MBW Reacts is a collection of remark items from the Music Enterprise Worldwide staff. They’re our analytical (and typically opinionated) reactions to main latest leisure information tales.
Final Wednesday (April 5), the wind modified.
William Packer, an influential media and tech analyst for BNP Paribas Exane, issued a analysis notice with a shocking conclusion: Packer had determined to double-downgrade Common Music Group‘s inventory, from outperform to underperform.
This in itself was uncommon: Based on TipRanks, of the 15 Wall Road analysts who’ve given views on the worth of UMG’s inventory prior to now yr, simply two are outright detrimental: Packer, and Joseph Thomas at HSBC.
All however one of many relaxation – a ‘Maintain’ score from CitiGroup – proceed to price UMG’s inventory as a ‘Purchase’, albeit with totally different ranges of optimism.
The ranks of this ‘Purchase’ group embody the likes of Andrew Uerkwitz (Jefferies), Daniel Kerven (JP Morgan), Julien Roch (Barclays), Lisa Yang (Goldman Sachs), Matthew Walker (Credit score Suisse), Michael Morris (Guggenheim), Omar Sheikh (Morgan Stanley), and Richard Eary (UBS).
Packer’s downgrade was additional shocking for the truth that, over the previous few years, he’s been an unwavering champion of the potential of UMG’s inventory worth.
So what was it that spooked Packer into demoting his view of Common’s industrial future?
Just a few issues contributed to his determination, together with doubts over the dimensions of TikTok’s funds to the music enterprise, and his assumption that essentially the most significant income enhance to UMG from streaming value rises might already be behind it.
There was, although, one issue that stood out above some other in Packer’s evaluation: The potential affect of synthetic intelligence (AI) on the music business within the years forward. And, specifically, on the main file firms.
Packer wrote that “AI is a brand new disruptive menace” the place “a glass half full business narrative is holding… for now”.
He added: “Whereas UMG is prone to stay a long-term winner from digital[z]ation, we expect AI music is to emerge as a difficult thematic and drag on elevated multiples.”
Packer particularly argued that huge volumes of AI-created music might “speed up [market] share loss” for UMG and the opposite main file firms.
As such, he touched on a recurrent conversational theme in at the moment’s enterprise: What occurs to main file firm streaming market share when the dam breaks, and we begin to see tens of millions of songs, all made by AI, distributed to Spotify et al every day?
Tens of millions of tracks, however is anybody listening?
Plenty of AI-created, or at the least AI-assisted, music is already washing up on the shores of at the moment’s digital music platforms.
The essential query, although, particularly when considering AI’s potential erosion of main file firm market share, is how many individuals are literally listening.
The stat that’s usually bandied round at this level – and I shall do the identical – is that which exhibits streaming quantity market share decline for the main file firms (in addition to Merlin members, aka most giant impartial labels) on Spotify over the previous 5 years.
In 2022, based on Spotify’s annual report, music distributed by the main file firms and their subsidiaries, plus Merlin members and their subsidiaries, misplaced 2% annual market share of complete quantity of music performs on Spotify (see beneath).
And throughout the 5 years from 2017-2022, these entities misplaced a complete of 12% market share on the service, down from 87% to 75%.
But right here’s one thing that’s hardly ever remarked upon: The above represents astonishingly sturdy market share upkeep from the majors (and Merlin) within the face of an awesome glut of music being launched.
As MBW reported final month, there are – based on market monitor Luminate – round 98,500 separate ISRCs (aka separate tracks or movies) launched to music streaming providers each day at the moment, a combination, little doubt, of human-created and AI-created music.
The amount of music created prior to now few years is unfathomably giant: almost 50% (!!!) of the ≈100 million tracks obtainable on Spotify et al at the moment had been created and uploaded prior to now three years alone.
Conclusion: The main file firms (and their ‘indie’ distribution subsidiaries) are releasing a weeny fraction of the ≈35 million new tracks/movies now hitting music streaming providers each 12 months.
How is it, then, that these main file firms (together with Merlin-affiliated indies) solely misplaced 2% of general quantity market share on Spotify in 2022?
In a phrase, curation.
Why the main file firms could also be well-positioned for the approaching AI onslaught
The statistics show that the overwhelming majority of music being launched at the moment – whether or not by robots or people – isn’t solely barely being listened to; in loads of circumstances, it’s not being listened to in any respect.
MBW pointed to this reality final month, once we cited Luminate information displaying that some 38 million tracks on music streaming providers didn’t appeal to a single play in the entire of 2022.
This story is even higher illustrated by Spotify’s Loud & Clear, the information web site revealed by Daniel Ek‘s firm yearly.
Based on the most recent replace of Loud & Clear, Spotify estimates that there are round 200,000 “skilled or professionally aspiring artists” working on its service at the moment.
Spotify characterizes a “skilled or professionally aspiring artist” as somebody who has (a) launched greater than 10 tracks on its platform to this point (5.6 million artists haven’t, says Spotify), and (b) who averages greater than 10,000 month-to-month listeners.
These two qualifiers mixed, says Spotify, apply to only 213,000 artists in complete.
Spotify brings down that 213,000 determine in its “skilled or professionally aspiring” calculation to 200,000 by mixing in information relating to the variety of artists who hosted a gig or digital reside occasion in 2022, a determine it says shakes out at 189,000.
These are the precise form of artists, in fact, that main (and indie) file firms are desirous about signing/investing in – one thing we’ll come again to.
Anyway, the necessary bit: The 200,000 “skilled or professionally aspiring artists” on Spotify at the moment characterize simply 2.2% of the 9 million music artists whose music the agency says it hosts on its platform.
And but, based on Spotify, the identical 200,000 “skilled or professionally aspiring artists”, in 2022, generated a whopping 95% of all royalties on Spotify.
It’s a surprising stat, but in addition one which speaks to one of many flaws of over-analyzing the Spotify market share chart on the prime of this story.
That chart, keep in mind, is for complete performs on Spotify; it doesn’t differentiate between the high-royalty performs going down on Premium Spotify accounts, and the low-royalty performs going down on ‘free’ Spotify accounts.
The truth that 95% of all royalties on Spotify are being generated by simply 200,000 “skilled or professionally aspiring artists” proffers an necessary reality: ‘Skilled’ artists, with correct fanbases, are being listened to by extra precious (i.e. paying) listeners on streaming providers.
As such, the main file firms might be much more involved about sustaining market share amongst these 200,000 acts (and others who’ll be part of their league in years to return), than they are going to the opposite 8.88 million artists / ‘artists’ on Spotify at the moment who, I remind you, solely generate 5% of all royalties.
(As exemplified within the chart above, the final official annual royalty payout determine we now have from Spotify, throughout information and publishing, is for 2021, when the service distributed ‘over EUR €7 billion’ to music rightsholders. Rounded down, these 200,000 ‘skilled or professionally aspiring’ artists would have generated €6.65 billion of a €7 billion royalty payout in 2022.)
What does all that should do with AI?
These stats recommend that the actual menace to main file firms from AI music isn’t, as is commonly proposed, certainly one of sheer quantity.
The fashionable music business, as evidenced by the above, already has a big imbalance between provide and demand. An explosion in AI music would definitely super-size the provision. However in music listening, as was ever the case, it’s the demand that actually issues.
That’s to not say AI-created music can’t or received’t have a critical industrial affect on the music enterprise within the subsequent couple of years. However its largest affect, particularly market share-wise, will in all probability be felt the place there’s a pure listener demand for it – and, for my part, that’s not in pop music. It’s in music for utility (or as Sir Lucian Grainge would have it, “purposeful” music.)
One solely want look to Berlin-headquartered Endel for proof. The corporate, with backing from the likes of Amazon Alexa and Avex Group, positions itself as a “sound wellness firm”, whose proprietary AI “create[s] customized, purposeful soundscapes to assist individuals focus, chill out, and sleep”.
Outdoors of Endel’s intelligent tech (with soundscapes that change primarily based on a consumer’s motion, coronary heart price, location and different components), Endel’s doing one thing else uncommon for an AI music startup: It’s really measuring its success on the variety of listeners it’s reaching.
Endel says its “ecosystem” of soundscapes already has over a million monthly users, who hear, cumulatively, to one million and a half hours a month of Endel’s sounds.
Stats like this will not be, to my thoughts, an endemic menace to Common Music Group, whose strategic bedrock continues to be music made by artists with identities which have significant real-world recognition. (Aka: Stars.)
I’d be much more involved by Endel, and certainly the ability of AI-generated “music for X” / “purposeful music” soundscapes, if I used to be, for instance, Epidemic Sound – the $1.4 billion-valued Swedish firm that’s amassed a really profitable business place by creating instrumental music for YouTubers and podcasts, and in addition by creating instrumental music that’s stuffed Spotify’s first-party ‘music for rest’ playlists.
(Sidenote: There might be some irony ought to music created by “pretend artists” – to echo MBW’s now-infamous flip of phrase – in the end be outmoded on mentioned Spotify playlists by full-blown robots.)
The true menace?
So, in conclusion, the main file firms are seemingly doing higher when it comes to Spotify worth market share – quite than Spotify quantity market share – than most sometimes perceive.
That is taking place as a result of the majors (and their strongest indie friends) are persevering with to give attention to a few hundred thousand “skilled or professionally aspiring” artists. These artists, in contrast to robots, (i) construct fanbases by taking part in reside, and (ii) are in the end eager to be acknowledged and appreciated over and above the music they add to faceless streaming providers.
(An ideal analogy I heard the opposite week: Report firms try to signal artists who cross the ‘pencil and paper’ take a look at, i.e. if I gave you a pencil and paper, might you draw an honest approximation of the artist merely out of your thoughts’s eye?)
The huge bulk of demand – monetized demand, at the least – on the planet of music listening continues to reside with these “skilled or professionally aspiring” artists.
So the concept a deluge of AI-generated songs (until mentioned songs are carried out by “skilled” artists!) will inevitably wash away main file firm market energy on streaming sounds to me just a little hyperbolic.
“Lucy is [our] first super-realistic digital idol.”
Tencent Music, talking in December 2022
The larger existential problem for the majors from AI music isn’t certainly one of impersonal quantity; for my part, it’s certainly one of human empathy.
We’re already seeing makes an attempt from giant tech firms, and certainly music firms, to market AI-created characters that resonate, emotionally, with human patrons. Avatars with sufficient mythology, sufficient Homo Sapien-esque traits – or a plausible facsimile of them – to elicit true fandom from a large-scale viewers.
I’m speaking “skilled or professionally aspiring” artists… who by no means perspire a single bead of sweat.
That is the place issues might ultimately, theoretically get scary for the majors.
Witness ‘Lucy’, a precedence mission for China’a Tencent Music Leisure (TME), which/who was first premiered in December.
TME’s AI capabilities in music are already underneath little doubt: The corporate boasted final yr that it had usual 1,000 tracks with computer-made vocals that replicated human voices, together with the voices of useless Chinese language superstars. Certainly one of these tracks had drawn over 100 million performs, mentioned TME.
Now, with Lucy, the agency has put an AI-generated face on this sort of machine-generated product – an avatar that it calls “[our] first super-realistic digital idol”, constructed by TME’s in-house Lyra Lab’s LyraSinger Engine.
On an earnings name in March, TME Govt Chairman, Cussion Pang, mentioned that Lucy’s staff had already “created three chart-dominating unique songs throughout totally different kinds inside only one month of [her] debut” and enticed industrial partnership curiosity from “a broad array of world manufacturers” together with Coca-Cola, and KFC.
We’ve seen this sort of factor earlier than, in fact.
In 2021 – full with an MBW headline presaging the approaching AI revolution in music – Warner Music China unveiled its personal first-ever “digital idol”, Ha Jiang.
Mentioned WMG’s now-co-President of Asia, Jon Serbin, on the time: “‘Digital idols’ are already an enormous phenomenon in China, in addition to different components of Asia.
“They’re attracting massive following on social media, significantly Gen-Z followers. Individuals turn out to be actually engaged with the idols’ each day lives, very like they’re with actual movie stars or fashions.”
“I imply, certain, there’ll be a robotic popstar sooner or later. After all there’ll – every little thing that may occur will occur. However that’s not going to be the way in which the [majority of the] business goes. It’ll be a one-off.”
Ed Newton-Rex, creator, Jukedeck
May these “digital idols”, replete with AI-composed music and human-mimicking vocals, actually someday problem the specialty of the main file firms: Globally marketed, human superstars with tens of millions of ticket-and-merch-buying followers worldwide?
It’s right here I flip to the phrases of a person in a greater place than me to reply this query: Ed Newton-Rex, the creator of one of many first AI music-making platforms, Jukedeck, which he bought to Tiktok/ByteDance in 2019.
Newton-Rex just lately advised me on an MBW podcast: “What individuals actually care about is a connection to the artist.
“I imply, certain, there’ll be a robotic popstar sooner or later. After all there’ll – every little thing that may occur will occur. However that’s not going to be the way in which the [majority of the] business goes. It’ll be a one-off.
“That reference to human artists has acquired to be at the least half of the rationale that we love music. It’s not nearly what we hear.”
That, in itself, will little doubt be music to the ears of Sir Lucian Grainge – recent from his signing of a brand new 5-year deal to steer Common by way of what might turn out to be essentially the most disruptive technological interval in main file firm historical past.
Grainge clearly isn’t taking the menace from AI-powered ‘purposeful’ music calmly: At the least a portion of his name for brand spanking new “artist-centric” royalty fashions at music streaming providers looks like agitating for an business bulwark in opposition to the prospect of tens of millions of high-quality instrumentals being spat out by machines at a price of knots.
In that case, Grainge isn’t alone in pondering the best way to trammel unhealthy parts of the development of AI in leisure: Witness the expansion of the AI-skeptical public coverage physique, the Human Artistry Marketing campaign (HAC), whose members span music, sports activities, movie, journalism and different disciplines of IP, and which simply elevated its membership base by 50%.
(Curiously, Common Music Group is a significant impetus behind HAC; the US trademark for the org was filed by UMG itself.)
Finally, although, as set out on this article – and within the face of doom-laden narratives round AI’s affect on music’s largest firms – there does appear to be just a few strong causes for that “glass half full business narrative” to carry just a little longer, at the least.Music Enterprise Worldwide