One aspect of the recent hype about blockchain applications for music is the spate of conferences and panels about it. I went to one of these last Wednesday: the Music 4.5 conference at the offices of the ReedSmith law firm in NYC. The event was cleverly organized: the first half was devoted to startups who touted the virtues of their blockchain-based solutions, while the second half featured seasoned music industry experts who injected doses of reality by discussing limitations, challenges, and the need to start small by solving a few pain points instead of trying to boil the ocean. The overall effect of this event was to give attendees a very worthwhile overview of the state of the art of blockchain solutions for media as well as a dose of reality about it.
Executives from six blockchain startups presented at Music 4.5: Mine (Mediachain), Revelator, Songtrust, Songspace, Stem, and Ujo Music. (PledgeMusic CEO Benji Rogers, spokesman for Dotblockchain, was conspicuous in his absence; he was speaking at yet another conference on this subject in London the next day.) Mediachain is focused on cultural content, such as digitized museum or library objects. Ujo Music is creating a digital music ecosystem of its own, based on the Etherium blockchain, smart contract functionality, and cryptocurrency (Ether); its collaboration with artist/songwriter Imogen Heap has gotten a fair bit of publicity.
Apart from those, most of these startups appear to be trying to do roughly the same thing: when someone plays or buys a song on Spotify, iTunes, Google Play, or some other music service, they want the service provider to store information about the action on a blockchain and then kick off a series of transactions that pay everyone what they are owed, all automatically, securely, and instantaneously, using smart contracts.
This is definitely a suitable application for blockchain technology. But as these startups proliferate, it could lead to a situation where DSPs (music service platforms) will have to feed N different royalty tracking services N different sets of information, including N different sets of metadata. That’s a recipe for a mess. This would nullify any advantage of blockchain technology to the DSPs and thereby make DSPs uninterested in participating. In particular, it will make DSPs uninterested in building the real-time transaction reporting infrastructure necessary to participate.
This brings to mind the mess that needed to be cleaned up when e-commerce companies started to sell physical media products online during the first Internet bubble. For example, when Amazon started to become a popular retail destination for print books in the late 1990s, it objected to having to ingest thousands of different sets of metadata about books. As Amazon’s market power grew, so did the need for publishers to accommodate the retailer by standardizing on book product metadata. The result was the ONIX product metadata standard, which is widely used today — for e-books as well as print books.
That’s the kind of thing that needs to happen with music blockchain applications, and it needs to happen quickly. The timing is excellent: these startups will likely know better than to try to impose proprietary metadata schemes on big established companies like Google, Apple, and Spotify. Instead, they will want to “increase the size of the pie” by promoting overall growth in the acceptance of blockchain technology for royalty processing and competing on features, service, and price.
In fact, the second half of the Music 4.5 conference — with veterans such as David Hughes of the RIAA and Michael Simon of SESAC/Harry Fox/Rumblefish — spent at least as much time on the need for standards as on blockchain-related issues. Many people in the room were aware of the Global Repertory Database (GRD), a broad and ambitious Europe-based attempt to create metadata standards and a rights registry for music that started in 2010 and officially failed a couple years ago. To avoid repeating the fate of the GRD requires trying to solve a much narrower problem; not attempting to gain the consensus of every player in the industry or cover every conceivable use case; and offering tangible incentives for all participants, not just some of them.
Some of the necessary standards are already in place, such as ISRC identifiers for sound recordings and ISWC identifiers for musical compositions. Their use throughout the industry is inconsistent, but (as I’ve said previously) the interest, activity, and funding around blockchain solutions should help solve that problem.
However, there aren’t any accepted standards in the music industry for metadata at the level required for this type of transaction processing — including for names of entities that could be rights holders (musicians, songwriters, publishers, record labels, etc.). One of the panelists mentioned the idea of letting creators use their Facebook or Twitter IDs. I mentioned ISNI (International Standard Name Identifier), which Wikipedia uses and which is gaining traction in publishing for authors and other content creators. Jesse Walden of Mine/Mediachain suggested that multiple identifier standards could be handled by translating among schemes. Translation among metadata schemes is doable, but it doesn’t seem practical for unique identifiers because of the potential for conflicts and redundancies.
Things get more difficult when considering metadata about rights and compensation: for each action on each music track, who gets paid and how much. Even though this limits the scope of rights process automation to actions that are covered under license agreements (as opposed, say, to uses of copyrighted material that may not require licensing, such as fair use situations), that’s potentially a lot of data. There are also practical limits on the amount of data that can be stored in each blockchain transaction.
If this standards problem can be solved, then service providers like the startups at the Music 4.5 conference should have success with indie artists, publishers, and labels. Bigger-name artists should become interested; for example, the managers of the likes of Katy Perry and Justin Bieber are already investing in Stem. This in turn should put pressure on DSPs and major labels to integrate.
Yet the magnitude and complexity of rights metadata remains a problem, especially if it’s not possible to put it all into each blockchain transaction. The Dotblockchain folks have a solution to this problem which they call MVD (Minimum Viable Metadata). I’ll discuss that next time.