Everything nobody
tells you.

Up to eight, and they are collected by different organisations that do not talk to each other. A single streamed song can generate a recording royalty paid by your distributor, a mechanical royalty collected by The MLC, a performance royalty collected by your PRO, and neighbouring rights collected by SoundExchange if it is played on digital radio. The same song can also earn from YouTube Content ID when other people use it, from sync when it is licensed to picture, from live performance royalties when you play it in a venue, and from foreign societies when it is played abroad. Registering with your distributor covers exactly one of those eight.

They are two different pieces of property and they pay through different channels. The composition is the melody and the lyrics, and it pays through your PRO and The MLC to whoever wrote it. The master is the specific recording of that composition, and it pays through your distributor to whoever owns it. A producer can own half of one and none of the other, which is why a split sheet has to state both percentages separately. Agreeing one does not agree the other.

No, and this is the single most misunderstood thing about streaming income. Spotify and most other services pay on a pro-rata pool basis. All subscription and advertising revenue for a territory goes into a pool, the platform keeps roughly thirty percent, and the rest is divided among rights holders according to each one's share of total streams in that territory. Your effective per-stream rate is therefore an outcome, not a rate, and it moves depending on how much everybody else was streamed. Any figure quoted as a fixed per-stream rate is an average after the fact, which is why honest estimates are ranges.

It means losses on one project can be recouped out of the earnings of another. If a deal covers two records and the first one loses money, a cross-collateralised agreement lets the other party recover that shortfall from the second record's income before paying you anything. The practical effect is that a successful release earns you nothing because an unsuccessful one is still being paid off. It is a normal clause to see and a reasonable one to push back on, particularly where it reaches across different income types such as recording against touring or merchandise.

Paying a third party to place a song on a playlist is prohibited by the major streaming services, and placements sold this way are frequently driven by artificial streams. The consequences fall on the artist rather than the seller: streams can be stripped, income withheld, and in serious cases the release removed. Editorial playlist consideration, which is the kind that actually matters, is free and is requested through the Spotify for Artists dashboard before a song comes out. If somebody guarantees placement in return for money, what is being sold is either against platform rules or worthless.

The unavoidable costs are small. Distribution runs from nothing to about fifty dollars a year, and every registration that collects your royalties is free for a writer; registering your own publisher entity at BMI is a one-off fee of about one hundred and seventy-five to two hundred and fifty dollars, worth leaving until the songs earn. Everything above that is production quality and promotion, which varies enormously: independent mixing commonly runs about one hundred and fifty to eight hundred dollars a song from an experienced freelancer, and more from a name engineer, mastering from about five to twenty-five dollars a song for an online service to fifty to two hundred for an engineer, and artwork from about fifty dollars upward. The honest summary is that collecting your money properly costs almost nothing, and the expensive part is entirely optional.