Data Analysis of the Apparent ‘Fake Artists’ of Spotify

Rahul Rumalla
Paperchain
Published in
5 min readJul 12, 2017

Earlier this week, the publication Music Business Worldwide(MBW) published a story about Spotify denying that they are play-listing fake artists. There was some harsh criticism and claims that Spotify was creating its own recordings and putting them on their playlists. What struck me the most about this story was that the statement about there being a large amount of music on Spotify that was created by fake artists and that the “artist name” under which these tracks are listed cannot actually be identified. A list of 50 artists were published on MBW.

15 of the 50 “fake artists” listed by MBW

At Paperchain we’re all about finding right owners of music and large part of what we are trying to do is identifying artists & songwriters ( right owners in general ) even with minimal amount of information. Ultimately, we’re trying to solve the problem of unclaimed royalties, popularly referred to as the ‘black box’. I took this up a challenge and to test the extent of our systems’ capabilities. Surely enough, we were able to produce a comprehensive digital audit trail of these 50 “fake artists” and their music on Spotify by linking their recordings to the song compositions. Here’s how we did it and what we found …

Following the Breadcrumbs — Digital Audit Trail

Given just the information we had i.e., the artist names, the first thing we produced was the list of tracks on Spotify that were under these names. To mainly fetch the ISRC code(s) that were associated with these tracks. For those who are unfamiliar, ISRC code is a unique code assigned to a music recording. It was important to fetch and validate this because the ISRC code reveals 2 pieces of very important information

  1. The country which the code was assigned from
  2. The ISRC agency that assigned it

EXAMPLE #1

Our first example is on the artist named ANA OLGICA that appears on MBW’s list. The artist has 2 tracks listed on Spotify, both piano based instrumental tracks that appear of the highly popular ‘Peaceful Piano’ playlist. The track titled ATOMS has the ISRC code SEXGF16157. The first 2 letters of the ISRC code is the country code i.e., SE = Sweden. And the next 3 letters indicate the ISRC agency that assigned them i.e., represented by the code XGF. ISRC Agencies are members of The International Federation of the Phonographic Industry(IFPI), responsible for assigning unique ISRC codes to recordings. Each agency themselves are given a unique 3 character code like the above. We were able to find out that the alphanumeric code XGF corresponds to a company named Catfish Recording. A quick google search reveals that there are some artists in Stockholm, Sweden under this record label but there is absolutely no link to the website of this business. Another strange thing is that this artist Ana Olgica has no tangible identity. No facebook page nor a soundcloud profile. However, the tracks are distributed on multiple platforms for purchase.

Now that we had the recording information, our next step was to link it to the composition copyright to fetch the music publishers and writers involved. We were able to dig further into this using Paperchain’s proprietary technologies to link up the composition to the recording. This song had T920356326 as the ISWC code (a code used to uniquely identify the composition part of the song or a ‘work’) and it was listed under 2 writers and no publisher i.e., BOSTROM FREDRIK HAKAN and NORDENSTROM JESPER OLOF. Both of these writers are actually indeed swedish music producers and composers (Source ASCAP). The data also reveals that they are affiliated to the Swedish collection society STIM.

EXAMPLE #2

This track titled NORDIC LIGHT is under the artist named CLAY EDWARDS. This is an 3 minute instrumental track, released as a single. Clay Edwards seems to be gospel music artists hailing from Kansas.

Nordic Light by Clay Edwards on Spotify

This track has the ISRC code as SE4RG1600705 (again SE = Sweden ). 4RG in this case corresponds to Firefly Entertainment AB, a Swedish music publishing and television company with representatives and music studios in Karlstad, Gothenburg and Stockholm. The song was attributed with the ISWC code T9203563262 and the writers associated NORBERG ROBERT PAR and WIGELIUS ANDERS SVEN. The publisher in this case is also listed as FIREFLY — quite possibly the company listed above. All of the 3 parties mentioned also are registered under the Swedish collection society STIM.

SUMMARY

For the 50 artists mentioned in MBW, we found 197 recordings on Spotify, of which 96 have ISRCs assigned from a Swedish agency. We were able to connect them to 693 songwriters and publishers. Only 69 of them were affiliated to the Swedish PRO STIM while most of them were affiliated to the American society ASCAP and German society GEMA. Some of these songs are also published under major publishers like Universal Music Publishing.

Recordings with their metadata
Compositions with their metadata

CONCLUSIONS

While there is definitive indication that a lot of this activity pointing to Swedish music producers creating instrumental music under pseudo artist names and coincidentally on playlists with high traction, there is certainly no indication in the digital trail that Spotify is a right owner of any kind. Which means, they are NOT generating royalties for themselves on these tracks. Additionally, this practice (rather strategy) of creating music under pseudonyms also seem to be popular in countries outside Sweden too.

What is however controversial is that these popular mood-based playlists (peaceful piano, songs for sleeping etc) are indeed owned by Spotify. And these artists do appear on them. This combined with the movement of AI-driven movement (Source: MusicAlly) of generating playlists AND music itself, can possibly significantly impact the royalties distributed to regular labels, artists and other right owners.

Feel free to reach out to me on rrumalla(at)paperchain.io to access the data sample or for further research or even discussion. Would love to hear your thoughts.

Love this post? Please like and share below.

Originally published on blog.paperchain.io on Jul 12, 2o17.

--

--