back to article The Great Bulgarian Streaming Scam may well have been scummy, but Spotify got paid

An unknown Bulgarian has become the talk of the music industry after a sophisticated and apparently successful attempt to game Spotify. Thanks to a fascinating exposé by Tim Ingham at Music Business Worldwide we learn that the perpetrator may actually have been playing by the rules, bringing to light a little-noticed …

  1. Crisp

    Payola

    Streamed over and over.

  2. Voland's right hand Silver badge

    He did not break the rules

    All I can say is that Spotify should define a better model.

    Applause to the guy who noticed it and did it.

    1. Korev Silver badge
      Pint

      Re: He did not break the rules

      Yep, very much a “Why didn’t I think of this” thing

    2. Skwosh

      Re: He did not break the rules

      Yes. My understanding is that the existing model is (in rough outline):

      (1) Put all the money from all subscribers in a big skip every month

      (2) Spotify takes 30% of the money in the skip

      (3) The remaining 70% of the money in the skip is given to each artist (well, rights holder actually, not the same at all necessarily) in proportion to the number of plays that artist got from the platform total.

      Bad points of this way of doing things include:

      (a) Every month a large amount of my hard earned subscription cash goes to top 50 artists some of whom I never listen to and whose music I loathe with deep and abiding intensity.

      (b) This model is susceptible to being gamed by the method described in the story.

      Actually, another way of looking at this is that the major record companies are currently pulling a similar scam, but on a much larger scale, by mind-controlling impressionable young people with too much free time on the their hands to listen over and over again to a small number of artists almost all of whom are on major labels.

      An alternative, which is often suggested when this comes up is:

      (1) Put the money from my subscription in a little pot every month

      (2) Spotify takes 30% of the money in the pot

      (3) The remaining 70% of the money in the pot is given to each artist in proportion to the total number of tracks *I* play from those artists this month

      (4) Repeat the above for each subscribing user

      Then, if the only tracks I play this month are by The Delve (a hypothetical skint band of which I may hypothetically be a fan) then only The Delve gets my sub money this month – and no one else gets any of it.

      I am led to believe that the major record companies are not keen on this model.

      1. Anonymous Coward
        Anonymous Coward

        Re: He did not break the rules

        I'm sure there's no technical reason why they couldn't do this...as they can track exactly what you listen to through your account...

        ..out of interest, what would you have them do if for whatever reason you didn't play anything for that month of subscription? Would Spotify get the whole 100%?

        1. Skwosh

          Re: He did not break the rules

          ..out of interest, what would you have them do if for whatever reason you didn't play anything for that month of subscription? Would Spotify get the whole 100%?

          Well - if I didn't play anything then I don't think I'd have any objection if they just applied the existing model (so 70% divvied up to artists in proportion to total plays on the platform). I'd rather my cash was thrown to the vagaries of aggregate public taste so that artists would get something, rather than Spotify taking 100%. Ideally I suppose I'd like them to hold it back for next month, or not charge me at all, but either introduces too much complication I think.

      2. Anonymous Coward
        Anonymous Coward

        Re: He did not break the rules

        Lots of people are not keen on the model you propose, including Google, the collecting societies, and the UK Intellectual Property Office.

        What you've just described as the current state of play is Collective Licensing. This is/was justified when the cost and bother of figuring out who should get paid for what exceeded the pay available. It's a relic of the analogue era. Nowadays its possible, via metadata, to log exactly what got played to whom, when, and for rightsholders to be paid accurately.

        How curious, then, that the trend in copyright is towards more collective licensing - throwing everything in the pot and only really paying the biggest players - and away from what digital actually makes possible: paying people accurately.

        As always: cui bono? Follow the money. It mostly ends up in Mountain View.

      3. luminous

        Optional

        It's worse than that. The more plays an artist gets per month, the more the artist is paid per play. It's not a flat rate per play for all, like you would assume would be fair.

        He was probably getting 0.0004 p per play since he was in the top 50. Others get much less. If you generate 10s of millions of plays a month, sure you can get a decent return, everyone else is stuffed.

  3. Fruit and Nutcase Silver badge
    Thumb Down

    Wot not airbags?

    Click-bait!

  4. Anonymous Coward
    Trollface

    "72 million plays, which at a per-track estimate of $0.004p per play (no wonder musicians are starving) may have earned the perpetrator $288,000 a month."

    I'd probably stave on $288,000 a month too. I mean, how many White Tigers Claw Soups could you buy with that?

  5. gnasher729 Silver badge

    Could someone explain to me how much $0.004p is supposed to be? I understand that $0.004 is four thousandths of a dollar, and 0.004p is four thousandth of a penny, but how much is $0.004p?

    1. Fruit and Nutcase Silver badge
      Pint

      gnasher729

      Could someone explain to me how much $0.004p is supposed to be?

      It must be the $1/£1 technology price equivalence!

      Have a beer, not exceeding 4% AbV for spotting it

  6. 27escape

    The record companies know this already

    I understand tht some popular artists use the multiple short track trick too, as a way of having a larger percentage of their album listened to, this then makes the album more popular, jumps it up the charts and they then get a bigger cut of the overall revenue.

    What is surprising is that a small timer has found a way to get in on the act and benefit and self fund the future income stream

  7. JeffyPoooh
    Pint

    Did it attract other listeners?

    Let's assume it was only his own bots listening. If so, then please explain Spotify's business model. Bring in $X in subscriptions, pay out $20X in royalties.

    The Bulgarian must have attracted other listeners, thus being very nearly legitimate.

    1. Halcin

      Re: Did it attract other listeners?

      @JeffyPooh, In short: No - not necessary. Please scroll up and read the post by Skwosh for details.

    2. Robert Carnegie Silver badge

      Re: Did it attract other listeners?

      Apparently the two fake playlists appeared in Spotify's album chart at No.11 and No.22 if I correctly remember from the story. Charts were invented by the music industry to increase sales - the chart is a form of advertisement (don't tell the BBC, they're not allowed advertisements) - so logically some other people were intrigued enough to play at least some of the tracks. I wonder if the BBC were... I don't listen to "The Chart Show" any more.

      1. Trygve

        Re: Did it attract other listeners?

        "Apparently the two fake playlists appeared in Spotify's album chart at No.11 and No.22 if I correctly remember from the story...

        - so logically some other people were intrigued enough to play at least some of the tracks."

        Or alternatively, 72 Million plays from some fake accounts is enough to get you a fair way up in the charts....

  8. Potemkine! Silver badge

    Right holders != artists

    no wonder musicians are starving

    Oh, the poor little ones!

  9. tin 2

    "but said it works hard to detect artificial manipulation of its streams"

    except of course that it doesn't

    cos 1200 people playing the same nearly 30s a track playlist over and over would NOT be hard to spot.

    1. anothercynic Silver badge

      To be fair...

      ... I have a specific playlist at work that I cycle over and over and over every day. Why? Because it contains tracks *I* like. Would that mean that Spotify would suddenly tell me "sorry mate, but play something else, otherwise you're a bot"?

      1. Robert Carnegie Silver badge
        Joke

        Re: To be fair...

        "I have a specific playlist at work that I cycle over and over and over every day."

        I think a Turing Test at some point wouldn't be unreasonable. I hope you pass. :-)

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon