top of page

Popped Culture 

What shapes our unique music tastes?

Impregnable Question 

Here's my dilemma...

I'm not the smoothest talker—I've stumbled on words and underbaked thoughts enough to concede I should always come armed with talking points. Of course there's the weather, homework if I'm with a classmate, Timothée Chalamet or America's Next Top Model if I'm with my friends (peak culture!), or mild topics like pollution or disease if I'm feeling bored. Regardless of my company, there's one question I never skip over: "Listen to any good music lately?" Raising this question —through experience—invokes some weird quantum force field that only allows one of three answers to emerge from the brain:

    1. "Not really..."

 

    2. "Yeah, kind of. But it's nothing you would know or like."

 

    3. "Yesmusicismylife!DoyouhaveaSpotifybecauseweshouldfolloweachother.

         Iwenttothe[xyz]concertlastweekanditchangedmylife!Imetthemafterthe

         showandtheyfollowmeonInstagramnow.Haveyouheardofthem?

         ThebandisalittlenichebutIlivefordiscoveringnewartists!"—By far the least             common response

 

The...trinity...of man?

 

I guess I fall somewhere in between categories 2 and 3. I love to talk extensively about music when I feel like I'm in safe company—someone who has a similar "vibe" to me (clearly doesn't eat vegetables or sleep enough to appease any doctor; frizzy, bewildered hair; inexplicably melancholy). In reality, I feel like if someone wanted to get to know me, I should just throw a playlist at them instead of spewing inevitable nonsense. You know what?...that's what I'll do right now. Who is this Palehound you may ask? Or maybe, what's a Silver Jew? Yucky Duster, Junglepussy, Cheekface, Lala Lala...dude, these are some strange names! All valid points. And all reasons why I'm both nervous and thrilled dive into this vibrant and formidable world of music.

I'm sure you can see that music is a deeply personal and prominent factor in my life. And I'm guessing it's that way for you as well. Great...so we're joined by this common experience of loving music. But at the same time, we also obscure or sometimes even lie about the songs/artists we love out of a mutual consensus that our true preferences won't resonate with others; we are aware that our tastes are likely singular, or—at the very least—quite unique. This didn't used to be the case: think back to ten years ago and you may recall belting out ubiquitous tunes with your friends in a ritual induced by a mutual love of Katy Perry or Beyoncé or the Black Eyed Peas or...Nickelback?! (an undeniable Dark Age.) Where did this cultural cohesion go? I could simply state "the internet atomized a shared experience into a sea of AirPod-wearing-niche-genre-loving-strangely-isolated loners" and drop the mic there. But that's just skimming the surface of a deeply complex and somewhat nebulous shift in the music industry.

 

Frustrated that you can’t find a receptive group to share your favorite songs with? Me too.

Music in the Money Industry 

How charts changed music from an art form to a numbers game

While Splish Splash is for sure an enduring masterpiece, I can’t say that these songs didn’t all eventually put me to sleep—not because they were bad, but because popular music in 1958 was gentle, staid, and fairly simple. Jaunty? Yes, you could say that, but in the same way that…like…a carousel is jaunty. (I wonder what these people would think of their fellow Hot 100 alumna WAP, but I digress.)

 

Here’s the thing, though—I’m equally bored by the today’s top songs, but for a different reason. They’re redundant, flashy yet somehow also soporific, and—more times than not—vapid; the sad, sad rubble of a sonic golden age that ended not too long ago. (Of course there are exceptions. I lend my full faith and admiration to artists like Lizzo, Megan Thee Stallion, Chloe x Halle, etc. Heck, I’ll say it…I even liked "Driver’s License!") The difference between 1958 and now is that we have free access to nearly all of the greatest songs made over the past century (and then some). The ceaseless pursuit of hit-making seems trivial amid our current proximity to the perennial works of David Berman, David Bowie, and David Byrne—just to name a few. My point is not that artists should give up, but that Billboard should. Ranking music now is much more consequential and misleading than it was pre-internet.

 

Is this to say that the Hot 100 was ever an accurate barometer of popular music taste prior to the digital age? Absolutely not. Music charts were dismayingly fraudulent in their early days, relying on human accounting rather than any sort of objective machinery. Record labels viewed this honor system as an opportunity to augment the success of their own music by flippantly bribing music store employees with free albums and concert tickets (…and random appliances such as refrigerators and microwaves!); this ensured that the biggest artists with the biggest representation remained on top. On the flip side, music store employees sought to wield their own influence on the industry by underreporting sales for music they disliked. As this corruption persisted, the legitimacy of the Hot 100 suffered. Fans grew disengaged due to their minimal influence over the rankings, which were dictated by reported sales and radio airplay.

 

Seeking greater credibility and prestige, Billboard made the transition to concrete data analytics with their 1991 Soundscan partnership. Soundscan—a reliable bellwether to this day—functions as a digital bookkeeper of countrywide music consumption. By recording almost all scanned music purchases, Soundscan developed a database that accurately reflected the purchasing habits of American listeners in real time. Less than a year later, radio analytics (the other variable in the Hot 100 equation) turned digital as well upon Billboard’s acquisition of Broadcast Data Systems. For the first time, music charts were cleansed of the shameless corruption that record labels so frequently employed. The result? A transition of power that would shift the music industry rapidly and indelibly.

 

The advent of digital accounting brought authentic representation to music listeners by ensuring that their purchases were spoken for. In this sense, the notion that money begets influence remained a constant theme within the industry—whichever songs fans purchased the most were the ones that landed at the top of the charts. Much to the dismay of record labels, this proletariat taste proved to be pretty darn divergent from their own agenda.

 

You see, the word of mouth system that preceded Soundscan was flawed for more reasons than just bribery, among them being the fact that music stores often feigned data to skew toward what they assumed Billboard and labels wanted to see. This meant a lot of guesstimating took place in order to align with the misconception that popular music was synonymous with the pop genre; stores would inflate their reported pop sales in an effort to appear consistent with the Hot 100’s strong pop lineage. It should come as no surprise, then, to see how pop-saturated the charts were prior to the implementation of Soundscan.

 

The introduction of Soundscan negated this relative homogeneity almost instantaneously. Shop owners’ book cooking turned out to be grossly unrepresentative of their customers’ listening habits—it turns out that pop wasn’t the only genre people wanted to purchase. Rather, data analytics unearthed America’s entrenched love for country, R&B, hip-hop, metal, and alternative music. Charts once topped by the likes of Madonna, Cher, Bon Jovi, and Mariah Carey were soon infiltrated with the grunge and grit of Metallica, Nirvana, Dr. Dre, Coolio, and Beck. The industry was witnessing the most dramatic upheaval of popular music taste to ever occur, and everyone was freaking out.

 

What happened next was a dizzying cycle of charts influencing radio influencing TV influencing sales influencing charts, etc. To be clear, the whole country didn’t start liking Nirvana right at midnight on January 1, 1990 (although, having barely lived through the 90’s myself, that’s just kind of how I assumed things worked back then). Like many celebrated grunge acts from that period, Nirvana gained momentum in Seattle’s underground scene in the late 80’s. Their label, Sub Pop, sought to circumvent record stores by distributing singles through a monthly subscription service instead. This tactic enabled Sub Pop to influence what their target audience was exposed to while forgoing the price tag of record store bribery.

 

The release of Nirvana’s first Sub Pop single (“Love Buzz”) set off an unprecedented snowball effect. Gaining quick notoriety among Sub Pop’s network of subscribers, they soon evolved from being a niche regional act to a steady presence among the college radio landscape. Their freshman album Bleach added an eponymous fuel to the fire of this growing fanbase, with consistent sales helping the band emerge on Soundscan’s radar and ultimately attain widespread attention. By the time their canonical single “Smells Like Teen Spirit” was released in 1991, Nirvana’s grungy rage had penetrated an industry recently characterized by staid, poppy ballads.

 

In this and many other instances of genre breakthrough, the driving influence of data analytics is hard to overlook; without machine-based accounting, Nirvana’s divergent sound would have never stood a chance against the shiftiness of Big Pop (I’m talking about record labels here, not, like, Coca-Cola or some genial grandpa with a nickname). Cold, hard evidence was exactly the wakeup call the industry needed to embrace the fact that audiences were not exclusively interested in pop. It was also the impetus behind the expansion of genre fanbases. Nirvana saw their work propagate from punk pop-ups to the airwaves of impressionable mainstream audiences eager to keep up with accelerant chart-climbers. A similar effect could be seen in country acts permeating coastal cities, and in hip hop migrating to landlocked suburbs.

 

Of course, at this point industry executives were having a fit. No money-hungry corporation is interested in preserving fairness—or real artistry, for the matter. It’s no surprise, then, that major record labels rebounded from this spurt of democracy in a more calculating fashion than ever before (literally, the difference was that they now relied on quantitative data to chart out their warpath rather than defer to past instincts of offering glamorous subtleties such as kitchen appliances in exchange for exaggerated music sales).

 

Tacticians came up with a long list of ways to manipulate record sales above the table, yet still below any reasonable standard of common courtesy. For example, it became common for labels to send a promising single straight to radio and delay commercial release until it gained significant airplay. Seems like a missed sales opportunity, right? Perhaps, but keep in mind that securing chart real estate was considered the precursor to momentous sales—prestige first, then profits. Labels would harangue radio stations to play their singles over and over to induce the benefits of repeated exposure (the more you hear something, the more of a chance you’ll like it) and gauge regional interest. This latter objective was important for the next step of the equation: deployment to record stores. If feedback indicated that a certain area was especially receptive toward a nascent hit, labels would concentrate their efforts on growing sales throughout that locale.

 

After a radio single was deemed fully baked into the American consciousness, labels were more than prepared to serve a piece of the piping-hot pie to as many record stores as possible. Oftentimes a single would be offered to record stores at a discount so it could be sold to eager fans at an irresistible price. This tactic optimized a song’s chance of reaching the top of the charts during its debut week—a feat impressive enough to mobilize indifferent consumers to invest in an evidently unmissable piece of culture, and thus preserve a single’s high standing. Radio stations followed suit by granting the hit even more airtime, with MTV trailing closely behind on the contingency that it was accompanied by an equally dominant (hell raising) video. Of course, this tactic only really favored the best of the best, i.e. singles coming from top artists in the top genres, backed by the top labels. Music from specialty genres such as reggae, New Age, and Latin, on the other hand, experienced profound neglect due to technological disparities; primarily sold by niche vendors who could not afford Soundscan equipment, they floundered beneath a lack of representative data.

 

In the grand scheme of things, it seems like peddling new music to consumers by way of discounts would be a win-win; an equilibrium of eagerness was sustained by deals that aligned with hit-seeking consumers and hit-hungry distributors alike. If history is any indicator, though, it’s clear that the music industry is averse to any sort of stasis. Keeping with the enduring theme of record labels pursuing a blatantly unnatural course of action in order to promote a blatantly unreasonable agenda, the musical landscape of the late 90’s soon saw a new type of “science” overtake the grounded marketing tactics of years’ past. I am, of course, referring to the Frankenstein-esque maneuvering that led to the likes of the Spice Girls, Backstreet Boys, and ‘N Sync.

 

In pursuit of the next big thing, labels conjured up their own uber-marketable-megastar-chart-toppers through auditions that seemed to target looks and personality over trivial factors such as…musical talent. Yes, the admonishments of poster-weary parents around the world may have been onto something—like the foods that populated the lunchboxes adorned with their faces, these musicians were saccharine and over-processed beyond (reasonable) palatability.

 

I’m pretty sure it was Sir Isaac Newton who proposed that a force of nature such as The Spice Girls may only be vanquished by an even more formidable, opposing influence. In accordance with this First Law of Motion, it’s clear that there was only one juggernaut large enough for this undertaking—the internet! (Wait…did you think I was gonna say Destiny’s Child or something?!) Digital music sharing was galvanized by the distribution of WinAmp (1997), a free software that enabled users to download and play MP3 files on their personal computers. WinAmp and other similar services helped resourceful fans circumvent restrictions imposed by record labels designed to protect albums from being broken up and shared as individual songs. Labels witnessed a waning autonomy over which songs could be accessed as singles vs. which resided within the bulk of an album.

 

Through the development of peer-to-peer networks (think of it like Spotify’s database, but commercial songs are illegally uploaded by users rather than artists), tech-zealous fans quickly constructed an extensive catalog of free music. The most consequential of these platforms was Napster, whose 1999 launch presaged the decline of physical media. Napster’s model was simple and attractive—with a little computer magic and human sweat, local MP3 files could be extracted from users’ computers and aggregated onto an open database…all for the low, low price of free.

 

This democratization of music distribution wriggled the industry’s airtight grip on listening patterns; consumers were no longer beholden to the confines of an LP, or even a preordained single for that matter. Rather, any song could now be accessed on its own, meaning there was really no need to buy a full album anymore. Between 1999 and 2003, yearly CD sales would drop by over 200 million. By this point over a quarter of American adults had shared music files online; college campuses were banning peer-to-peer sites to alleviate network congestion; labels and musicians were suing Napster left and right for copyright violations (most famously, perhaps, was Metallica’s 2000 lawsuit), and eventually fans as well.

 

This slew of legal drama helped nobody. Napster was portrayed as an irreverent hindrance to honest moneymaking; fans were shamed for their complicity; and the higher-ups came off as greedy and desperate. Much to everyone’s relief, Apple, in its recurring role as the fairy godmother of the early 2000’s, brokered peace with its introduction of iTunes in 2001. iTunes was considered win-win for record labels and fans alike—royalty-sharing agreements helped revive sales previously stunted by the emergence of free music. On the consumer end, listeners welcomed the restoration of a centralized musical outlet following the shutdown of a heavily litigated Napster that same year.

 

And of course, who could forget Billboard. Characteristically late to the party, Billboard added digital sales to its Hot 100 algorithm in 2005, and online streams in 2007. This overdue adjustment heralded a period that extends into present day: the age of online listening. Of course, there are many ways in which music consumption differs now vs. a decade-and-a-half ago. But in the grand scheme of things, any evolution taking place post-Napster pales in comparison to the analog-to-digital switch in the late 90’s. It was during this fundamental transition that albums were atomized into widely accessible singles, giving fans unprecedented autonomy in cherry-picking exactly what they wanted to listen to. This shift still defines the state of the industry today—I mean, what is Spotify, or Apple Music, or Tidal, or Bandcamp if not platforms that give us the ability to access individual songs (alongside specially curated playlists)?

 

Quantitative analysis of chart data supports this grouping of post-Napster years. Economists Andrea Ordanini and Joseph Nunes set out to determine whether the introduction of digital music facilitated greater homogeneity in the Hot 100 . Their answer is… complicated. First, let’s imaging the Hot 100 is composed of two sub-lists: the Top 10, and the losers (the Lukewarm 90, i.e. songs 11-100). These first ten spots are typically occupied by the crème de la crème of the industry—Brittneys and Whitneys, an assortment of Justins, people with dubious titles such as Lady and Dr., etc. And the next 90? Truly random.

 

Ordanini and Nunes argue that these two groups followed diverging yet consistent paths during a period that spans the birth of Napster in 1999 to the completion of their research in 2013. While the yearly number of unique songs and artists gracing the Top 10 decreased, the Lukewarm 90 saw the exact opposite. This outcome reflects how we wielded our newfound access to the largest musical archives in history: the slowed turnover in the Top 10 group speaks to diminishing cultural cohesion, with new stars and new hits emerging at a lowering frequency. While this data suggests that we are no longer actively chasing after the Next Big Thing, it also implies that those at the top stay at the top longer. Basically, as a culture we’ve decided to take a collective breather in our search for pop idols; we let the big names grow bigger out of habit, choosing only to coronate a newbie when their star power provides enough momentum for us to exert the bare minimum. We look to Miley Cyrus, Cardi B, and Lil Nas X as prime examples—whether through preexisting fame, a persistent fight for exposure, or a notorious app, we are most amenable to developing collective reverence for things that are brought to us. As a result, this hallowed cohort is able to work the charts to their advantage by leveraging their reliably large fanbases. Taylor Swift and Lady Gaga sustain their chart standings simply by continuing to be their behemoth selves and mobilizing a widespread army of fans. They stay at the top because we are comfortable holding them up.

 

The remaining 90 songs, however, seem to be collectively amorphous. The rapid influx and removal of songs from this group hints at the dissipation of a once monoculture; we are overwhelmed by the enormity of content at our disposal, jumping from new song to new song too quickly for anything to effectively stick. Of course, our inability to commit to the majority of these quick-hits is not purely a symptom of fast pacing, but also lackluster quality. A 2017 study lead by computational analyst Matthias Mauch identified a plateau in the functional diversity of Hot 100 songs propagating from the Napster days onward. Essentially, the songs that reach the Hot 100 can often be derivative of those at the very top in an attempt to replicate their success.

 

It may be puzzling to consider how this Lukewarm 90 came to fruition during the age of streaming. When we can listen to any of Spotify’s 70+ million songs, how do we all seem to encounter the same handful in passing? Record labels have essentially devised a series of methods to “hack” the charts through the infiltration of promoted content on apps. With Spotify this involves labels trying to land their artists on central playlists such as Today’s Top Hits, New Music Friday, “Essential [insert genre name here]”; releasing multiple versions of the same song in different styles (there are reggae iterations of Ed Sheeran songs out there in our evidently screwed universe); and producers avoiding slow buildups in order to prevent a song from being skipped. For the depraved, burgeoning TikTok, songs gain popularity through app executives compelling creators to develop and share accompanying dances (did I mention we’re screwed?). And this is where we went wrong. Charts are now meaningless avenues for greedy executives to exploit an algorithm in order to retain notoriety in the industry. They no longer reflect what we want to hear, but rather what the labels want us to hear. In this digital climate where we’re so easily influenced, the survival of charts means the sacrifice of artistry over strategy.

 

At the end of the day, we can only be manipulated so much before running out of patience. Sure, the agenda of record labels may make us stop and glance at the songs they’re trying to promote, but ultimately we no longer need to be told what to like—Spotify and other services now offer us personalized guidance via the generation of user-specific playlists. They facilitate exploration by appealing to our specific preferences rather than treating us as a collective target for backhanded manipulation. Thanks to streaming, we have finally reached a Nirvana in which chart-induced shiftiness no longer dominates what we listen to.

Lists can be annoying—they attempt to impart order and reason onto things that are naturally complex. Think about the Best Colleges list from U.S. News, the handful of movies nominated for Best Picture at the Oscars, or even the New York Times’ hotly-contested bagel rankings. While the logic behind all of these is nebulous at best, there’s no denying the influence that accompanies a declaration of quality: “definitive”, neatly packed data is convenient, and convenience begets exposure, begets money, begets power. The allocation of this power, however, is where we get into real trouble; authority is bestowed onto the listees by the listers. But what about the third party—the list readers? We’re taken along for the ride, our opinions siphoned through an equalizing funnel of simplified thought…the prevailing consensus is universally palatable because it doesn’t taste like anything.

 

Mass culture is often shaped by a law of averages, through which the most neutral iterations are also the most celebrated. And herein lies the fundamental issue—our gravitation to the same shortlist of mediocre art is blaringly unnatural. Nowhere is this more evident than in the music industry, where the presence of charts has calcified an overbearing emphasis on hitmaking over quality. The perpetrator at large? Billboard charts, most notoriously the Hot 100.  

 

To understand the impact of the Hot 100, it’s best to start by taking a look at how it’s evolved in parallel with the music industry. Let’s indulge ourselves in a timeline:

 

  • August 1958: The Billboard Hot 100 is born!

  • October 1982: Billy Joel’s 52nd Street becomes the first music CD released

  • November 1991: Billboard’s implementation of Soundscan introduces computing to record sales accounting

  • November 1998: Non-retail (radio-only) songs become eligible for the Hot 100

  • June 1999: Napster launches, offering the first centralized digital music library

  • January 2001: Apple launches iTunes

  • July 2001: Napster shuts down

  • October 2001: Apple strikes again! The iPod is released, enabling users to carry “1000 songs in their pocket”

  • February 2005: Billboard adds digital sales to their Hot 100 chart criteria

  • September 2005: Pandora Radio opens paid subscription service (free version followed in November that year)

  • July 2007: Hot 100 adds digital streams to its formula

  • July 2011: Spotify launches to the public

  • February 2013: YouTube views are added to the Hot 100 formula

  • September 2016: Tiktok undergoes its initial launch

 

It’s a chore to keep up with the mutability of Billboard’s criteria. To be honest, putting together that timeline made me want to tear my hair out or—at the very least—get a (weak) Twitter campaign going. Complaints aside, it’s hard to place blame on Billboard alone. I give them a lot of credit for attempting to keep up with the daunting mobility of the music industry during its most concentrated period of technological advance. This isn’t to say that Billboard’s relationship with new technology is completely reactionary, but rather that a synergy exists through which either one could incite the other to evolve.

 

Billboard’s approach to filling charts continues to be chaotic yet potent: chart criteria change in response to new technologies; avaricious record labels find a way to exploit these new criteria in order to surge to the top of the charts; music becomes more homogeneous and stagnant as different labels pursue the same shortcuts; technology inevitably shifts again; rinse, repeat.

 

The one constant in this storm is the list itself: while the accounting has changed, the very act of ranking and publishing a weekly list of top songs has not. We’re talking about the same charade propagating from generation to generation of industry executives, while the rest of the world invariably and monumentally changed. I mean, come on! When the very first Top 100 was released in 1958, we were still a decade out from the invention of the internet, forty years out from the advent of streaming, and, most importantly, 37 years removed from Thee birth of Megan Thee Stallion. There was no such thing as hip hop, the Grammy’s had yet to be established, and America’s top three songs were deemed:

Extraordinary Machine  

How do streaming services facilitate discovery?

There is no modern invention I value more than Spotify (maybe this would be different if I had a tighter grip on reality, or if I wasn’t allergic to penicillin). It’s how I begin and end my day, how I fill the silence of my single apartment, how I seek solace when I’m down, and how I tolerate any road trip longer than five minutes. According to their records, I spent 32,018 minutes on Spotify in 2020—that amounts to over 22 full days of listening, a moderate timespan relative to the numbers reported by my friends. 

 

I’ve been on Spotify since 2014, when I was a mere 14-year-old disciple of artists such as Vampire Weekend, the Black Keys, Regina Spektor, Kimya Dawson etc. My taste was quintessential 2010’s indie, indistinguishable from that of my quasi-hipster middle-aged peers. But that’s not the case today. In my seven years on this app, I’ve migrated from alternative to folk to singer-songwriter to punk. I’ve devoured their tailor-made playlists, attended recommended concerts, and revered their featured artists. Basically, I’ll eat anything they feed me.

 

It’s no secret that streaming services play a huge role in determining what we listen to. Spotify is undoubtedly the most potent of them all, amassing a user network larger than the U.S. population. Being the most popular platform suggests that Spotify is also the most effective at enticing and retaining users with music that truly speaks to them. Through a steady diet of both new and old music, Spotify keeps our musical curiosities nourished.

 

Uncovering the alchemy behind Spotify’s appeal allows us to understand exactly how our music tastes are targeted and shaped. To start, Spotify embraces the disorienting reality of our multiple personalities—rather than take the perfunctory approach and categorize our listening patterns by genre, Spotify’s algorithms aim to sort our behavior using emotion- and behavior-based criteria. This approach is driven by three synergistic models: collaborative filtering, natural language processing (NLP), and audio models.

 

While collaborative filtering is not a novel concept, Spotify’s implementation is considered exemplary. Collaborative filtering facilitates an “I’ll have what she’s having” phenomena, without any of the characteristic subtle stalking. Spotify is constantly updating user taste profiles, with every stream, skip, and like helping them to paint a more vivid picture of who we are as music listeners. Our personal profiles are then compared to existing playlists to identify overlap; if the filtering algorithm identifies common ground between a user’s profile and a playlist, it will extract recommendations from the playlist’s remaining undiscovered songs. This process iterates across the hundreds of millions of playlists and users on the platform, ensuring that suggested music aligns with our multidimensional tastes. Unsurprisingly, it is the impetus behind Spotify’s myriad of personalized playlists and radio stations (this is how your Discover Weekly stays fresh!).

 

The ubiquity of Spotify’s collaborative filtering mechanism suggests that similar listeners will be fed similar recommendations—someone with the same music taste as me (bless their heart) will likely find compatibility with the same set of playlists, leading us both to glean comparable recommendations. Essentially, our personal predilections lead us to be grouped into specialized communities. This may explain why you sometimes encounter accounts with disarming resemblance to your presumably niche tastes…I can’t emphasize how unsettling it is when I encounter a fellow Sidney Gish-Mitski-Fiona Apple stan out there (pseudo-lachrymose, faux-indie millennials and Gen Z-ers abound). It’s a mix of “darn, I guess I’m really not special,” mixed with “woohoo! My people are out there somewhere.”

 

Again, Spotify’s algorithm is especially effective in scoping out recommendations from resemblant pairings. They infuse their algorithms with humanity by deferring to user- or employee-made playlists for generating a database of preferences—tastes are shaped using experiential evidence rather than pure computation. This is where Spotify stands out from its peer streaming services, such as Pandora Radio. Once dominant in the streaming sphere, Pandora’s user-base was merely a quarter of Spotify’s in 2019. Their demise? An algorithm that treats listeners as isolated data points rather than contributors to/beneficiaries of an archive of music listening patterns. In other words, Pandora treats the preferences of each user as if they were stored in their own black box—recommendations are configured from the ground up, rather than derived from the habits of subscribers with similar tastes. While it is true that this causes users’ playlists to grow increasingly singular, it also means that we’re insulated from receiving suggestions inspired by the listening history of our like-minded peers.

 

Aside from their effective filtering techniques, Spotify also seeks to understand users through the deployment of natural language processing and audio models. The former explains how Spotify is able to transcend designating songs simply by genre. Playlists are converted to text files, from which lyrical affect can be deciphered. This gives Spotify’s database an idea of what kind of moods a listener seeks from music, and what time of day/week these moods correspond to. It’s part of the reason why we’re fed peaceful songs on the weeknight, and hardcore rampage anthems on Friday and Saturday.

 

Audio models step in when filtering is stunted by a lack of data, such as in the case of deciding when to recommend a new song by an obsolete artist. Because the obscurity of a smaller song hinders natural discovery, Spotify cannot empirically determine the type of listener who it would appeal to (and should be marketed to). Instead, neural networks are used to gauge characteristics such as key, tempo, and dynamics, and identify listeners with proximate preferences. 

 

It can sometimes feel like I’ve sold my soul to Spotify—but, truly, in the most delightful way. I provide my honest reactions to their steady influx of neatly packaged, mystery-flavored playlists, and in return I get more personalized content. It’s one of life’s purest joys…encased in a halo that dims ever so slightly upon realizing that their benevolence is fueled by underlying commercial motives. Yes, Spotify’s mission to entice users with pitch-perfect, soul-peering recommendations is (at least partly) driven by a desire to retain and grow their population of subscribers. We’re pushed deeper and deeper into our comfort zone under the presumption that safe users are happy users. Regrettably, Spotify primarily recommends content similar to the things we currently enjoy.

 

While the logic behind this is certainly defensible, it ultimately prevents us from expanding far beyond our incoming tastes. Of course with any exposure to new music, our opinions and preferences will evolve and expand—hence my afformentioned transition from alternative to punk. But this algorithm is likely incapable of helping me make the long trip all the way to the rap universe (no matter how much I try to get an explicit song on my Discover Weekly, I almost never can!).

 

It’s not surprising that Spotify, a multi-billion dollar corporation, isn’t exclusively in our corner—while their complex, thoughtful algorithms certainly lend extensive depth to our tastes, the same cannot be said for breadth. Cynicism aside, I firmly believe that the good outweighs the harm. I mean, we have free access to a personalized music library after all! It almost makes me want to sing along.

Coming of Age  

Could our disparate music tastes be a birthright?

Anchor 5

The final piece of the puzzle is a bit more clinical, but compelling nonetheless. First, a little bit of introspection: say you’re making a playlist of your all-time favorite songs—the comfort food that stirs you fundamentally, that somehow is deeply personal despite coming from a complete stranger. When in your life did these songs become important to you? Or, what time period does each song make you nostalgic for?

 

I’ll let you have your moment.

 

While this exercise may simply serve as a warm (or bewildering) trip down memory lane, I’m hoping that is also helps introduce you to the psychological theory of the Reminiscence Bump—that our tastes are indelibly defined throughout our teens and twenties. Basically, we solidify lifelong traits, beliefs, and tastes during the period in which our brains undergo the greatest development. I’ll be honest—when I learned this concept a couple years ago, I was completely alarmed. As a then nineteen-year-old, I felt pressure to suddenly develop fantastic, refined taste so I could set myself up to impress fancy people for the rest of my life. I was also horrified to learn that my teenage Gossip Girl fixation would probably be a lifelong affliction…why didn’t anyone warn me that I should be watching, like, The Wire, or listening to Bob Dylan or some other sage old man, or—I don’t know—reading?!

 

Our Reminiscence Bump defines who we are; as we trickle into adulthood we become real people by confronting tough decisions, forming and ending potent relationships, moving away from home, etc.—basically all the things songs are written about. Naturally, we often seek music during this time to regulate, complement, or even stimulate the inevitable mess of emotion; it meshes with our internal monologue so profoundly that neural pathways associated with our favorite songs are permanently established. It makes sense, then, that the majority of Spotify users fall within the 16-34 age bracket. Lending to a conflation of free time, angst, and “disposable” income, we seek out music to help explain and process momentous change.

 

In line with the Reminiscence Bump theory, a 2014 survey of Spotify data found that users listened to music released between 13-14 more than any other age. While this result is fascinating, it doesn’t tell the whole story. Think back to 2014: sure, streaming was on the rise, but it was much less prominent than it is today. In fact, I would classify this year as an inflection point, after which radio quickly became obsolete and streaming finally took over. The fact that users preferred songs that were released at age 13 and 14 points to the presence of a centralized music culture during their teenage years: they subsisted on a diet of new songs spoon-fed to them by radio DJs. This cultural cohesion fomented the development of a cohesive generational reminiscence bump; the shared soundtrack to their youth continues to influence their listening habits as adults. The same can’t be said about teenagers after, say, 2014.

 

Growing up during the streaming boom was both exciting and confusing. My love of music was already strong up to that point—I got lucky with a dad who has what I consider to be the coolest taste in music for any middle-aged dweller of Midwestern suburbia. For as long as I can recall, car rides consisted of the Beatles, Talking Heads, Elvis Costello, Fiona Apple, Kanye, Vampire Weekend, Cat Power, and many, many more. Again, I was very lucky. But I was also hungry for more, so I decided to branch out and try Pandora Radio in 2012 (I was 13). After feeding a few artists into my page and giving eager likes, dislikes, and skips, I soon was fed a personal radio station consisting of artists like Kate Nash, Lilly Allen, Jaymay, Regina Spektor, and Kimya Dawson. And I’ve been listening to them since. While these artists are not my all-time favorites, their influence on my taste today is crystal clear: my playlists are now saturated with angsty, eloquent female singer-songwriters.

 

Whenever a song from my Pandora days comes up, it’s like sweet yet jarring trip back to middle or high school, when I spent more time with headphones in than without. This music is my security blanket—when I listen to it, I feel at home. However, my personal roster of beloved songs probably doesn’t invoke the same ardent feelings for most of my Gen. Z peers. The rise of streaming urged the dissolution of a centralized music culture, meaning that what I was listening to during the pivotal Reminiscence Bump years differed drastically from those around me. The allure of personalized playlists often trumped the shininess of new releases that would have penetrated any other generation.

 

Consequentially, post-2014 teenagers lack cohesion in their foundational preferences, which extrapolates to even an even greater disjunction in adulthood. In other words, our modern Reminiscence Bumps are like fingerprints—each one unique to the body it belongs to, intrinsic to our identity. We therefore miss out on the musical bonds prevalent among past generations.

 

In a very 2020 twist, TikTok (to invoke my curmudgeonly inner-parent) has come to wreak havoc on life—and music—as we know it. The app’s recent insurgence has shaken the very foundation of our modern music industry, teasing the rise of a once-again monoculture—whereas Spotify exists to deliver music to listeners, TikTok promotes a handful of flashy songs as an important but subordinate means to heightening content. Complicit in the breakthrough of omnipresent artists such as Megan Thee Stallion and Olivia Rodrigo, TikTok established its grip on the music industry by brokering intimate relationships with record labels.

 

In the case of our friend Ms. Stallion, success was gleaned through a collaborative effort involving her representation at 300 Entertainment and TikTok strategists: a handful of songs were selected to be featured on in-app playlists to encourage distribution via (eventually) ubiquitous user-made videos. Hence the popularity of "Savage". Artists also commission TikTok celebrities to develop dances, or “challenges,” for a particular song. The hope is that users will replicate the dance in mass, rendering the song a fixture on everyone’s feed.

 

TikTok has over 689 million users as of 2021, which makes it around twice as large as Spotify. As TikTok continues to incubate hit after hit, it’s clear that the individualism of Spotify may be facing its downfall. If this comes to fruition, the days of atomized listening may be supplanted by the return of a centralized culture à la the 90’s. Reminiscence bumps will once again converge, heralding the first generation with collective taste since the rise of streaming. On the one hand, I pity the kids who may lack the impetus to branch out on their own and uncover the weirdness and singularity of their musical preferences. But I also marvel at how cool it would be to have a widespread community of people who loved the music as me. Think—which would you rather have? I’m not sure that I have a definitive answer myself.

The Conclusion  

What's the value of a shared music culture?

This section’s title may be a bit of a misnomer. As you’ve seen, our taste in music is the product of numerous powerful influences—self-promoting record labels vying for their artists to penetrate pop culture; powerful algorithms designed to know us better than we know ourselves; our (often unfortunate) teenage preferences. After being processed through this turbulent conveyor belt, what do we have left? Which of these three forces prevail?

 

I don’t, and will never, have a definitive answer to this. The landscape of music listening is, as I warned from the start, vibrant and formidable. As one traverses it, they invariably fall in love, grow enraged, succumb to sadness, feel inspired or understood or uneasy or undone—or all of the above. We ask for music to be there for us in any time of need, to feel intimately and completely seen, to lend us the words we can’t seem to come up with ourselves. But, if music today is such a personal experience, we miss out on establishing connections through the ties of a shared culture.

 

As we broach this oncoming phase of music listening ushered in by TikTok’s unifying influence, we need to consider which side of the coin we prefer—being personally influenced by an algorithm, or seeking connection among a central culture that ignores our individuality. Meaningful art comes at a price. So does meaningful human connection. While we’ve seen either extreme prevail in alternating periods, I’d like to languish in this inflection point—grateful to Spotify for revealing to me who I am through song, but excited to sing along with friends in the near future.

bottom of page