Streaming learns to weight a play

Streaming royalties were built on an assumption so simple it went unexamined for fifteen years: every play is worth the same fraction of the pot. That assumption is being dismantled, and the interesting part is not the arithmetic but who the arithmetic is designed to move money away from.

The model that replaced it applies weights. Descriptions of the artist-centric approach adopted by some services set thresholds below which a recording earns a reduced share, and multipliers above which it earns an enhanced one.

The stated purpose is to stop paying out on material that is not really being listened to. A pro-rata pool divided by total plays pays functional audio, silence tracks and algorithmically generated filler exactly as it pays a record somebody chose, and at scale that is a real transfer.

The unstated effect is a transfer in the other direction. Weighting by thresholds moves money toward catalogues large enough to clear them, which describes the major labels and the established independents rather than the artist releasing a first record to two hundred listeners.

Both things are true at once, which is why the debate is so bad-tempered. Removing payouts for fraudulent and functional audio is straightforwardly good. Doing it with a threshold that also removes payouts from small legitimate artists is a policy choice wearing an anti-fraud justification.

The pool itself has not grown, and that is the constraint every version of this runs into. Subscription prices moved slowly for a decade, so any change to the distribution rules is a redistribution rather than an increase, and every winner implies a loser.

Per-stream rates remain the most quoted and least useful number in the business. What a rights holder receives depends on the service, the country, the subscription tier, the label's deal and now the weighting, which is why published averages differ by a factor of five.

Fraud detection is the part of the change with no real opposition. Streams bought to inflate a payout come directly out of everyone else's share, and services investing in detection are protecting the pool rather than reallocating it.

Artists at the bottom of the distribution have limited recourse, because they have no leverage in a negotiation between platforms and major rights holders. The terms are set between the parties who control the catalogue and the subscriber base, and communicated to everyone else.

What would actually change the arithmetic is a bigger pool, through higher subscription prices or a tier that costs more, and the services have moved on that far more slowly than they have moved on weighting.

So the fight over how to divide the money continues to substitute for the harder conversation about how much of it there is. Weighting a play is a real improvement over counting every play equally, and it does not put another cent into the system.