Has the Spotify Algorithm Changed?

Speaker A: Hey, I'm Ian Austin. I'm joined by my co hosts William Toll and Haddon. We are Indie Roots, a team of independent artists with, give or take, 700,000 monthly listeners combined on Spotify. And today we're talking about how the algorithm has changed, I think in the last two years. Um, but before we get into that, we do offer coaching services. This is one on one advice that we'll give you over a zoom call. We just want to transfer everything we know as our experience as independent artists and we want to make you self sufficient and we want you to have a thriving career. If you would like to hire us for that, please reach out@, uh, Indieroots llcmail.com and we'd love to hear from you. All right, so we think, at least I think I just talked to them about it, but I think the algorithm has changed in the last two years and I'll explain why. So I was talking to a guy who, he used to be part of a label and he kind of like has all the connections but he left the label. And um, I'm friends with him now. So I was talking to him and I was talking to him about like, um, I was like, man, I've been, I've been releasing so much this year, sometimes I've been releasing back to back. So like I'll release one song the week before and then the next week I'll release another song. And I'm seeing diminishing returns where just to kind of give you a rough like, um, reference point, like a good, a good song for me will get probably 2,000 to 5,000 streams a day. And so I was thinking, okay, my songs this year, I think they're just as good, they'll probably be getting just as much. But the songs, um, for the most part have been getting like under a thousand. Some songs like 300 a Day. And I'm like, what? This song is awesome. And uh, really confused about that. But I was talking to him about that and he was saying, well, because of the introduction of AI music, I'm pretty sure it's 70,000 songs. AI songs are uploaded a day. Because of that, he's saying they would have had to change their algorithm. So basically it's just a logical necessity because of this new environment that the algorithm is not going to reward, uh, releasing like a bunch of music at once. Because if you look at AI artists, I don't know if you've seen their profiles, but there will be just a ridiculous amount of songs in such a short amount of time. Like there will BE like in 20, 25, you look at an artist, they'll have like six albums out, six full albums. Some of the albums have 30 songs on them in one year. There's no way a human can create that much. But what they're doing is like, you'll see they'll have like 200,000 monthly listeners. And maybe those songs are only getting like a little bit of traction. But if you have like, I don't know, 200 songs out in one year, you're gonna go up pretty high. So anyways, what, what, what I'm seeing with my music is if I'm trying to do like a similar version that the AI artists are doing, just on a smaller scale, let me just release like 50 songs this year. The algorithm does not reward that anymore. But the reason I'm saying this has changed in the last two years is because I think there was a thing two years ago where the former CEO of Spotify, he said if you want to make more money, release more songs, just release more music. And so that seems like he is, he's promoting the idea of release more, you get more money. But now that, that seems like it's changed. What are you guys thoughts on that?

Speaker B: Yeah, I think, um, uh, several, uh, years ago, you know, um, I think Russ was kind of the original one to popularize like the, the Weekly. So. And um, he had a lot of success with that. And then, I mean we've all watched a lot of Nick, uh, D's advice for music marketing and he's a big believer in releasing um, very consistently. And so yeah, that's kind of been my mindset as well, is just, you know, throw more darts at the target. Um, but I've kind of seen the same thing as you Ian, because, um, the first like starting off this year, I basically released a song every two weeks and each song uh, performed worse. Um, and instead of like gaining more monthly listeners, I've just been like losing listeners. And so, and not only that, like, um, I've mentioned this before, but the moment I would release a new song, like if it came out on a Friday, my previous stream like, or my previous song that same day would start losing streams. So it's like the algorithm was cannibalizing my old song to make space for my new song. And, and as a result, you know, it's just the legs are getting uh, cut off before they have a chance to, you know, find some long term traction. So one, one guy that I've been following a lot and it'd be cool if we could like, get him on one sometime. But, you know, Abe Parker, we've talked about him, but, like, he's been promoting this one song since like, December and it just dropped last week. And I mean, it seems to be like, blowing up for him. And I've listened to the song a few times and like, it's an amazing song too. Like, you can just feel the passion and it just gets me in my feels. Um, and so I, Yeah, I just think it's more about just finding, finding the heavy hitters and just capitalizing on those. Um, which I guess the, the friction in my mind right now is how do I keep creating on a regular basis so I don't get rusty and, you know, um, and how do I objectively decide, like, which songs I should release? So I don't really know that. But, um. But the last thing I'll say is

Speaker C: that

Speaker B: I think a great song will transcend any algorithm. Um, and one example I love to look at is Coldplay's top song, Yellow, which came out, um, in like the year 2000. And that was before any algorithms existed. They released that during the CD era, and that's still their top song. And that just goes to show me that you just need to focus on making a great song that you're really passionate about. And as the, the algorithms come and go, you know, you'll. You'll still be left with like an evergreen product, an evergreen song. So, um. Yeah, yeah, yeah.

Speaker C: My thoughts are that I don't understand how, like, the algorithm itself changes as a result of a preponderance of AI music. To me, it just seems like the simplest explanation of what's happening is that AI music is, like you said, they're releasing so much of it, just a huge quantity. And what that does is it just waters down the supply or. Yeah, it waters down your chances of being heard, um, because you've got so much supply and only the same amount of demand. But the supply is absolutely skyrocketing because there's such a bigger supply of music on the platforms. So, like, I don't know, however many. Let's say there's a hundred million people listening to, um, music on Spotify and Apple Music, like on streaming. Streaming music. And, and you've got 50 million songs per year. But now with AI, you've. You're basically double or tripling that amount of songs per year. So the, the number of songs, the sheer number of songs in the pool is so great. I've. I've heard somebody propose Like a policy change where it's like, well, maybe, maybe you have to verify the authenticity of the music. You have to verify that it wasn't just. Just straight up produced. Like straight up generated by AI. Ah. And then, then, then you can monetize, um, M. Or I've also heard it said that all you need to do is put like a little, um, icon next to the song title that just says AI. Just like they do on, like, like if you're on a stock photo website, you're like, if you're looking for some album art or something, you're on. You're on Pexels. Um, some of this stuff will be AI generated, but there will be a little icon in the corner of that picture that says AI. And so, you know, um, and then, and then like in music, you see, you see, like if there are swear words in the song, you see the ex it says. Right? Like, I think, I think it's just E. Just E? Yeah, just E. A little icon that says E next to the song title. Well, the proposal there is like, just put a little icon there that says AI. That way people at least know. So you're not necessarily, you know, banning AI, but you're making it. Making the listeners aware because. And, and so with that, in my mind, I'm thinking, like, people generally want the. They value the human created music over computer generated music. Um, especially people that just like, following artists and like to hear a story and stuff. Um, I think that that would help because people would just naturally gravitate towards real art and away from fake art. So, yeah, I think those make sense to me. But I'm not sure exactly how like, the algorithm itself would be, you know, changing its nature. It would just be like, now it's dealing with like a whole lot more songs, which, which just means that, you know, more songs, same amount of ears, that means less streams per song. But yeah, go ahead, Ian.

Speaker A: Yeah, so to answer, uh, just what we were talking about, like, can the algorithm change or has it changed? I. I know in the past, I don't know about Spotify, but I know there was a thing a few years ago where the. I think the Instagram algorithm changed and the YouTube algorithm changed, but it was almost like entering a new era. Maybe it was 10 years ago, but it was first. I think YouTube's algorithm was primarily for kind of the thousand fan rule. It was like, we'll just show it to people in your area or your friends. It's not like the algorithm did not emphasize videos going viral, but then it kind of changed where it no longer prioritized your subscribers or your friends or things like that. It was more like, we'll just throw it out into the void. And. Which actually, it made it. It made it easier for your video to go viral, but it made it harder for, like, if you're just trying to get it to your people. And I, um, heard someone talking about that was like. That was like a big shift. And so I do think algorithms can change, and it's not. It changes it by its own initiative. I think the YouTube people, they probably reprogrammed it. Um, so, yeah, I think Spotify could manually change it. I don't think the algorithm's like, oh, I'm seeing this data now. I'm going to change my nature. Um, but another thing is, I kind of want to talk about what do we do in response to. Regardless of if the algorithm has changed, what do we. What do we do in response of this AI

Speaker C: oversaturation?

Speaker A: Uh, and I want to bring up something that Ben first. Or hadn't. Sorry, hadn't first told me about, and it was in this video, uh, of Ryan Tedder talking about songwriting and producing. And he said, the. The music industry now is like, you're standing on a big mountain looking across at, like, a hundred different fireworks shows. There's all these different fireworks going off, and the only thing that's what's going to really catch someone's eye is a firework needs to be more interesting, so it needs to be unique or it needs to be bigger or taller. It just needs to do something different than all the other fireworks. And so when I heard that, I thought, yeah, okay, I can really resonate with that, because when I was living in Utah, it was like that, uh, you could stand on this big hill and there's all these people doing their own fireworks shows. Like, there was probably actually a hundred fireworks shows I was looking at, and it was just. It was really interesting. And, uh, something that he didn't mention, but that I want to add in that caught my eye was when there was fireworks, like, consistent fireworks in one area that would really catch my attention. Like, oh, I'm going to look at this one. Um, but that doesn't really apply to our situation because just being consistent doesn't work because all the AI guys are being consistent. So, like, really the landscape is all these fireworks shows. There's 100 fireworks shows, and they're all consistently putting out, like, four fireworks a second. And so I don't think you're going to stand out just by trying to, um, put out more volume than the AI guys, because they'll always beat you in that. So then it comes to, well, you just need to have a bigger firework or a more interesting firework, a more unique one. Or here, the secondary path, you could have a private firework show and say, I'm not even going to try to compete in this public sphere. I'm going to invite a hundred of my friends to my backyard and they're all going to look at my fireworks because I got their attention. I don't need to compete anymore. And that's what I would say is 100,000 fan rule. And that'd be like Patreon.

Speaker B: Your.

Speaker A: Uh, the problem is how do you get them to Patreon? But still, that is still another viable option. What do you guys think about that?

Speaker C: I think, yeah, I think we should close it out here. But, um, but I do think that the, what you said is true about the problem is how do you get them to Patreon in the first place? And that goes back to our conversations about top of funnel, middle of funnel, bottom of funnel marketing, which is kind of like the. Just the law of nature in the marketing world. Um, but yeah, Patreon is a great bottom of funnel. And if you can get a thousand people over there, that's absolutely amazing. I do substack. Same thing, same concept, I guess. Um, but yeah, it's all about where's the top? Where's the top of the funnel? Because you do need a top. But we can maybe expound on that in a future episode or the next episode. Um, but we're trying to keep our episodes a little shorter now, so we're going to keep them around 15 minutes. Um, but again, we are indie roots. Please like and comment and subscribe to the podcast. Leave a five star rating if you found the content valuable and share it with a friend. We'll see you on the next episode.