Spotify's Algorithm Throttling Indie Artists?

3:06has caused these particular artists to be dramatically affected. And this is going from like a,000 listeners a day in the algorithm down to less than 100. So it's a 90% reduction over a short period of time. Contrast that with this. This artist did I looked at their saves before this which which I can actually flip to here. They did a bunch of promotion back here and they haven't released a song in the last 12 months. And their algorithm has been kind of slowly falling over the past year. And this this person says that they've been algorithmically throttled just as much as the other ones that I just showed you. So I I want to just I want like everyone that's thinking that they've been throttled to first just give yourself a reality check. are do you fall into the camp of you've had a

3:52severe drop off on a on a ideally well not ideally but specifically around the beginning of August August 9th August 10th is the drop off very significant where all of a sudden just the average drop by 90% and it hasn't recovered um has it been more than a month so that any temporary changes to the algorithm like Spotify discovery mode might have changed um and you know stuff like that like honestly do you think the fallout is natural Because to me, this just looks like natural algorithmic falloff and it's just the person doesn't want to accept the fact that it's just a natural thing and they they want to blame a change in Spotify because that that's easier. And I'm not I'm not blaming them for that, but it's just that is in my opinion the reality of of what's happening for this particular artist. And there's a couple

4:38cases like this. This one here I think is more of a middle ground. like there's been a pretty clear reduction in what they were experiencing here, but it's their algorithm is already all over the place and it doesn't correlate with that date. I mean, it there's kind of does a little bit here, but the they're kind of they're almost back up to the levels before it here. They're only a little bit reduced. It's not like a 90% over week and their algorithm is already all over the place. So, same thing here. I don't I don't think this is an example of that here. This one I is a maybe in my book because this one there's a pretty clear drop off the beginning of September, but this could be like discovery mode. So, if you're not familiar with Spotify discovery mode, it rerolls at the beginning of the month and your algorithm can pretty dramatically change

5:25at the beginning of the end of the month because of discovery mode. Like maybe some songs were opted out or opted in, something was changed and that was what caused this, but it doesn't line up with the same date. And this drop off here in July is I mean one this increase is pretty dramatic but then the fall-off is equally dramatic. So this drop off I don't think is significant. This drop off I think is more significant but it's only been a couple weeks and it correlated with the beginning of a month. So this is a maybe in my book. And then we get to the ones that are I think are more clear-cut like this one here. They got the August 9th thing. It's a 90% reduction over the over a couple weeks. And same thing here. Same thing here. This one is an outlier. Uh this one seems to correlate back with the beginning of the year. And there was

6:12a pretty dramatic falloff from January to February, like going from 2,000 listeners a day down to less than 200. So 90% reduction over over the course of a month and then it just kept falling. Now it's less than 100. So it's been a it's been an immediate fall followed by a very slow fall. This one looks like the artist is getting lumped in with AI generated music. Um, and I don't know for a fact if they're AI or not, but I checked a bunch of the people that are replied that say they're not making AI music. And one, I listened to it didn't sound like AI. Two, I ran it through an AI music detector that historically I've found to be super accurate. And like publicly they're 97% accurate. And none of the examples here reflect as AI. So, I don't think anyone's lying in this

6:57case because it it would have I'd have more false alarms or I'd have more alarms in that case, but every everything seems to be pretty clean. So, as you can see, pretty clear two camps. Uh, this one I'd say is included even though the date doesn't line up and then the other three and then we have three examples that I think are I mean there's one maybe and there's two others that are definitely not on threads. There was also a bunch of people who said that they were being throttled like the AI artist and they were definitively not. It was much more the case of of this where it's been a slow gradual decline. Now, one thing to note here, um there's a lot of like things where you're like, "Oh, that that almost looks like a trend, but it's not a trend." Because I'm trying to look at this data to figure out like, well, what is the problem here? Why why are some artists

7:43affected and some artists that are not? because uh I didn't even know about this until this video that I made and the comments came under because all the artists that I work with haven't been affected. I haven't seen any trend across the artists I was I was working with. [sighs and gasps] And that doesn't mean it didn't happen. It just didn't mean it didn't get surfaced to me, right? I work with a lot of a lot of artists, but all the ones I worked with directly that I'm looking at regularly, none of them were affected in the same manner. And [music] um that means it's not all artists. It's only a certain percentage. So why are some affected and some are not? And like when you look at some of these um some of these are like well look at how much third-party playlisting they're using here. Like that's that's a lot. They're most of their audience is thirdparty playlisting. And same thing with this one. Uh this one's not the case. They have more activives than third party. They had more algorithmic than third party. And then this one's also the same boat. They

8:29had a lot but it wasn't like a crazy amount. And this person did have a lot but they weren't affected. And this person did have a lot but they weren't affected. and and I the other artists that we're not looking at here, same boat. I've seen some that that had a ton and they haven't been affected and some that had none and they also weren't affected. So, it almost seems like a trend. Like, is third party playlisting related? I don't think so. Um, one thing that I thought might have been a trend was that I started looking at monthly active listeners and I noticed that for for a few of these ones that were affected, the percentage of monthly active listeners very small. So like 6% of their audience being active and only 36 people out of 60,000 listeners were super listeners. So incredibly low engagement. Like this artist doesn't

9:15really have that many fans. They're mostly held up by third party playlisting and and algorithmic playlisting and other stuff. It's all passive. Um we look at this artist who's very clearly been affected. Um not so much the same case. 20 to 30% active is a lot more natural. Uh 250 super listeners isn't great, but it's not horribly bad, right? Like that other artist had way way way way way way way worse. So, um this number is not that bad. Um this one here is also not that bad. It's between 20 and 30%. And this one here is is very bad. So, there's some here that are seen like very low active audience. And at

10:00first I was like, "Oh, maybe that's it." And then I look and it's like, well, no, these other ones don't seem to be affected. But I'm just bringing it up is maybe that's a factor. Maybe it's like a combination of things where one of them is you have a lower than average monthly active listeners. Maybe that's one variable. Maybe super listeners play into it. Um, if we look at this one that was not affected, they also have a very low active monthly listen account and very low super listener. an example of an artist that's not affected but had a super low 7% monthly active listener count. So I don't think that's a trend either. One thing that that is a trend across all three of the like very clearly affected ones but not the fourth I believe that that was um a different time scale was all all of the ones that were affected directly on August 9th

10:46were some type of electronic music um some type of house basically and I don't think that's necessarily a factor by itself because I looked at other artists who were not affected that were house but all three of the clearcut ones that were dropped on August 9th were. So, I don't know if Spotify has been doing like a rolled a rolling stage out of algorithmic changes where like maybe they tweak this variable to counter some like thing they think is a problem in this genre and then a couple months go by and they're like, "Oh, well that's there's a different problem in this genre, so they tweak a variable there." Um, that's my guess. I don't think it's the case if you're doing like chill house that you're screwed. But, uh, it is it is a thing worth noting. And I'm just bringing this up because I I want to see if like if anyone here has any

11:32more information. I've been talking with Jason from SubmitHub who has access to infinite amount of Spotify artist accounts and he's been looking through this a ton to try to figure out if there's a trend and we chatted about it. Um and then also Sam Dubov from Spotify is aware of this on threads. Someone tagged him in the thread where all these artists were talking about it and he's forwarded information to the internal team. So, there's people working at this and and unfortunately I don't have like a good solution for you if you're if you're affected from this. But the biggest takeaway that I want you to have is if you are affected, one, you're not alone. Uh and then two, there are people potentially the the best group of people that could possibly be looking into this, looking into this to try to see if they can learn anything about why some artists are affected and why others are not. And then hopefully eventually we can come up with with is like well if

12:19you are affected here's what you do to kind of get out of algorithmic jail if that's even what the problem is in the first place. [sighs] So if you are affected let me know in the comments if you have any other insight. Uh also drop them in the comments and make sure you subscribe and ring the bell on this video uh on my channel so that you don't miss my inevitable future update when I learn more about what might be happening. And if you want to watch that video I mentioned at the beginning with the AI music getting hitting throttle, check that out right here. And if you want to see whatever YouTube thinks you should watch, check that out right here. Anyways, thanks for watching and I'll see you the next one.