>, and your artist will get 1/1000×15000=£15 and the other one will get 999/1000×£15000=£14985. The denominator of 1000 in your fraction is artists but the pro

>, and your artist will get 1/1000×15000=£15 and the other one will get 999/1000×£15000=£14985. The denominator of 1000 in your fraction is artists but the pro rata calculation in the paper has total streams as the denominator. If Spotify paid out by directly mapping userid to (unpopular) artist, and a user only listened to 1 track that month, that 1 artist of that 1 stream would receive the entire user's subscription revenue. (I.e. Assuming a rough calculation that Spotify pays ~70% of revenue to license royalties, that 1 artist would receive 70% of $15 which would be ~$10) But with the total streams aggregation, the more people listen to Lady Gaga / Katy Perry / Taylor Swift / etc, the more it makes the denominator bigger with popular streams. Therefore, the more it reduces income percentage of the unpopular artist. (I.e., that's why an unpopular unknown indie artist receives a fraction of a cent for each stream instead of $10.[1]) It's a calculation that favors artists of popular songs (streams). In contrast, I believe the Amazon's subscription for Kindle Unlimited payout is based on author's pages read without aggregating all other authors' pages into the denominator. Therefore, an Amazon's subscriber's revenue goes more towards the particular authors the subscriber read and is not affected by millions of others reading the more popular JK Rowling or Stephen King. [1] https://www.google.com/search?q=spotify+indie+musician+payout+cents https://www.google.com/search?q=spotify+indie+musician+payou...