Fraud Detection Software Is Coming to the Music Industry — And It Will Change Everything

Music Industry Fraud Detection Software Reading
  • Save

Music Industry Fraud Detection Software Reading
  • Save
Original data vs. Manipulated data. Can you tell the difference?

Up until this point, it’s been extremely difficult to detect fraud in music royalty statements.  That’s about to change in a very serious way.

Since the early days of the music industry, there’s always been a very serious problem.  And that problem is outright fraud.  Simply stated, if a label hands an artist a statement, that artist has never had a way to investigate the veracity of the numbers reported.

Even worse: an audit against a label (or publisher, or touring company, or manager) is incredibly expensive to conduct.  For starters, the auditors themselves need hundreds of hours to properly scour a massive flow of royalty statements.  Then, those auditors (and lawyers) have to do battle with accountants and lawyers on the other side.

Those opposing accountants and lawyers are paid to make that process more difficult and expensive — and give up the least amount to the auditing artist.  It’s really hard to win — or even start the battle.

Now, the problem has compounded in complexity.

In the analog era, there were only so many things to count and report.  LPs, CDs, and other physical items like t-shirts and ticket sales comprised the limited list.  Now, labels and artists are forced to manage billions of streams, while trusting the platforms that report it.

On the publishing side, the math on public performances and mechanicals has always been incredibly complex, with entire organizations like ASCAP created just to count one license (in one country).  Now, that complexity has become arguably unmanageable, with formats like non-interactive streaming (i.e., digital radio and satellite radio) requiring entirely brand-new organizations (like SoundExchange) to keep track.

Admirably, tech-focused companies like INgrooves, Kobalt, STEM, HAAWK, Audiam, and oneRPM are developing solutions to help organize and manage this data influx.  But what if the numbers themselves are fudged?  Or, simply reported in error — because of simple human error?

Enter Austria-based Rebeat Innovation, which has now developed a fraud detection software solution to solve this issue.  The solution is currently in beta, with version 1.0 expected later this year.

Rebeat’s system uses complex algorithms to determine if fraud has likely occurred.  The company demonstrated the beta-stage solution to Digital Music News this week in Los Angeles, and it’s pretty fascinating.

It turns out that fraud is very difficult to detect.  In fact, most fraud is purposefully hidden to avoid obvious detection.

In the demo, I was presented with lots of different royalty graphs showing a variety of combined royalty data.  For example, one involved a combined graph showing Spotify, Apple Music, Google Play Music and other streaming platforms mapped against iTunes downloads.  And every spike and aberration that looked like fraud actually turned out to be legit.

In almost every case, the fraud was buried into smaller, less-detectable crevices, ones that the naked eye would never pick up.

Rebeat’s CEO, Guenter Loibl, has been managing digital music royalties for more than a decade.  His company has built massive systems to ingest, analyze, and calculate royalties from surging streaming plays.  But he told DMN that he’s never had a way to inspect the actual streams and royalty statements handed to him — even though clients were often wondering if these numbers were legit.

“In our study, we show that the analysis of the fluctuations of the proportions of these revenues over time is a highly efficient way to identify fraud.”

That led to a major, multi-year collaboration with researchers at the Vienna University of Technology (Technische Universität Wien).  The result was a massive relational dataset of billions of inputs, all thoroughly analyzed to form a comparison benchmark.

  • Save

  • Save
Vienna University of Technology

For those into data geekery, the researchers at Technische Universität Wien employ a concept called ‘Compositional Data Analysis’.

This is already a cornerstone of fraud detection solutions being deployed in other industries.  “Compositional data (CoDa) analysis has been applied in quite diverse areas such as biology, demography, ecology, economics, geology, politics, chemometrics or even sports,” the research team explained in an unpublished white paper shared with DMN.

“CoDa is applicable when the structure of the data is such that the relevant information is contained in the relations between the compositional parts.  In the digital music industry, these parts represent the revenues from different platforms for a given artist, song, or label (entity).”

Basically, when millions of data points are analyzed, the relation between those points become extremely predictable.  “In our study, we show that the analysis of the fluctuations of the proportions of these revenues over time is a highly efficient way to identify fraud,” the team continued.

“In other words, even though the total revenues for a given entity varies over time, on average in the non-fraud case the relationships (the platform proportions) stay stable.”

What YouTube, Pandora, Spotify, Apple Music & Amazon Are Paying Artists In 2018

The results are pretty fascinating.

As mentioned, fraud detection is actually a serious discipline in other industries, so a number of interesting discoveries are transferrable here.  And lots of this stuff is candy for the tech-geek mind.

For example, if a certain number (like ‘9’) appears too many times in a statement with a certain regularity, it’s almost always fraud.  There’s such an incredibly small chance of it happening naturally.  In the world of non-fraud statements, certain numbers simply appear less frequently.

Another red flag comes from movements on one platform that aren’t accompanied by another platform.  For example, a surge in streaming on Spotify is almost always accompanied by similar movements on related platforms like Apple Music and Amazon.  If there’s a Spotify exclusive, then a big disparity is explainable.  Otherwise, it’s probably doctored data.

And by the way: labels, publishers, and venues aren’t always the ones committing the fraud here.

In many cases, artists themselves are buying fake streams or purchasing huge amounts of their own music on iTunes.  ‘Playlist payola’ is another emerging fraud, with Spotify one giant target.  All scams that get detected within two seconds using this software.

Ultimately, Rebeat Innovation estimates that their fraud detection solution will carry a 94-5% accuracy of detection rate.  Over time, that percentage will climb much higher.  But there’s still some work to be done.

At this stage, Loibl says that the software can detect ‘irregularities,’ or aberrations that require further investigation.  It could be fraud, or simply an unexplained surge.  In one case, for example, the system triggered an alert based on a calendar-counting disparity involving iTunes.  One service was counting February one way, another service was tracking it differently — but no fraud.  It’s just another aberration that the system needs to learn.

Rebeat is planning to release its first iteration for recordings later this year.  Then, it’s on to solutions for publishing royalties and potentially other areas.

How I Got 10,000 Spotify Plays For a Totally Fake Song

Already, this is incredibly valuable.  That’s because in current auditing processes, forensic investigators are typically looking for signs of fraud, then digging deeper.  So irregularity checks already address the first step.

In the future, Rebeat envisions an industry in which every statement can be checked for fraud.  Effectively, this could spell the end to doctored, dirty statements.

Usually, when I hear bigger claims like these from entrepreneurs, I roll my eyes.  But in this case, I could actually see this technology making the music industry a better place.  And just like cameras in a bank, the presence of this technology will reduce fraud before it begins.

Right now, discussions are already happening with a few major partners.  But down the road, it’s easy to imagine that a large group of labels, distributors, publishers, and management groups will use the system to vet every statement received.  You could even ‘fraud-check’ statements with a monthly subscription.

But before that happens, I can also imagine a few lawyers who might want to employ this system for their clients.  Previously, fraud detection software solutions have been absent from high-profile royalty disputes.  But this could make all the difference in a multi-million dollar case.


Send confidential tips to  Written while listening to Joey Bada$$.

6 Responses

  1. lol

    that fritzl nation at it again lol.
    it won’t change much, actually next to nothing. Nobody is ripping you off anywhere, bigger labels = no.. distros = no… stores = no, accurate
    ‘shady labels’ = the only real candidate for this
    artists buying fake streams = stores will ban = no statements at all
    artists buying ituens downalods = depends, but technically buying songs is allowed, apple will probably pwn you andas with fake streams you will lose money

    what will change for the average artist, listener, bigger label, stores is almost nothing.
    Nothing revolutionary here paul is excited about checking out data, but at the same time trashing every store so much seems like he wants to have no data to check at all

  2. dontsmoakdmn

    Also Aliens n Flat Earth….
    we need more conspiracy theories bruhs

  3. Atherean

    (damn captcha, if double post, delete)
    advertise on deezer , spike = fraud
    advertise on spotify, = fraud
    get featured on a huge playlist = fraud
    advertise with google ad words google play = fraud
    bunch of offices listening = fraud
    advertise iTunes with iAd = fraud
    Advertise youtube with just one link = fraud

    So anything that doesn’t look ordinary gets an audit , legit traffic hackers … This means extra work, that is going to pay who exactly.
    And let’s say the data looks doctored , but you have no proof , how are you going to prove that it isn’t an anomaly /rare/ unknown ad techniques or that is actual fraud? You gotta have proof other than just stats.
    Based on the average statement? It’s easy to do this as a human… Distros will ban music and account as soon as they get 2 or more reports from stores about hacking.
    Plus you could always distribute to just one store, that way there’s no all stores comparisons (spotify bump but nothing on other stores example).
    Then you have stores that actually have only 2 pay rates (family or single sub- nothing else)
    Spotify payola though, that’s pretty hard even for this tech… Those all look legit streams, since they are legit, except for being payola, which isn’t legit. But someone explain to me how u gonna find out the streams were from a payola playlist, with legit people from all over the world listening .. Only way to prove that is to have proof of transaction to payola services or direct proof your song was on a payola playlist (both utopia) . This can easily mix with legit playlist pitching. The answer is not this, but prevention. Shut down all payola services.

    This sounds nice, however it’s far from as easy as this article is trying to say.
    Much better thing as in any field , medicine whatever, is prevention. And prevention is already in place , much better betting on preventive measures than getting a bunch of data that has gone thru already and all the mess that it brings with it.

    • Paul Resnikoff

      I understand the skepticism, but I think it’s overblown. The first stage is detecting an irregularity, the second is identifying fraud with a high level of accuracy. Keep in mind: ad buys, playlist updates, etc., these things happen all the time. And they are embedded into the dataset that is being used here. That dataset will tell you that even a major playlist inclusion on Spotify will bleed into other formats, often within specifically-defined delta ratios.

  4. Charles

    See what happens Paul? You write a solid article on a pretty spectacular technology and it gets trashed and discarded. All because you put so much focus on making your brand a gossip and nonsense entertainment site rather than news. Maybe you don’t care, or maybe you should split the sites into 2 so you can split the audiences. I for one would gladly unsubscribe here and follow the news site.

  5. LDAFL

    Jul 24, 2018 iTunes Match – Americas Sale $0.00 2 plays
    Jul 24, 2018 iTunes Match – Americas Sale $0.00 2 plays

    4 plays = $0.00

    = Fraud

    In law, fraud is deliberate deception to secure unfair or unlawful gain, or to deprive a victim of a legal right.