Every songwriter has a version of this story: a session that felt completely equal in the room, followed by a much less equal conversation about percentages once the track actually goes somewhere. Nobody wants to talk numbers while the song is still exciting — which is exactly why the numbers need to get talked about then, not later.
There's no universal formula
Despite what forum threads and old industry folklore suggest, there's no single "correct" way to split a song. What matters is that everyone involved agrees on a method before anyone starts assuming. Three approaches cover most sessions:
- Equal splits. Everyone in the room gets an even share, regardless of who did what. Simple, common for small writing teams, and avoids relitigating every contribution.
- Role-based splits. Lyrics, top-line melody, chord progression, and beat/production are valued separately — often close to a quarter each, adjusted for who did more.
- Contribution-weighted splits. A frank conversation about who brought the hook, who wrote three verses versus one, who spent six hours programming drums. Less tidy, often more accurate.
None of these is more "correct" than the others. The mistake isn't picking the wrong method — it's not picking one out loud.
The complications that catch people out
A few situations turn a simple split into a slow-burning dispute if they're not addressed early:
- Session musicians vs. writers. A session player might be paid a flat fee with no ownership stake, or they might be a genuine co-writer if they contributed a melodic or lyrical idea. Which one it is should be clear before they leave the studio.
- Samples and interpolations. If any part of the track uses another writer's material, that's a separate conversation from the split between the people in the room — but it affects everyone's eventual share once cleared.
- Features. A guest verse is usually a smaller, defined percentage rather than an equal share — but "usually" isn't the same as "agreed," so say it out loud.
Why the moment right after finishing matters
Right after a session, everyone remembers who did what — who came up with the hook, who was scrolling their phone for twenty minutes, who rewrote the bridge three times. Six months later, if the track hasn't gone anywhere, nobody thinks about it again and it doesn't matter. But if it takes off, memory quietly rewrites itself in everyone's favour at once. That's not bad faith — it's just how memory works under incentive. The fix isn't trust, it's a record.
A practical process
- Before the session ends, list everyone who contributed — including anyone who left early or joined late.
- Next to each name, note what they actually did, in plain language.
- Agree percentages that add up to 100%, using whichever method the group prefers.
- Write it down somewhere everyone can see, and get a lightweight confirmation from each person.
This is exactly the gap Collaborro fills: a five-minute process to log contributors, their contribution, and their percentage, then send it out for everyone to confirm with a single tap — no accounts required for collaborators, no spreadsheet, no "I thought we said 60/40."