Get on the mailing list

Spotify's "narrative fallacy"


First to get this out of the way: I use Spotify, I pay for a family tier subscription, there's no other way I could listen to thousands of tracks that I wouldn't play often (or ever again). When I've done sound post-production or composed music to video, Spotify has been a great tool for making editorial decisions. I'm also on Spotify as an artist, mostly so that I have an informed opinion of the platform.


With apologies for literally shitposting this and this (and this) article, I have some thoughts on Spotify CEO Daniel Ek's recent statements regarding the music industry.

First let me quote The Fader, really you should read that in full:

Ek denied criticisms that Spotify pays insufficient royalties to artists, and insisted that the role of the musician had changed in today's "future landscape".

Ek claimed that a "narrative fallacy" had been created and caused music fans to believe that Spotify doesn't pay musicians enough for streams of their music. "Some artists that used to do well in the past may not do well in this future landscape," Ek said, "where you can’t record music once every three to four years and think that’s going to be enough.” [...] What is required from successful musicians, Ek insisted, is a deeper, more consistent, and prolonged commitment than in the past.

Was he on drugs?

Ek, according to Pando, in his teens, would bribe his co-students with video games to code websites that he sold for $5,000 a pop to various clients.

I happen to have been in the sound and music industry, the music distribution industry, and the web industry, and have the same view on all three:

Do not exploit other people.

Sweden has some history with treating music as a commodity, not art, and it angers me deeply to see a billionaire entrepreneur talk trash about his own content suppliers. Ek might think that Spotify – the product – is a great recommendation engine and a set of technologies and tools that work with music as just one possible use case. But music and podcasts (sometimes videos) is what Spotify's content is.

Thus, I find it disrespectful to skirt the central problem of the long tail while basing your business heavily on it. For an artist, it's only viable to be in the "head", not in the tail. Then there's content like lo-fi hiphop - all listens, no fans, which Spotify seem happy to enable. The number of plays required for music to "perform" is ridiculous. I think if Spotify eliminated all such "underperforming" music from their platform, they themselves would cease to exist.

Also, a website for $5k is junk. I've worked at a digital agency before, around Ek's 20s. It costs ten times that to sweat the details on a website.

Do not exploit people.

What Ek won't point out is that Spotify loses money every year. So technically this is venture or public capital going into a business model so aggressive it isn't concerned with profit, only with growth. Like Uber.

on scheduling

Regarding the "narrative fallacy" statement, more specifically Ek's: I feel, really, that the ones that aren’t doing well in streaming are predominantly people who want to release music the way it used to be released (Ek), I offer you Depeche Mode's schedule from 20 years ago:

  • 03 Feb 1997 Barrel of a Gun single - 12 formats
  • 31 Mar 1997 It's no Good single - 13 formats
  • 14 Apr 1997 Ultra album release
  • 16 Jun 1997 Home single in 8 formats
  • 20 Oct 1997 Useless single in 7 formats
  • 07 Sep 1998 Only when I Lose Myself single in 14 formats
  • 28 Sep 1998 The Singles 86-98 album release on CD and DVD
  • 02 Sep - 22 Dec 1998 The singles tour - 65 shows in two legs
  • 23 Apr 2001 Dream On single - 12 formats
  • 14 May 2001 Exciter album release
  • 04 June - 05 Nov 2001 Exciter tour - 84 shows in two legs
  • 30 Jul 2001 I Feel Loved single in 15 formats
  • 11 Nov 2001 Freelove single in 10 formats including DVD
  • 11 Feb 2002 Goodnight Lovers single in 2 formats

What the fuck is Ek talking about?

Release often they did. Let's ignore the idiotic view that albums are mere promotion of live performances. There was no world tour around Ultra - the frontman nearly died of substance abuse around that time, and the band nearly died with him. Recording and releasing to a regular schedule, touring heavily and staying on top of their game, nearly destroyed DM.

Anyone who gets it that being a successful performer is its own toxic positive feedback loop, would have also arrived at the idea that an artist needs space and time, to not get hurt by their own art.

So what does Ek want? "Music the way it used to be released" was already often enough, so he doesn't even know what he's talking about, not to mention disrespecting the craft so deeply. But one can't release music every few months unless they release low-density (low-intensity) content. I want less mediocrity. Ek is optimising Spotify to be a conveyor belt, not the Library of Alexandria of music. This doesn't look dissimilar to his entrepreneurial past. I suppose all is fair in love, war, and ruthless capitalism. It's just that a green elephant happens to be in the same room in which we're discussing modern slavery.

I have made my process visible on Twitch. The fork in the road is this: unless that work (play) is full focus, backed by means that don't respond to KPIs, bookended by breaks as long as necessary, it is not sustainable to make good music. It ends up in the (extremely) long tail's tail, and the stress from making ends meet makes the music worse. There's enough bland and generic music out there, and Ek is pushing people like me to stay at that level and work his polluting coal plant. For $0.0045 per play.


« 65 streams later
» VA - Eastern Depths IV