Discover more from Genesis Block
Protocol Analysis #6
This is the sixth article in our Protocol Analysis series. In this series, we use our investment template outlined in How to Analyse a Web3 Protocol to fundamentally and critically analyse a variety of crypto protocols, aiming to separate the ‘real’ from the hype. Let’s dive in!
Protocol: Audius, a decentralised music streaming service that connects fans directly with artists. Artists on Audius can upload any and all audio they want, and users can stream, comment, like, and perform other actions one would expect on a music streaming platform.
Overview of Audius 🗒️
Founded in early 2018 (with the mainnet going live in October 2020) by Roneil Rumburg and Forrest Browning, Audius is a decentralised music streaming service. Think Soundcloud on a blockchain.
Currently, the platform is free to use. However, when Audius does begin to monetise, it intends to split the revenue 90:10 in favour of artists, with the balance being accrued to node operators. The underlying principle the founders are approaching monetisation with is that instead of monetising streams, they want to capture value via artist-fan relationships and interactions. The reason for this is two-fold:
Monetising the content directly via familiar subscription models will require navigating various legal obstacles with music labels. This could not only lead to less value accruing to the artist, but also is resource intensive. Audius could begin signing artists directly, but then it would have a vested interest to push those artists on the platform, sacrificing neutrality.
Over time, the cost per stream asymptotically approaches zero. This can be seen in how we went from purchasing vinyls, cassettes, and CDs, to iTunes’ $1 per track, to today’s ~$10 for unlimited streams. Music has become commoditised, thus limiting the value generation potential by artists and the platform.
There is no specific method of monetisation the project wants to implement. The thought is to allow artists to monetise however they deem fit (e.g., subscription, fixed fee, per stream, NFT gates, etc.), and over time, the best models will naturally emerge.
The team has raised ~$15M via token sales since inception, across three rounds – ~$6M in early-to-mid 2018, ~4M in mid-2020, and ~$5M in mid-2021. Investors include General Catalyst, Kleiner Perkins, Pantera, Lightspeed, Multicoin Capital, Coinbase Ventures, Binance, Standard Crypto, and tier-1 music celebrities like Katy Perry, Nas, The Chainsmokers, Jason Derulo and Pusha T.
How Does Audius Work? 🤷
On the front-end, users experience Audius via its native web/mobile app, or via 3rd party applications, all which leverage the Audius protocol’s back-end infrastructure.
On the back-end, the protocol is made up of three pillars – content nodes, the content ledger, and discovery nodes.
The content nodes leverage AudSP (an in-house developed decentralised storage protocol built using IPFS) to store and maintain the availability of all content across the network. Once the artist uploads a content piece and metadata to the content node, the artist receives an IPLD link (a reference pointing to the content piece stored), which the artist then publishes to the content ledger alongside the content metadata.
The only way artists can program or ‘gate’ the experiences they offer to users is by running a content node and hosting their own content. For example, they can attach conditions for fans to listen to their music, like, for example, proof of their fandom. However, setting up the node and adding these conditions requires a large amount of technical knowledge, which almost all artists do not have. If this process is not abstracted away for the artist, i.e., all the artist needs to do is upload their content piece to Audius and the protocol takes care of publishing the IPLD link and content metadata to the content ledger, the process is definitely too technically difficult and time-consuming for the artist and may disincentivise them from using Audius.
The content ledger is an amalgamation of all smart contracts across Solana and Ethereum (and any future L1/L2 blockchains that Audius expands to) that host different parts of the Audius ecosystem. Solana hosts the user-facing features one would need to run a music streaming service, while Ethereum primarily hosts the governance and node hosting functionalities. Collectively, the content ledger includes:
A consistent audio content and metadata format specification to ensure accessibility.
A decentralised process for artists to control track content, revenue splits, and content ownership structure.
A registry of all nodes reachable in Audius.
The social graph of all users interacting with Audius.
Implementations of the token and governance systems.
The discovery nodes index the content ledger, such that users may efficiently query and discover content. These nodes can be used to fetch a fan’s feed, a playlist, song and artist metadata, search the corpus of Audius entities, and execute other queries about the network.
The journey of a content piece and the flow of interactions between the content node, content ledger and discovery nodes can be seen in the diagram below.
Audius offers open API access to their entire catalogue. This allows other developers to leverage the Audius service in other third-party applications. For example, Star Atlas, a Web3 game, has taken the design choice to upload game soundtracks to Audius and access the tracks in-game via Audius’ API.
Audius has also integrated with Tik Tok, allowing creators to port their tracks from Audius directly into Tik Tok. At the time of writing, Audius is the only streaming service that offers this functionality.
The native Audius client-side application leverages Hedgehog, an open-source, client-side wallet developed by the core Audius team. Hedgehog uses a username/password system and links it to a non-custodial wallet on the back-end and does not prompt users to confirm transactions. Essentially, this system allows users to interact with decentralised applications just like they would any other website, without centralising control of private keys.
The Trends Supporting Audius’ Emergence 📈
Audius is trying to solve for a fundamental problem that persists in the music industry today – only ~12% of the ~$45B generated in revenue by the music industry eventually goes back to the artists. The two main reasons for this are:
There is a plethora of middlemen in the music production to consumption value chain.
Artists have an incredibly high dependency on advances.
After revenue is generated by the end consumer, middlemen, including but not limited to labels, rights societies, distributors, and digital service providers (like Spotify, Apple Music, etc.), end up taking a significant slice of the pie before any value accrues back to the artist – the classic Web2 problem. Audius aims to solve this by directly connecting artists and fans – artists can directly upload their work for fans to consumers, and fans can directly compensate the artist using cryptocurrency, all while providing artists with the reams of data generated about their fans and their content consumption patterns.
With respect to artist dependency on advances, artists end up taking cash up-front from labels to finance their early career expenses in exchange for a hefty percentage of future revenues. An industry standard contract structure is the ‘360-degree’ deal, where labels provide artists not only cash advances but also assistance in other areas like distribution, legal, marketing, etc. However, in return, labels are entitled to a cut of revenues generated by any and all forms of the artists’ work – not only limited to music consumption. From a music label perspective, such terms are necessary as their business model is akin to that of a venture capitalist. They must manage their risk by capturing a large percentage of artist revenues because only a small percentage of their artists will actually generate any significant revenues. Audius (and the broader Web3 movement) wants to solve this dynamic by enabling different forms of value capture for artists such as NFTs and social tokens.
Launch & Future Roadmap 🚀
There was a pre-mine of one billion tokens at launch in October 2020. The token distribution is explored further below.
The team has not provided us with any form of a formal roadmap, but from the documentation and media interviews, we can infer that the following are what the team is working on in the near term: creator monetisation, stablecoin integration, and social tokens.
It is not a good sign that there has been no information provided about the future roadmap, or even the fact that there has been no concrete development on creator monetisation – which is the primary purpose of building this Web3-based music streaming app!
Roneil Rumberg, the co-founder and CEO, is a 2014 Stanford alum with a bachelor’s degree in Computer Science. He began mining various cryptocurrencies during his college days, and later co-founded a cryptocurrency wallet start-up called Backslash. He then joined Kleiner Perkins as a founding team member in their early-stage seed fund, where he was predominantly evaluating and investing in blockchain companies, like Lighting Labs.
Forrest Browning, the Co founder and Chief Product Officer (CPO), is also a Stanford alum with a bachelor’s and master’s degree in Computer Science. He was in the same undergraduate class as Roneil, and Stanford was where where the duo met. He has been listed in the Forbes 30 Under 30 list and has previously founded and exited StacksWare, an enterprise data centre management platform that provided analytics to Fortune 500 companies.
Their advisory board includes DeadMau5, 3lau, Bing Gordon (co-founder of EA), Tom Schmidt (partner at Dragonfly Capital), Greg Hazel (Chief Architect at BitTorrent), Ash Pournouri (Avicii's former Manager), Justin Kan (co-founder of Twitch and founder of NFT-based GameFi platform Fractal), Robb McDaniels (CEO of Beatport), Jesse Grushak (Senior Strategy and Product at Consensys), and Cooper Turley, a Music NFT influencer.
A primary concern is with the transparency of the team, which is usually a warning sign. For one, neither is there any mention of Ranidu Lankage (an ex-cofounder of Audius) by Roniel or Forrest anywhere in their media interviews, nor is there anything in the project documentation. In my opinion, a falling out at the leadership level is something noteworthy and must be communicated to the broader community. Additionally, there are long gaps in the project's blog posts, one from September 2018 to May 2019, and second from August 2019 to July 2020.
Thoughts from Using the Protocol 🖱️
The experience of using Audius is extremely disappointing. Application functionality has been suboptimal. Errors while performing basic actions like logins, track streaming, claiming rewards, etc. were all too common. While logging in, the page repeatedly did not load, and when we were finally able to log in, the music player did not even work. This is a huge red flag. If a project that is meant to be used to stream music cannot even fulfil its most basic premise, it casts significant doubt on whether it can meet its stated aims. An idea is only the very first part of developing a business – execution is 99% of winning the game. If Audius has not managed to even get the very core and basic features of its product correct, then its future prospects seem markedly pessimistic.
However, once we were able to log in and play music, the UI of the app turned out to be pretty sleek. The sound quality was good, and the colour and font scheme are attractive.
Users pay tips to artists in the form of $AUDIO tokens. While this hints at a use case for $AUDIO, in reality this is not so since it just adds friction to the user experience. Users should be able to tip artists in stablecoins/any currency they prefer and hold in their wallet, and the artists receive tips in stablecoins/any currency they prefer, with a module inside of Audius converting the tokens.
Below are some screenshots of the Home Feed, Explore, Trending, and Tip screens.
AUDIO is the protocol's native token, built using the ERC-20 standard, and was launched in October 2020.
The AUDIO token serves 3 purposes: network security, feature access, and governance. Besides speculation, these represent the main reasons why an individual would be incentivised to buy and hold AUDIO.
It facilitates network security as node operators must stake AUDIO (minimum of 200k) to earn the right to perform node services. The higher the stake, the higher the chance for a node to be picked to perform the work. Of course, these stakes are subject to slashing if the node maliciously behaves to the detriment of the network. Also, individuals can delegate their tokens to node operators to avoid leaving the tokens idle in their wallet.
Depending on the amount of AUDIO artists hold, they can get access to various features within the Audius application. For example, holding 100 AUDIO gives one access to a silver badge and the ability to showcase NFTs on their artist page. You can see the tier offerings here. AUDIO is also used to tip artists.
Lastly, staked AUDIO gives the holder governance rights, which we will discuss further in the Governance section below.
All in all, in its current state, the AUDIO token does not provide much utility beyond governance rights. Due to high hardware requirements, staking is out of the question for most. The features provided from holding the token are very rudimentary, however this does indicate a way in which the token could have future utility, by being the instrument that needs to be held in order to access token-gated experiences. Earning yield via delegation is commonplace across Web3 protocols, and the team has not indicated any revenue being accrued to the token when the protocol starts monetising. Lastly, using AUDIO to tip artists puts immediate and constant sell pressure on the token unless artists are heavily incentivised to hold the tokens – which is definitely not the case here, and hence represents misaligned tokenomics design.
Currently, therefore, there is not much use to holding the AUDIO token beyond speculation – and given the extremely high token inflation that we will talk about below, speculation too seems like a bad bet.
At the token generation event, 1 billion AUDIO tokens were allocated:
~5% to the top 10k artists and listeners
~41% to the Audius team and advisors
~36% to investors
~18% to the community-governed treasury
The first and most obvious thing to notice is the sheer percentage of tokens that have been allocated to investors and founders. If we use the token distribution highlighted in ‘Optimizing Your Token Distribution’ by Lauren Stephanian and Cooper Turley (ironically, an investor in Audius), a maximum of 35% should be allocated across the founding team and investors. Audius wildly diverges from this number, providing a staggering 77% of its initial token supply to these two entities.
Such a poorly weighted initial token distribution may engender a lack of trust from its core users and artists, since the protocol is meant to be creator-friendly and have artists at its very core. Given that community is a core tenet of Web3, the protocol users and project treasury must be given a higher allocation. Allocating a large majority of the initial tokens to the team hints at the team being in it for the money, not having appropriate incentives to develop further features in a timely manner, and in reality, not placing artists and creators at the core of the protocol since they only received 5% of the initial token distribution.
The token follows a fixed rate inflation of ~7%, out of which ~6% emissions occur via node rewards to stakers/delegators, while ~1% occurs via token rewards to incentivise protocol activity like uploading content, sharing, profile completion, etc. As of September 17, 2022, annual rewards for nodes are ~25%, with the delegators cut being determined by each individual node operator.
However, due to the token unlocking the effective annual inflation rate is much higher. You can see the supply entering circulation in the image below. For context, in October 2020, the AUDIO circulating supply was ~150M. In September 2022 it was ~825M.
As can be seen, even with the community supply increasing significantly over time, the level of dilution suffered by the team and investors is not high, indicating that they will still govern the protocol. As of mid-September 2022, ~60% of the circulating supply was held by the top 2 wallets, and ~67% was held by the top 3 wallets. Out of this, ~44% is held in non-smart contract wallets, making price manipulation a possibility. You can see the latest distribution here.
This is a huge centralisation risk and risks exposing artists to the same issues that have arisen in Web2 music companies like Spotify – namely, more money goes to the treasury/team, and less to the artist.
The Open Audio Foundation is the entity formed and endowed with the community treasury tokens allocated at Audius’ token generation event. For any of these funds to be used in protocol growth programs, or any update to be made in the protocol, an on-chain vote must take place. Currently, only nodes can propose changes, and all proposals require a 5% of staked AUDIO quorum and 50% majority, for it to pass. Every 1 AUDIO staked provides 1 vote to the staker/delegator. As we’ve discussed previously, this one-token-one-vote model has a lot of flaws.
As of September 2022, the top 8 nodes make up 50%+ of staked $AUDIO, while none of the top 5 have voted on any proposals. Latest distribution of voting weights can be seen here.
It is worth mentioning that the team has inbuilt emergency powers (referred to as “short circuiting” in the whitepaper) to a community multi-sig which need six out of nine signers to pass any protocol update/veto a proposal without any community vote. The nine signers are not explicitly mentioned in the documentation, but it is likely the founding team/advisors. These powers can be voted away via a community vote, but there is no system in place for the multi-sig controllers to not veto this proposal.
The governance structure in its current format is centralised – the team can pass or veto any update without community involvement, which completely nullifies the impact of the community. This again serves to lessen trust in Audius and its team and reinforces that under the hood, Audius is not really a protocol that holds the values of decentralisation and community empowerment close to its heart.
Traction & Performance 📉
In May 2022, Audius passed 1 million song uploads on the platform. While Spotify has ~82 million tracks, 1 million in less than 2 years of mainnet launch is nothing to be sneered at.
As of mid-September 2022, total AUDIO staked has declined to ~260M from ~315M in October 2021. However, in the same period, total API calls increased from ~35M to ~160M.
Interestingly, in October 2021, unique plays and unique users was ~6.6M and ~1.2M, respectively. In August 2022, the same was ~6.4M and ~3.5M, respectively. In other words, while unique monthly users tripled, song plays per user fell ~66%. This indicates that users joined the app, played (or tried to play) a song, and left the platform. This is not a good sign at all for user retention and stickiness, and highlights that it is critical for Audius to focus on improving their UX, song selection, and curation, among other improvements, in order to maintain their customer base.
Since October 2020, apart from a few peaks in August 2021, daily, weekly, and monthly active users have actually trended downwards, reflecting minimal uptake of Audius by both listeners and artists. The ‘old’ users are also trending downwards, which shows the difficulties Audius has had in user retention. The chart for monthly active users depicts this:
Note, however, that there is a significant discrepancy between the data provided by the Audius Dashboard, which says that the number of unique users in August 2022, for example, was ~3.5 million, while the above Dune dashboard indicates that the number of monthly active users in August was less than 1,000. We are not sure about the reason for this discrepancy.
The protocol dashboard provides high level data on protocol usage, but it must be more detailed.
To begin, when calculating data points, a definition of what is being captured should be mentioned. For example, on the Audius dashboard, it tells us that Audius had ~3.5M users in August 2022, however on the Audius application, we couldn’t find a single track with more than 500k streams. The artist profile with the highest followers was Audius itself, with ~1.3M followers.
Allocating the initial supply to only 10k users (when Audius reportedly had ~500k users) seems overly restrictive. On top of this, many of the tokens went unclaimed. All these instances suggest Audius’ user base is either much smaller than we are led to believe, or the definition of monthly user has been too loosely defined. It is also possible that the bulk of protocol usage is coming from third party applications. But even if this is so, it must be clearly communicated.
The most jarring aspect about Audius’ monetisation model is that, currently, there is none. As we mentioned in the Future Roadmap section above, the most we can do is infer that Audius is working on creator monetisation as one of its next goals. For a protocol that aims to be about the creators and the wealth of possibilities for creators with smart contracts and tokens, it is a big, flashing red flag that Audius has not even published a simple roadmap for creator monetisation, let alone ship any features.
Audius has also said that artists will be able to monetise in any manner they see fit. It is critical that when (if) this feature is rolled out, the UI and UX are extremely clear and simple for artists to use without any friction. Artists should not be expected to have any technical knowledge and set up their own monetisation models. Moreover, the options for monetisation should be very clear, with the benefits and drawbacks of each clearly outlined, and the selection of the appropriate model for the artist should be as easy as ‘click-and-select’. Everything needs to be geared towards making it as easy as possible for artists to do what they are best at – make music and reach as large an audience as possible.
Content Rights Management
Another very fundamental point of concern has been loosely attended to at best – Content Rights Management. The magnitude and importance of successfully tackling this hurdle will eventually dictate the success of Audius as a whole.
We clearly saw the decline of Soundcloud (arguably the most direct comparable to Audius), due to its inability to reconcile, one, its ethos of being an open audio platform where artists have full autonomy of what they can upload and users who can facilitate grassroots discovery, and two, the constant need to effectively manage and monetise platform content, especially when it came to label owned content. Audius will likely face the same issue as it scales. Currently, the team has mentioned that it has a community moderation process/tool, but neither do we have clarity on how this works nor is it clear how this will scale.
Smart Contract Risk
Lastly, while regulation and smart contract risk is present in all Web3 protocols, it is worth mentioning that the protocol was victim to a hack in July 2022, where ~18 million AUDIO tokens worth $6 million were compromised. While the bug has been fixed, we have no guidance on if there will be any remediation for the missing funds.
Closing Thoughts ⌛
All in all, the problem the project is trying to solve is in much need of attention. For too long have artists had to be subject to predatory deals, resulting in them unable to make a sustainable living of their craft, in spite of their work resonating with hundreds of thousands, if not millions, of people.
Audius has made some progress towards this goal. It has assembled a formidable core team, advisors, and investors, their user interface alone is indistinguishable from a Web2 application, it showed commendable agility to respond to the July 2022 hack, and the Tik Tok integration will give it much needed attention.
However, if Audius has any chance of achieving its vision, its offering will have to be magnitudes greater than what is being offered today for users to make the jump from the Spotify and Apple Music’s of the world. As it stands today, Audius serves a very niche population, and it is unclear (refer to the above concerns) how it will be able to cater to the mainstream. It also has quite a few red flags, and it would be remiss not to mention these:
Lack of transparency from the team;
Lack of monetisation;
Lack of future roadmap;
Significant UX issues with the app;
Too much technical knowledge expected from artists, which feeds into the difficulties with UX;
No real utility for the token;
Poorly weighted token distribution;
High inflation via the token release schedule;
Centralised governance processes;
Poorly defined data standards for daily/weekly/monthly active users;
Potential future concerns with Content Rights Management in the vein of Soundcloud.
This is a personal blog. Any views or opinions represented in this blog are personal and belong solely to the article authors and do not represent those of people, institutions or organizations that those authors may or may not be associated with in professional or personal capacity, unless explicitly stated. All content provided on this blog is for informational purposes only. The owner of this blog makes no representations as to the accuracy or completeness of any information on this site or found by following any link on this site. The owner will not be liable for any errors or omissions in this information nor for the availability of this information. The owner will not be liable for any losses, injuries, or damages from the display or use of this information. Any views or opinions are not intended to malign any religion, ethnic group, club, organization, company, or individual.
👇🏽 please hit the ♥️ button below if you enjoyed this post.