Editorial Versus Algorithmic: Spotify's Playlist Power Has Shifted

RapCaviar was once powerful enough to change a rapper’s career overnight, a single placement on the playlist could push a song into rotation with millions of listeners who had never heard of the artist before pressing play. That kind of editorial power built Spotify’s reputation as a genuine hitmaker. It is also, by most accounts from inside the labels that track it, fading.

The Editorial Playlist That Used to Move the Needle

RapCaviar and Today’s Top Hits are the two most cited examples of Spotify’s in-house editorial curation, playlists built by human editors selecting songs for a defined audience. For years these were treated as the closest thing streaming had to a modern equivalent of radio’s most powerful stations, capable of driving a track from obscurity into heavy rotation.

That influence is now measurably smaller. Label staff who track streaming sources report seeing plays attributed to RapCaviar drop somewhere between 30 and 50 percent, a decline tied directly to Spotify’s own strategic shift toward algorithmic and personalized discovery rather than fixed, human-curated lists. Spotify has also cut staff on the teams that build these playlists, and shifted some previously fixed editorial playlists into personalized versions that differ from listener to listener, which by definition dilutes the singular, career-making moment a fixed playlist placement used to represent.

Why Spotify Made the Shift

The company’s own engineering team has described its current approach as “algotorial,” a blend where human curators still shape the initial pool of songs and tag them for context, but algorithmic systems handle the actual placement and personalization for each listener. The stated logic is that personalized recommendation keeps listeners engaged longer than a one-size-fits-all list, and engagement is what Spotify’s business model runs on.

For an artist, the practical effect is that there is no longer one gatekeeper meeting to win over. Getting flagged and correctly tagged for the algorithm to surface matters as much, or more, than convincing a human editor to add a track to a flagship list.

Discover Weekly is the clearest illustration of where Spotify’s investment has actually gone. Every Monday, the playlist delivers thirty previously-unheard songs to each of roughly 200 million listeners, individually assembled, not shared across users the way RapCaviar or Today’s Top Hits are. It runs on collaborative filtering, matching a listener against others with overlapping taste, layered with content-based audio analysis and behavioral signals like skips and saves. No single placement on Discover Weekly is as visible as a RapCaviar cover slot once was, since every listener effectively sees a different playlist, but in aggregate it reaches far more listeners than any fixed editorial list ever could, which is precisely the trade-off Spotify has been making.

Discovery Mode: Promotion or Payola

Running alongside that algorithmic shift is a more contested Spotify program called Discovery Mode, launched in 2020. It lets artists or their labels flag songs for algorithmic promotion, wider placement in personalized radio, autoplay, and recommendation surfaces, in exchange for accepting a lower royalty rate on any stream generated through those specific features. Reporting describes the discounted rate as landing in a band of roughly $0.003 to $0.005 per stream, a real cut from the artist’s normal payout.

The comparison to payola, the mid-century practice of labels secretly paying radio stations for airplay, has been made explicitly and often. The Recording Academy has publicly criticized Discovery Mode on those grounds, and the Future of Music Coalition has argued the program contributes to broader wage suppression across the platform. Music journalist Liz Pelly, whose book “Mood Machine: The Rise of Spotify and the Costs of the Perfect Playlist” examines the company’s inner workings through more than a hundred interviews with insiders and former employees, frames Discovery Mode as a cost-saving mechanism dressed up as an opportunity, one that mirrors the undisclosed-payment structure that made traditional payola illegal for radio in the first place. In November 2025, Spotify was named in a class action lawsuit alleging Discovery Mode operates as an undisclosed pay-for-play scheme that misleads listeners into thinking they are hearing organic recommendations.

Where That Leaves an Artist Weighing the Trade

Spotify has defended Discovery Mode as opt-in promotion, not a mandatory toll, comparable to any advertising spend an artist might otherwise make to reach new listeners. The distinction that keeps the payola comparison alive is disclosure: radio listeners in the payola era did not know a station had been paid to play a song, and Discovery Mode critics argue that listeners hearing a track through algorithmic radio or autoplay have no way to know that placement came with a royalty discount attached, rather than being a neutral recommendation based purely on listening patterns.

The Practical Calculus Now

For an artist or manager deciding whether to use Discovery Mode, the honest framing is a trade of certain, smaller royalties now for uncertain, larger reach later, the same bet independent radio promotion always represented, just automated and priced in cents per stream instead of dollars per spin. What has genuinely changed is that the fixed editorial playlist, the single win that used to guarantee a career bump, carries less weight than it did even three years ago, which makes betting entirely on one editorial placement a weaker strategy than building a broader footprint the algorithm can pick up on its own.

That footprint is also what the algorithmic side of Spotify is explicitly built to read. Saves, repeat listens, and playlist adds a listener makes on their own all feed the same collaborative filtering and audio analysis systems that power Discover Weekly and algorithmic radio, which means an artist with genuine, sustained engagement from a smaller audience has a real path into recommendation surfaces without ever needing a human editor’s attention. The skill an artist needs now is less about pitching one gatekeeper and more about generating the kind of organic listening behavior the algorithm was built to detect.