Five music-tech forecasts that misfired in 2019

The fact that the 2020s is fast approaching has thrown the entire media industry into collective introspection.

Music publications are no exception, as nearly all of them at this point have openly reflected on the biggest musical moments and trends not just from the past year, but also from the past decade. Music Ally’s 42 Trends of 2019, The Atlantic’s 14 Best Music Moments of 2019, Billboard’s Top 100 Moments of the EDM Decade, NME’s 10 Artists Who Defined The Decade, Rolling Stone’s 50 Most Important Music Moments of the Decade … the list goes on and on.

For my last essay of the year, I want to do something slightly different: I want to focus on what didn’t happen in the music industry in 2019.

Of course, this year brought forth several monumental changes to how we understand music economics and technology as a whole — e.g. Spotify throwing money at podcasts, TikTok cultivating “Old Town Road” and the Music Modernization Act getting signed into law.

But I also think 2019 tempered many of our positive, idealistic forecasts about digital innovation in music with a healthy dose of realism.

We learned, for instance, that distribution is actually a difficult business to succeed in; that blockchain hype is nothing without applications that are usable, let alone practical; and that a variety of other emerging technologies, including voice and virtual/augmented/mixed reality, have a long way to go before they really alter the ingrained, traditional ways in which we consume and monetize music today.

With that in mind, below I lay out five, tech-forward music-industry predictions that ended up flopping in 2019, and a few reasons why I think each prediction didn’t come to full fruition. (Side note: I last wrote an article in this format for Forbes in 2016, and am excited to bring it back this year.)

tl;dr —

1. Voice and smart speakers didn’t disrupt the music industry.
2. Fan clubs didn’t come back, even though a lot of people still want them to.
3. VR/AR/MR for music didn’t take off, again — although a new vocabulary is emerging around “virtual entertainment.”
4. Tidal and SoundCloud didn’t get acquired — but their core brand cachet continues to erode swiftly, amidst ongoing consolidation.
5. Musical A.I. didn’t top the Billboard charts and steal artists’ jobs.

1. Voice and smart speakers didn’t disrupt the music industry

Seemingly everyone in music was getting so excited about voice last year.

Not only are smart-speaker sales projected to see double- and even triple-digit growth year-over-year in several markets around the world by 2020, but music streaming is also consistently a top use case for voice-enabled devices.

This has led many music-industry insiders and commentators — myself included — to suggest that voice experiences could transform not just how we consume and discover music, but also how we package and market it. For instance, Eamonn Forde predicted that “titles of songs and names of acts will have to be recalibrated to ensure the right key words [sic] put them at the top of the [search] results”; music marketing expert Kara Mukerjee separately referred to this phenomenon as “voice SEO.”

Yet, almost no one I met in the music industry this year prioritized voice in their marketing campaigns — let alone in a way that truly took advantage of the transformative nature of voice-driven discovery and interaction.

Interestingly, those who did talk more about voice were consistently in genres outside of mainstream pop and hip-hop, such as classical and country, which both tout relatively older listener demographics. And I think it says wonders that one of the most-shared voice-related music moments of the year ended up being the viral video of a toddler requesting “Baby Shark” — a heartwarming family story, not quite a mainstream music-industry story.

Amazon did invest in several music-related marketing campaigns for Alexa this year, such as the campaign “A Voice is All You Need,” which featured the likes of Taylor Swift, Katy Perry, The Beatles and Chance the Rapper. But these campaigns were as much for the Amazon Music service as they were for Echo devices, and ultimately promoted a more traditional, lean-back, hits-driven listening experience that didn’t really move the needle beyond what we were all familiar with already.

Turns out, even Amazon itself is struggling to figure out how to increase engagement and monetization with Alexa beyond simple requests for music, weather and news. According to a recent report from The Information, even though Amazon’s internal Alexa unit has over 10,000 staff members, the software brought in only low six-figure revenues in 2018, far below the company’s internal goal of $5 million.

The company has tried and failed to compensate for its financial under-performance through advertising — testing audio ads tailored for smart speakers with a select number of brand partners, and even discussing a potential rev-share deal with Spotify in which Amazon would get a cut of ads on Spotify’s free tier that were streamed through Alexa-enabled devices. But those talks either fell through or led to lackluster results.

The core problem is that even though the smart-speaker hardware is everywhere, the underlying software — encompassing conversational A.I., context awareness and perhaps voice synthesis — still has a long way to go before people find it user-friendly and compelling enough to explore deeper use cases.

In the meantime, an interesting trend I see happening in the music industry is that artists and labels are merging voice strategy with podcast strategy.

For instance, the “microcasts” that K.Flay is releasing through Interscope Records fundamentally constitute a podcast series, but are also clearly targeting voice, as episodes are released exclusively on Alexa and Google Home first before a wider release on YouTube a week later.

The reasoning here kind of makes sense: if podcasts and voice interactions are both built on audio, why not use one piece of audio content to tackle both areas in one fell swoop? That said, I think conflating voice strategy with podcast strategy can run the risk of not actually taking advantage of the disruptive, dynamic possibilities of smart speakers and voice interfaces.

2. Fan clubs didn’t come back, even though a lot of people still want them to

One consistent, and somewhat surprising, complaint I’ve heard from artist managers over the past few years is that it remains difficult to assemble a clear, consolidated picture of who fans actually are across platforms.

Indeed, a crucial feature of the shift towards an all-you-can-eat streaming model isn’t just that subscribers don’t own their music; it’s also that artists don’t own their audiences. At worst, this can end up mirroring the world of pay-for-play social media, with particularly frustrating implications for the very community that many of these streaming services claim to serve on the backend — namely independent and emerging artists.

In light of this information asymmetry, some people in the music industry predicted last year that “fan clubs” would make a comeback, as more artists sought direct control over their audience data and relationships. A handful of high-profile bands launched their own fan-club equivalents — such as The National’s Cherry Tree and the Jonas Brothers’ Jonas Vinyl Club, which are both structured as monthly memberships with perks spanning exclusive merch, records and digital content.

Yet, I don’t think these launches really did anything to drive more conversations around how artist-direct payment and fan-engagement models could actually change the economics of being an artist as a whole, whether a megastar or an emerging act. A handful of celebrities, like Taylor Swift, Jeremy Renner and the Kardashian sisters, even shut down their own fan clubs and social networks altogether.

In fact, I found that global interest in the term “fan club” has dramatically declined over the past 15 years (see graph below). Perhaps the concept of a closed “fan club” is more difficult to implement in the modern era — in which streaming and social media allow for unprecedented scale, and fans around the world increasingly leverage that scale in the form of decentralized “armies,” rather than insular memberships.

That said, the tech and venture-capital worlds are betting big on smaller social networks, and on platforms that allow everyday people to build sustainable businesses around more niche communities (e.g. Substack, Patreon and Glow).

It’s inconclusive whether the music industry is really on board with this wider social and financial shift. On one hand, I heard virtually no one mention “chatbots” or Messenger bots at music-industry conferences this year. On the other hand, hundreds of celebrities are now posting “their” phone numbers on social media and asking fans to text them directly — i.e. signing up for the artists’ SMS blasts on Community.

To me, this signals that there’s still high demand in the music industry for better tools for customer relationship management (CRM) — or, more appropriately, fan relationship management (FRM).

Whether this takes the form of superserving only superfans, or of taking a more holistic approach by tapping into multiple APIs across mass-market streaming services, remains to be seen. And one uphill battle for founders in this sector is that ushering niche social networks into the music industry will also require a change in stubborn behavior.

For instance, I’ve personally signed up to a bunch of artists’ Community SMS feeds, but I can count the number of them that are regularly active on one hand. The remaining majority still turn to mass-market platforms like Instagram or Snapchat to deliver updates like tour dates, album releases or press placements that could be more targeted based on fans’ interests and locations. A slick product won’t make an impact if human behavior doesn’t also shift as a result.

3. VR/AR/MR for music didn’t take off, again — although a new vocabulary is emerging around “virtual entertainment”

At the end of 2018, SuperData Research predicted that 2019 would be the year that VR broke into the mainstream for good — in part because of the launch of the wireless Oculus Quest headset, which has a starting price of $399.

Needless to say, that hasn’t happened. Some of the sector’s biggest companies continue to struggle with sales. For example, Magic Leap — a shamelessly futuristic mixed-reality company that’s currently raising a Series E round, after previously raising $2.6 billion from investors — reportedly sold only around 6,000 devices in its first six months on the market, falling short of its founder/CEO’s internal goal by 99%.

With a few exceptions, musical experiences in virtual reality have suffered from what I once dubbed the “science fiction payphone problem,” in which we falsely assume the continuity of our previous experiences when trying to build the future.

What is growing rapidly, however, is a new vocabulary around virtual experiences in general, which can proliferate far beyond the confines of a headset strapped to our faces.

Around a year ago, I heard a speaker on a music-industry panel say that music will not be the leader of cutting-edge virtual-reality applications, but rather will co-opt the best experiences from other industries like gaming, healthcare, construction, retail and sports that are arguably further ahead with the technology.

We finally saw that relationship unfold this year, specifically with gaming. I found it fascinating that Marshmello’s concerts in Fortnite made the concept of “virtual” worlds really click for many people in the music industry, independent of any VR hardware. Riot Games also incorporated augmented and mixed reality into its live music performances at League of Legends world championships, which featured bespoke K-pop and hip-hop groups formed by Riot Games with the specific purpose of promoting new LoL features.

Unlike what some people predicted, audio-first streaming services like Spotify and Apple Music haven’t actually expanded much into virtual experiences around gaming. Instead, these services are still relying more on lean-back integrations with gaming consoles like Xbox or Playstation, prioritizing their software footprint over one-off, ephemeral experiences.

The rise of “virtual influencers” like Lil Miquela (who also has over 350,000 monthly listeners on Spotify) has also helped audiences warm up more to the concept of virtual worlds and economies that can thrive across multiple media formats, instead of being confined to just one piece of hardware.

Last month, United Talent Agency recently hosted a Virtual Beings Summit in its Los Angeles theater, featuring speakers from virtual-entertainment startups like Toonstar as well as from big-tech companies like Google, Facebook, Verizon and Netflix. I hope music-industry decision-makers will get to take part in similar discussions more often in the future.

4. Tidal and SoundCloud didn’t get acquired — but their core brand cachet continues to erode swiftly, amidst ongoing consolidation

I heard several murmurs online and offline over the last year or two about Tidal and SoundCloud, now among the underdogs of music streaming, potentially getting sold to a bigger buyer.

Turns out, both are still swimming as standalone companies — but are also pivoting away from the only advantages they really have going for themselves in an increasingly crowded streaming market.

Tidal launched and built its brand not around cutting-edge technology, but around exclusive content from some of hip-hop’s biggest stars. Yet, 2019 marked the year that the streaming platform’s most coveted sources of content exclusivity abandoned their deals in favor of omnipresence. Beyonce’s Lemonade became available on multiple streaming services in April 2019, followed by Tidal owner Jay Z’s entire discography on the rapper’s 50th birthday a few weeks ago.

In an interview for Billboard’s latest Women in Music issue, Roc Nation COO Desiree Perez, who also oversees Tidal, claimed that the streaming service has turned down multiple M&A offers because the prospective buyers didn’t align with “the purity of [Tidal’s] naiveté,” in terms of the company’s focus on artists and content.

“On Tidal, we could have sold five different times. We could have merged five other different times. We could have taken a check very early … But the only way we would do a deal for Tidal would be if the vision is about the art and the artist,” she said. “If not, we couldn’t do it, and that has been the reason, actually, why we’ve turned down a lot of deals, because the different companies we’ve talked to just were not thinking the way we’re thinking.”

An open question: now that Tidal no longer has a stronghold over its most valuable content, will the platform be forced to start thinking differently?

Now for SoundCloud. The company laid off over 40% of its staff two years ago, after failing to build a compelling, and fully-licensed, consumer-facing subscription product. In the past, there have also been numerous rumored acquisition talks the likes of Spotify and Google.

2019 solidified a significant pivot for SoundCloud away from consumer subscriptions towards monetizing artist services instead. The UGC-powered streaming platform acquired distribution and label-services startup Repost Network in April 2019, although some sources tell me it’s still not clear how exactly the two companies are integrating together on the inside.

As I discussed earlier in this piece, there is a demand for better artist analytics tools and fan-engagement services, so SoundCloud isn’t just shooting in the dark. But the company has faced so much criticism and financial downside in the past, that I think this coming year is really their last chance to get things right before they completely lose public trust.

And arguably SoundCloud’s biggest selling point historically — fostering real, organic communities among artists and fans — continues to erode with each passing day. Rarely do I hear artists nowadays even mention SoundCloud in their music rollout plans, nor do I hear fans turn first to SoundCloud for music discovery.

In general, this year has seen ongoing consolidation of low-margin businesses in recorded music, especially in streaming and distribution, that’s far from over. SiriusXM now owns Pandora, while a growing number of distributors are shifting towards serving the “middle tier” of independent artists as they seek to improve their bottom lines.

Tech giants with money to burn are flexing their muscles on streaming bundles that “pure-play” services can’t afford to compete with on their own. Consider, for instance, how Amazon Music launched a hi-fi tier AND a free tier this year, or how Apple is reportedly considering a mega-bundle that would combine Apple Music, Apple News+ and Apple TV+ services into one discounted offering. In this landscape, underdogs like Tidal and SoundCloud are particularly vulnerable.

5. Musical A.I. didn’t top the Billboard charts or steal artists’ jobs

I’ve witnessed and partaken in several speculative conversations claiming either that we will soon have a hit song with its beat composed entirely by A.I., or that major labels and/or other tech companies will lean more into creative A.I. to automate tasks that would otherwise be assigned to human musicians, pushing the latter out of a sustainable income.

Sensing a high-growth opportunity, a ton of startups and music companies announced new products and deals related to musical A.I. in 2019. Warner Music signed a distribution deal with adaptive music startup Endel. ByteDance acqui-hired the automated music-composition startup Jukedeck, whose CEO now runs an A.I. lab at the parent company.  Boomy launched out of beta and has been used to create nearly 300,000 tracks to date. Popgun launched two new products (Gloss for mastering and Splash for consumer-facing experimentation). Even Amazon launched an A.I.-driven keyboard called DeepComposer (about which I have yet to hear a positive review).

But — surprise, surprise — A.I. didn’t end up topping the charts and taking over the music industry’s aesthetics or livelihoods.

There are a few reasons why I think this is the case. Firstly, like the voice-adjacent technology we previously discussed, creative A.I. is still an emergent, scrappy field with relatively low-fidelity results.

In this vein, a second obstacle is that there are so many different approaches to musical A.I. today that there’s little consensus on what “success” really means — i.e. on what it means for creative A.I. to “grow” beyond its current, “nascent” stage. Some are pursuing pure, indistinguishable imitation of works that already exist, while others are simply pursuing a sense of meaning for their users.

Thirdly, as of today, creative A.I. still requires a lot of manual labor and input from humans. To understand this dynamic, we can look at what I think are some of the most compelling experiments around music and A.I. from this year, which are coming from independent artists and projects on the fringes of the mainstream like Portrait XO, Holly Herndon and SKYGGE.

It’s not like these artists are just producing entire albums immediately with the press of a button. If you talk to any musician incorporating A.I. into their creative process, they’ll likely tell you that it still involves several days of training an algorithm, several more days of sifting through the algorithm’s output to figure out which sounds are actually interesting, several weeks of patchworking those pieces together in a way that sounds coherent for the artist and their fans — and then the more traditional process of working with mixing and mastering engineers.

In my opinion, before A.I. even has a chance of breaking into mainstream sounds, it first has to permeate mainstream discourse. Grimes is the best-known artist in my recent memory to discuss the impact of A.I. on music creation publically, with a rather dystopian tone.

But in order for “A.I. beats” to enter the upper echelons of the charts, I think a Billboard top-10 artist — so, as of today, someone like Lizzo, Post Malone or Mariah Carey — would have to talk publicly not just about creative A.I. in general, but also about how they would apply the technology specifically to crafting their own future hits. Then producers and aspiring hitmakers will interpret that discourse as a signal of future opportunity and demand, and will then incorporate it concretely into their recording sessions, perhaps even giving software formal songwriting credit.

To borrow a phrase from the world of leadership and sociology, we can’t be what we can’t see.