• Tech
  • Media

These Are the People Picking Your Next Internet Radio Song

4 minute read

It’s never been easier to listen to music–and never harder to decide what to listen to. First, Apple let us digitize our music libraries and put them in our pocket with the iPod. Then streaming services like Spotify made the idea of a “library” obsolete, promising unfettered access to tens of millions of songs at all times. Now, tech companies are trying to make sense of this massive sea of songs with an idea ripped straight from the analog era: getting actual human beings to offer up curated selections of music.

“No one on the Internet is like, ‘But where do I get more songs?’” says Elias Roman, product manager for Google Play Music. Roman was a co-founder of Songza, a music service focusing on curated, contextual playlists that was acquired by Google in July 2014. Since acquiring Songza, Google Play Music has placed context front and center in its streaming interface. When booting up the service on mobile or desktop, Google Play Music provides a list of potential activities you may be doing based on timing. On the 4th of July, for instance, the service offered different sets of playlists for hosting a barbecue, watching fireworks or celebrating with patriotic tunes.

“What I’m most excited about is having people feel like they’re paying to make their workouts and commute better,” says Roman. “The value prop is around lifestyle enhancement.”

Though Google calls these playlists “radio stations,” they’re built differently from the way Internet radio services such as Pandora operate. While Pandora uses listeners’ past preferences and song metadata to algorithmically build stations based on artists or genres on the fly, every track in Google’s contextual playlists are hand-selected by teams of music curators and editors (Play Music has a separate, Pandora-like radio feature with algorithmically driven stations). That human touch helps give the stations a quirky specificity that’s hard for computers to match—there are playlists for Ron Burgundy’s bachelor pad, Kanye West’s soul-sampling “Pink Polo” era and sipping tea with Drake.

Google pays musicians, DJs, music journalists and other experts to build these unusual playlists. A small team of editors permanently employed at Google (and previously Songza) are in charge of managing the curators, tweaking their playlists and framing each collection of music with the right headline and description. They also analyze the way users are interacting with individual lists, swapping out songs that are being skipped too frequently for ones that might be more appealing.

While the curators often have highly specific taste—Google has worked with an expert in Spanish-language children’s music, for instance—the editors have to be generalists who understand the appeal of many different styles of music. As part of the hiring process, some editors had to make a playlist for Susan Boyle fans to prove they could pick songs that don’t necessarily align with their own taste.

“Even if it’s done by a super expert, it’s still for a general audience,” says Jessica Suarez, a product marketing manager at Google who serves as one of Play Music’s editors. “We’re trying to reach as many people as possible.”

Only Google knows exactly how well this focus on human-curated content is paying off. The company declined to disclose the number of Google Play Music subscribers, though it said the figure has doubled since it acquired Songza. A recently launched free, ad-supported tier centered around these playlists indicates that Google believes in their appeal. Newly launched Apple Music, meanwhile, has a similar emphasis on human expertise, with numerous playlists made by Apple editors and a live radio station geared toward highlighting lesser-known gems. Spotify also pays curators to make popular playlists on its service (curators across the services can expect to make a couple of hundred dollars per list, according to The Wall Street Journal).

For online music services, the increased importance of these human experts may be proof the algorithm’s role as the final arbiter in all digital decisions may be fading. “At this point, it’s not cool if you’re listening to Green Day radio,” says Roman. “Like, what does that even say about you? But if you’re listening to ’90s Aggro Anthems’ at the gym, you’re feel like you’re listening to something that’s thoughtful and representative of you and your aspirations and it’s not just pure, lowest-common-denominator background music.”

More Must-Reads from TIME

Contact us at letters@time.com