Mood Machine

To understand why streaming flattens music, Pelly says, follow how the money moves. Spotify pays rights holders from a shared pot, roughly proportional to each track's slice of total streams. This is the pro-rata model, and its logic is unforgiving: every play of your song is competing with every play of every other song for the same finite pool. A track that keeps a session going — that nobody skips, that nobody notices enough to stop — earns exactly as much per stream as a masterpiece. The system doesn't reward greatness. It rewards not being turned off.

From there, Pelly documents a set of practices that fall out of the math. The most striking is what she calls, drawing on internal sources, the platform's interest in cheaper music. If the company can fill a chill or ambient playlist with tracks it pays less for — commissioned instrumentals, library-style recordings, music from producers working on flat fees rather than full royalties — then the same listening minutes cost the platform less. The listener hears "peaceful piano." The balance sheet hears margin.

The human cost of this arithmetic is where Pelly's reporting turns concrete. She talks to working musicians — not stars, but the mid-list players who once made a modest living from records — and the picture is consistent. The per-stream payout is tiny, fractions of a cent, and it takes an enormous volume of plays to add up to anything resembling rent. The catalog is infinite and the pot is fixed, which means most of the money flows to the very top while the long tail earns close to nothing. Access for listeners has quietly become precarity for the people who fill the shelves.

What makes it harder to see is that the platform frames itself as an ally. It offers dashboards, analytics, tools for "building your audience," the language of empowerment borrowed from the wider gig economy. Pelly is sharp on this: the artist is recast as a small business responsible for their own growth, endlessly optimizing releases, chasing playlist placements, feeding the algorithm content on its schedule rather than their own. The work of being heard expands to fill the time once spent making the thing worth hearing.

Step back from the royalties and the playlists and a larger shift comes into view, and this is where Pelly's book widens. Streaming didn't only change how music is paid for. It changed what music is for. The mood machine assumes that music's proper role is functional — to set an atmosphere, smooth a commute, fill a dinner party, mask the silence of an apartment. Music as utility, delivered on tap, optimized to stay pleasantly out of the way. And when the reward structure favors that use, the whole ecosystem drifts toward supplying it.

The flattening Pelly names is not that all songs now sound identical; plenty of loud, strange, specific music still gets made. It's that the center of gravity has moved. The safest place to be, economically, is the background — smooth, moderate, emotionally legible, easy to tag. Music that demands to be foreground, that interrupts, that asks you to stop and reckon with it, sits at a structural disadvantage, because the system pays best for the music you're least likely to notice. Attention was always music's currency; the platform quietly discounted it.

We started with a Tuesday evening and a rack of moods, and that scene is where Pelly's argument comes to rest. The mood playlist feels like a small kindness — the platform sparing us the labor of choosing. But every convenience she traces turns out to be load-bearing. The pro-rata pot, the cheaper background tracks, the artist recast as a self-optimizing business, the drift toward music that never interrupts: all of it hangs together, and all of it follows from treating a session, not a song, as the thing being sold.