There ’ s no exact number , but the most commonly-cited figure is $ 300 million USD per year . That estimate originates with Louis Posen , founder of indie label Hopeless Records , who in 2019 did his own investigation and told Rolling Stone , “ My sources think that three to four percent of global streams are illegitimate streams … That ’ s around $ 300 million in potential lost revenue moved from legitimate streams to illegitimate , illegal streams .”
That $ 300 million figure is conservative , Hayduk thinks , and he also says the percentage of illegitimate streams on certain services and in certain regions could be much higher than 3-4 %.
Here ’ s how Hayduk breaks down the potential cost , beginning with the IFPI ’ s most recent numbers that show music streaming earned $ 16.9 billion in global revenue in 2021 : “ About 10 % of all streaming is manipulated , and that can be higher in certain regions where we ’ ve seen data , and we ’ ve heard anecdotes from DSP execs of rates exceeding 20 % of all activity . Being very conservative , let ’ s assume 3-4 % of that 10 % is not caught by existing filters ( and it may well be more ), which means that between $ 500 million and $ 670 million in actual dollars was lost by legitimate artists and rights owners last year . If the industry hits the Goldman Sachs 2030 [ streaming ] growth targets of $ 89.3 billion , and nothing is done to slow down streaming manipulation , we ’ re into the billions of lost revenues .”
Going back to 2018 , the original business proposition by Beatdapp ’ s co-founders – Hayduk along with CTO Pouria Assadipour and Co-CEO Andrew Batey – was to audit streaming royalties for labels and other rights holders .
“ We started there and built some really cool tech to help facilitate auditing at scale . But something we discovered along the way was that one of the biggest drivers of discrepancies between the sales reports a label sees and the underlying data that exists on a streaming service ’ s servers is fraud . And the reason it exists and the reason that it shows up in those two reports is because each DSP is kind of responsible for policing this themselves . There are different tactics and techniques applied , and obviously different size and composition of datasets , but every DSP was trying to police this stuff themselves and finding significant percentages of fraud happening on their platforms ,” says Hayduk . “ When we discovered that , we realized that if we were ever going to be really good at audit , we ’ d have to actually be great at finding fraud first . That took us down the path we ’ re on right now , which has been , I think , really welcome from the industry .”
Beatdapp won ’ t reveal which streaming companies or how many they ’ re currently working with – Hayduk simply says “ a handful ” – but they do say the focus on the DSP side has been with the larger second-tier streamers that have 50 to 100 million monthly active users , meaning not yet Spotify or Apple Music .
“ No one is yet ready to be the standard bearer of the industry as first-to-market to adopt the solution . I hope that changes over time . As a streaming service , I think there ’ s a real benefit to acknowledging that you ’ re working with someone like us , in part because it ’ s a great signal to your content partners that you ’ re taking this issue really seriously . And the relationship , obviously , between content partners – major labels and the major independent labels – and the DSPs is a really valuable one . I also think there ’ s value in it on the customer side . It might not be a customer-facing marketing message to say that ‘ we ’ re the best at fighting fraud now ,’ or ‘ we ’ re getting better ,’ but it shows up in lots of subtle ways that actually influence the customer experience on your platform .”
The secret to Beatdapp ’ s technological and methodological approach – which exists on the blockchain and utilizes machine learning / artificial intelligence to identify suspicious listening behaviour in the data – is what they term “ cross institutional analysis .” Essentially , monitoring listener behaviour in the data of all partner DSPs in order to find trends , rather than each company ’ s data being viewed in a silo . ( Sidenote : here ’ s our favourite testimonial quote on Beatdapp website , courtesy of Scott Cohen , chief innovation officer at Warner Music Group ; “ Beatdapp is a bright spot in a sea of blockchain shit !”)
“ As we started building that tool out , I think we gained credibility and trust , because you get one partner on board and you find all the fraud they found , plus an additional X percent that they didn ’ t know existed at a 90 % confidence , and the next logical question from them is , ‘ If we give you more data , how much more can you find ?’” adds Batey to Canadian Musician . “ The creativity of fraudsters never surprises me anymore . Sometimes you ’ re like , “ hats off , that was pretty genius .’ But we ’ re getting way better at catching them . I think we ’ re so far ahead of what any individual entity can do in terms of having their own team because we ’ re looking at billions of streams and our models are learning faster and faster and faster . And if you ’ re just one solo DSP , you only have your data set to look at and you can only catch what you see in your singular view . Our models are just far more sophisticated at this point than what , I think , a lot of DSPs or other data partners can possibly build .”
The biggest obstacle to ending streaming fraud is not technological ,
ANDREW BATEY
Hayduk says , attesting that their 17-person company could handle every DSP ’ s data tomorrow if they needed to . Rather , in order to collect the data from streaming companies to identify fraud , it requires
MORGAN HAYDUK
a level of cooperation and trust that is not currently seen between labels or streaming companies .
“ Over the longer term , this has to be a collaboration between DSPs and labels with us , ideally , in the centre of it all , because there is only so much you can do when you ’ re only looking at your own slice of the pie . And so , it ’ s great because we get to be like an outsourced R & D team for rights holders , and help them deploy more sophisticated models , and give them access to a team of data scientists and engineers . Even if you ’ re the biggest label , it probably doesn ’ t make sense to have a team of 20 data scientists and engineers , internally , who are solely focused on your data and looking for streaming fraud ,” says Hayduk . “ In a lot of ways , music is just a big pot of money that ’ s exploitable on the internet , when you reduce it down through the eyes of someone who ’ s engaged in fraud and other verticals . And so , we ’ ve tried to think about it through that lens and ask , ‘ How can we make these platforms harder to attack ?’ Then hopefully divert the attention of these criminals elsewhere into other corners of the internet and keep music as a high-integrity space for creators and artists and labels and everybody else who makes a living in music .”
For Canadian Musician ’ s full conversation with Beatdapp Co-Founders Morgan Hayduk and Andrew Batey – which goes into much greater detail about how music streaming fraud occurs , its impact on the industry , and Beatdapp ’ s approach to combatting it – listen to the Aug . 17 , 2022 episode of the Canadian Musician Podcast .
Michael Raine is the Editor-in-Chief of Canadian Musician .
CANADIAN MUSICIAN 11