Special Edition: SXSW

Hey everybody,

Thanks for your patience over the last week and a half. I usually blast out my newsletter and podcast on Fridays, but last week was SXSW and I was on the ground in Austin, Texas taking it all in.  

The good news is that I have a bit of a special edition for you today. In this newsletter, we'll explore several themes that stood out to me while I was at the festival. So, if you missed SX this year, or even if you attended and spent your entire time day drinking and listening to your new favorite bands, this episode will get you caught up on what ideas and themes were buzzing throughout this monster of a conference.  

If you've never heard of SXSW, it's a massive cultural conference in the heart of Austin Texas. It was started as a new-music festival in the late 1980’s but has since exploded into a multi-track interdisciplinary meeting of the cultural zeitgeist. Today it hosts a film festival, technology exposition, and speakers that touch on everything from public policy to gastronomy. There have been gaming expos, visits from world leaders and celebrities, and of course, it still hosts its music festival and conference. SXSW week also hosts hundreds of unofficial events like brand activations, art installations, and efforts by foreign countries to boost their cultural capital. Each year SXSW draws hundreds of thousands of curious and thoughtful minds to downtown Austin. You’re as likely to walk past Nicholas Cage on the street as you are to pass by Michelle Obama. If you’re interested in exploring cutting-edge music, tech, film, and other ideas, SXSW is the place to be each year. 

This year was no different. Unsurprisingly, discussions around the future of AI in music took center stage, but there were other themes that I came across repeatedly at SX. We’ll cover interesting discussions around AI, marketing music, compensating rights-holders, and the impact of glocalization on the music industry. At first glance, these themes live in silos, but the beauty of SXSW is that its interdisciplinary nature allows you to place these topics in context with each other as well as with everything else happening in the world. 

 The Inevitable AI Explosion

Regardless of if the discussion took place over a cozy lunch meeting or in a packed keynote speaker session, everybody had something to say about how AI would impact their work. One thread that emerged throughout the week was that very few people believe that AI will fully replace the human element in music. I heard it described multiple times (including once by Grimes’ manager and cofounder of CreateSafe, Inc. , Daouda Leonard ) as being similar to the synthesizer—the logic being yes, it greatly expands access and the sonic options available to musicians but does not replace the need to make creative decisions. After all, the invention of the hammer did not presage the death of the carpenter. I’ve personally been a bit worried about the effects of AI in music, so it was refreshing for me to dive into the other side of the argument. 

Another application of AI that was widely praised was the potential for AI to help artists and labels market their music. Many of my colleagues at labels are now treating marketing music as a nearly impossible task. Most of the label owners that I chat with have resigned themselves to simply recommending that their artists make TikToks every day and see what sticks. For obvious reasons, this strategy doesn’t resonate with artists. At SX there was an effort to grapple with this problem and find a better way. 

To summarize the problem, music creation and distribution have never been easier and more affordable. Even access to traditional marketing channels has become democratized. These dynamics have contributed to a flooded market where there is more music than there is demand for it. The question has become, how do artists break through the noise and establish a career for themselves?  

Any good answer to this question is complicated and involves multiple factors each moving in tandem with each other. One of the principles that was discussed at length at SX included shifting away from an industry focused on selling music as a commodity, and towards an industry focused on selling music as an experience.

This thesis opens up new doors for artists. There is more money to be made in brand deals, fan activations, etc., than there is in convincing people to stream new releases. Gen Z and Gen Alpha are not sitting down with a CD and listening to an album all the way through. They crave the attention and interaction of their favorite artists. They’re interested in using music, not simply listening to it. In this environment, the question for artists and label marketers is not: how do we promote our new record? Rather, it has become: how do we craft meaningful experiences for our fans?

On the other side of this debate is pushback that working in music is no longer about music. Successful career artists now need to become de facto influencers. As someone who does not want to be online 24/7, I’m sympathetic to this argument. When artists—by necessity—focus less on perfecting their craft, the music will most likely suffer. Perhaps some of the Gen-AI tools will mitigate the damage by cutting out some of the "tasks" of music creation. Still, after this years SX, I’m nearly certain that AI will not be able to capture the broader socio-political-cultural environment and distill it into a hit record. For example, it’s unlikely the AI could have captured all the anger, energy, and violence of the Vietnam War in the same way that CCR's “Fortunate Son” was able to. 

At SX, it also became apparent that the toothpaste is out of the tube. It’s unlikely that the macrostreaming economics will reverse themselves. It’s also inevitable that AI will be used to help solve these problems. After hearing about some of the new tools available to artists, my unequivocal recommendation is to dive into how you can utilize new technologies to help you market your music. If you haven’t played around with it yet, ChatGPT is the most accessible large language model to start with. From there, there are dozens of more streamlined and specialized LLMs to explore. One such AI application that was particularly impressive to me was SymphonyOS . If you’re an artist—check it out. It’s an uber-affordable automated marketing and fan engagement tool that could potentially save you hours of work each month. 

 How Do We Pay Artists More?

Running parallel to conversations around marketing music were conversations around how to make sure we can fairly compensate artists and other rights-holders. These discussions were set against the backdrop of Universal’s struggle with TikTok over usage of its catalog. For the most part, music copyright law—despite being somewhat confusing—is settled. There is no controversy over what kinds of actions represent the usage of a master, or what constitutes the public performance of a composition. However, the development of new technologies in need of music from companies with little understanding of the music business has created challenges for music companies. The leverage that certain tech platforms hold, and the technical difficulties inherent in tracking an unprecedented amount of data across countless new usages are fueling new debate about how we can ensure that artists get paid when their music is used.  

We can frame these questions in a historical context. The accessibility of music in its digital format has undoubtedly contributed to losses in revenue for the industry. According to a recent study, nearly 70% of small businesses in the US do not properly license the music that they play. I love my local coffee shop, but the fact is that the baristas who work there play whatever music they’d like from their personal Spotify accounts. Now, before all the small business owners out there chime in about how their margins are so slim, I’m obliged to point out that Pandora & Sirius XM’s business plans are only a few hundred dollars a year and eliminate your liability by playing music that is properly licensed for use in small businesses. The reality is that businesses lack the knowledge of music rights and are also searching out the easiest solution—simple streaming. Users of music need to know what is legal and illegal usage, but perhaps more importantly, they need an easy and abundant solution for their problems.  

With the rising popularity of interactive media, we are facing a similar inflection point. We must ensure that there is an infrastructure in place for users of music in digital spaces. When GTA launched GTA radio it was simple. The developer would license music as a traditional sync, and then they’d ship disks with a set number of songs on it. This model is complicated as the lines between users and developers/creators blur. We’ve seen multiple platforms grapple with this over the past decade—YouTube, Twitch, Twitter—but newer platforms like Roblox, and other metaverses, have not figured out their own best practices. There is an opportunity here, and at SXSW there was a robust discussion about how we can ensure that virtual spaces don't end up as under-monetized as small businesses.  

The problem seems to center around two topics. Firstly, how to best track the usage, and secondly, who is ultimately responsible for paying rights-holders? I met roughly a dozen data-oriented folks working on solutions to ensure that we can identify what, when, where, and how long music is used online. Trying to explain their solutions in detail wouldn’t do them justice, so I’ll just say many of them involve matching ISWCs to ISRCs and using fingerprinting technology to determine the necessary payments and payees. One such company working in this space was Copyright Delta founded by Daan Archer . Daan’s panel was a fiery discussion of how to implement some best practices across industries. His panel was complete with audience interaction that boarded on interruption and differing opinions from various stakeholders including PROs, Developers, Lawyers, and music-tech types. One interesting side-note here—conversations about blockchain technology made a subtle comeback in conversations in this space. It was interesting to see how much the once polarizing technology has begun to mature with usages that solve real-world problems instead of imagined ones. 

Revenue also stole the stage at a featured session that brought together Universal, SACEM and Deezer to discuss their experiments with artist-centric streaming models. Artist-centric streaming models are an alternative to the pro-rata streaming model that is currently the status quo at most DSPs. In the pro-rata model, artists get paid from a pool of royalties set aside from subscriptions and then paid out in proportion to their overall popularity on the platform. Every stream is essentially weighted the same (there are some differences based on the country of origin and if the listener was on a paid or free tier) and paid out as such. Under this model, an artist who uploads 31 seconds of noise is paid the same amount as an artist who uploads a symphony. Similarly, when a listener listens entirely to Taylor Swift, the portion of their subscription fee that is designated for paying artists is held in the royalty pool and divvied up amongst all artists in the same way as a user who listens to twenty different artists. 

Deezer’s artist-centric model makes some small tweaks to this formula in order to try to increase payments to artists who already earn some significant amount of money from their music. It’s important to note that Spotify announced a version of the artist-centric model with slight differences, and it’s also useful to mention that most video streaming platforms have similar models with “premium” content and “non-premium” content.  

Deezer’s artist-centric model operates on five pillars:   

1. Firstly, professional artists—who Deezer defines as any artist who receives more than 1000 streams per month from at least 500 unique listeners will receive a “double boost” in their royalty payments. 
2. Secondly, any song that is actively searched for instead of algorithmically served will receive the same “double boost.” (yes, these boosts stack). 
3. Thirdly, non-music noise audio will be removed from Deezer and replaced with content that Deezer creates and then de-monetizes. 
4. Fourthly, Deezer will double-down on streaming fraud. Removing content suspected of manipulation. 
5. And lastly, insofar as the royalty pool goes, every user’s contribution to the royalty pool will be capped at 1000. Now, this last one is a bit confusing. It just means that if a user streams 2000 tracks in a month, each track will be weighted as ½ stream. If a user streams 3000 tracks a month, each stream will be weighted as 1/3 stream. Simply, this means that no single user will generate more than hundreds of thousands of streams fraudulently on a single track. 

Deezer has been experimenting with this model in France for about six months, and at this panel, they discussed some of their findings. 

Most importantly, there appeared to be very little negative impact on any of the market segments. Serious artists got paid more, unserious artists got paid less (though we are talking a matter of dollars here), and fraudulent and noise projects got paid nothing at all. There was very little difference between the effects seen on artists at the majors and artists and indies. Most Genres saw an increase in revenue going their way, but hip-hop (arguably the genre with the most content distributed by unserious artists) saw a very small decline in royalties. Overall, the biggest losers here were fraudsters and creators of white noise.  

Spotify has also launched a version of an artist-centric streaming model, but they do not make payments to any artist earning less than 1000 streams per year. On principle, I like Deezer’s model a bit better as it grapples with the idea of non-premium content, but still acknowledges that there it has some value (regardless of how little) and needs to be fairly compensated for. All in all, I left feeling confident that the Artist-centric model is a net-positive shift for the music industry and I’m excited to the experiments to expand and see other DSPs take it on.  

 The Effects of Glocalization

The last big takeaway from SXSW is that we are now seeing spin-off effects of glocalization. Glocalization is an idea conceived by the music economists Will Page and Christopher Dalla Riva that streaming has allowed for more robust local music scenes around the world. For example, there are more French artists singing in French who are charting in France, than there were ten years ago. It used to be the case that most worldwide charts were filled with American and British artists singing in English. This is simply no longer the case. Over the past few years, we’ve seen the rise of robust niches like regional Mexican music, and French hip-hop have serious moments within their home countries. In the case of regional Mexican music, some of those stars have even gained international recognition.  

This trend manifested itself in various ways at SXSW. Will Page himself was a featured speaker, but there were several other panels devoted to discussing the intricacies of regional markets. I sat in on an interesting discussion of regional South American music and dove deep into Brazilian trends. I learned the names of roughly a dozen music genres that are household staples in Brazil but were entirely new to me. I also sat in on a fascinating discussion of the Balkan music industry where apparently, there is a very different music infrastructure as a result of the Cold War, and as a resault foreign acts can tour with very little overhead cost.  

There were also more regional activations than I’ve ever seen at SX. Small EU countries like Finland and Ireland had exciting showcases, while even US states and cities like Tennessee and Tulsa rented out entire bars.  

Overall, this contributed to a greater sense of regional and country-specific pride. In years past, SX’s music showcases have been about exporting their best artists to the American market, but perhaps in reaction to these recent changes in technology and listener behavior, this year’s country showcases occasionally seemed to be equally interested in importing American artists into their region. I heard at least three times a cultural export officer extol to American bands the benefits of touring in their country. 

What Was Missing

To cap off my summary, I do want to mention a few topics that we noticeably sparse or entirely missing from this year’s conference. I’ll start with the obvious—there was very little talk about NFTs. Now, as I mentioned earlier, multiple presenting companies utilized blockchain technology (CreateSafe, Music Delta, and PIXELYNX ), but the focus was not on the possibilities, rather, it was on the specific service each of these companies were offering. NFTs were almost entirely ignored.  

Also only lightly discussed were the benefits and differences of various distributors and DSPs. Every year, without fail, discussions surrounding the trendy new distributor, or a flashy new feature from Spotify dominated discussions that I had with label owners. This year, label owners seemed to focus much more on AI, and alternative marketing methods than the platforms used to feel so beholden to. 

Lastly, there was very little discussion—outside of a great panel with the JKBX CEO—of music as an asset class or tech-enabled artist advances. I’m not entirely sure why this is. All indicators point towards this being a space that is ripe with opportunity. Two companies that last year had big visibility, beatBread, and Labelcoin, seeming had no activations, or representation at this year’s conference.

The Great Unknowns 

Overall, I left the conference with a sense of unease. It felt like change was on the horizon. I twice heard label owners reference damns bursting bringing forth massive change in the industry. The coupling of discontent stemming from the current state of marketing, and compensation, combined with the uncertainty and waring feelings of skepticism and optimism that surrounding AI advancements, made for an occasionally tense environment where nobody seemed to have a grasp on what was next. The only measure of agreement was that something big would happen, and it would be happening here shortly. In many ways, I suppose the question that I (and all the other attendees) was grappling with was how can we be ready for the paradigm shift when it comes, and how can we ensure that we mitigate any fallout for the artists who make our industry possible?

Unfortunately, I don’t have an answer for that right now. So, subscribe to the Indie Insider podcast or newsletter to stay up to date as things happen in real time. I’ll be tackling this question in my next few interviews, so stay tuned for info on interview announcements.  

Of course, it wouldn’t be a SX newsletter if I didn’t mention all the great music that was down there. Congrats to all the artists. For me, the best act was a British band called Fat Dog. They had an ELECTRIC set at the Licks Mag showcase. Their music blurred the lines between British/Irish post-punk bands like IDLES and Fontaine D.C. with Ska influences, and even some flavors of Balkan folk music. Their frontman is electric, and their show borders on David Byrne-level theatrics. If you have a chance, check them out.  

I’ll be back tomorrow with the news, until then, cheers everyone!