Part I — Own it

Splits, before the session

Have the splits conversation before a session instead of after a release, and file a song correctly even when one of the people on it has not got their side together.

The conversation is easy on Tuesday and impossible in December

A split sheet is the signed record of who wrote what share of a composition. It is a small document, and the reason it goes unwritten has nothing to do with the document.

Before a session, nothing exists. Nobody is defending a contribution, and the shares are a matter of intention. After the record is finished — and especially after it starts working — every split conversation is a negotiation about a thing that exists, between people who can now estimate what it is worth. The same sentence, "so how are we splitting this," is administrative in the first case and adversarial in the second.

That is the whole timing argument, and it is why the practical move in this material is unglamorous: ask for everyone's writer and publisher information before work starts. Look it up yourself if you already know their names. Get it out of the way while it is a formality.

The social cost is real, and pretending otherwise is how people talk themselves out of it. The person who raises splits is briefly the least popular person in the room. That is the price of being the one who gets paid, and it is cheaper than the alternative — a dragged-out conversation that damages a working relationship between people who all understood the stakes.

This chapter is US-shaped. The registries, the affiliation conventions and the forms described here are US institutions. Splits themselves are universal; the mechanics below are not.

What actually goes on the sheet

The document is not a vibe. The sources want specific fields:

• Every writer's legal name — not their artist name.
• Their publisher, if they have one, and their identifying numbers as a writer and

publisher.

• The agreed percentages, adding up.
• Evidence of agreement from every party.

That last one is the one people get wrong, because they treat the sheet as a note-to-self. Use a signing method that makes refusal visible — a shared document, or email. The point is not formality. It is that at a glance you can tell who has signed from who is stalling, which is information you need long before you need the sheet itself.

One shortcut in this material is flagged and should not be taken. Where the source expects a fight over a co-writer's details, he suggests putting your own address on the claimant line, on the reasoning that nobody checks and it can be corrected later. Filing a registration that misstates the claimant is a different category of thing from leaving a field blank, and "nobody checks" is a statement about enforcement, not about accuracy. The underlying point survives without the shortcut: the split sheet is what makes the record correctable at all.

A named-and-reserved share is recoverable; an omitted one is a hole

This is the single most useful rule in the chapter, and it solves the problem that actually stops songs being filed: your collaborator has not set up their side.

Do not wait, and do not leave them off. Register the work now, with their name on it and their share reserved — the US organisations have a marker for a party who is not yet affiliated, which the source gives as NA or NS, "not available" / "no society". When they do affiliate, they submit a change request and correct their own entry. The split sheet is the record they claim against.

The reciprocity is the elegant part: the person you filed without is the person that document later protects, because they need it to claim their share.

There are two more patterns worth knowing, both about mechanism rather than about money:

• A writer with a writer account but no publisher account. Rather than stalling, route their

publishing percentage through their writer account, using the "no publisher" option the form provides, so the money still reaches them.

• Splitting your own publisher share with a manager or an indie label. Attach two original

publishers to yourself as the writer — you and them. Do not add them as a writer. Adding a non-writer as a writer silently dilutes every other writer on the song, which is a change to other people's money made through a form.

Two constraints ride along with all three patterns. First, the numbers must come from the contract that governs them; these fix the mechanism, never a disagreement about what the split actually is. Second, this is the shape of specific organisations' forms at a particular moment. Check the current form. A chapter that hardcodes a checkbox label ages badly and confidently.

And the law under all of it, which generalises past any form: filing incomplete-but-honest beats filing late. A share that is named and reserved is recoverable. A share that is omitted is a hole somebody else's registration eventually fills.

Inconsistent registration is worse than no registration

Most people assume the failure mode is a missing song. It is not. The failure mode is one song registered three different ways at three different bodies.

An unregistered song produces payments that are missing — visible, and fixable. A song whose owner sets disagree across registries produces payments that are wrong: money moving to the wrong people, quietly, while everything looks like it is working. It is also harder to unpick years later, when the people involved have lost contact, fallen out, or forgotten what was agreed while the record kept earning.

The source claims the different bodies run their registrations on different scales, so the same 50/50 song is entered one way at one organisation and a different way at another. This book does not tell you which is which, for a specific reason: the library's own review found that claim unverified and the speaker's worked example partly garbled — and a scale error propagates silently to every registry at once, which makes a half-remembered rule worse than no rule. Take the warning and none of the specifics: open each registry's own form and read what it is asking for before you type a number into it.

The cost is not only royalties. A catalogue that cannot be proved cannot be valued, borrowed against or sold, because what a buyer or lender prices is provable ownership, not the music. The registration is not a record of the asset. In commercial terms it is the asset.

The trade most artists refuse, and probably shouldn't

Here is the most contrarian claim in this part of the library, and it cuts directly against the instinct the rest of Part I builds: give the writer the bigger publishing share, and keep the master.

The argument runs like this. Writing every word yourself is often an ego position rather than a strategy — it forces you to work in your weakest craft while your actual strength, the voice and the performance, goes under-used. You supply the story, the message and the direction; a professional writer executes the words. Because you shaped the direction, you can still hold a share of the publishing without having written the lines — but only if that is on the sheet, which is why this belongs in this chapter rather than a craft one. Influence earns nothing that nobody wrote down.

The hinge is one line: 100% of zero is still zero. Keeping full ownership of a record nobody connects with is the expensive outcome, not the safe one.

Two honest caveats. The source asserts that the master makes more money than the publishing anyway, and that shows and merch pay out faster — both stated as universal, both dependent on your particular revenue mix, and neither supported with figures. And the sequencing he prefers, conceding publishing now and buying shares back later out of profit, has an obvious predatory reading when the other party is an up-and-coming writer who needs cash. He raises that himself and does not resolve it. Neither will this book.

What survives is the decision frame, and it is genuinely useful: concede the share that buys you a better record; keep the asset you can exploit yourself. If you cannot stomach giving up publishing at all, the fallback is to pay the writer up front to keep the rights — which is cleaner and much more expensive, in that order.

If the room deadlocks, the source reaches for a named convention as a fallback and does not define it. Do not read a particular division into a name you have only heard in passing; ask what it means in the room, out loud, and write down the answer.

One thing this site will never do

There is no split-sheet tool on hiphop.world, no percentage field and no registration path, and that is deliberate. Capturing splits would imply we do something with them, and the conventions above are contested enough that a helpful-looking form is the most damaging thing we could build. Learn it here, file it at the real registries, keep your own copy.

Claiming a profile proves you are the artist. It says nothing about who wrote the song.

What to do this week

1. Pick the next session on your calendar and send one message before it: ask everyone attending

for their legal name and their writer/publisher information.

1. Take one released song with an unclear split and write the sheet retroactively, today, while

people are still reachable. Reserve any share you cannot confirm rather than dropping it.

1. Pull up one registered song and check that the same owner set appears everywhere you

registered it. If it does not, you have found money going to the wrong place.