How Streaming Fraud Affects Artists ft. Andrew Batey (Beatdapp) – The Label Machine Podcast Ep. 22

On this episode Nick sits down with Andrew Batey, co-founder of music streams auditing and tracking company Beatdapp and professor at Trinity Western University. Batey is an experienced start-up founder who has raised over $50M for his start-ups and filed over 40 patents in seven different countries. He has been featured in publications such as Forbes, The Wall Street Journal and the New York Times and is a frequent guest speaker at conferences focused on venture capital, marketing and blockchain.

Join us for an hour-long discussion on digital music distribution, DSP's, streaming fraud, content auditing, music marketing tips and much more.

NICK SADLER: Welcome to The Label Machine Series, where we discuss with our guests how artists and record labels monetize music. Today's guest is Andrew Batey. Is the current CEO and co-founder of Beatdapp, a platform that music labels and artists use to track their songs and collect royalties by providing real time audit reports of streaming play counts. Andrew has a huge amount of experience running and raising capital for creative and high tech startups.

Having raised over $50M for his own startups and filed for over 40 patents across seven countries. He's been featured across all the top entrepreneurial publications The New York Times, The Wall Street Journal, and Forbes. At the helm of Beatdapp, he has guided the company to numerous awards, such as one of the exploring Blockchain Breakthroughs competition and getting to number two as a seed stage company in Canada by Crunchbase.

To top it off, Andrew is also a professor at the Trinity Western University teaching advertising, marketing and entrepreneurship to startups and future thought leaders. Absolute honor to have you here today, Andrew. How are you?

ANDREW BATEY: Good. Thanks for having me. One quick correction.

NS: Yes.

AB: I am the co-CEO. My best friend, Morgan and co-founder. We actually share the responsibilities together, so it's kind of a unique setup. I think not a lot of people have it, but when you've been best friends as long as us, it kind of works out. So kind of both.

NS: Co-CEO and co-founder, gotcha.

AB: Just, you know, for everyone out there wondering what it's like, happy to go into more detail there. But it's a, we sort of help each other a lot because I'm the guy that runs off the cliff and he's often the guy making sure that we have parachutes.

NS: So I know you mean it's you know, it's funny with with the titles as well like that, as well as when you're when you're a small, smaller company and… you know, so First Flights, small company, there's four of us, right? You know, five of us. And, you know, technically I'm the CEO, but I never really say CEO because I'm like, am I like CEO for me is like, you know, you've got 30 employees, you're quite big, you know?

There's people underneath you, there’s people underneath them. So I understand the thing that it is. And my business partner as well, Phil, I think like you, like we balance each other really well, we’re both really the CEO’s, like we both we both call shots together, like a CEO would normally say “I have the final say,” whereas it's not really like that.

I mean, yeah, how have you always just gone? “Yeah, I'm CEO and that's it.” Even if this sort of company is four people or you know, what's your take on that?

AB: Yeah I think I agree with you. I've always thought that CEO sounded weird. I also hate when people say they're like, and this is my own pet peeves. But like, I hate the word “entrepreneur” sometimes because I feel like people use it and they're not actually like building anything. You know, they're more they should rename themselves as “Dreamer” because they haven't actually tangibly done anything yet and stuff like that, like entrepreneur or sometimes is another word for “unemployed.”

NS: I agree. But the funny thing is, you know, if you go to a party meeting somebody and you've got multiple companies and you enjoy studying out, you know, you enjoy starting up companies, selling them, then you are an entrepreneur, by definition and a really easy word. If someone say I'm a doctor, I'm an entrepreneur, but I am like you, I cannot say I just feel feels too cringe.

AB: But you're right, though. It's like people say, “What do you do?” I'm like, “Well, you know, I mean, technically there's these five things that I've done mostly, but I have started 20 companies.” Yeah. So it's like, it's like pretty difficult to, you know, where do you draw the line? Like, is this just a company with one project or agenda they made money on?

Or like, does it not count as a company unless you have 20 or 30 employees and you're like, what's the, where's the line with that? Because there's lots of stuff that I start that I just think is fine.

You know, like recently this is music related. A couple of us got together and we decided to release a song to see if we could hack the charts. Like, could we, could we become like a charting artist? And we did that in April and my big focus on that, I know we hadn't gone into exactly a deep depth as yet, but at some point I lost… I feel like we lost focus on the artist itself. We're like an enterprise company and so I was like, It'd be really cool to go through the journey of the artist and just make a song, figure out how hard it is, figure out on our own how hard it is to promote.

Just go through the hero's journey that artists have to go through. So we better have we have more empathy for them and then look at our product and say, “Where on that journey could we help them?” And so I you know, that that was the initial kind of seed of this whole idea. And then it was crazy, like we got-

NS: So tell us, what was that? What was the, what was the song like? Start there. What kind of genre was it?

AB: We decided to start with country because I figured there's a lot of publishers already in country. It's easier to get going. You don't have to… I was trying to figure out how do you eliminate having to write the song, you know, because obviously I'm not that talented. So like, is there a way to take a song first? I love country, country’s like a genre I really, really, really like, work with a lot of country people.

So I thought to myself like, that might be the easiest one to go in. At least there's pre-done songs or there's a whole industry of publishers and writers and people that you can listen to a near complete song and get a get an idea of what your vibe should be or what direction you want to go versus if you rap.

Like when we were looking at rap songs, for example, because my concept was, I’m all over this right now but the concept originally was “Let's do five songs, five different genres, and I want to call the album undefined” because I don't think a fan is truly one genre. I think fans love everything and I think if you have a vibe for the entire album, people will probably listen.

So my vibe was going to be Wake Surfing with your Friends. I wanted a song, an album you could play end to end no matter the genre, that was just fun to listen to as you're awake surfing with your friends and like, that's that was the vibe I wanted to go for. And so we started with country because it felt like the easiest one to get out the door and start building momentum with and see if we could just get this done because I didn't really know what I was getting into.

So the song is called “Drinkin’ Drinks.” The album was called the sort of group’s name is Urban Outlaws. I used NFT’s for everybody so that we, no one had to use their actual brand. Like any artists that helped us didn't necessarily feel like we were stealing their brand, their brand equity to do so.

So we had a lot of really famous artists help us with the song, and the benefit was that they didn't have to go out there and promote it. I just wanted to make sure we didn't have a crappy songs. I went to them. I was like, “Look, I need your help, but your name doesn't need to be on it. And all the money goes back to charity.”

So we actually wrote a pretty big check to MusiCares, which goes, which goes back to artists. And we didn't take any money from it at all, but the long and short of this whole rant is that it's really hard to be an artist, which was super confirming.

And we were lucky like, we were… We did something, some cool stuff that helped us break through some of the noise we ranked. We went number four, we went number four Country music single in the United States the week of April 22nd.

NS: Wow. That's the national…

AB: Chart. The national charts, the United States.

NS: And so it was on release day. Essentially you went to number four…

AB: That whole week. Yeah. So I think the big focus we had was how do we break the noise for the week? I think where I messed up was I didn't have a plan for after week one. I sort of assumed, I hadn't, I hadn't really tried to promote an artist, so I cut my teeth back in the day breaking artists, and that's what I used to do a lot of.

And I was really well known for sort of growth hacking platforms early. So I was one of the first people building on Facebook back in 2006-2007 when it was still .edu based and you had to have an .edu address in order to like, do anything. And then I started building audiences on Facebook and helping launch artists. And that seems intuitive and normal today.

But back then, nobody was building audiences on Facebook. So like being able to leverage this this very viral platform that no one was really using with their college students to launch an artists was very novel at the time. And so when we did it, we actually found this this artist at a bonfire from Nantucket, and we were like, Hey, what if we made him a college rapper, like a frat rapper? And we launched him and he went number seven Billboard and just like, blew, like, blew up and that and that was really interesting.

And we did a couple other things earlier with Twitter and earlier with YouTube. And so I was kind of that guy that was really well known for breaking artists in. But it had been, I stopped doing that couple of we sold in 2011.

So it's been over ten years and now in the day of TikTok… plus those things have changed, back then used to really try to hack visibility, like you're trying to break into someone's attention. And then once you got their attention, they decided where they're like, you think about YouTube, you can have the front page of YouTube and at the time everyone go to the front page of YouTube. That’s where everyone went to start their journey. There wasn't like an app yet and stuff.

And what's interesting is that if people didn't like you, which happened to us quite a bit, like we would go to launch an artist and it didn't work, you know. Because if they start playing the music video and they dislike the music video, they're not staying for 3 minutes like they're out.

And so it's pretty interesting. Like you can hack the platform to get the exposure, but if your product, didn't like, actually meet the consumer or the customer, there was no product market fit there. Like it was obvious and that was it. I think in today's world it's more interesting that the habits have shifted and there's so much passive listening, so people are just listening to playlists and if they don't like that song, they kind of just more than not let it ride through like that, you know?

NS: As long as it’s not absolute rubbish.

AB: Yeah, it's not screaming in their ear. They're like, I don't really like this one, but it's too much effort to reach my phone and hit next. So like, I'll just let this it's like 2 minutes or whatever. It's not that bad that I do it all the time. Yeah, it's not that bad. Sometimes I feel bad. I'm like, Maybe they deserve an extra play. I'll just keep letting it go.

And so it's interesting to me because that's not really fan building like in the other context, the way that I did this in the past, you would launch, then they would have to decide if they liked you or not. And today's sort of TikTok and playlisting era, just because you got exposure does not mean they're paying attention, but you have a captive listening audience or that they even have decided they like you or not.

And that's really hard to translate. And so I think that's the miss I had. I thought if we break into all of these playlists and if we could get charted, we would get noticed. And I assumed that if there's a spark, maybe a fire will catch. And I think what I didn't have an appreciation for at the time was how much additional stuff you really need to do in weeks 2-10, you know, and really following that up with with more just sort of blocking and tackling that's required.

And so anyway, we did really well the first week. We had everything geared for the first week and then you kind of saw us peter out.

NS: So just, just on that point though, because I think that's, I think listeners would be really interested in, you know, you're saying like, you know, tackling and blocking for those weeks, you know, 1 to 10 like you know, and new analysis afterwards, what would have you what would have you done? Like what three things would be... You said this is what we should have done to keep that going and try and get to that number one point.

AB: Yeah, I would say that the first thing we did is we had like no emphasis on capturing or reengaging with our fans. We just tried to push everything to the listen, to start that first. Everything was about pushing as many downloads and listens as possible for the first week and there wasn't really like a follow up plan. So like, how are we going to engage them?

Should we think about any kind of tour dates? Should we try like, what are the things that we can do after they've decided they're going to listen to us to sort of engage with them? We just did none of that. It was really just like, get them to the song, get them to listen, get them on, get on playlists was like the biggest thing, like how do we get on playlists? How do we get people's attention? I think we could have done way more. And also beforehand, there's a bunch of stuff we didn't do.

Like I chose not to focus on TikTok at all. I felt like, I felt like getting all the influencers on TikTok was going to be really difficult, and had I had more time to plan and a bigger team, maybe I would have done that. But I have. I talked to a couple of TikTok influencers and immediately realized this was going to be hard because like, then their managers were involved. Yeah. And I'm like, Oh man, this is going to be like a whole thing. And I probably needed way more time for this and, and a creative strategy too, around like what they're going to post. And so that became difficult. I tried this platform where they help you like do like a TikTok incentive and it was like basically they have a contest for people to TikTok and it was like the worst thing. Our song is called “Drinkin’ Drinks.” And it's about all the types of drinks you can drink and just how fun it is to drink with your friends.

And like the group or segment that picked it up on TikTok were pregnant moms, and I was like, I think this is the wrong market! This is like not the direction we were going.

NS: I mean, they're the ones that need to drink. Probably. They're the most stressed out, but they're definitely not the ones that should be drinking.

AB: It was so crazy. I was like, Oh, man, this is not, this is not what I was going for here. And so, you know, it was…

NS: So you created this, you created this this big storm essentially but then, you didn't have I guess you didn't have anything to catch all the rain that was falling down and and then and reuse that and recycle it and create something.

Yeah, that's. Yeah, that's really interesting. Just, just to stay on this as well because I know it's just something that comes up a lot as well. I haven't even really gotten to your backstory yet. We're nearly halfway through. This is crazy.

Okay, so I do want to ask though, you you, you know, you got to do number four and you really focused on getting on Spotify playlists. What are couple of strategies that you did to get on playlists, I mean, if someone's listening and they're like, “Yeah, I need to get on playlists, like what would you recommend doing today?”

AB: I mean, like the reality is, it's really hard to get on a playlist.

NS: Did you use it? Did use any actual platform?

AB: Yeah, the one that I thought was the best was actually Now, I didn't get like massive playlists that drove hundreds of thousands of plays, but I got consistently, I think I got 70 or 80 playlists that probably drove, you know, ten or 20,000 plays so far. So I would say and it's consistent. Like we still get plays. Like if I logged in right now, we probably style plays coming from them and it's it's pretty cool. It's like real people who have real playlists and like, you know, one of I think that, there was one called like “country road trip” that someone put us on and we probably got like 8000 plays from that, that playlist. So other people were clearly like listening to that playlist and listening to us.

And that was pretty cool. And I think I also know the CEO fairly decently well and just from being in the music industry and they have a real effort on kicking out fraudulent playlists. So they like really actually sort of diligence all the playlists that come onto their platform. So before they, so how it works is you upload your music, it goes to all these editors, they basically rate it whether they like it or they don't like it, and then they get to choose if they're going to add it to their place or not.

But it's like a discovery platform for them. And so you get feedback on if they like your song or they don't like it, and then you end up getting often playlisted. And so what I thought was really interesting about that is you at least know it's going to real editors and not bot farms. Like not somebody who's just running the numbers up on a playlist and you're going to get in trouble because honestly if you get blacklisted and you do the wrong things and you pay the wrong… a lot of people go like, “How do I get playlists?” And they go pay some company $200 or $300 to get on playlists like that.

That's the fastest way to getting banned. And if that happens, like it's going to be very difficult for you in your career to sort of tell Spotify, “Hey, that wasn't me.” And then like explain why. Get a hold of anybody there to talk to them about it, because it's sort of just like it's just done and so it's it's pretty difficult, I think, to come back from that even.

And it's not just artists, it's labels too. Like I get labels all the time that say, Hey, do you know any way they can help us at this at this DSP, you know, one of our artists was unintentionally flagged for fraud, which is not fraud. And we can tell you exactly what we did in this unique promo.

And it's hard if you look at it from there, and there's a lot of fraudulent plays going on. So they're going, hey, this looks abnormal. Ours must look really abnormal. I mean, we had just crazy amounts like our focus and how we got, how we broke through was I went to a bunch of band clubs that are not country related.

I went to like the BTS fan club. I went to all these fandoms that are related to like Harry Styles and like Nicki Minaj and like all these other fandoms that that we know and work with all the time. And they said, “Hey, I'm doing this song, this is going to sound crazy. Could you get your sort of audience to help, like help us promote it?” And all of them were like, Look, this is awesome. You've been a huge champion of us for a long time. We're happy to support you. So what I did, what the unique superpower we had was we had access to fandoms and what I, what my hypothesis was that if I went to them with a clear ask, which is help us promote the song that they would and, because of our relationship.

And so we were able to line up a little over a million fans across different fandoms to like really promote our song the day that it released. And so the day that it was leading up to the release and the day it did release, we use this app called Renaissance App, which is like a fan engagement app that focuses on this thing called like a streaming party.

And what we did was we told all the fandoms come to this app, set up streaming parties for “Drinkin’ Drinks”, and tell all of your fans to come here and stream this song for us. And so that worked crazy well, I mean, we had just I mean, that was honestly like 80% of our first week place, just fans coming from everywhere.

And so if you looked at it at the time, it would be like, Oh, “Drinkin’ Drinks,” it's like song artists related to this that other fans like would be like BLACKPINK, BTS and you're just like some some Latin and some like, you know, like Latin artist and it was nothing to do with country. And so I'm on the other end if I'm, if I'm Spotify or if I'm one of these DSPs I'm looking at going like, “this is clearly fraud.” Like if I'm on their end, I'm probably thinking this is fraudulent.

Now, I had seen kind of I saw this happening and I reached out preemptively and said, “Hey, guys, this is what we did. This is not fraud.” But just because I was actually worried we'd get our account banned, not because it was wrong, but only because it looked so like an outlier.

And I think when things look strange, you know, if you're just looking at the data you have to make a decision like, do you think this is fraud or you don't think this is fraud? And that's kind of the two-edged sword here, is like as an artist, you want to be able to break out, which means you have to do something different. You need to get noticed. But you also don't want to risk getting in trouble for that thing. And so there's this kind of like duality to it where it's hard to, it's hard to navigate that and say, okay, what can we do that makes us get noticed? But is it a violation of terms and conditions and isn't like scamming somebody?

So how do you like, how do you do that? It's kind of cool because you can be creative. But anyway, my point is, if you're fraudulent or you engage in fraud, you will get caught. There's companies like, Beatdapp is what we do, like catching, catching fraudsters and anomalies. But I think that sometimes it's just hard to tell that story when it's in data, like there's some analysts or some safety and security team on the other end saying this is weird. All their streams are coming from Sao Paulo, Brazil, Ecuador, that doesn't seem right. And in my case it was and I could prove like, hey, look at the, look at the 40,000 retweets. Like, like you could see each pocket like a community and engaging and so yeah, that was... It's just hard. I think if you're going to do something crazy, you should keep documentation so that way you have something to prove, otherwise it's going to be a struggle for you.

But yeah, that's not to get down on people with that. And then the other thing I did really heavily is I started using, we used NFT’s for all of our photos because it was important for me that we went through the journey, but none of us got recognition. Like I wasn't trying to be a famous artist. I wanted to really understand what it's like to be an artist.

And so from that perspective, everyone who hopped on the song we gave NFT’s to, we used NFT’s as our art. And because of that, coincidentally a lot of NFT communities bought our song. They purchased our song for download because of the NFT’s. And I think the miss eye I had was, had I had more time, I would have spent less time doing TikTok exploration and done more time reaching out to NFT communities and trying to get the communities there to buy.

Because, you know, one community bought like 3000 downloads of our of our song. And like you, Morgan Wallen that week had done like 5000 downloads. So just one community was so responsible for a ton of our downloads and that was pretty consistent.

"as an artist, you want to be able to break out, which means you have to do something different. You need to get noticed. But you also don't want to risk getting in trouble for that thing."

NS: And is a download as well worth more than a stream. Do you know how that works? Like if you're trying to get to number one, there's like one download equivalent to like 500 streams or how does that work?

AB: I think it’s 100 streams for every download. And then which 100 streams is a lot like that's a lot of streams, like someone listening to your music all day long. And then the other thing is that but that's only on a single, I think on an album, if you release an album and they had bought the album, it would be 1500 streams.

So the trick well, the thing I didn't know at the time was I thought you could release a single and then get the 1500. I didn't realize that it was only 100. That's a learning curve for me. I also didn't realize that you can't release an album with a single. You basically need to release the whole album or like it doesn't count.

For example, like on Apple, the original plan was, “Let's send people to Apple and download and purchase the tracks so that we can chart on Apple (Music).” Apple won't let you do that. Like you can't presell anything until it's already live like, the old trick used to be you loaded on there, you send real fans there to prebuy all these pre-purchases and the day of, you chart, because you've racked up say 2000 purchases, and you cannot pre-purchase a single anymore.

NS: Oh so they stop that, you can't it doesn't, it doesn't all count towards your day of release? What do they do?

AB: The day it releases people can buy it so you then you don't get the benefit of like working say, two weeks to get the buys to go live on that day.

NS: Is it just to Pre-Save now.

AB: Yes, just a pre-save and then they have to go back. So they've added this level of friction and that's to keep people from gaming the chart that way. The only caveat to that is if it's an album, but we didn't have a whole album to release. So part of my strategy, I didn't realize wasn't going to work until like a week before, and that was like a mistake on my end, for sure.

NS: I would've made the same mistake because I've just learned today that you can't do that anymore. It was so standard for so long, but it makes sense why they've done that, really, isn't it? Because yeah, people could just get by, buying their way in their charts. So this was April this year. Is that right?

AB: Yeah, April 22nd.

NS: So you did that. You know, you're talking about, you know, making sure things weren't fraudulent. Had you started Beatdapp then or did that come off the back of that?

AB: Oh, we started it. So we started Beatdapp in the very end of 2017, beginning of 2018. And the premise for Beatdapp, which is the company that I co-founded and co-CEO, came from actually at one of our labels. So my co-founder and I used to work in the music industry. He worked on the lobbying side. He often jokes that, you know, he's a recovering lobbyist, but he helped extend copyright protection for the Canadian, for artists in Canada from 50 to 70 years.

So he worked on the File for Music Canada that helped extend copyright protection. More on the government regulatory side, I was obviously an artist promotion, breaking artists. We met in grad school and decided at some point we wanted to build a company together. He moved to L.A. at the time, I was living in L.A. and we started going to football games together, like all the time with a bunch of other music guys.

And one of those executives at a label said, Hey, guys, we have no idea how many times this song is played on streaming services. Like we get a report from a DSP that says, you know, artist A did 100 million streams, but we don't actually know if that's 100 million. Everyone in blockchain and everyone in art rights management is focused on who owns what percentage of the song or the royalties. No one's actually checking to see if the number is correct. Like is the number of times that song was played actually real.

In the old days we would know because we'd have trucks distributing CDs. We know how many pallets went out. We know how many CDs fit on a pallet. We know how much gets returned. We know how much breakage there is.

In today's world of digital, we no longer have a view into, like a real view into distribution. So the number they give us is like gospel. That's the number of we then invoice off. If your label and your artist has 100 million streams of 500 million total on the platform. It's called Spotify. You get 20% of the advertising revenue and the subscription fees from that platform.

20% is what you did of the total streams. There's a pool of money, you get that pool of money, your percentage of that. What happens if it was like 150 million streams and you should have gotten way more and like in numbers like, okay, it's an extra 50 million. But that's happening, you know, every quarter for lots of artists. And if you're a major label that actually amounted to billions of dollars potentially missing for the major labels like globally.

NS: Is because they the numbers are not being... because at the moment I'm saying well if they have overreporting surely that's a good thing because you're making-

AB: They’re always underreported.

NS: They're always underreported.

AB: Yes. Almost every, in almost every case they lost, they've lost 10 to 15% of their plays.

NS: Why? But why is that, just sort of get what I mean? Like I'm like, if you've got a million plays and then you come a long ago, actually 20% of those are fraudulent. So it's actually only 800,000. You know, from an honest point of view, you're like, well, that means I'm ending up with less money. So like, how's that?

AB: All right, let me back up. The first problem that we were trying to solve was just transparency and the number of plays that actually occurred. So forget about fraud to beginning. There's like this, this thing called an ETL script. It takes data and transforms it from like one database into a different language or into a different database.

It moves data along because these platforms are global. They're using different servers in different regions, they're using different types of database structures. There is like effectively digital breakage that occurs as data gets transformed or someone's phone dies, it loses all of the that data. So if you were playing an offline playlist and your phone dies, like all those plays go away. They don't get re-uploaded typically.

And so there's a big bunch of these plays that just do, they never make it back to the parent database, let's say that occurred.

NS: Okay.

AB: Now if you go to, if you go to the server log of the DSP, like if you go there and you pull all the logs every time the song was sent out and you match that to the times they said it was played, there's this discrepancy, and almost every time it's like 10 to 15% under what would they what should have been shown.

So like if you, if you're Snoop Dogg and it said 100 million plays, it's likely you probably had 110 or 115 million plays. And so then that changes what you would have invoiced for, because if you're the label, you're doing 100 million to 500 million. So that's a percentage. The fraud piece was interesting because as we saw this discrepancy, that 10 to 15%, we went in there and said, “this can't all be scripts being broken.”

So like, why else are you holding back money if you're a DSP? And their answer to us, well, we suspect a couple of these things as fraud. So even though the song played, we think this is a potentially a bot farm. We think these might be hijacking accounts where your accounts are actually stolen and someone else is playing it to play music to ramp up the number of plays on their songs.

Yeah, there's a number of these reasons why fraud exists. And so we then started looking heavily into how do we help solve the fraud problem, because in order to solve fraud, in order to truly audit, which is what we built an auditing product, you actually need to also solve fraud. So we started building fraud detection tools, anomaly detection tools, and I guess more importantly, setting around standards for the DSPs are like, this is what fraud is.

This is our confidence score on why it's fraud. This is the definition of this type of fraud. And really trying to set like a standard language for all the DSPs as to why they're holding back potential, you know, plays and sort of their justifications for it. So today we have this amazing blockchain product that can audit. By audit I mean like you turn it live and day, it goes live moving forward. We can track every single play in real time and we have, you know, 30 something patents and seven countries for high throughput blockchain processing.

We also have this fraud detection tool that helps DSPs fight fraud. And so with the DSPs, we have a lot of household names that we're helping and we literally, we literally go through we have over 100, up to 100 models running, supervising, unsupervised models, trying to detect all the types of fraud.

And it's like Whack-A-Mole. Like as soon as we discover one, it's always, I'm always I guess not surprised, but almost like in a little bit of admiration with how creative these fraudsters are, we find the craziest stuff sometimes. And I'm like, “Man, that was really brilliant. Like, I can't believe that they thought to do that.” And, you know, so part of me is like, “Wow, that was brilliant. The other part is like excited that we're the ones catching it” because, you know, it's it's a never ending process.

And at the end of the day, these DSPs don't want to be the police of of streaming. Like, that's not their job. Their job is to live to deliver content to the most of us, the 99% of us with frictionless experience and have a really positive relationship with the music we're streaming.

And so our job is to find the fraudsters. And so we have all these models running. We find fraud and it actually helps everyone. It helps the artists too, because a lot of times the fraud, the fraud is like shifting market share, not just for the majors, but for all of the artists themselves.

Like, you know, like you get a percentage of payout as an artist. Your percentage might be smaller, but you still get a percentage based on your plays. And if that is being artificially increased because you have a bot farm that’s running up, you know, sleep sounds like that does not help anybody. That's actually an artist, you know, because a lot of it's ambient noise or audiobooks or other types of things that they're just doing to financially defraud the system.

NS: When you're talking about models, what exactly do you mean by you're running models? Are these like algorithms that like AI’s running? Like, how does that work? And following up from that as well? Like what's an example of, you know, some fraudulent activity that you were kind of like most in admiration of when you saw it?

AB: Yeah. So models there's all kinds of like supervised and unsupervised models, supervised means you train the model, you tell it what to look for and unsupervised means you sort of let it loose with a bunch of data and it tries to find abnormalities or anomalies on its own through like large amounts of data. And so a lot of what we do is supervised in that, like the human eye can look at something and say, this looks like fraud.

Like if I'm looking at data and it shows that one device or why an account is connected to 400 devices in 18 countries in the same week, that's probably fraud. If I look at another one and it says that there's 33,000 plays this week, that's also likely fraud.

If we build another model, we'll look at similarities between users. So it'll look at how one user listens to music and consumes it and how closely related it looks to another user. And so all of them are exactly the same. That obviously is fraud because how you consume music is also like a fingerprint. You know, everyone listens to music differently.

NS: Yeah. Not everyone's going to like, play at exactly the same time on a song, 400 people. Like, that's beyond coincidence.

AB: Yeah. And do that song after song after song. Like that's eight different countries. Like, that's not likely. And then the other one that I think is the more problematic one, is the hijacked accounts. So we're seeing a significant amount of hijacked accounts. And what that means is to give you an idea, because I've said this on podcasts before, so I'll just describe how I would build this and this is how they are building it.

I can go on the dark web and buy 100,000 accounts. Most people will sign on to streaming services. A lot of people use single sign on. So like Facebook, Google, you know, when you sign in through the different application. And so what you can do is you can buy, let's say 100,000 YouTube SSO accounts and then go to different services and just see quickly which ones are connected to which services.

And then when you have that map, so you've now filtered in, you have like 50,000 on one platform, 30,000 other 20,000 another. Then all you do is use those platforms API to say, “Hey, call this, call this, call this account and find out the last 50 songs you played.” And if you do that, like every 2 hours, you end up being able to map how each individual account plays music over two weeks or something.

And then all you have to do is cluster them by inactivity. So put all the accounts together that don't play music in specific hours. That's also just a simple filter. It's not very hard. So now what I have is I know I have your login. I know when you don't listen to music and I know what platform you're on.

So then the last step then is just telling it, “Hey, when you're not playing music, when Nick’s not playing music, turn on his account and go play these songs.” And so what's really hard if you're the DSP, it just looks like Nick's listening to music.

NS: Like, he stayed up late on Tuesday night.

AB: Or 12 to 2 midday because you're at a lunch like you never listen between one hour. And so we see a lot of these days are just hijacked accounts which are… it's actually easier than what it used to be. That's an easier way to defraud the system because it used to be I would make 100,000 accounts and then I have to make a listening history for 100,000 accounts. So it's not obvious.

I can't just only play Nick or all 100,000 accounts become obvious. So if I want in the old days, if I wanted to drive a million plays to Nick, I had to drive 100 million plays to other artists and bury my activity amongst all this other stuff so that you don't know what needle is in my haystack. I'm trying to create a bunch of noise around me, but now you, Nick, you create my own noise. You create the noise for me, like your regular listening is my noise. I just have to slide in to an area where you don't realize your account’s been compromised. And then if your account does get, if you figure it out, I just go to the dark web and buy new accounts and I can buy them for pennies.

And so it literally takes like very little time to steal accounts, set up the infrastructure to do this and then play songs and very few people will know their accounts are being hacked. So much so that you know, someone you know, someone at a service asked us, “Hey, I think my account might have been compromised. Can you look?” And we look and we're like sure enough, your account is connected to like 16 devices in X number of countries or whatever it was, and you have been compromised. But like, you know, they weren't sure until we checked.

NS: They weren't sure even they didn't. That's amazing. Who are the people that are actually doing this? Like where are they based? So they ganged you know, it's a gangs doing this? Is it just bored like 17 year old programmers is like, what's your theory?

AB: It is just theory because I think no one really knows who's behind the computer. I suspect that there's two types of fraud. I suspect there's like bot networks out there that are just trying to sell artist packages. Like you go to a website and it's like, “Oh, you help promote my song, get 100,000 streams guaranteed.” Like that's likely going to a bot farm of some kind. And they all have unique infrastructure to evade getting caught.

So this is just another way for them to evade getting caught, to sell you on streams and make money off of artists who don't know better. I think that's one huge bucket. The other bucket is financially motivated fraud and that's like all of these shell corporations that are specifically trying to target DSPs who maybe they think they can get by.

And, you know, there's a couple of ways they can do it. They can steal an artist's song and slightly change it and then reupload it. So that way it gets past to sort of content ID mechanisms. And then now all of a sudden, as they're driving plays to it, they're getting paid. There's people that will take, you know, like sleep sounds or things that are ambient noise and try to drive plays up to those so that they get a percentage of revenue from the royalty pro-rata payouts.

So I think there’s a combination, there's people who are financially motivated to do the fraud because they can build a network, they can target the songs they want and they can recoup cash on the other end. It's very direct. You play the song, you get money, you play the song, you get money. It's very few other than advertising where like there's a click or there's a view you get money.

So I think, I think in a lot of ways music has just become large enough that now people who would defraud e-commerce or defraud, you know, you try to run fraud schemes on, you know, like advertising or PayPal or whatever have just moved into music. And I think as we start kicking them out, we become an industry that's harder and harder to scam and they just move to a different one.

Like these are just like professional. There's a huge bucket, I think, of professional scammers who see this as a gravy train and are just like pulling money out of the system and not getting anything back.

NS: I mean, you also, you know, there are a lot more third world countries that are coming online. You've got an Internet connection, you've got a computer, your brain screwed on, but you know, you're destitute. How can I make some money? What can I do? You know, this might also be a way of like, you know what, you know, talking around like, you know, the water pump or something. Hey, have you heard? This way you can earn some money and then you can get paid by crypto and download it like I would imagine, you know, if you're quite entrepreneurial, this would be a way to go and do it, you know?

AB: Yeah, there was, there was actually my co-founder Morgan, he’d be able to record, I’m horrible with names and stuff, but there was an article that came out about this kid in the Philippines and he had to hit a certain amount of plays per day, and if he didn't, his fans would come off and he's in this like hot sweatshop with all of these phones trying to hit, trying to get plays going. And if he didn't hit his targets, they turned his fans off on him as a punishment. And like, this is like a very real thing that's happening in these emerging areas and markets.

And there's ways to tell, like in that kind of instance, how we would tell is sometimes we get data from DSPs that's like phone battery, gyroscope. So if the phone's never been dead, that's a problem. Like if the battery never depletes, like that's also like indicative like who hasn't had their phone dropped below 50%? Like that's, that's like not normal.

If the gyroscope on the phone never leaves and so it's never been in a car or on a bike or anything walking because you can tell the difference, you know, like that's how all these apps work. You know, if you're walking, driving, walking, there's a gyroscope in your phone. If that's not moving, then why are you never going anywhere? Like, that's also indicative of fraud. So there's like an absence of things that also is an indicator.

"Someone at a service asked us, “Hey, I think my account might have been compromised. Can you look?” And we look and we're like sure enough, your account is connected to like 16 devices in X number of countries or whatever it was, and you have been compromised. But like, you know, they weren't sure until we checked."

NS: Yeah. Wow. That's some. Yeah, that's amazing. I didn't really think about those sides of it. So which labels and well I guess you know, I know you're working with some like you're saying household names, you know, but are you able to tell us about a particular artist or label you worked with that, you know, you showed them the real playground and they were like, whoa, you know, you're sort of almost quite proud of working with and showing them that sort of side.

AB: I think no one really wants to wear the Scarlet Letter right now. So most of our relations that are strongly in the aid and we actually can't public, it's like one of the hardest thing not be able to publicly share really who we're working with. So I would say that for DSP’s it's the summary I hope that people realize here is there's a lot of fraud. Like officially, I think we're now saying 10%. It could be much larger in terms of all streams like it is a, it's a large number, but it's not the DSP’s fault, like they are doing everything they can to stop this fraud. There's like fraud teams internally. They're having companies like us work on all this sort of anomaly detection, like we have spent years building these models.

I think we're the best in the industry at finding fraud and we help a lot of them find the fraud. But it's Whack-A-Mole. Like as soon as we close one pathway down, these guys find another pathway. And it's just about closing all the pathways that are the easiest. And at some point they switch into a different industry or majority of them try to leave.

So I think that, I think that it's not necessarily a DSP fault, but they're happy we're there. It's the first time like when we built the auditor product no one loved when you showed up and you're like, “Hey, I'm the friendly auditor from down the street just to make sure the numbers are correct.” When you show up and you're like, “Hey, I'm the sheriff to find all your fraud,” it's just a different conversation.

And the DSPs are way happier to engage with us and say, “Hey, we trust you.” You know, the other thing that was important that Morgan and I did, and Pouria, our other co-founder who runs all the tech like, none of this would be possible. He's like the genius behind all of this craziness in terms of the technology stack and what would be, what was crazy to be getting was we said, okay, we can't take any money from labels and we can't take any money from DSPs.

We have to stay completely neutral and be like a neutral third party no matter what. And that was really hard for us because initially all the money from people that were interested were from labels and from DSPs. And so trying to convince regular investors that there was a chance in music when there's this graveyard of companies that are all dead in music, investors hate music tech investing. And so we had to convince people that we were the real deal and that they needed to invest in us too, so we could maintain our neutrality.

And that was really, really hard to begin with. And I think now that we have it, it makes it way easier when we go to talk to a DSP and say, Look, we aren't owned by anybody.

We're here for authenticity, we're here for transparency, and we're here to help, like, we are the firefighters. And it is like all hands on deck right now. And that's the reality. I mean, there's services out there that have publicly said at certain periods they've had up to 50% fraud on their platform, like, it is a big problem right now.

And I think that it's just getting larger and larger and larger. And luckily, I think we're here to help a lot. And I think that, you know, everyone's receptive to it at this point. So hopefully my answer to you is that in another year I'll be able to come tell you a list of all the companies we've helped, you know, but I would say that, yeah, I think it's becoming more of a norm.

Like they're starting to show up on panels with us now. You know, where it's becoming more… people are becoming more open about it. But definitely we've been kind of, you know, everyone's favorite secret friend for like a couple of years. Yeah.

NS: So, you know, you talk, you talk about the, you know, the bigger household names and the bigger majors and, you know, that they do have to deal with Spotify where they get a percentage, etc.. What about, like independent artists and labels, you know, genre-based labels? What's, you know, why is this important to them? They're not getting that kind of percentage share.

Yeah. Why is it important to them and how can they engage? What's the best way of engaging with your platform? You know, what's the best strategy for someone like that? And I'm talking about an independent, maybe 6 to 7 employees, maybe turning, you know, around about a million a year.

AB: Yeah, I would say as of right now, we aren't the best at helping them. We hope to have a product for them soon. It starts with the larger labels because they control so much of the like negotiating power with the DSPs and making sure the data flows like there for us to analyze and or provide back. So I think that, I think we will have a sort of product for the smaller labels within the next year.

At least that's what we're working towards. And so we're a little early for them now, but it's because the problem isn't like the absolute dollar value for their problem doesn't materially change their like it might be $100,000 for which we can charge $10,000. But to support that, you know, might cost us $20,000 or something. So until we're at a place where we have sort of all the pipes laid and it's just an incremental like costs for us to turn on, it's hard for us to support these smaller labels, but I think that we're getting to that point. I think in the next year or so we'll have all of that dialed up. So I would say if you're an independent label and you want to use a Beatdapp service, then just hold tight. We're coming soon for you.

I think that there is value, though. Like the value is like if you're an independent label and you can get auditing correct, you're probably instead of a million making 1.1 million. And if you can reduce the fraud and shift market share back, you probably are making 1.2 million. So that's like a 20% lift for you, which is material when you're making $1,000,000 a year.

But in absolute dollars, it's still only 200,000, right? If you're a major label and you make $1,000,000,000 a year, $200 million is the number of… just the dollar numbers they make are so massive when it comes for the larger labels, which is why we're just trying to like fight one fire at a time as we sort of get all the groundwork laid out.

Just because yeah, it's hard to explain. But every label is different data and building models for that isn't just like you turn it on. It's not one time, it's not out of the box. Like we get their data, we train like thousands of like… we have to go tell it what to look for thousands of times per model.

And that actually takes like a month of all of our team's time. And so to do that at scale for smaller labels is really difficult right now. Once all of that has been laid and done and we've sort of, let's say, done most of that for all these big labels and all these big independent, the model learns from itself and it becomes infinitely easier to then turn on a smaller label without the human cost of like a whole month to train a model.

AB: And so that's, that's my point, is that we're nearing that as our technology, it's smarter and smarter, and that will unlock our ability to help all of these smaller labels.

NS: What it may be as well be where you work with the dealers, the distributors, and then they have inside the distributor an option to say, do you want to switch on like do you want to switch on Beatdapp and we'll do this for you, like, and then that you're sort of consolidating a big distributor that has like you know, a thousand labels, you know, especially if it's a genre-based as well, you know, like Label Engine, you know, all electronic music, say like it might be an easier way of doing that, especially if your models.

AB: Now I think that is a really valuable input. We also agree. I think the hard part is that some DSPs or some distributors works really well for because maybe they own the license for that. A lot of times they're just the pipe and the owner of the data and the owner of the content is still the label. So it just depends on the set up with the distributor.

But we are trying to we are trying to build products for distributors as well. We also have this like universal for distributors specifically, this idea of a universal list where like what happens now is if you're an artist that gets kicked off a platform or because you're fraud, say you will just like change a couple of metadata fields and then go to the different platforms.

You might go like TuneCore, Distrokid, Symphonic, just sort of like work your way through. And so what we're building is a universal list so that distributors can ping our database and say, is this artist fraudulent before they upload the music? So we become like part of their QC flow. So we're starting to work with distributors now on that kind of stuff and we're hoping to, at the end of the day, we just want to help everyone catch fraud and make sure the numbers are right so artists get paid correctly.

And you know, like I said, when you asked about independent artists, it made me think of the whole reason went back to doing Urban Outlaws, which was we started to help artists. We realized we needed a stall like the backbone of the industry's problems for us, which is these major labels and DSPs and stuff. I wanted to do a song, so I understood what it was like to go through that process so that I could say, okay, how do we make it easier for artists?

Like what of our technology stack? And we then unlock for these different artists and independent labels and stuff. And so again, part of me doing the song was not about me doing the song was me understanding their journey. And I think that's what I hope. I hope that that is the thing that leads us to being able to help all of these independent artists, because I think I've learned just so much about like, how do you get stuff distributed? Where do you go to promote your song?

Like already there's probably ten new partners I can think of for Beatdapp that I wouldn't have thought of in January. So it's just, it's interesting to me, like distributors is a good example where I wouldn't have thought of that necessarily because they weren't our core customer at the time. So yeah, I'm excited because I think that I think what we're building is going to be useful for the entire industry and I really think it's going to change everyone's lives.

And I think that's the if you work as hard as you work to make a song, you deserve to get paid. Like that's just the reality of it. And right now it's being eroded little by little from all these different angles. And I'm just hoping we can sort of combat one or two of them.

NS: Firefighter.

AB: Exactly.

NS: So, you know, you have you through the journey as well of what based all your industry experience and as well as moving as well as releasing the music as well. You know, big thing that we talk about here is, you know, how do you monetize music? I'd love to ask you, you know, there are so many platforms and tools that promise to help monetize, you know, music for artists.

You know, what are the what do you think are the biggest needle movers for artists that they should be focusing on to, you know, essentially monetizing music to get to the point where they can hopefully, you know, pay their bills, you know, and do music full time.

AB: I think I have come full circle on this. So I don't think there's a silver bullet like.

NS: I know, you're right. I don't think there is a silver bullet. But, you know, there's spokes to it, right? There's lots of little...

AB: Sure, I guess what I mean is nobody wants to do the sort of old school way that I remember back in the day when I was watching artists. I would talk to alt rock guys and I would say like, why do you send them on tour for four years? Like, they seem like it in all of our pop genre, you know, artists, for example, they'd go in the studio, they make an album, we gear up to promote the album, we release the album, we push sell through, help them sell tour dates.

But it's very like, you know, 24, 16 months, it's out, you know, like you're going and they make it or they don't make it. Why are you sending these alt rock groups on four years of tours? Like it doesn't make sense to me. I didn't understand. And the response back from they're kind of old school rock guys was like, look, fans especially in like alternative rock, need to feel like they found something they could really discover with you. And part of that is the relationship they build with you on these tours, like when you're doing a show and there's 100 people in the in the room and you stay after to meet 100 people and befriend them. They're like, “Man, I, I knew Chad before Nickelback blew up. Like, I fucking knew that they were going to be huge.

I called it like, love those guys.” And that never goes away. Even if they try to cross over and become big later. And so what I thought was really interesting to that approach is in a lot of ways it's just hedging their bets. Like if you're a label, you're putting them on the road because you're building a true, authentic fan base of all of these real fans.

And if you don't make it commercially, like, let's say, when you push to radio and try to do a bigger push, you still have 32 markets to go sell tickets to once a year for, you know, for a tour and you can live still touring and make the money back, the label at least will make their money back from that initiative.

So I think the same thing is true in artists today. Like if you think about it, all the distribution platforms are so passive right now. It's like maybe other than TikTok where they create music or create something for your song, but like playlisting and they're just listening and it's ephemeral. It's like it comes and it goes. And no one I couldn't even you what songs I knew, I knew I liked two or three songs yesterday. I have no idea what they were, you know, they just were playing and I, you know, so that happens all the time to people and I think that, like, it's all of the sort of like build the true thousand fans for yourself. Like if you can build a thousand true fans that will pay your bills and I think that's true. Like can you go and do these shows? Can you rock out with eight people on stage with eight people you know, at your show when you're on stage the same way you would if there's 40,000 people and make them feel like they've had an experience. Talk to all eight after the show and make them your biggest fan and do that over and over and over and over again until you have a thousand fans lined up.

I think if you can build a thousand true fans through blocking and tackling, which is just like going and creating the shows, meeting the fans, building rapport with them, shaking everyone's hand, taking photos, all the hard stuff that no one really loves to do. It's exhausting. You get sick a thousand times because you've shook so many people's hands and everyone's coughing on you and hugging you. It's like actually, like difficult. But I think if you do those things, that is how you build a lasting career. It's not how you quickly became popular for one second, you know, and then hopefully took off like that can work too. But that's less likely. I think you just you keep building, you build the building block, you keep adding a layer, adding a layer, adding a layer, adding a layer.

And over time, I think you you build this really massive true fan base. And so to me, when you ask what tools I would be looking at tools that help me tour, I'd be looking at tools that help me manage my merch. I'd be looking at tools that do like the physical thing that I'm engaging with those fans and I'm a digital person saying that, and I'm saying that because I really truly believe that is the way to building like true economic value as an artist.

I know, I know an artist. I won't name him because he'll be upset. But he was really famous for a couple of years and he, you know, it's funny is that he's still selling out shows everywhere. Like everywhere he goes, it's small shows, maybe it's 500 or a thousand seats, but he still makes a living today, 15 years later, because his fans love him so much. And that's like, that's incredible to me. He hasn't had a hit in years, but he can provide for his family. He can, you know, he owns a house, he pays for his kids, like is he's like a legitimately working artist who maybe just did it the hard way, you know, like brick by brick by brick.

But I think at the end of this, when you say who's lasted 40 years and this guy's made money for 40 years touring like not yet, but in that time frame, I think you look back to that and say that was probably a better journey than somebody who went really, really hard, really fast and burned out quickly because it comes and goes. Even the artists we broke, like I told you about that I think I told you went number seven on Billboard. You know, he was gone within like two years. Yeah, like we sold the publishing and all that kind of stuff. But, you know, he's irrelevant now. And so I doubt fans are buying his tickets, you know what I mean?

Like, that's just the reality. And so I think that's what I, if I'm an artist trying to build a real career and real economic value for myself, I'm thinking, how do I set up tours? How do I reach venue, talent, talent buyers? How do I book my own shows? Easier, because I'm doing it myself, likely as an independent.

How do I get my own press write ups to try and get exposure? How do I make cool merch that people actually want to wear so that people talk about me? Like I'd be focused on that level of stuff. You know, that's about this is my soapbox, but that's, that's I honestly think that those are the things no one wants to do. Everyone wants to like, release a song, get on a major playlist and become famous and like, that's just not how most of this works. And the ones that are getting on the playlist have probably done all the hard stuff also. I think, you know?

NS: No, I mean, I completely agree that is the way you have to do things. And I think the one thing that's stuck out is build authentic fans, you know, and it's still true. I still say a better thing to fans. You can get a thousand fans to give, you know, “superfans” is the term I use, $100 a year, which isn't a huge amount of money. Like if someone's a superfan, I spend, you know, I love The Prodigy and The Chemical Brothers. I spent $100 on them at like more than $100, especially if they're coming in. I'm going to a show. I've already spent that on just going to the show, let alone buying t-shirts and vinyl and all that kind of stuff. That's 100 grand straight away. And, you know, and like you said and you know, I will I mean, they're not in the charts anymore, but I'll support those guys until I'm dead. So yeah I completely agree with what you're saying, like build up those fans, those super fans and use the tools that do that.

Right. Well, we're bang on an hour. I wanted to quickly ask you, the artist that you launched in April when you were an artist because it came off the back like, you know, when someone asks you what you do, it's like, well, what day of the week is it? Because technically, at one point you could have said, Well, I'm an artist. That's, you know, #3 in the charts at the moment, #4 in the charts. What was the name of the, what was the name of the act?

AB: It's called Urban Outlaws. And the song was called “Drinkin’ Drinks.” Drinking Without the G, D-R-I-N-K-I-N’.

NS: Alright. And Andrew, where can we're like, where do you hang out online? Where can we find you?

AB: You can find me on LinkedIn or Twitter @andrewbatey. You can find me… I'm also a crypto “degen” and so you could probably find me on Telegram and Discord. I just whatever, just, you know, I'm always open if people want to ask me questions or hit me up. I love helping people. So if anyone's listening and wants to reach out, feel free. Like I'm happy to lend a hand.

NS: That'd be awesome. I mean, we didn't even get to touch on any of the crypto, NFT stuff in it. We briefly mentioned that would be, I think, a whole other hour. So maybe next year I will. You know, we can get you back on and we can, you know, especially if you've got something new for the independent artists. So yeah, thank you so much for being here. It was an absolute honor. Yeah. Cheers

AB: That was rad! I appreciate you having me. Cheers.

Available in all major podcasting platforms.

Other Podcast Transcripts

Latest The Label Machine Posts