The Secret Sauce Behind Your Productivity Playlist

Hint: it’s often a human.

When I was in college, I used to write my papers in the music library. Not because I was studying music, but because the music library was literally under construction. The study rooms there were cubes of bare white cement — a jail cell, basically — with a desk in it. I liked (and like) working in complete isolation, but heaps of basically-inconclusive-but-moderately-compelling-as-a-whole studies suggest that there’s another way that might be more effective.

There is a growing cottage industry devoted to figuring out what kinds of music make people more effective, more efficient workers. Certain genres — especially those without vocals — are frequently linked with “working.” But the actual scientific evidence concerning the links between music and productivity are varied and tend to be hyper-specialized.

Some studies indicate that music, or at least some sort of organized noise, can raise efficiency; one such study found that those performing repetitive office-like tasks were more efficient while listening to music than while working in silence. Other studies have gone more specific, focusing on specific genres like baroque classical music or even simple nature sounds, analyzing productivity’s correlation with how much the worker likes or dislikes the music (the results: if you want efficiency, don’t listen to music you particularly like or dislike), or whether the listener is an introvert or extrovert (introverts do better with silence, extroverts do better with music).

Muddling the science on this topic even further are the nearly equal number of studies indicating that music doesn’t, in fact, help you concentrate more efficiently. The amount of variables in any of these correlational studies are enough to crush any confident conclusion.

Despite their inconclusiveness, these studies have dovetailed with modern music streaming services in some interesting ways. For one thing, streaming services give listeners access to music they’d not normally buy — which, let’s be real, includes popular “productivity” music like classical and jazz. But because the libraries are so vast, they necessitates some sort of curation to surface music best-suited to getting work done.

Thus, Pandora, Spotify, Tidal, Apple Music, and others all boast these so-called “productivity playlists,” curated selections of songs designed to help you work — and they’re often very popular. According to Dayle Dempsey at Pandora, ‘Classical for Studying’ has been their top classical station since August 2013.

Most services offer a wide variety of productivity playlists. Google Play has classical, jazz, and instrumental pop stations, but also trip-hop, dubstep, instrumental trap, and quite a bit of metal. Tidal’s playlists under the “Focus” category include “Autumn Piano,” “Blips and Blops,” and “Post-Rock Essentials” along with more expected playlists like instrumental jazz and classical. Spotify’s productivity playlists also fall under a category titled “Focus,” and include instrumental hip-hop, indie folk, a fair amount of low-key electronic music, and even white noise.

The major attribute that seems to signal “focus music” to Spotify is a lack of vocals, as there is a bit of science suggesting that intelligible lyrics can have a negative effect on productivity, and quite a bit of anecdotal evidence suggesting it’s annoying. Furthering this point, Gregory Ciotti, writing for HelpScout, writes:

Research from Applied Acoustics shows that “intelligible” chatter — talking that can be clearly heard and understood — is what makes for a distracting environment. Shifting focus to figure out what someone else is saying is the reason why speech is often considered the most troublesome element of a noisy office; in one study, 48% of participants listed intelligible talking as the sound which distracted them the most.

Consistency between songs is another key attribute. You want a long enough playlist — several hours long at least — with similar-sounding music. Not necessarily mellow music, but music without major shifts in tone or beats-per-minute or genre. Big peaks and valleys can shock a listener out of the productivity zone, which you don’t want.

Given the utilitarian and somewhat robotic nature of productivity playlists, you might think they’re created by algorithms. Not so much. “All the stations are hand-curated,” says Jessica Suarez, who leads a team of in-house playlists editors at Google Play Music. In fact, her team crafts them based mostly on anecdotal information, personal preferences, and determined browsing of comments on the internet. “I’m a big Lifehacker reader, so I go through a bunch of the comments from people about what they like to listen to at work,” Suarez explains. “I also go through Reddit threads on what coders like to listen to at work. That informs a bunch of the stations.”

Even though the playlists are made by humans, Spotify, Tidal, and Google obviously have tons of data on what people are listening to: how much of a song they play, how often something is skipped, when volume changes, when a song is repeated or added to a user’s playlist, all kinds of stuff. “If people are skipping songs you know you’re taking them out of the moment, so we tend to weigh that more heavily and be more aggressive with how we edit [the productivity] stations,” Suarez elaborates.

Pandora — the media company that developed the Music Genome Project—takes a blended approach, drawing multiple data points from humans and machines. What a listener actually ends up hearing on Pandora is the result of a complex interplay between the two. Rhodes Kelley — a product manager at Pandora who works on their radio algorithm — elaborates: “There are more than 60 ‘strategies’ that can fire in tandem every time we need to find a song to play. A human curated strategy can nominate a song at the same time as a personalized data strategy nominates it alongside a musicological content based one. The layers of strategies picking tracks simultaneously allow the algorithm to blend all of our strengths together.”

Ultimately, the divide between algorithms and human curation comes down to how one perceives what music is. Music is different than other forms of data — for example your weather or fitness apps — in that it elicits emotional and cultural responses. For the techno-utopian inclined, the solution is simply to build smarter algorithms to parse the data — thus providing the best possible outputs for the user; but for those who see musical data as an incontrovertible cultural/emotional artifact, algorithms can never match the intentionality of human curation. The truth is likely somewhere in the middle.

Sponsored by Bose, The Sound of Innovation explores the ways in which sound shapes our experience of the surrounding world. Learn more about the new wireless QuietControl 30 and QuietComfort 35 headphones at Bose.com.