Non-music companies know more about music than musicians do
ANOUK DYUSSEMBAYEVA | SEPTEMBER, 17 / 2020
IMPORTANT TECHSTARS MUSIC LINKS: APPLY FOR TECHSTARS MUSIC 2021 // RSVP HERE FOR TECHSTARS MUSIC 2021 RECRUITING EVENTS // RVSP HERE FOR TECH TOWN HALL WITH PELOTON & WARNER MUSIC GROUP
Talking to Bob Moczydlowsky, the Managing Director at Techstars Music, we discussed the future of the music industry, startups, and why the people that dominate the culture are from outside of the system.
Here is the second part of our interview, exploring the possibilities of predicting what type of music people need, why non-music companies know more about music than musicians do, and what opportunities are opening up for vertical marketplaces.
Here is the second part of our interview, exploring the possibilities of predicting what type of music people need, why non-music companies know more about music than musicians do, and what opportunities are opening up for vertical marketplaces.
I've been reading a couple of interviews and I saw that a lot of the interviewers were quite interested as to why Peloton recently joined, but there was a great article on Andreesen Horowitz about Platforms vs Verticals, and that the future basically belongs to verticals. I feel that while many people thought it was interesting that Peloton joined, it comes as no surprise, because all of their content involves music …
Literally 18 hours a day, music, all of it. Every single thing they do. Everytime we touch the product: music.
… Peloton will soon know more about music than musicians will know about music …
They already do — they have internal tools that can already tell you which music performs better, which music makes people go to classes, what music you're listening to when you set your personal records. They know an enormous amount about what's going on.
… and they have much more insight than musicians do. So if a musician posts their music, they have no idea how much traction their music gets, what people do after they listen to their music. Peloton has this unique access to that information
So does Spotify, Apple, and Amazon. They have different access and different touch points at different times.
The music business in general has always been wholesalers. They make a product and then they sell it to somebody else, who then sells it to the customer. Spotify, Apple, and Amazon have effectively kept the music business as wholesalers. [This model] is okay if you are a record label because what your job is to get as much of the revenue from Spotify as possible. Spotify is paying 70c out of every dollar out to rights holders.
If you're an individual musician though, you're just one slice of that stack, [and] the more people are taking a piece of your income in that stack, the less you're left with at the bottom and the farther away from your customers you are. Your job as a musician is to get as close to your customers as possible. The path to a lot of customers is through a label or Spotify or Apple or Amazon, but the path to that direct relationship is through the internet.
Peloton is one small example where they're skimming the music off the top — they're not playing classical concertos or undiscovered music — they're taking the hits, the next layer below, and they're figuring out functional uses for that music. What they know though, is that, and this is often true in a lot of ways, the functional use of a piece of music creates way more economic value than the primary use of that music in terms of [people] want[ing] to just listen.
The song from the 1980s Eye of the Tiger. Super cheesy, big pop song. Originally released, it was a serious album, [created] to play on the rock-and-roll radio. That song, functionally through sync licensing and through exercise, has become like an ironic joke and proxy for a certain specific kind of moment. That song is a hundred times more valuable because of all of its cheesy, goofy, ironic, comedic, exercise-style uses than the original intent of the song. And there's a bunch of songs that are like that.
Peloton has an unfair advantage, in that specific vertical, over what are the songs that deliver real functional value. [Therefore], they can afford to get more of them and create opportunities for them, but they can also do things with those songs that give them leverage with the artists and the labels around what it costs them to access those songs.
The music business in general has always been wholesalers. They make a product and then they sell it to somebody else, who then sells it to the customer. Spotify, Apple, and Amazon have effectively kept the music business as wholesalers. [This model] is okay if you are a record label because what your job is to get as much of the revenue from Spotify as possible. Spotify is paying 70c out of every dollar out to rights holders.
If you're an individual musician though, you're just one slice of that stack, [and] the more people are taking a piece of your income in that stack, the less you're left with at the bottom and the farther away from your customers you are. Your job as a musician is to get as close to your customers as possible. The path to a lot of customers is through a label or Spotify or Apple or Amazon, but the path to that direct relationship is through the internet.
Peloton is one small example where they're skimming the music off the top — they're not playing classical concertos or undiscovered music — they're taking the hits, the next layer below, and they're figuring out functional uses for that music. What they know though, is that, and this is often true in a lot of ways, the functional use of a piece of music creates way more economic value than the primary use of that music in terms of [people] want[ing] to just listen.
The song from the 1980s Eye of the Tiger. Super cheesy, big pop song. Originally released, it was a serious album, [created] to play on the rock-and-roll radio. That song, functionally through sync licensing and through exercise, has become like an ironic joke and proxy for a certain specific kind of moment. That song is a hundred times more valuable because of all of its cheesy, goofy, ironic, comedic, exercise-style uses than the original intent of the song. And there's a bunch of songs that are like that.
Peloton has an unfair advantage, in that specific vertical, over what are the songs that deliver real functional value. [Therefore], they can afford to get more of them and create opportunities for them, but they can also do things with those songs that give them leverage with the artists and the labels around what it costs them to access those songs.
It is quite similar to UberEats: they know the demand on certain foods and where the unmet demand is, but the restaurants don't know that. Peloton also knows a lot more about music for fitness than musicians do, because musicians don't have access to that information. Do you think we can predict what music is needed for people and how, in your opinion, will music be produced and made in the future?
Good question. Yes, I think we can. Even from my mediocre music nerd perspective, if I spend two or three hours with somebody, I can recommend music that that person would like. If you gave me access to your Spotify account, I could recommend jeans, sneakers, movies, and books to you, just because I consume a lot of culture as a human — it's a natural thing we do as humans. That will eventually become algorithmic, and the algorithms are already getting really good at that.
In 2008-2009, I worked on products that had recommendation engines inside of them and they used to drive me crazy. I would listen to the Buzzcocks and then it would recommend Green Day to me. That's not how it works. I know Green Day exists, [and] I've obviously chosen not to listen to them if I'm listening to the Buzzcocks, but the algorithm couldn't deal with that. That problem has been solved. If you go into Spotify and start listening to the Buzzcocks, it is going to recommend you Stiff Little Fingers, The Futureheads — it's going to give you the things that are in that lane.
That will get better and better, and when you combine [all the below-the-line features of music being taken over by software] with the desire to connect with people, the music will get made for specific purposes. I think music will start being made on the fly or in very short cycles where the music itself doesn't have to last forever.
It could be that you and I are having an experience. We're going to go for a hike, and because of what the algorithm knows about you and what the algorithm knows about me, it can generate music in that scenario that is perfect for the combination of people, our surroundings, and the feedback we're giving it, but it doesn't necessarily have to record it and keep that information permanently forever.
One of our portfolio companies is called Popgun, it's based in Brisbane — they are by far the world's best AI music engine, but that's not the product of the company. What they've realized over the last three years is that kids don't want to make music. The objective of their user is not to make a song, the objective of their user is to perform and have an audience attached to them: the objective is to get attention.
One of the things that I need so that I can perform and get attention is some music, and I need to assemble it in an awesome cool-sounding way so that people will listen and be attracted to me, but it's not a primary asset. The asset is the audience. I think music is going to be made on the fly and in an interesting combination of ways for that purpose. We've only thought about the music business in terms of these permanent fixed pieces of artwork that last forever and have to be assets. When it's my friend's birthday, I can make a song for my friend's birthday. The objective is to make my friend happy; it's not to make a song that goes and competes with 99 Problems.
If hundreds of millions of people around the world are making songs on the fly to communicate with each other, perform, and get attention, that will lower the value of existing songs, because you can only listen to one song at a time. You have a finite amount of attention every day that you're spending: you're giving some to Facebook, you're giving some to Apple, you're giving some to newspapers, the New York Times, or wherever you get your news from, you're giving some to Wired, you're giving some to blogs, you're giving some to Instagram, but there's only so much of it. That's the asset that everybody's fighting for.
In 2008-2009, I worked on products that had recommendation engines inside of them and they used to drive me crazy. I would listen to the Buzzcocks and then it would recommend Green Day to me. That's not how it works. I know Green Day exists, [and] I've obviously chosen not to listen to them if I'm listening to the Buzzcocks, but the algorithm couldn't deal with that. That problem has been solved. If you go into Spotify and start listening to the Buzzcocks, it is going to recommend you Stiff Little Fingers, The Futureheads — it's going to give you the things that are in that lane.
That will get better and better, and when you combine [all the below-the-line features of music being taken over by software] with the desire to connect with people, the music will get made for specific purposes. I think music will start being made on the fly or in very short cycles where the music itself doesn't have to last forever.
It could be that you and I are having an experience. We're going to go for a hike, and because of what the algorithm knows about you and what the algorithm knows about me, it can generate music in that scenario that is perfect for the combination of people, our surroundings, and the feedback we're giving it, but it doesn't necessarily have to record it and keep that information permanently forever.
One of our portfolio companies is called Popgun, it's based in Brisbane — they are by far the world's best AI music engine, but that's not the product of the company. What they've realized over the last three years is that kids don't want to make music. The objective of their user is not to make a song, the objective of their user is to perform and have an audience attached to them: the objective is to get attention.
One of the things that I need so that I can perform and get attention is some music, and I need to assemble it in an awesome cool-sounding way so that people will listen and be attracted to me, but it's not a primary asset. The asset is the audience. I think music is going to be made on the fly and in an interesting combination of ways for that purpose. We've only thought about the music business in terms of these permanent fixed pieces of artwork that last forever and have to be assets. When it's my friend's birthday, I can make a song for my friend's birthday. The objective is to make my friend happy; it's not to make a song that goes and competes with 99 Problems.
If hundreds of millions of people around the world are making songs on the fly to communicate with each other, perform, and get attention, that will lower the value of existing songs, because you can only listen to one song at a time. You have a finite amount of attention every day that you're spending: you're giving some to Facebook, you're giving some to Apple, you're giving some to newspapers, the New York Times, or wherever you get your news from, you're giving some to Wired, you're giving some to blogs, you're giving some to Instagram, but there's only so much of it. That's the asset that everybody's fighting for.
I feel like that would change the way people produce music so much. And speaking of verticals — we see that marketplaces move from being horizontal to being vertical. What opportunities for vertical marketplaces do you see in the music industry?
Oh my gosh. Thesis-wise I think this is still an opportunity, though I am now uncertain how long it will take for it to become true — I think there's going to be a marketplace for experiences which are not concerts, movies, [but] interactive experiences where I can both create content and consume content at the same time.
In Paris, there is L'Atelier des Lumières. You buy a ticket and it's a bunch of projection mount classic artworks, high resolution images, scans, and sets of music, and you go in and you're totally surrounded by it: it's like a giant warehouse full of cool projection mounts. Very other-wordly in the way it makes you feel. In New Mexico, there's Meow Wolf, which is like a crazy museum, where you crawl through the artwork and you end up feeling like a little ant. Modern art museums around the world have created infinity rooms — you see a bunch of Instagram influencers taking their picture and it's nothing but stars and mirrors behind them — that create these experiences. [This] is an incredible opportunity to have them be very music centric, because music and visuals combined together is such a powerful thing.
I also really like the way Splice is working: the marketplace for parts of songs or sounds is really interesting, [although] I think it's kind of temporary. Something like the Popgun engine grows up and the whole planet has the ability to make whole songs anytime they want, parts of songs and those assets individually become less valuable. But right now, a good portion of the top 100 songs on the planet have parts in them and samples that were purchased in the Splice marketplace. The pre-market components of songs as [a] marketplace is super fascinating to me.
The other thing, [and] we've already made a couple of investments about [it], is the marketplace for partial ownership of rights. There are more copyrights being created today than ever before. There are more people getting a little of YouTube ad-share revenue, getting some SoundCloud money, or getting some Spotify income, where there's enough attention to these pieces of artwork [and] not enough capital necessarily for them to have a liquidity event where they get a loathsome of a large amount of capital to go onto the next level. [At the same time], there's a predictable amount of attention where you could make advances against that, or sell a portion of that income upfront for a larger amount of capital that you could reinvest in your business and grow forward.
I think there are going to be pretty vibrant marketplaces for future earnings. If you were to win a lottery, and be paid $1 million dollars a year for the next fifty years of your life, someone would come along and say: "You just won $50 million, but that's going to take you fifty years to collect it. We'll give you $35 million in cash right now in exchange for your $50 million. We're going to make a return of $15 million on our deal, but now you have $35 million all at once rather than a million dollars a year." That's going to be a vibrant marketplace that doesn't involve lawyers, managers, and paperwork. That will all be automated and musicians are going to have access to that capital in a pretty incredible way.
That'll change the way copyright deals work, too, because right now you have songwriters doing really bad publishing deals just so that they can get a $25,000 advance. Optimistically, I think that the marketplace [would] create a better and more just economy for musicians. Musicians do better with that as their capital source than they do as record labels and publishers.
We'll see. I'll probably be wrong and there will be something else, and that's how it goes. (laughs)
In Paris, there is L'Atelier des Lumières. You buy a ticket and it's a bunch of projection mount classic artworks, high resolution images, scans, and sets of music, and you go in and you're totally surrounded by it: it's like a giant warehouse full of cool projection mounts. Very other-wordly in the way it makes you feel. In New Mexico, there's Meow Wolf, which is like a crazy museum, where you crawl through the artwork and you end up feeling like a little ant. Modern art museums around the world have created infinity rooms — you see a bunch of Instagram influencers taking their picture and it's nothing but stars and mirrors behind them — that create these experiences. [This] is an incredible opportunity to have them be very music centric, because music and visuals combined together is such a powerful thing.
I also really like the way Splice is working: the marketplace for parts of songs or sounds is really interesting, [although] I think it's kind of temporary. Something like the Popgun engine grows up and the whole planet has the ability to make whole songs anytime they want, parts of songs and those assets individually become less valuable. But right now, a good portion of the top 100 songs on the planet have parts in them and samples that were purchased in the Splice marketplace. The pre-market components of songs as [a] marketplace is super fascinating to me.
The other thing, [and] we've already made a couple of investments about [it], is the marketplace for partial ownership of rights. There are more copyrights being created today than ever before. There are more people getting a little of YouTube ad-share revenue, getting some SoundCloud money, or getting some Spotify income, where there's enough attention to these pieces of artwork [and] not enough capital necessarily for them to have a liquidity event where they get a loathsome of a large amount of capital to go onto the next level. [At the same time], there's a predictable amount of attention where you could make advances against that, or sell a portion of that income upfront for a larger amount of capital that you could reinvest in your business and grow forward.
I think there are going to be pretty vibrant marketplaces for future earnings. If you were to win a lottery, and be paid $1 million dollars a year for the next fifty years of your life, someone would come along and say: "You just won $50 million, but that's going to take you fifty years to collect it. We'll give you $35 million in cash right now in exchange for your $50 million. We're going to make a return of $15 million on our deal, but now you have $35 million all at once rather than a million dollars a year." That's going to be a vibrant marketplace that doesn't involve lawyers, managers, and paperwork. That will all be automated and musicians are going to have access to that capital in a pretty incredible way.
That'll change the way copyright deals work, too, because right now you have songwriters doing really bad publishing deals just so that they can get a $25,000 advance. Optimistically, I think that the marketplace [would] create a better and more just economy for musicians. Musicians do better with that as their capital source than they do as record labels and publishers.
We'll see. I'll probably be wrong and there will be something else, and that's how it goes. (laughs)
In the music industry we have a lot of people that come with little background in music. We see a lot of new players coming in that have never been there before …
Outsiders who just connect with people and now here they are because they have attention and audience. They're not from music by training or craft, they just have attention.
… Do you think these new players are going to influence the industry more and more?
Yes, I think they're in. I think it's over. The value is in the outsiders who can create audiences that they win. Record labels don't sign artists, they sign audiences. Publishers don't sign songwriters, they sign successful track record of making hit songs. Everything is data-driven at some point and it is harder and harder to get financing to develop an artist, because the cost of reaching your audience has fallen to zero.
There are only two things that are scarce: a genuine talent to make incredible sounds that move people, and then the ability to connect with those people. There are definitely musicians who are insanely talented, make really interesting music [that] is not finding its audience. There are hit songs out there that have not had vessels to connect with [the] audience. There are songwriters who cannot be artists, who are just not interesting or compelling enough as people to be artists that people will follow, but they are great musicians and amazing songwriters. The concept of a virtual artist, the ability to perform online, or team up with people to create entities for them to attach their attention and audience affinity to, is brand new and very exciting, and a lot of traditional talented musicians who are not artists, or performing artists, will benefit from that.
In terms of kids going to music school, I think the business has already moved on. You're better off going to a computer science school and becoming a developer, because you're already talented enough as a musician if you can connect with people. There is value in the craft, there's definitely more value in the application of it and the connection with people. If in your studies you are learning how to connect with people, and share ideas and communicate at scale, then you're getting a thing that's going to attach to the craft. But if you're only getting the craft, you're probably better off pair[ing] up with some nerd kid around the world who's making a movie right now and composing the music for the movie.
The very classically trained skilled musicians end up playing what other people tell them to play — their job is to make the sounds, and it gets easier every single day to make the sounds. It becomes harder and harder every day to get people to listen to the sounds. If you want a career in music, be good at figuring out which sounds people want to listen to and get them to listen to the sounds. Don't be the person who makes their living making the sounds. The laptop can make the sounds, you can hire a musician to make the sounds, but there's very few people who know what sounds we're making, why we're making them, and who we're making them for.
The visual artist Damien Hirst doesn't make any of his artwork. The artist Jeff Koons doesn't make any of his artwork. He communicates with the audience, he creates the concepts, and he has fabricators who make the artwork. Two of the most successful and influential artists on the planet today, neither of them make the art. Ai WeiWei, the Chinese artist, is incredible. Same thing: [he] has a staff of 80 people, his artwork is incredible. There's no part of the artwork he's not the artist, but he very rarely makes the artwork.
In hip hop, it's the model. The producer is the artist, the rapper is the vessel. Metro Boomin is one of the most influential people in hip hop and he doesn't even perform. [He] knows what sounds to make, makes the sounds, and assembles the compositions.
There are only two things that are scarce: a genuine talent to make incredible sounds that move people, and then the ability to connect with those people. There are definitely musicians who are insanely talented, make really interesting music [that] is not finding its audience. There are hit songs out there that have not had vessels to connect with [the] audience. There are songwriters who cannot be artists, who are just not interesting or compelling enough as people to be artists that people will follow, but they are great musicians and amazing songwriters. The concept of a virtual artist, the ability to perform online, or team up with people to create entities for them to attach their attention and audience affinity to, is brand new and very exciting, and a lot of traditional talented musicians who are not artists, or performing artists, will benefit from that.
In terms of kids going to music school, I think the business has already moved on. You're better off going to a computer science school and becoming a developer, because you're already talented enough as a musician if you can connect with people. There is value in the craft, there's definitely more value in the application of it and the connection with people. If in your studies you are learning how to connect with people, and share ideas and communicate at scale, then you're getting a thing that's going to attach to the craft. But if you're only getting the craft, you're probably better off pair[ing] up with some nerd kid around the world who's making a movie right now and composing the music for the movie.
The very classically trained skilled musicians end up playing what other people tell them to play — their job is to make the sounds, and it gets easier every single day to make the sounds. It becomes harder and harder every day to get people to listen to the sounds. If you want a career in music, be good at figuring out which sounds people want to listen to and get them to listen to the sounds. Don't be the person who makes their living making the sounds. The laptop can make the sounds, you can hire a musician to make the sounds, but there's very few people who know what sounds we're making, why we're making them, and who we're making them for.
The visual artist Damien Hirst doesn't make any of his artwork. The artist Jeff Koons doesn't make any of his artwork. He communicates with the audience, he creates the concepts, and he has fabricators who make the artwork. Two of the most successful and influential artists on the planet today, neither of them make the art. Ai WeiWei, the Chinese artist, is incredible. Same thing: [he] has a staff of 80 people, his artwork is incredible. There's no part of the artwork he's not the artist, but he very rarely makes the artwork.
In hip hop, it's the model. The producer is the artist, the rapper is the vessel. Metro Boomin is one of the most influential people in hip hop and he doesn't even perform. [He] knows what sounds to make, makes the sounds, and assembles the compositions.
Why do you think there are so little entrepreneurs with a background in music, and do you think that will change? What can music schools do in order to change that?
I feel like there are a lot, I just don't feel like they're going to music schools. I feel like the ship has already sailed on whether or not the people who are shaping the music businesses are going to music school. They're not.
The people who go to music schools are the people who are makers of sounds and protectors of traditions. They are preservationists and they're keeping artforms alive, in the same way that art historians keep our knowledge of painting, sculpture, and fine art alive. They're archivists and documentarians. They're not driving or shaping the culture.
I only work in the driving and shaping culture side of it, so I see a bunch of entrepreneurs who are coming from the fan perspective of a consumer-driven business, [and want] to go help consumers have a great experience. Also, every year on the planet there's 300 million teenagers who all want to express themselves musically, and it's too hard. Almost everybody at some age, usually young, decides [they're] going to make a song. Some incredible amount of time passes, practice, failure, practice, failure, practice, practice, practice. At some point, you have a song that's decent, [but] it's literally years away from the moment that you originally had that inspiration … to make some music as a kid. The place that's creating entrepreneurship is a bunch of talented Machine Learning engineers saying: "What if the time difference between want[ing] to make a song and [creating] a really proud song that you want to communicate with people, was not five to ten years, but five minutes? What if we remove tuning, practice, theory, composition, chord structure, key, all these things that are not part of the communication piece, and we replace all of that with software?"
If you go to music school, you're not learning that. You're learning the old way to do it.
The people who go to music schools are the people who are makers of sounds and protectors of traditions. They are preservationists and they're keeping artforms alive, in the same way that art historians keep our knowledge of painting, sculpture, and fine art alive. They're archivists and documentarians. They're not driving or shaping the culture.
I only work in the driving and shaping culture side of it, so I see a bunch of entrepreneurs who are coming from the fan perspective of a consumer-driven business, [and want] to go help consumers have a great experience. Also, every year on the planet there's 300 million teenagers who all want to express themselves musically, and it's too hard. Almost everybody at some age, usually young, decides [they're] going to make a song. Some incredible amount of time passes, practice, failure, practice, failure, practice, practice, practice. At some point, you have a song that's decent, [but] it's literally years away from the moment that you originally had that inspiration … to make some music as a kid. The place that's creating entrepreneurship is a bunch of talented Machine Learning engineers saying: "What if the time difference between want[ing] to make a song and [creating] a really proud song that you want to communicate with people, was not five to ten years, but five minutes? What if we remove tuning, practice, theory, composition, chord structure, key, all these things that are not part of the communication piece, and we replace all of that with software?"
If you go to music school, you're not learning that. You're learning the old way to do it.
What advice would you give to music students who are just starting on their music startup?
If you're going to build a startup [or] a company, spend all of your time on how many people on the planet have the problem that your company solves.
If you want to pitch to an investor like me, here's your cheat sheet: if this is a consumer product, for kids to make music, listen to music, or have a culture experience, and I look at your product and I can't see an easy path there being 100 million users of your product, I'm not interested. If it's a product between two businesses (B2B), I need to be able to believe that you have a path to $100 million annual revenue. That's me, because I'm a venture investor — for every dollar I invest, I need to get $10 or $100 back — I need extreme multiples of my investments.
If you are ... going to start a startup and want to teach kids how to play the flute better, the first question you ask each other is: are there 100 million kids who desperately want to learn to play the flute better? [That] is a lot, there's definitely a million or two, though. The biggest first mistake is thinking that a million people is a lot of people. Seven and a half billion people on this planet, five and a half of them have one of these (Bob points to his iPhone), how are you going to make a hundred million of them care about your thing?
If you want to pitch to an investor like me, here's your cheat sheet: if this is a consumer product, for kids to make music, listen to music, or have a culture experience, and I look at your product and I can't see an easy path there being 100 million users of your product, I'm not interested. If it's a product between two businesses (B2B), I need to be able to believe that you have a path to $100 million annual revenue. That's me, because I'm a venture investor — for every dollar I invest, I need to get $10 or $100 back — I need extreme multiples of my investments.
If you are ... going to start a startup and want to teach kids how to play the flute better, the first question you ask each other is: are there 100 million kids who desperately want to learn to play the flute better? [That] is a lot, there's definitely a million or two, though. The biggest first mistake is thinking that a million people is a lot of people. Seven and a half billion people on this planet, five and a half of them have one of these (Bob points to his iPhone), how are you going to make a hundred million of them care about your thing?
