The Spotify Problem

I was engaging in the wonderful (soul sucking) world of Instagram the other day, and I scrolled to a disturbing piece of news. An account by the name of @laura_horowitz_narrator was claiming that Spotify had just updated its terms of service on its audiobook distribution platform, Findaway Voices. After looking into it more, I was shocked to find that it was true. The terms, "required authors using Findaway's services to grant Findaway and Spotify rights to 'translate, modify, [and] create derivative works' from submitted audiobooks. The terms further stipulated that Spotify could use the audiobooks (termed 'user content') in 'training' and 'modeling' in connection with Spotify services." (authorsguild.org) Scary.

View this post on Instagram

A post shared by Laura Burrows Horowitz (@laura_horowitz_narrator)

Luckily, Spotify has since decided to revise these terms. However, I believe this has very much to do with the backlash more than anything else. Had people not spotted the change, I think it would still be in place. "The new terms remove the rights to 'translate' and 'creative derivative works' from the grant of rights, make the grant of rights revocable by the authors, and further make it explicit that the grant of rights does not authorize Spotify to use user content to create new ebooks, audiobooks, or use it for voice narration AI training." (authorsguild.org)

The thing that really scared me most about this prospect is the potential for this rights abuse to be complimented through the use of AI. The generative process that AI uses is still notorious for its often comical flaws and inability to create complicated original ideas that hold water. However, imagine if it was able to be fed existing works to base its creations off of? In the case of fiction writing, it could be provided with a fully flushed out world, and even entire plot lines. In that case, all the AI would need to do is to create something derivative--a much more manageable task. This could essentially cut the writer fully out of the creation process. With the terms they had originally stipulated, it would be legal. Imagine the horror of spending years crafting an entire world only to have it stolen by a faceless corporation. Reading words that wouldn't exist without you, but that you have no control over. You could watch your own work rise to prominence with absolutely nothing to show for it. Yikes!

Now it's true that this ended up not being the case, and there is an argument to be made that something like this could be contested in higher courts, even if it is stipulated in terms of service. For musicians, this might even seem like a distant issue, we don't do audiobooks, we make music! Yet, I think it's important to think of the arts as a collective whole. When a portion of us suffers, we all suffer. Such a potential loss of rights in the writing field sets a precedent that would surely come to haunt us shortly after. It's by standing together that we have a chance to prevent abuses like this. Even so, this fiasco isn't so important in terms of its direct impact, but more for its spotlight on the psyche of the folks running Spotify. This platform is currently not here to help us. As much as they insist otherwise, it's here to profit off us and cut us out at the first chance it gets. I want to dig into just how harmful it is, and how helpful it could be.

Streaming Royalties:

If you look up how Spotify decides what artists get paid, you'll be pointed to a couple of very neat and polished looking videos. They feature that loveable AI DJ voice we all care for so much, describing how the sausage gets made amidst a collage of colorful and creative animations. Except, this doesn't actually tell you how things get calculated. It leaves the question of why royalties are so low for streaming to the vague notion of distributors and publishers impacting this. In fact, they really throw them under the bus quite a bit. While yes, this can definitely impact the royalties that an artist might receive, it isn't the full story.

The video above is missing a little term called "streamshare" which is used by Spotify to describe the amount of streams that a given distributor received through their artists over the duration of the month. When Spotify looks to pay musicians, it looks at the total number of streams and then the total number of artists and divies up a percentage of the total amount of revenue earned (Spotify claims this is 2/3rds of every dollar of revenue) amongst their rights holders (distributors, publishers, etc.). This essentially creates a system in which only artists with a massive following who can take up the biggest percentage of the total stream pie will see any kind of substantial income through streaming royalties. For example, because Taylor Swift might be getting millions of the billions of total streams while smaller artists may only be getting in the hundreds, Swift could get over $100 million in streaming royalties for her massive percentage of streams, leaving the smaller artist with pennies, if that. This rubs me the wrong way, especially because hundreds of streams is nothing to scoff at!

Another Way?:

I've been contemplating solutions to this problem over the last few months. After discussing with friends and colleagues (special thank you to Immanuel Mellis and Finn Penzo), I think the best system we've been able to dream up is one in which royalties are paid on an individual consumer basis rather than a percentage of total streams. Here's an example comparing the current system with the proposed solution:

Willy pays $10.99 every month for Spotify Premium. A fan of Taylor Swift, they streamed her music 76 times over the course of the month. Willy also streamed the music of their struggling musician friend, Sean, 24 times. Under the current system, that $10.99 would be collected, a portion would be used to pay taxes and expenses, and then 2/3rds would be put into a pool of revenue to be distributed to artists based on a percentage of the total streams that month. This means that it doesn't really matter that Willy was only a Swiftie for 76 percent of the month and a good friend (I'm kidding) the remaining 24 percent. The total streams Taylor Swift received would drown out Sean's considerably smaller audience and that would be reflected in the royalty distribution. But what if that wasn't the case?

Under this proposed concept, Willy would pay their $10.99 every month, but instead of that money being thrown into a pool after expenses are paid, it instead stays with Willy's listening stats and is distributed that way. After taking that 1/3rd out for taxes and expenses, we're left with $7.32. 76 percent of that goes to Taylor, so $5.56. Then 24 percent goes to Sean, $1.75! You might be thinking, that's barely $2, so what? But I'm thinking that's almost $2! For 24 streams! What?!
​

I love this concept for two reasons: it would increase small artist revenue considerably (which would actually benefit their distributors and publishers as well, as they'd be the first to see that money--something I might want to discuss in a later post), and it would create a direct way for fans of an artist to support them, just by listening! Whenever I'd stream my favorite artists, I would know exactly where my money was going. It'd let me support them in a way that was tangible for both myself and them. Powerful stuff. Yes, this would cut into the millions that Swift and other big artists make off Spotify, but I think it's a small price to pay to ensure another revenue stream for aspiring artists.

What are your thoughts? Would this system work? Why or Why not? I'd love to hear from you! Thank you for reading, and happy jamming!