Where the industry is heading

Genre momentum

What is accelerating, and what is fading

We follow the same artists month after month — 430,719 of them, every one with a Spotify monthly-listener reading in all 8 snapshots — and add up their listeners by genre. Because the panel is fixed, a genre can only move when people actually listen more or less, not because we started tracking more artists. 359,330 of those artists carry genre tags, down to the sub-genre.

Sub-genres gaining the most

Ranked by growth in panel listeners over the window. The 2027 band is the window rate and the last-quarter rate, both compounded twelve months out — a projection, not a measurement.

Sub-genres losing ground

The big genres

Balanced panel of artists with a Spotify monthly-listener reading in every snapshot. Genres are Soundcharts artist tags (root:subgenre); an artist counts in every genre it carries, so genre totals overlap and do not sum to the catalogue.

Artists to watch

Who is breaking out into 2027

The same panel, read one artist at a time. To make this list an artist had to start the window small (between 25,000 and 2,000,000 monthly listeners), grow at least 50% across it, go up in at least 5 of the seven months, still be climbing in the last quarter, and never multiply their listeners by more than 4× in a single month — a jump that big is a change in the source data, not a career. 7,964 artists cleared those rules; these are the fastest.

Artists with a Spotify monthly-listener reading in all 8 snapshots, between 25,000 and 2,000,000 listeners at the start of the window, up at least 50% across it, rising in at least 5 of 7 months and still rising in the last quarter, with no single month multiplying their listeners by more than 4 (a jump that big is a change in the source data, not a career). Not mainstream or superstar. This is a watchlist, not a forecast: it says who is climbing fastest right now, not where any one career lands.

Markets

Where the listening is moving

The same panel again, grouped by where the artist is from — not where the listening happens. Across the window the panel grew 5.3% in total, so a market only gains share by growing faster than that. The shift is in percentage points of the panel's listening, which is why a country can grow and still lose ground.

Gaining share

Losing share

Balanced panel of artists with a Spotify monthly-listener reading in every snapshot, grouped by the country on the artist's profile — where the act is from, not where it is listened to. Share is that country's slice of the panel's listening; the shift is in percentage points, so a country can grow and still lose share.

Who controls editorial

The majors lost half of editorial playlisting

Every track that entered an editorial playlist in our scope, counted in the year it was placed — not the year it was released, because the question is who is being put in front of listeners. In 2018 the three majors took 77% of the placements we can attribute. In 2026 they take 38%.

Shares are of the 62% of placements, on average, whose distributor we have on file. Independent releases are likelier to be missing one, so the independent share here is a floor rather than a ceiling — the real shift is at least this big.

And this is who took the room

The majors did not lose to one rival. They lost to distribution infrastructure: Fuga, AudioSalad, Believe and DistroKid went from almost nothing to thousands of placements a year.

Every track entering an editorial playlist in our scope, counted in the year it was placed and attributed to its distributor. Major means Universal, Sony or Warner. Shares are of placements whose distributor we know; the unknown share is reported alongside. Independent releases are likelier to be missing a distributor, so the independent share is a floor. This scope leans towards background and functional playlists, so it describes editorial placement, not the whole market.

The sound of 2027

Editorial music is getting quieter, calmer and wordless

The audio of every track placed in an editorial playlist, averaged by the year it was placed. Two numbers carry the story. Instrumentalness — how likely a track is to have no vocals at all — went from 0.28 to 0.58, and the average placement got 6.7 dB quieter. Editorial playlisting is drifting from songs people listen to towards music people put on.

Solid line: measured, 2018 to 2026. Dashed: the 2027 projection. 58,375 placements, all of them with audio features on file. Audio features of every track placed in an editorial playlist in our scope, averaged by the year it was placed; a track placed several times in one year counts once. Every placed track has features on file. The 2027 point extends the least-squares trend of the last five years by one year — a projection, not a measurement. The scope leans towards background and functional playlists, so this is the sound of editorial placement, not of all music.

Fake artists

The PFC wave peaked in 2025

Of the 29,907 artists we scored, 531 look like PFC with high confidence and 1,600 with medium confidence. Counting the years in which those artists were placed in editorial playlists, their share climbed from nothing to 6.1% in 2025 — and then fell to 3.6% in 2026, the first drop in the series.

Read this one carefully: artists are classified as they look today, and the series then asks when they were placed. So it is how much editorial placement today's PFC artists accumulated per year, not the PFC share measured at the time. The 2027 range is deliberately wide — its low end carries on the 2026 reversal, its high end the climb that came before — because one year is not enough to tell a correction from a turn.

The playlists most exposed

Sleep, calm, covers, noise. The exposure is concentrated exactly where listening is least attentive — which is also where the sound of 2027 is heading.

Artists are classified by the PFC score as they look today; the series then counts the years in which those artists were placed in editorial playlists. It therefore reads as how much editorial placement today's PFC artists accumulated per year, not as the PFC share measured at the time — a second tracklisting cut would be needed for that. Shares are of placements we can attribute to a scored artist.

The calls

Three things we expect to happen in 2027

Every projection states its method, its window and its error band, and is labelled a projection, not a measurement. The editorial-playlist universe this report is built on leans towards background and functional music, so genre levels are not a picture of global listening — the movement over time is what the projections read.