Time to jump off the algorithm highway

Life is perpetual change, so it is perhaps overdoing it to suggest that the music business is at a cultural pivot point. Yet, what comes next has the potential to be looked at, years from now, as a dividing line between before and after. For more than half a decade, the music business has been hurtling down the algorithm highway, repurposing artist development, marketing, fan engagement, and even the structure of the song itself in order to stay the path. Everything is splintering, from attention to remuneration, with creators and rightsholders alike finding themselves feeding a beast whose hunger is never sated. Much like an addict who wants to quit but cannot, the music business understands the problem and the costs it incurs them, yet they dare not jump off the algorithm highway for fear of being left behind by those who do not. And yet, jumping is exactly what is needed, to halt the perpetual commodification of both music and creators. It is a leap of faith, but onto a welcoming crash mat: scenes.

At Future Music Forum this week, myself and fellow MIDiA analysts Tatiana Cirisano and Kriss Thakrar talked a lot about MIDiA’s new research into scenes and identity (MIDiA clients can read our latest report on the topic here). Regular readers will be familiar with our work on fragmented fandom and how the splintering of consumption has created a parallel splintering of culture, with new hits becoming smaller and more short-lived. In this song economy environment, it is the song, not the artist, that is the central currency, thus making nurturing smaller fandoms mission critical. But fandom itself is the symptom, the cause is identity, and this, along with the scenes in which it manifests, is where the future of music marketing lies.

Algorithms have assumed a central role in the success of artists in today’s music business, with marketers forever trying to improve their understanding of their inner workings in order to gain advantage for their artist. It is, in many respects, a fool’s errand, as it is in the platforms’ interest to continually evolve the algorithms in order to ensure it is themselves that determine success, not third parties. Nonetheless, there are ways to succeed in the song economy: you may not be able to beat the algorithm, but you can join it. This means thinking and behaving like an algorithm, to hold virality by the hand. Just like an algorithm, this means real-time multivariate testing within target segments, and progressively expanding only to next-level associated segment, resisting the ability to go big as soon as something fires. But using the algorithm as a marketing discipline truly effectively entails a degree of ruthlessness that many artists and labels would find unpalatable. Algorithms find success by casting out failure instantly, instead only amplifying that which resonates within target segments. So a label pursuing this approach would need to be willing to ditch a campaign incredibly early if it does not, however much the label might believe in the release or however big a priority the artist might be. Artist rosters would become a production line of bets, as quickly discarded as signed. Failing fast is as important as succeeding fast in the song economy. 

This ruthlessness does not sit well with the traditional model of building an artist but, as dystopian a vision as it might be, is the exact path that labels already find themselves on. Scenes represent an alternative way forward.

Scenes and identity

Scenes have always existed, but now there is a growing proliferation of online scenes that allow a degree of specificity that was simply not possible previously. As Tatiana puts it:

“Not only can people find people across the globe with the exact same interests and values, algorithms actually push those people closer together”.

Though scenes can be transitory and ephemeral, subject to fast-shifting cultural trends, the really valuable ones are those that are rooted in identity, that speak to who people are about. The eBoy scene, with Young Blud as an icon, is a case in point, reflecting the values of a tribe that does not identify with the Instagram-perfect archetype of appearance. 

These scenes sometimes revolve around music, but most often, music is simply the soundtrack, with a number of artists emerging as icons, not because they have cynically targeted them but because they come from those communities and reflect their values. Fandom is an output of this shaping of identity. It is simultaneously a way of showing how much identity matters to you and of reinforcing that identity. In fact, fandom is identity’s virtuous circle of influence, with people’s fandom reinforcing their identity and communicating it to their scene community, thus reinforcing their bonds within it.

Identity is fandom’s ground zero. Music marketers that are able to identify and nurture it (rather than simply attempt to harvest it) have an opportunity to forge a depth of artist-fan relationship that will endure far beyond the whim of any algorithm, survive both hit and miss singles, and will not disappear into the black hole of lean-back consumption. 

Streaming put fandom on hiatus. Scenes represent an opportunity to reforge fandom for the modern era, an incubator for artist careers. In short, an antidote to the song economy.

WMG is moving beyond superstars – and that is a good thing

Warner Music Group’s (WMG’s) Steve Cooper recently stated that the major is no longer financially dependent on superstars – which is, of course, quite a different thing from not being culturally dependent on them, but we’ll get to that. For a major’s CEO (exiting or not) to make such a claim is both bold and a reflection of the reality on the ground. In fact, it is a natural milestone in a trend MIDiA identified years ago: fragmented fandom. As streaming audiences and consumption fragment, so does the impact of superstars. As with any transition, the shift is not linear and there will continue to be more Olivia Rodrigos and Billie Eilishes, but they will be fewer and farther between, and crucially, they will be smaller than their pre-fragmentation peers.

Superstars getting smaller is music to the ears of independent labels and artists alike, but it is far from the death knell for big labels. Instead, it simply reflects the new environment in which they will operate. Indeed, Cooper said WMG is pursuing a “portfolio” strategy “across a bigger number of artists” to reduce financial “dependency on superstars”. This comes after BMG’s CEO, Hartwig Masuch, said of their latest results: “The extraordinary thing about our first half result is that we grew revenue 25% with virtually no hits”. Having no superstars does not mean having no hits, instead it means more, smaller hits.

In 2019, MIDiA wrote that “Niche is the new mainstream”, that the water cooler moments of the linear era were being replaced by cultural moments. Audiences are in different places at different times, with algorithms delivering them different personalised content. Concepts, such as ‘song of the summer’, are becoming different for everyone. Each listener has their own song of the summer. In the era of fragmented fandom, water cooler moments across the masses, where everyone heard the song on the radio at the same time, are replaced by smaller groups of people finding pockets of likeminded fans across the world.

The consequence for artist marketing is a progressive shift from ‘carpet bombing’ mass media in order to build artist brand reach, to campaigns that, instead, reach real fans with laser-focused targeting. In the old model, a superstar artist was a household name, with mum and dad just as likely to know them as their kids. But what was the value of mum and dad knowing the artist if they were not the target audience? It might play to the artists’ egos, but it was an inefficient spend of marketing budget. Now, targeted marketing reaches the consumers who care. The result is smaller, but more passionate, fanbases. This is marketing to build fans rather than audiences. It is just a shame that western DSPs are built for passive audiences rather than fandoms. That will need to change. DSPs paradoxically triggered the fragmentation, but they do not provide the mechanisms for artists and labels to benefit. A cynic might argue that that is by design.

Indeed, the fragmentation of listening that streaming is pushing consumption towards the middle, away from the superstars, as Music Business Worldwide’s Luminate chart for streams of the US top-10 tracks shows.

For the superstars who are used to mega-fame from the pre-fragmentation days, a new release’s performance can look like diminishing success when measured by traditional metrics – just ask Beyonce. But, because fragmentation means it is truer fans that engage with the music, the cultural relevance of these smaller hits can actually be bigger – again just ask Beyonce.

There are, however, extra complications. As we are currently in a transition phase, pre-fragmentation hangovers are muddying the streaming waters. Pre-fragmentation hits stick around for longer on streaming because they had the pre-fragmentation brand reach. Since they benefited from the old-world mainstream media exposure, their hits cut through on streaming in a way that newer ones often struggle to. These pre-fragmentation hangovers have the effect of fragmenting new hits even further as they take up so much of streaming’s consumption. The result is that streaming is not so much a level playing field as a field of all levels.This transition phase will play out, and while it does, there is a world of opportunity for artists and labels that can harness the deeply held fandom that fragmentation creates.

The rise of scenes

The most exciting knock-on effect of fragmentation is the rise of scenes and micro scenes. In the old world, consumers had a limited range of things with which they could identify themselves, as everyone was watching the same TV, reading the same magazines, listening to the same radio, and shopping in the same shops. Now, consumers can build their own identity from an ever more diverse set of attributes, across fashion, music, TV / film, games, politics, etc. 

As my colleague Tatiana Cirisano put it:

“The result is that scenes are becoming more complex and splintered. Consider the seemingly endless range of subcultures on TikTok, from #cottagecore to #EGirl, or the Instagram account @starterpacksofnyc, which has garnered more than 64,500 followers by crystallising super-specific, yet eerily-familiar, personality types.”

This rise of scenes is what will shape the future of marketing, with scenes becoming the new territories, transcending borders and cultures. Superstars will get smaller, but they will get better at monetising their superfans (this is why Taylor Swift’s Universal Music Group deal includes a broader range of rights than just recordings, as her sales were only going to go in one direction). Superstars are not dead, they are changing, become smaller and less, well, super. It is an inevitable second-order consequence of streaming splintering listening and the smart labels will harness the trend rather than try to fight it.

Fake artists are what happens when fandom dies

The topic of ‘fake artists’ refuses to go away. For those who have been on Mars for the last couple of years, fake artists refer to artists who release under a streaming pen name but do not build any artist profile around the music. Most of this music comes from production music libraries (typically ‘royalty free’) and is seen by the traditional music business (record labels especially) as a means of gaming the system – especially as the assumption is that DSPs pay less for such music (even though record labels have started playing the game themselves). Although the ‘if you can’t beat them, join them’ might seem like a pragmatic solution, it, of course, only exacerbates the problem. Because the problem is not fake artists, but it is, instead, the way in which streaming is killing fandom.

Streaming is racing to be radio, not retail

Streaming is fast becoming more of a replacement for radio than it is retail. Retail used to be where (engaged, smaller scale) fans went, while radio was where (passive, larger scale) audiences went. As streaming got bigger, there was always going to come a point in which its focus would be the large passive audience segment rather than the smaller engaged fan segment. But what has happened is that streaming is turning everyone into the passive massive, even fans. Streaming has turned music into a utility, like water coming out of the tap. This might have helped drive global scale, but it came at the cost of fundamentally eroding the cultural impact of music, by making it about consumption rather than fandom. 

Streaming music soundtracks our everyday lives. There are playlists for everything we do (study, fitness, relaxing, cooking, working, etc.). By becoming pervasive, music has lost some of its magic. The fandom that was inherent in people buying music because they loved it is gone. The biproduct of ubiquity is utility. In the immortal words of Syndrome from the Incredibles: “When everyone is super, no one will be…”

The problem is that, from the ground up, Western streaming is geared for consumption not fandom. From playlists through to economics, streaming is all about consumption at scale. Songs fuel consumption, not artists. Which is the breeding ground for mood music, of which ‘fake artists’ are but one sub-strand. 

Streaming’s torrent of ubiquity

This is not to say that there is anything inherently bad about consumption, after all, radio has been a corner stone of the music business for, well, pretty much forever. Labels have had a love / hate relationship with radio, but they valued the way in which it drove sales and delivered exposure for songs and artists (especially as DJs talk about the music being played, interviewing artists, etc.). With streaming, though, the discovery journey is the destination. So, the post-consumption part of the equation just disappeared. And a consumption-first environment, tailored to individuals’ daily lives and shorn of the artist context delivered by DJs, is fertile ground for mood music. In fact, mood music is the natural evolution of a consumption-first system. A system in which artists get washed away by streaming’s torrent of ubiquity. 

Add poor remuneration for mid and long-tail artists into the mix, and you have a perfect storm. Why? Because artists are compelled to diversify their income mix to eke out every extra dollar they can get from their creativity, with production music libraries being eager customers of their ancillary work.

Fandom has moved up the value chain

Streaming may have killed off fandom within its own environment, but fandom itself has not died. It has gone elsewhere (Bandcamp, Twitch, TikTok, etc.). It is TikTok that has arguably done the most to reinvigorate fandom in recent years. But, crucially, it has inserted itself before consumption instead of after it. You will be hard pushed to find a mainstream music marketing campaign that does not include TikTok as the place to kick start discovery and (if all goes well) virality. TikTok has thus become the top of the funnel for consumption. Yet, rather than filtering out what is valuable, the process is more like panning for gold, i.e., filtering out what is not valuable – consumption. Fandom, identity, recreation, engagement, and connection are all left with TikTok, while consumption flows through to streaming. Little wonder, then, that TikTok is diluting streaming’s cultural capital. 

It does not have to be this way. Chinese streaming services demonstrate that streaming can be fandom machines too. Tencent Music Entertainment makes around two thirds of its revenue from non-music, fandom revenue. But perhaps the most startling example of just how much is being left on the table by Western streaming services, is found in NetEase Cloud Music’s inaugural earnings release. 212 million music users generated RMB 3.6 billion. 0.7 million social entertainment users generated RMB 3.7 billion. Yes, that means an audience that is 0.32% the size of the music audience generated more income in fandom-related revenue than the music audience did in music revenue. Right now, if anyone in the West is going to be streaming fandom machines, it is probably going to be TikTok (a Chinese company) and Epic Games (a company 40% owned by a Chinese company).

Fandom remains the under-tapped resource in the West, but its value is not simply in the revenue potential. Fandom is the essence of what makes music move us. Under-invest in it, and music will continue on its path of commodification. Which might serve the streaming platforms well, but not the wider music business. ‘Fake artists’ will become the norm, not the exception. To misquote syndrome “when everyone is fake, no one will be…”

Tribes are the future of fandom (and that may or may not be a good thing)

At MIDiA, we spend a lot of time exploring the fan economy and how new forms of fandom are redefining media businesses. The most significant underlying dynamic is the fragmentation of fandom: the dynamic whereby we move ever further from mass-reach media, where everyone is exposed to the same content, to a world where entertainment exists in a complex mesh of filter bubbles. Niche becomes the new mainstream. Whereas, as Asian entertainment companies have become adept at industrialising fandom in this new paradigm, Western companies are less so. In music, big record labels still have a mindset of wanting to create mainstream, global hits. But the fandom playbook is changing. Global success now depends less on how wide your message can reach, and more on how deep it can go. Mass reach is becoming superseded by conversion, and mainstream is become replaced by tribes.

What do Donald Trump and tribal fanbases have in common?

Tribalism is, for better or for worse, central to how humanity functions, and, of course, underpins millennia of armed conflict. How tribalism can manifest among strident fanbases is simply a lighter shade of the same dynamic that can send countries to war, part of the same sliding scale on which Trump operated. In fact, the tactics of the Trump movement often bore more resemblance to those used by K-pop acts than those used by political opponents.

Just like a contemporary, always-on artist, Trump fed his audience with an immense amount of content, access and engagement, and even did a live national tour. Like some of the most tribal of music fanbases, Trump nurtured a sense of otherness among his supporters – it is them against the world and the system is rigged against them. There is perhaps no better way of strengthening a tribe’s sense of oneness than strengthening its sense of otherness.

Tribalism and the role of us-vs-them

Fandom trades on basic human instincts and psychology, particularly two of the ‘deficiency needs’ in Maslow’s hierarchy of human needs: belonging and esteem (political fandom also trades on a third need: safety). Any kind of fandom depends on people feeling like they are a part of something, and how that something plays a role in identifying who they are. Tribal fandom takes this one step further, distilling this shared identity to an ‘us-vs-them’ mentality. Fans become focused on identifying, not just what makes them, well, them, but also what makes the rest of the world…not them. At its worst, that can be used by politicians, like Trump, to stoke fear against immigrants, liberals, people of colour, or basically anyone that is not part of their tribe. In less severe forms, it can manifest as artist fanbases mobilising en masse on social platforms against something they do not like. The sense of oneness, defined against the otherness of the rest, enables them to feel like the establishment is against them, even when they become the establishment, whether that be becoming the political party in power or the band that sets YouTube streaming records.

This flavour of tribal fandom has been made possible by social media and the broader way in which it facilitates online conversation. Often, social media facilitates a hyper-defensive style of discourse, with the loudest voices winning out, even if they are not the majority. Indeed, the fragmentation of fandom means that the whole concept of ‘majorities’ are becoming a thing of the past. In a political and business environment, now defined by the claims of the Facebook whistle blower, Frances Haugen, there is a growing understanding that a recalibration of social media is required, and ideally before Facebook Meta simply migrates social into the metaverse, warts and all. 

Fandom’s industrial revolution

But there is also vast positive opportunity within tribal fandom. As Chartmetric identifies, what sets K-pop artists from big Western artists is that they convert the vast majority of their audience into fans. There is little wastage. By contrast, big Western artists often have bigger total audiences, but a much smaller share that are ‘fans’, e.g., artist follower and active listener counts. This is simply taking Kevin Kelly’s 1,000 true fans conceptand recreating it on an industrial scale. In fact, you could argue that we are entering fandom’s industrial revolution phase. As this epoch plays out, the most effective marketing and monetisation across all forms of entertainment will be that which focuses on tightly defined tribes, rather than mass market reach. The shift from cultural moments to cultural movements is only just beginning. Politics (e.g., Trump and the Brexit campaign) have learned the tricks well, as has a select group of entertainment companies (e.g., Netflix and Big Hit Music). The coming years will be shaped by everyone else playing catch up.