Revisiting “Post-Royalties”; Open Editions, and New Digital Rights
Market developments imply that we are accelerating towards the need for new Creator business models
We recently moved back out to the mountain cabin in Norway. My happy writing place. There’s a lot to talk about…
I recently did a couple of podcasts, with Yash and Maartin on Appetite for Distraction and with Joe on Music Ally Focus, which sit quite well alongside this article.
18 months have passed since I first put out “The Case For a Post Royalties Music Industry” through Water & Music. A lot has happened since and I think it’s worth revisiting the arguments I made in the context of a number of key evolving narratives:
OpenAI inspired hype is sweeping pretty much every sector, with music very much in focus for language model development
Major labels are starting to get very very worried about the preservation of their marketshare within streaming services
The wider financial outlook for tech companies, especially platform businesses, is starting to bite DSPs
Music Language Models
As every tourist VC quietly switches its focus from web3 to investing in language model / AI startups, we are setting up for a very noisy year of hyperbole.
In some ways, it’s not without warrent. Anyone that has used the notionAI tooling to summarise an interview or article that they’ve written, can see the potential time-saving power. Microsoft is rolling out AI features across its product suite. And OpenAI itself recently announced hitting one hundred million users within two months of go-live (TikTok, not a perfect comparison, took nine months, Instagram two and a half years).
Things, equally, are moving quickly in the creative industries, as described by Water & Music as part of their research stream on the subject. Google made headlines recently with “musicLM”, but as detailed by W&M, this is only one of ten similar models announced since the start of the year already.
I am no expert in AI or language models enough to comment on the significance of these developments in isolation, but, viewing this through the higher-level lens of strategic positioning, and medium-term potential implications for incumbents, provides some indication as to where we are heading.
Streaming Market Share & Economics
There is undoubtedly going to be an unprecedented influx of music once the music language models start churning out tunes.
My “post-royalties” article centered on the impact that lower barriers to entry have had on the market share driven, zero-sum, calculations for money distribution from streaming incomes. And surely these latest developments are going to have orders of magnitude further effect.
One of the concerns of focus, when I was working inside a rightsholder, was the growth of library/production music, which is created on commission and often through aliases, eating into market share. The industry has a particularly gross description for this kind of repertoire, “fake music”, and it’s been a hotly debated subject for years. The fact that has been seen as an attack vector belies the weakness of the streaming model already and sets up a very uncomfortable evolving situation.
Sir Lucian Grainge, CEO of UMG, the Godfather of music, and ever-present winner of the “Billboard Power 100”, certainly sees this, and made the case for a new streaming economy as part of his annual update to his company. Days later, Universal and Tidal announced that they were developing a “more artist and fan-friendly streaming model”.
This is obviously significant as Universal has gained more than any other party from the streaming age. Through deft implementation of the major rightsholder playbook, countering platform aggregation by holding control of enough of the supply, Universal has seen years of plenty in an age where music itself transitioned from a product to a service.
And so, for Universal to publicly announce that it is facing an innovators dilemma, and seek to use its weight (while it has it) to bend a re-formation of streaming economics to something that it finds agreeable, in the face of exploding market share that it cannot control, is a major event that everyone should take note of.
In reality, the Tidal partnership is likely posturing ahead of relicensing negotiations with DSPs across the board. But there is a chance, probably more than we’ve ever seen, of a future where DSPs move more towards exclusive repertoire rather than the current expectation of everything being everywhere.
DSP business model
The Covid years, in hindsight, were somewhat of a make hay while the sun shines period for tech companies. A time to aggressively hire and set ambitious targets based on work-from-home inflated projections.
Spofity partly spent this time doubling down on its strategy to diversify the content that its platform relies on, particularly through podcasts, as well as deeper monetising of creators through things like Marquee. Anything to try to decouple its linear relationship between revenue and cost of goods and services.
However, now that the dust is settling, the vibecession is setting in, and most tech companies are seeking to reverse some of the fever-dream excesses (mostly through lay-offs), we can look at the state of Spotify’s business model, and by extension, the wider streaming market (with the caveat that almost all of Spotify’s competitors are part of larger groups that can run their music streaming arm as a loss leader).
Spotify’s Q4 2022 results weren’t great. Subscriptions are up, monthly active users are up, but operating profit came in at -€230m. Q3 2022 saw an operating profit of -€228m, and Q1 2023 is forecasted at -€194m. Despite continued growth, Spotify is currently unprofitable to the tune of c.€1bn a year at its current rate.
But really, it’s the steady gross margin line of 25% that tells the true story. Despite its recent positioning, and efforts to move into “audio” rather than only music, Spotify isn’t scaling into profitability. And probably won’t unless there is a fundamental change.
We are approaching a very interesting flash point, where generated content could genuinely positively shift the economic model for DSPs by reducing their cost of goods and services…royalties… But this would be to the detriment of “non-fake music” assuming that there is no step change in the top line of the industry.
I see the dynamic of an increasing volume of generated music entering the streaming marketplace as an accelerant towards the end state that I described in depth in the Post-Royalties essay.
There will probably be some tweaks to the royalty equation once Universal has executed its new DSP licensing strategy. For example, part of Lucian Grainge’s proposal hints towards a structure that can reward and monetise an artist’s more engaged fan base. However, for the vast majority of artists, the deck will remain increasingly stacked against you with regard to earning sustainable money from streaming.
And this is fundamentally what the Post-Royalties thesis rests on - if the majority of creators can not build a sustainable career through the payout from streaming royalties, then, what are the emerging business models that will provide a better option?
It’s worth saying, as an aside, that the historical response to this question has often been, what about live? I am no expert in live music, but when speaking to those that are, the overriding consensus is that live music is if anything in an even worse spot than streaming for building a sustainable career.
I expect to see a continued resurgence of Kellyest 1000 true fans, creator economy, narrative as we build forward. For example, my friend Vaughn recently reminded me of an article that Chris Dixon wrote on NFTs and a 1000 True Fans, which is a great starting point for thinking about how crypto can enable better monetising under the demand curve.
More accurate monetisation, not leaving money on the table, is only part of the story though.
I’d argue that a major failure of the creator economy vs. the traditional business is that trades scalable effort to cash (copyright & royalties) for unscalable endeavours (subscription & continuous service).
Software can help with this of course, and actually, it’s a great use case for AI; using the algorithms to scale creators (see Water & Music’s experiment to “scale” their founder Cherie here). I expect there to be significant developments in this regard as the Post-Royalties economy evolves.
The other side of the coin to monetisation for a creator is always going to be reach and cutting through the noise. For this, I can’t see any reasonable short-medium-term strategy that doesn’t utilise DSPs and social platforms. Finding new fans is always going to involve some degree of getting in the mix in the big digital meat market.
One narrative that I do find encouraging though is the notion that the next iteration of the internet will have a more “local” community bias. Digital scenes, that are less centered on location as they were traditionally (by necessity), but on common values and interests. DAOs have provided some testing ground for this, but I feel like we have a long way to go.
As per my Productive Asset Model essay, I believe that the unlock will be where digital tribes can bring in an effective and sustainable self-governed economic layer. There have been some interesting developments, such as the Zora-driven Nouns Builder model (for example how The Park is using it to build out its own DAO).
In summary, I think that the most successful model moving forwards for a creator will be to generate capital and early supporters through the crypto markets, transition that into relatively oversized attention and exposure in the traditional streaming market, while, building their tribe and coordinating with other growing communities to build the next generation of music companies.
The death of Social Tokens and rise of Open Editions
Rally.io recently rugged its userbase, which for anyone watching had been a long time coming; the model had failed a long time ago and the company had become a zombie.
Rally sucked. And its failure has burned the reputation of crypto as an option for many artists, which is very unfortunate. The real question though is whether the concept of “Social Tokens” is still viable or not.
One of the challenges with Social Tokens, essentially an economy that a creator floats around their career, is that once deployed and initially monetised, there is little room for the creator to experiment moving forwards. They have duty to their token holders to direct value back into that token, which, obvious securities regulation aside, can provide extremely restrictive. Failure to do so risks rugging your most loyal fanbase.
In contrast, musicians work in releases, a chain of projects that are individual expressions and experiments. Social tokens aren’t a great fit, in a similar way to how the Creator Economy-inspired subscription-service models don’t work that well for many artists.
The final nail in the coffin for Social Tokens though is likely to be the rise of Open Editions; project-based tokens that have no cap and a specific buy-in price. Through an Open Edition, a creator can get many of the benefits that they would through Social Tokens, but without the future restrictions.
But, and this is where I beat the drum that a few of us have been sounding for a while now, in my opinion, Open Editions are only at their most powerful when they are not linked to copyright.
There’s been a debate for a while now around “music NFTs” and I am firmly in Team: “keep the music out of music NFTs”. NFTs offer the potential of freedom from the incumbent industry, but by linking in copyright you are instantly handing over that freedom and opening yourself up to future liability. You are turning something that could be disruptive into something that is sustaining.
I hope, and expect, that this mentality will prevail as we move forwards. Through clever use of Open Editions an artist can establish their community of early supporters, and then work together with them to create, distribute, and expand into the industry. Open Editions can work alongside the music as campaigns, supporting its development, promoting and leveraging the artist and their community’s brand. They can also provide the unlock for new digital rights that are better suited to the internet.
New Digital Rights
Vaughn and I first spoke about sound.xyz about 18 months ago, around the time of its launch. His first reaction was something that has stuck with me since, “this is cool, they are putting comment rights on-chain”.
The idea of taking Soundcloud-style comments and linking the right to make them to a token may seem a relatively low-value proposition. But the important thing here is that it is a new right, that is enforceable through code, and something that is not open to litigation through the courts. There is no copyright.
Sound’s model has subsequently become more focussed around curated music releases, which is understandable, but under utilises the potential of this genuine differentiator to the traditional industry. Hopefully, future development will look to expand more in this direction.
Comment rights are just one example of a basically endless design space of new digital rights that a creator can establish with their token holders. Rather than off-chain benefits (e.g. access to a live event) or access to off-chain royalty streams, which can’t be enforced through code and reduce the token itself to being a form of marketing.
Outside of music, a good example of New Digital Rights can be found through DAO governance tokens. Putting plutocracy concerns to one side for a second, and the inherent problems of voter apathy, governance tokens represent a New Digital Right that are valuable for certain entities, enforceable, and, not easily replicated off-chain. And they are starting to become extremely important, as in the recent Uniswap “bridge” decision, for example.
Token-gating is in some ways building around this idea of New Digital Rights, though I think there is a subtle but important distinction between the token-gating of access (e.g. chat/community), vs. the token-gating of an event/service (e.g. livestreams), vs. the token-gating of something that isn’t directly reliant on the efforts of the artist or their community. The latter to me feels the most valuable and scalable, though harder to envisage for the music industries.
This though is the area that I am most interested in terms of future development. I think that token-gating as it has been rolled out to date is certainly useful for building up initial communities, and provides some basis for monetising under the demand curve, but the real step change will come through an expansion of what comment rights hint towards, a digital right offered by a creator that provides value that is scalable beyond the linear efforts of them or their community.
Recent and evolving developments in the streaming economy provide further confidence to me that the Post-Royalties arguments are directionally correct. The speed at which this becomes a reality will in large part depend on the moves of the major incumbents over the coming 18 months, though in the end, it feels inevitable.
The evolving toolset being created within Crypto Music will provide disruptive opportunities along this journey, if, builders and creators are smart about how they navigate the existing business landscape and, specifically, build in parallel to copyright.
The areas of most potential are in the creation of New Digital Rights, capital and community formation through Crypto, and, using these advantages to leverage position in the increasingly competitive environment of social and streaming platforms.
This was great. Agree on all counts here. I must say, it's hard not to have a bit of schadenfreude about the big record labels facing doom from DSP generative music. 😅
My understanding is that Open Edition NFTs don‘t pass the Howey Test because of the Copyright/art involved.
(1) People are investing their money
(2) in a specific NFT community/market.
But since the NFT centers around copyright, they say: „I want to collect this art because it makes me feel something“ and this means they‘re buying it for its present value and not to obtain a future profit (3) through the efforts (4) of others.
Now, when you take the „Copyright“ or the „art“ out of those things, I think it changes the test significantly.