When McDonald listens to music, he sometimes thinks about how it'll ripple beyond the immediate moment.
"I think it's funny to hear individual songs and imagine, 'Is there going to be a whole other thing that spins out of that?'" says McDonald. He cites Drake's massive hit "God's Plan," and specifically the "fluttery orchestral samples" within its beat. "It's just different enough that you could imagine a bunch of kids who hear that song and say, 'Yes! that's what I want to do—muttering hip-hop over looped orchestral samples,'" he theorizes. "And then later you'll have somebody reciting abstract poetry over [Krzysztof Penderecki’s 1960 composition] 'Threnody to the Victims of Hiroshima,' and you'd read an interview with the artist and they'd say 'Well, you know, I heard 'God's Plan' when I was 13, and it was the only Drake song I liked it because it was like this.' There would be a thousand of them, all in separate bedrooms, and then they'd form a subreddit to talk to each other. And then you have a subgenre."
Subgenres are McDonald's business. By going over listener data and identifying patterns, McDonald and his co-workers can identify clusters of artists who might coalesce into a genre—something he’s been doing since his earliest days at the music-intelligence company The Echo Nest, which Spotify bought in 2014. Today, his work with Spotify's data helps listeners discover artists that may have been hiding in plain sight. McDonald’s “data alchemy” helps populate the Fans Also Like sections of Spotify's artist pages, as well as Daily Mix; it also provides a real-time chronicle of how music is developing and splintering into different styles.
Creating scene shots
McDonald’s plotting out of genres began as a debugging tool when he was at The Echo Nest, which used machine learning to evaluate different aspects of music—tempo, mood, energy—but which needed to have a human double-checking its distillations of music. Since the 2014 acquisition, the combination of The Echo Nest’s evolving analysis of musical’s “digital signatures” and Spotify’s ever-growing repository of listener data has allowed McDonald and his team to surface and identify different genres and styles, as well as “scenes” of artists who have a collective rapport with a particular audience, from all around the world. “If we can identify an audience, if we can identify a scene or cluster of artists, or a rationale ... then we're interested,” he says of how he and his team pull new genres from their massive amounts of data. Because of Spotify’s relative youth and vagaries in the metadata it receives from labels, the genre list gives a snapshot of what’s happening in music right now, with artists organized in groups and not chains of influences. "We have a lot of data about the present moment, and it gets fuzzier in either direction—but it gets fuzzier in the past almost as much as it gets fuzzier in the future," he says. But McDonald's focus on the present means that he and his team are always interested in those artists who have reached a certain critical mass in order to "build the stairway from obscurity to greatness," as he puts it. "If you get 300 people to listen to you, that's cool," he says. "It ought to be potentially the beginning of something for you…. I'm always looking for ways to find those people with hundreds or thousands of listeners, particularly when there are a bunch of them that they can make into some kind of cluster."
Seeking out the overlooked
Take Trap Queen, which McDonald recently added to his Every Noise at Once genre map. The term (one of 1,932 as of this writing) refers to a host of MCs including the youthful Bhad Bhabie, the experimentally minded Mykki Blanco, the Bronx star Remy Ma, and the raunchy Chicago upstart cupcakKe. "It's almost—not entirely, but almost—all female rappers," says McDonald "A few are fairly well-known, but there are tons and tons of women making this music."
In part, the genre came out of trap's gradual dominance of hip-hop's sound. "Originally, there was the great trap schism between electronic trap and the hip-hop version of trap," McDonald recalls. "But at this point, it's basically come to be the name for what the core of hip-hop is doing. That happens in everything, I think. You go through stages—right now, we're coming out of an era in which Auto-Tune dance pop was kind of the exclusive definition of 'pop,' and into one in which that plus hip-hop is what we mean by it."
While McDonald realizes that "gender is not a genre," slicing the biggest musical styles in a loosely demographic fashion doubles—and this is important to him—as an opportunity to elevate artists who might be overlooked by categories with a larger scope. "Popular hip-hop is so male-dominated. If you're not deeply into the hip-hop scene, you could easily hear [only] two female rappers if a hundred hip-hop songs are all you ever hear—Cardi B, and maybe a Nicki Minaj song, and that's it," he says. "For me—who's not immersed in hip-hop—being able to find this stuff is fascinating and revelatory." Artists like Brooke Candy and Lizzo have found success in the mainstream, but McDonald delights in unearthing artists like the Vanessa Carlton-sampling St. Louis artist Sexyy Reds while poring over stats and looking for patterns. "All her songs have less than a thousand listeners," he says while looking at her artist page. "Monthly listeners, 260. This is what I'm trying to find."
McDonald's work with musical data is aware of the mainstream, but that's also a way to guide his line of vision to the fringes. "I think almost everything I do is trying to get below the averages," McDonald says, "and find something that's true of some significant group of people, but not necessarily true of everyone.… It's always interesting if there's a particular group of people who like something. They probably cultivate it; they probably have somewhere they hang out, and there are probably other artists deliberately or accidentally falling into that orbit. We can find them."