From Platform Risk to Protocol Empowerment
Upstart protocols are beginning to power an open-source creator economy
Platform risk is hardly a new concept. It is somewhat self-explanatory, and describes situations in which a company becomes overly reliant on a platform for one or more core facets of its business: content, distribution, growth etc. Dependent companies incur risk as the platform they rely upon either ceases to exist, or makes critical adjustments to its access points (e.g., sunsetting an SDK) or moderation efforts (e.g., changing an algorithm). These adjustments are often motivated by a variety of factors for the platform, but the key point is that the platforms can make these changes on a whim, potentially rendering dependent companies obsolete in the process.
Recent history is abound with cautionary precedents of platform risk. Web2 ushered in a ‘power to the platform’ era, giving rise to behemoths that subsequently cultivated vibrant developer ecosystems to extend their reach. As platforms evolve, so do their priorities. But one thing remains constant: a platform will always prioritize itself and its core users over its associated developers and their users. Several examples underscore this:
Access: In April 2015, following its acquisition of live streaming product Periscope, Twitter shut down rival product Meerkat’s API access. Meerkat had essentially built its business on the back of Twitter, relying on the platform and its social graph for sign-in authentication and broadcast distribution. In a particularly ruthless manner, Meerkat was given just two hours’ notice ahead of Twitter’s access restriction. With much of its initial scale contingent upon Twitter, Meerkat was never able to fully recover and closed its doors in September 2016.
Moderation: In January 2018, Facebook announced dramatic changes to its newsfeed algorithm in an attempt to prioritize user-generated content over news publisher and branded content. The move was in response to the increasing scrutiny on Facebook as a media company rather than a platform. There were a bevy of brands that relied almost exclusively on Facebook for distribution. Namely, LittleThings, a brand promoting ‘feel good’ content that scaled to 58 million unique users at its peak, lost 75% of its traffic from the algorithm change. It was out of business a month and a half later.
Spotify & Social Listening
There is a similar platform reliance occurring in the music tech industry between ‘social listening’ companies and DSPs (e.g., Spotify, Apple Music, SoundCloud). Social listening companies are predicated on the (correct, IMO) notion that music discovery and consumption through DSPs are largely siloed experiences. These social listening companies (omitting their names as I’m actively rooting for them to buck this trend), fueled by streaming SDKs through DSPs, aim to provide more seamless content sharing, discovery, and interaction experiences for both music fans and curators. In the process, they’re empowering new tastemakers and have scaled to hundreds of thousands of users. Ultimately, they’re designed to replace the text threads that music fans like me have relied upon for years to share tracks.
As part of their licensing arrangements with music rights holders, DSPs can sub-license commercial catalog to their developer ecosystems. On its face, the relationship is a win-win for all parties: rights holders see an expansion of revenue through developers (each stream on a social listening platform is counted as if it took place on the DSP), DSPs can use developers as a subscriber acquisition funnel, and developers can tap into commercial catalog in a more economical manner vs. a direct licensing deal.
Despite the seemingly symbiotic relationship, the leverage is still lopsided in favor of the DSP. In July, Spotify announced it was sunsetting prior iterations of its mobile streaming SDK, which is relied upon by thousands of developers. The forced migration to a different endpoint caused temporary pauses and bugs for some developers.
Beyond endpoint updates, developers must be authorized and get comfortable with usage parameters that are determined unilaterally by Spotify. Commercial catalog is understandably a sensitive asset; it requires extensive safeguards to ensure it’s being tastefully and legally exploited. But the fact that these usage parameters are subject to change without any required notice, as stated in the Developer Terms, puts developers in a precarious position:
Thus far, Spotify’s direct investment efforts in social have largely been focused on live audio (e.g., Betty Labs). These investments culminated in the development and launch of Spotify Live (fka Greenroom). But, to the extent one or more of the social listening companies scales to millions of users, it’s not inconceivable to think that Spotify, which has proven to be very acquisitive (27 total acquisitions), would bolster its social efforts through an acquisition. And what do we think this would imply for the rival/other companies still using its SDK?
Power to the People…Via the Protocol
In contrast to the platforms that have defined web2, decentralized protocols are permissionless and credibly neutral. In what has quickly become a seminal text (‘Hyperstructures’) on crypto protocols, Jacob Horne, co-founder of Zora, describes the gold standard as a ‘hyperstructure;’ these supercharged protocols are defined by the following characteristics:
This year marks a seismic shift - from platform to protocol - in the architecture of some of the most prominent companies that facilitate both the creation and collection of crypto assets. Namely, Zora, Sound, Decent, and Manifold have launched decentralized protocols. Prior to this shift, these companies largely resembled their platform predecessors as centralized intermediaries providing more exclusive creation access via higher touch activations. Their refreshed orientation as open, self-serve toolkits enable any and all artists to create, distribute, and own crypto assets with complete creative flexibility. This egalitarian access aligns with the decentralization ethos and could represent web3 music’s ‘killer app,’ according to Dopr co-founder Rob Abelow.
In ‘Hyperstructures,’ Horne articulates perhaps the most important implication in this shift:
‘…being permissionless means that everyone can confidently build their own platforms, applications and economic models on top of the hyperstructure without the risk of de-platforming. There are no API keys required, and no fear that a platform can suddenly upend your entire project on a whim.’
Amazing, but how does this help the social listening companies that want commercial catalog without direct deals? Well, it doesn’t just yet. While we might be a long way from music rights holders granting licenses to open protocols, one can imagine these protocols assembling a library of openly accessible content for artists to remix and developers to integrate (similar to Arpeggi). This content would either need to be designated as CC0 by the creator, or contain an enforceable royalty upon usage with effective systems in place for proper attribution and content moderation (which might still require human intervention). In any case, I expect these protocols could evolve into open-source content hubs that fuel a more composable creation era.
Of course, the protocols’ open nature begs the question of whether there can be preventive measures in place to deter bad actors and copyright infringement. Web3 streamer Audius has been publicly maligned for its apparent inability to remove infringing tracks. Spotify also deals with infringing uploads on a daily basis, but has a team in place to rectify.
Closing Thoughts
The irony of web3 is that the majority of activities - transactions, communications, marketing, etc. - are still conducted through centralized platforms. OpenSea, Twitter, and Discord all present some variation of platform risk, a phenomenon I hope can be reduced in this new paradigm. The actions of Zora, Sound, Decent, and Manifold represent a meaningful leap forward in the embrace of a purely open-source creator economy.
Sources (of inspiration)
Platform Risk by Startup Illustrated
Hyperstructures by Jacob Horne
Startup Anti-Pattern: Platform Risk by HighContrast
Meerkat Founder on Getting the Kill Call from Twitter by TechCrunch
Little Things Shuts Down, a Casualty of Facebook Newsfeed Change by Digiday
Introducing Sound Protocol by Sound.XYZ
Powered by the Decent Protocol by Decent
Enter Zora V3 by Zora
Yet again, NFM reveals the underlying vulnerabilities (and strategy) hanging out in the fine print. What else can we learn from reading the terms & conditions from Big Tech and Web3 startups?