an event in Providence at this DIY venue called Lost Bag and a few days before the event, Mike Sugarman, who was hosting the event, contacted me. He said, “I have this extra server at my house, what if I brought it? A couple days before the event, we could just post online and ask people from Providence if they want to send in some tracks by local artists, and I can just real quick code a streaming interface, and at the event, the house music can just be playing from this local music streaming service of all Providence artists.” And I obviously said yeah that sounds cool. And then he did it, and now it's permanently at that venue. There's a local music streaming server in the corner filled with tracks by local artists, and when you're in the venue there's a piece of paper on the wall that says type this into your browser, enter this code, and then you can access this browser-based streaming service of all local music.

Damon Krukowski That points to another thing about Spotify, which is that mystifying the technology is a part of what they've done. It’s the opposite of DIY. They make it seem like nobody else could possibly make this magic happen. But actually streaming is all over the place online.

Liz Pelly I spend ten pages in the book talking about public libraries that have created local music streaming programs for local artists. Library card holders and just the public in general can access these collections of local music streaming through their websites. And those are also really technologically uncomplicated programs. It's more about libraries and local institutions being willing to do the work of organizing it, and getting the budget to pay musicians to license their music for these collections. It's actually not that big of an investment that has to happen in order for these small local collections to take hold, which is part of why I point to it at the end. This is something that is already happening across dozens of libraries in the United States, Canada, and probably other places too.

Damon Krukowski The library model is remarkable, because as you say, it's locally based, as public libraries are — but you also make this excellent point, which I want to read, about streaming through libraries: "Like libraries, music benefits from solid positioning as a public good." That’s so important. Were we to readjust how we're thinking about music and say it's a public good, it's not just something to exploit the way that the corporations are looking at it all the time, but it's a public good. And then as you said its delivery becomes not a technological problem to solve or necessarily even a financial one, it's more about defining community and serving community, which is what these libraries are doing. You should all check if it's happening here in Portland, it may well be. There's a program called MUSICAT that a lot of libraries are using as the back end for various approaches to what they stream.

Liz Pelly Yeah, so each of these libraries sort of runs it differently, but there's information and resources that have been shared across libraries, so some of them run their programs pretty similarly. For the most part, it's a situation where anyone can stream the tracks, but you have to log in with your library card to download them.

Damon Krukowski The group I'm involved with, UMAW, United Musicians and Allied Workers, we're like a proto-union of mostly independent artists coming together to define what we feel could work better in the music industry for creators. And we've been working very hard pushing back on Spotify and to be paid more by the corporations that are exploiting our work. But many of us at UMAW are also very interested in this idea of streaming through libraries. And I think you cut to the heart of why when you talk about it as a public good. Because I think all musicians understand that our work and the work that we enjoy by others is not just a commodity.

Liz Pelly A sentence from the conclusion that is just like ingrained in me now is where I write that “music is too important to be left to the marketplace.” I think all of the work being done to make streaming services less exploitative, and to make the music business less exploitative, is really important. But we also need to just try to minimize the reach of corporate power more broadly. It feels like there also needs to be an acknowledgement that we rely way too much on the corporate music business to look after music as an art form. There's no incentive for streaming services to have any sort of archival goals or preservation perspective on music. That work really requires looking outside of private companies.

Damon Krukowski Which is another reason why libraries are a counterbalance to what the corporate approach to music is or has become. Because there’s a value in preserving things that people don't necessarily even want to hear right now. I mean, that's the point of the Library of Congress, that's the point of so much of our system that looks at information as a public good. But if you look at information as these tech companies do, which is information as power, information as money, information as profit, you're not going to hit accidentally on the idea that we should preserve it as a public good.

That may be a good place to pause and take questions, but I wanted to read one more statement to give us something to think about from Liz's conclusion, and how it leaves us in a hopeful place. So Liz writes – and I think this is a key paragraph to the book:

"On a collective level, we have to be active participants in the cultural economies we want to see flourish. We have to validate the culture we want to see in the world. The corporate culture industry entrenches its power, not just through controlling the marketplace, but also by controlling the popular imagination, by convincing us that there are no alternatives. The alternatives are growing all around us, though."

That’s your positive DIY sensibility there. People are already doing it. It doesn't mean that they can defeat Spotify and all the corporate powers that be, especially with so much political power being wielded against so many of us right now. But the idea of expanding the possibility — cause there's no way the marketplace can control our imagination, right? I mean, try as they might. And I think that's to me why Spotify is doomed, because in the end that's their goal and there's no way they can win that battle. They can make us all broke, but they can't stop us from thinking of the alternative.

So with that, shall we take some questions?

[applause]

Liz Pelly Thank you.

Listening to: Cancionera by Natalia Lafourcade

Cooking: ojo de cabra beans