Learn
A contribution-weighted split needs a number for "how much did this person do". Royalty points are that number — derived from work you completed, not from what everyone remembers at the end.
Fixed percentages require predicting, on day one, how much each person will contribute over the following two years. Teams are bad at this, and the error is not random — the people who over-deliver are the ones who notice, and they're the people you can least afford to lose.
Contribution weighting inverts it. Rather than guessing shares in advance, you record work as it happens and let the shares follow. The team still decides the rules; they just decide them before knowing who benefits, which is when people reason most fairly.
Every completed task awards points on three components:
a base amount for completing the task, plus the task's difficulty × a difficulty weight, plus logged hours × a time weight.
Your team sets all three weights. A studio that cares mainly about throughput weights the base higher; one doing deep technical work weights difficulty higher; one where effort varies wildly weights time higher. There's no universally correct setting, which is why it's a setting.
Keeping the three separate matters. A single blended number can't distinguish someone who did one very hard thing from someone who did twenty easy ones, and those are genuinely different contributions that teams weigh differently.
Most teams adopt a system mid-project. Historical claims are converted into task-equivalents using the project's own observed history — its average hours and difficulty per completed task — so earlier work is credited on the same basis as tracked work rather than being lost or over-rewarded.
If a historical claim covers a period that also contains tracked tasks, the points already earned by those tasks are deducted from the claim. Without that, the same work is paid for twice.
Most teams log only a fraction of real working time. The time component is scaled by how much this project's tasks actually log, so a team with patchy time tracking isn't systematically underweighted against one that's diligent.
A project with six completed tasks doesn't have reliable averages. Estimates are shrunk toward a prior learned from cross-project history, so early numbers are stable instead of swinging on a single unusual task.
Points produce a suggested split. They don't force one. You can weight, adjust or override the result, and the tooling shows the impact of a change before you commit it.
This matters because contribution isn't the only thing teams value. Someone who brought the funding, owns the IP, or carries the commercial relationship may warrant a share that no task board will ever reflect. A points model that pretended otherwise would just be wrong in a more confident way.
What the points give you is a defensible starting number and a record behind it. The conversation moves from "I feel like I did more" to a specific disagreement about a specific weight — which is a much easier conversation to finish.
Set the weights, track a sprint, and look at what the split actually says.