Spotify For Artists Best New Feature

Spotify for artists just rolled out one of their best updates in quite a while and they're not even talking about it. The feature that they have been talking about is this artist profile protection where if music is getting pushed to your profile, you can turn on protection for it to make sure that other people can't push their music that isn't yours to your profile. So you can basically block it, which is amazing. This feature's awesome. I'll link to it in the description so you can learn more about it. But that feature's only available to like a thousand artists. It's in beta. It's going to get rolled out throughout the year. The feature that we're talking about today is actually just as awesome and it all has to do with the data you can see inside of Spotify for artists. >> [music] [singing] >> Now, you may be familiar with this feature if you go to the profile level of your artist account. You can go down to the segmentation and look at source of streams. And in here, you can graph active sources versus programmed. Maybe you want to look at specifically algorithmic versus third-party playlists separately. And now we can look and see, okay, cool. Over time, this blue is my active audience and this this pink or whatever is my algorithmic audience and the green is playlisting. And I can graph that over the last year and I can learn a lot about how my marketing efforts have impacted various things, right? I can see if my ad increases are causing more algorithmic activity. Like you can see my active growing over here and then all of a sudden that causes a bigger spike in algorithmic. I can see that around this time that I was using some playlisting. Did that impact anything algorithmically? And it seems like it actually may have a little bit. So you can correlate all those things together. The problem is that data's only at the profile level. And so if you just have a couple songs or you're getting your first song to get some momentum, this is good enough. But for example, if you're looking at an artist with a hundred thousand monthly listeners or million or even not, this artist account we're looking at now as my band. We only have thirty thousand monthly listeners. Even still, it's kind of hard to see what's going on for an individual song. Now, we can go to a song and you get all that same information for the song level, which might seem silly, but it's game-changing. This literally the first day it rolled out immediately was useful for someone I was working with on a call. So let me let me show you why. So if I go to segmentation source of streams, I can graph active. This is going to tell me what my ads are doing and then I can look at the algorithmic personalized playlist autoplay mixes. And so I can see we're dropping the song and we're marketing it. This blue line is my active and then at some point the algorithmic kicks in. And so I can kind of see like my ads and my algorithm developing separately over time. And I can even get more granular than that though and I can split apart my profile catalog versus listeners own pro playlist and library. So I can see that this this purpley line is my ads essentially. The green is the fans who are saving and adding the song to their playlist re-listening, which is caused by the ads, but it's a secondary metric. And you can see that green growing as the pink line kind of stays the same, right? Because the amount of new people I'm throwing at the song is staying kind of constant, but the amount of people who are re-listening to the song is growing. And then secondly, sometimes you look at this data because you're like you know, you're you're trying to track what we just did like when our algorithmic playlist is kicking off and all that. But one thing that's useful, a lot of times you aren't sure like is my stream per listener good, right? A lot of times I say it should be between a 2.5 and a 3.5 when it's just ads. Well, this isn't just ads, right? The algorithm is pretty strong in here. I can scroll down I can see that out of the the 15,000 streams that the song has so far, um out of like 6,000 have come from my active and 8,500 have come from the algorithm. So more than half of the streams have come from the algorithm, which pulls down the stream per listener. And so a lot of times I give people the caveat of of if your ads are most of it, you should look at 2.5 to 3.5. If the algorithm or playlisting is kicking in, it can be lower. But now you can just click source or streams per listener and look at source of streams and you can see that separately. So I know my active audience is giving me three streams per listener. So I'm right in that healthy range of 2.5 to 3.5. I'm exactly in the middle. Um and 1.2 is the algorithmic. So I know that the reason why this is so low here is because of the algorithmic streams pulling it down. And similarly, like you you might have better metrics here than I do too, right? You might be able to see, oh, my ads I'm getting five streams per listener and from my algorithm I'm getting something dramatically lower. Um and let's pivot and look at look at another song that I have pulled here. This song's been out for a lot longer. It's been out for like eight months or something like that, ten months, eight months. I don't know. Uh and and it's it's got 360,000 streams. So looking at this, we can segment it out, source of streams, active, personalized playlists, other listeners playlists, other. So this is all the categories. And I can already see right off the bat that other listeners playlist and other is is insignificant. So just to clean up the graph, I'm going to remove those. And now we're left between active and personalized. And so we can see these different algorithmic playlists developing individually. So we start off with marketing, we just got blue, which is our active. We get some immediate burst at the beginning, which is probably release radar and discover weekly happening. And then eventually later, I have a feeling this is when we opted the song into discover mode. Right? So we see radio going much more significantly than it was back here. And we get to see that kind of activity all happening in kind of real time. Now, similarly, I can break apart the artist profile and catalog, listeners own playlist and library. And we get to see, okay, cool. This is when we were running our ads. Pretty clear. This is when we're throwing new people on. This is the volume of streams happening after essentially as a result of the campaign. And another area this graphing has been super helpful already is if you're promoting your music in playlist, especially if you're doing like instrumental music or sleep music, etc. You're trying to gauge like what what is the amount of activity you're actually getting from your So all the streams are getting directly from the playlist, but then those playlists cause radio algorithms to grow and you're trying to kind of measure like, well, how much did we grow the radio algorithm? The way I used to do that in the past is I would just manually look, okay, we have these radio streams and then we go add this song to the playlist and then a month later we measure and now we have these radio streams. So we grew the radio streams last month from this up to this after adding it to the playlist. However, now what we can do is we can just go to a particular song that we've added to the playlist, source of streams, other listeners playlist, personalized algorithmic. And we can see down here this green is the playlisting. So we got this song that had like a hundred streams a day from the algorithm. We added it to the playlist that we were promoting. So it's getting 500 streams per day. And then um we can see the dramatic increase to the radio algorithm, right? So before, in fact, what we did do during all of this is right back here, we wrote down how much radio traction we were getting in just general algorithmic traction. And then like a month in, we did the same thing. And then we're doing manual calculations spreadsheet to gauge how much the radio algorithm is growing. Now you can just graph it. You don't have to do all that work. You still can. You can still write down data and stuff, but you don't have to. I can just immediately see, oh, cool. We were getting about a hundred a day before, maybe a hundred fifty, and now we're averaging out to like six hundred or something streams a day from the radio. And then after all this activity stopped, right? The the amount of money that they're investing into from the playlist dramatically goes down, the radio algorithm now hovers around two hundred to three hundred streams a day. So even afterwards, we doubled it. And like in this particular case that this data led to us thinking, well, should we go through this cycle of like rotating songs into this playlist to try to like kick off the radio algorithm. Then cool, algorithm's kicked off, let's cycle other songs up there and then come up with this crazy plan to kind of cycle songs in and out of this playlist that we're just always promoting to lift up the whole catalog of radio streams. And this allowed us to do that or would would have allowed us to do this with way less effort. It took hours manually figuring this out. And this would have cut down most of that time. Now, if you want to see how you can run a campaign where this data actually becomes really useful, check out this video to see how you can run an ad campaign to promote your music from start to finish. And check out this video down here to see whatever YouTube thinks you should watch. Anyways, thanks for watching. I'll see you in the next one. Bye.