Mass Adoption Narratives: Decentralized Social Media
Social media as we know it today was arguably invented in 1997. That was the year that SixDegrees.com launched, and introduced the concept of having a profile based on your real identity, to connect with your real friends. The platform garnered one million users before being acquired for $125M in the year 2000.
Today, established tech behemoths have brought the social lives of nearly the entire world online, and social media has settled in at the very heart of our society. As of November 2022, there were just above 4.5 billion users of social media worldwide, with each user spending an average of 2.5 hours daily on these platforms.
Time Spent on Social Media
Over the last 10 years, applications such as Tik Tok, Instagram, and Facebook have exploded not only in usage, but in their expansiveness - each has disrupted how business is done in multiple industries. This has been accompanied by massive value creation, but ceding this level of power to these companies has had an invisible cost, as well.
While user-generated content and data drive all of the revenue in the ad-based world of social media, the platforms seize ownership of these resources under the traditional Web2 business model - not the users. According to Chris Dixon, aggregation platforms (social or otherwise) first attract users with free services, which enables a rapid build-up of network effects as they eliminate all friction for a new user to join the network.
Over time, however, after reaching a critical mass of users, they begin to extract value. They do this by collecting data on users, then selling it to companies who subsequently use the data for targeted advertisements. Since only the social media behemoth has access to and ownership of this wealth of data and content, they command significant leverage amongst all industry stakeholders.
This model has a number of conspicuous drawbacks:
- It gives rise to a multitude of privacy issues, since these companies have spotty records with handling data responsibly. The Cambridge Analytica scandal is only one extreme example of this.
- It disincentivises innovation. This model makes it borderline impossible for newer apps to attract their first cohort of users (also known as the cold start problem.)
- It effectively makes the content on these platforms a “walled garden.” Companies don’t like to let content leave the platform, which produces a lack of composability. Creators are forced to create separately on each platform and cater to each audience in isolation. In other words, neither can you leverage a reputation built on one platform to supercharge growth on another, nor can you port over a following from one platform to another, or vice-versa.
This is to say nothing of the second order effects of centralization, such as de-platforming, censorship, echo chambers, etc.
As you can see, we're stuck in a vicious cycle: users have to use these companies' apps because they have a monopoly on the content, which forces creators to give their content up to them to get reach, which continues to empower these companies at the expense of creators and society as a whole.
These companies have managed to create a global network effect around a private pool of content that they solely monopolize. Granted, while there is value in combining all of the content into a single pool (since it allows for curation at a global scale), whoever we put in charge of maintaining the pool is ultimately going to become a centralised gatekeeper with vastly disproportionate power. A solution would arise if we had a way to shift the network effect to a public pool of content that no individual entity controls – but can it be done?
Web3 and the Social Graph
Social media built on Web3 wants to make the relationship between all stakeholders – creators, users, and the social media platform – more mutually beneficial, and make the presence of a highly extractive middleman obsolete. Web3 social is a fundamentally new approach at re-architecting the traditional Web2 social media stack, where closed databases and opaque algorithms become open networks, where users and creators own their data, and where there are transparent free market dynamics in algorithm and application feature development.
To understand how this is achieved (or being attempted), it is helpful to visualize the social media tech stack as shown in the diagram by Messari below. While a bit dated, the categorisation is informative.
The Nascent World of Web3 Social
Most critical to focus on is the social graph. To begin with, in the context of a blockchain, a graph is an indexing protocol. Indexing reduces the time needed to find a particular piece of information. An easy-to-understand example of this is an index in a book, which helps you find the page where that information might exist. Without the use of graphs, you would have to search each recorded block in a blockchain in order to find the information you’re looking for. Obviously, this is no good as it would take a long time. Graphs bring order, structure, and connection to data in a way that makes it easier and quicker to query and find the specific information you need.
It is useful to think of a social graph as a data structure that collectively refers to the global mapping of all users, their connections, and their interactions and behaviors towards others. It is the aggregate of all of the social relationships between accounts – every like, follow, comment, etc. – helping provide users with a more rich and engaging experience.
The problem with Web2 social networks today, as mentioned earlier, is that they all read and own user data from their own centralized databases. There is no portability. Your profile, friends, and content are locked to a specific network and owned by the network operator. This causes each network to fight a zero-sum game for your attention.
Web3 social protocols aim to create permissionless systems where users own and control their data, while maintaining a unified ‘social graph’ for applications to tap into – in other words, an open social graph, i.e., an uncensorable and interoperable protocol that tracks, on-chain, every user interaction. This allows developers to permissionlessly access user data, irrespective of where it originated, to create curation algorithms and novel front end clients to serve to users.
Contrast this with Web2 development, where developers need to work at a specific tech company to get access to the underlying data, and still can't leverage it to build new products. Now, user data has gone from being a commodity or precious resource to more of a public utility.
The result is a fundamentally different competitive landscape for social media platforms. In Web2, all social platforms faced the aforementioned cold start problem of attracting their initial cohort of users and validating the product being built. This process of acquiring new users is the most resource intensive aspect of launching and growing any new social media application. However, in a Web3 world, developers can freely build applications on open social graphs. Imagine a user using a new social application, but their experience is as if they have been using the application for many years. Thus, when users are not a moat, feature and curation algorithm development becomes akin to a Darwinian process of natural selection, as opposed to take-what-you-get dynamics with Web2 social incumbents.
Now users become the primary determinant of what features organically emerge, while also significantly reducing the upfront costs of customer acquisition. Additionally, by virtue of existing on the blockchain, Web3 social platforms can leverage the use of token incentives to further bootstrap application usage.
When creating a decentralized social protocol, there are four key problems that must be solved.
- Identity: Ability to issue a decentralized identity – a collection of a unique (username) and non-unique identifiers (profile picture) – that is mapped, owned and controlled by a real world entity (allowing the user to utilize this identity across platforms).
- Authentication: All data that is generated by a user is verifiably untampered.
- Availability: Ensuring data is accessible between applications (a user carries his/her data, while jumping between applications), while maintaining reasonable constraints on the size and volume of data, such that the data can be realistically indexed.
- Consistency: Ensuring all user activity recorded is done so consistently, as conflicting data can emerge due to use of several clients writing data to the same database.
Each protocol discussed below is trying to solve for these, in their own nuanced approach.
Launched on Polygon in early 2022, Lens Protocol is a composable, user-owned, and open social graph that any application can plug into. As we’ve been discussing, it is the ‘infrastructure’ layer for social media, rather than a specific application itself. It provides the tools that various applications can be built with, with each application using Lens Protocol benefiting the whole ecosystem and turning a zero-sum game into a collaborative one. This enables developers to design meaningful social experiences without needing to turn to increasingly extractive feedback mechanisms to lock in a user's attention.
Moreover, the protocol is currently overseen by a multisig wallet, which will be expanded to a broader DAO that can develop and vote on new modules and expanded functionality in the future. The proposed DAO governance will be an important factor to follow closely and assess.
Lens Protocol Features
With respect to users, Lens Protocol has seen a significant drop off in new user growth, with new profile creations peaking in June 2022 at 22,000, and dropping by 80%+ to sub 4000 in February 2023.
Number of New Profiles per Month
But wait - in contrast, in the six months from August 2022 to February 2023, Lens has witnessed a 200%+ growth in monthly active users, from ~17,000 to ~60,000 users.
Active Profiles per Month
Noteworthy projects building on Lens are Lenster, Phaver, and Orb.
Lenster, a browser based application with all standard social features (short form text, images, video, audio), was the first application on the Lens Protocol to get meaningful traction. The application has not yet introduced any token incentives, making all usage organic and feature driven.
Phaver, a mobile based application with all standard social features, is the second-most popular app on Lens. While similar in features to Lenster, Phaver differentiates itself by incentivising activity via tokens. Users can “bet” on posts by locking tokens on a post, which yields the user token rewards if the post does well. Similarly, creators can do the same for their own posts, which yields them rewards the more other users lock in their tokens on the same post.
We can attribute a large growth in the number posts due to these token incentives, making the growth quality questionable.
Weekly Post Counts of Lenster and Phaver
Orb, a mobile based application with standard social features, takes third place. While there is no tangible differentiation, the combination of a mobile friendly and intuitive UX to access all Lens content is a possible reason why Orb has seen traction in recent times.
Overall, Lens is built from the ground up with modularity in mind. The true unlock for Lens will be when all of its core features (i.e., everything is an NFT) are abstracted away and truly user-friendly applications are created on top of it.
Farcaster is a self-proclaimed sufficiently decentralized social network. It is another open social graph protocol that can support many clients (i.e., user facing applications). The team compares themselves to email – the underlying protocol that is used by the likes of Gmail, Hotmail, etc.
Built on top of Ethereum, as one would expect in an open social graph, users will always have the freedom to move their social identity between applications, and developers will always have the freedom to build applications with new features on the network (without the risk of API access arbitrarily being revoked).
The first client, and currently the most popular one, being developed on Farcaster is being done so by the protocol team itself. The eponymous application is most similar to Twitter, in that one can send (cast) short text broadcasts, alongside other standard social features such as resharing, displaying NFT profile pictures, etc.
In terms of traction, the all time cumulative user base for the farcaster protocol is ~11,000. If we look specifically at the month of March 2023, the protocol had ~3,300 monthly active users (defined as a wallet unique users who published a cast over the last 30 days). Of course, the bulk of this originates from the Farcaster client itself.
Today, the protocol has a total of 30 clients being developed, for which you can see an up to date list here. The go-to-market strategy for Farcaster is to attract all crypto native users, for whom alternatives like Twitter are insufficient (in ways like not natively integrating on chain activity). After this cohort, the protocol will focus on the larger mainstream cohort of users.
Launched in September 2021, CyberConnect is another open social graph protocol, on top of which decentralized applications can be built. Identity sovereignty is a core tenet for the team, keeping this central to their development plans. Built using Ceramic Network and IPFS, Cyberconnect is chain agnostic, and currently exists on Ethereum, Solana, BNB Chain, and Polygon, with a plan to cater to all EVM compatible blockchains.
Similar to the aforementioned social graph protocols, Cyberconnect aims to enable users to own their onlines social identities, content, and connections in a social network, while providing a composable social data layer for developers. While several third-party decentralized applications have integrated Cyberconnect in their stack, it is unclear which (if any) is the primary client that users utilize to avail social media features.
This is as the Cyberconnect protocol team has launched a product, Link3, that can be described as a Web3 Linktree. This can be seen in the traction figures for the Cyberconnect protocol.
The protocol claims to have ~30 applications utilizing Cyberconnect, and a total cumulative user count of ~750,000. Out of which, we can see below, the DAU, WAU, and MAU statistics, as of March 29th 2023.
Cyberconnect by the Numbers
As you can see, this is significantly larger than Farcaster and Lens. However, it is worth mentioning that it is unclear what is counted as an “active user”. If one link click on a Link3 site is considered an active user, these numbers may not represent the full reality of Cyberconnects traction.
DeSo (Decentralised Social), is a Layer 1 blockchain aiming to create and scale decentralized social applications. Founded by Nader Al-Naji, a former Google software engineer, DeSo has raised north of $200 million from the likes of Sequoia, Pantera, Coinbase, Andreessen Horowitz, and others. It aims to enable seamless storage and indexing for data rich applications (like social media) to exist on a blockchain. Once on-chain, companies can run nodes and create curated feeds, e.g., an ESPN feed, a Politico feed, etc. Thus, instead of being a black box accessible only by private companies, user data now gets stored on a public ledger.
Unlike other Web3 social networks discussed earlier, DeSo’s native coin (DESO), is already launched and publicly traded, most notably on Coinbase. The DESO coin’s primary function is most likely to maintain the security of the blockchain and incentivise user interaction on the chain. DESO can also be used to purchase native social tokens, NFTs, and tipping; these features effectively are the avenues of monetization for the content creators. While the specifics of the tokenomics are not explicit yet, especially considering the blockchain’s current migration from Proof of Work to Proof of Stake, we do know the total supply is fixed at 10.8M, out of which 10.6M is already in circulation.
With respect to traction, noteworthy applications being built on top of DeSo include Diamond (short form text app), Stori (short form video app), NFTz (NFT marketplace), and DAODAO (crowdfunding app). Further, as per OpenProsper, DeSo has ~2M wallets (public keys that have interacted with the DeSo blockchain) and ~1M creators (wallets with a username); these figures were ~50k at the start of March 2021. While commendable, it is worth noting that the metrics are loosely defined.
Wallets and Creators on DESO
A slightly better metric (although one that still has many weaknesses) to assess DeSo’s adoption is the number of transactions on the blockchain. Since March 2021, cumulative transactions have grown from ~150k to ~87M. From the graph below we can clearly see a plateauing of new transactions, which is likely indicative of the broader crypto slow down.
Number of Transactions on DESO
There are a plethora of other social protocols and standalone applications that are currently in development that we could not cover in this article, in the interest of time. These include protocols such as DeSo and SubSocial, and a longtail of applications like Minds, DTube, Peer, Chingari, Iris, Interface, SocialX, and many more – some who are trying to recreate existing Web2 services on Web3 rails (Eg: YouTube), and other creating entirely novel products (e.g.,: contextual and human readable blockchain explorer).
In conclusion, it is evident that Web3 social is in its infancy. Using conservative numbers, we’re looking at sub-100,000 monthly users; compared to multi-billions on Web2 social media platforms. Nonetheless, as alluded to earlier, there has been a Cambrian explosion of Web3 social apps within the last 12 months. As the winners emerge, it is likely we will see an increasing number of Web2 social users switch over to Web3. For this to happen, however, a few critical problems will need to be solved for.
- Inferior UI/UX: The two core pillars here are abstracting away Web3 technicalities and focusing on mobile native experiences. Only when applications have onboarding processes as simple as Web3 platforms, as well as them being accessible via smartphone, will we see adoption by the mainstream audience.
- Unscalable Infrastructure: Traditional social media generates upwards of 4 petabytes of data per day. Every minute, on Facebook alone, 510,000 comments are posted, 293,000 statuses are updated, 4 million posts are liked, and 136,000 photos are uploaded. Web3 rails simply cannot handle this kind of traffic, in its current avatar.
- Skeuomorphic Development: Today, the bulk of what is being built in Web3 social are replicas of Web2 applications. This is unlikely to catalyze migration of Web2 users to our new paradigm. We cannot rely on superior privacy protection, but will need novel use cases – those that Web2 cannot compete with – that will cause new users to try Web3 social.
- Token Design: The existence of tokens represents a significant and fundamental different tool the Web3 developer has in his toolkit. While the implementation of tokens has historically led to mercenary user behavior, cracking the code of customer acquisition and retention via tokens could expedite the migration to Web3 social. Phaver is one example of this. In addition, this is a key vector on which novel forms of monetisation could be explored as well, disrupting the traditional advertising-led model.
If these problems can be addressed, then over time we may see decentralized social explode to overtake the legacy social media world and all of its shortcomings.
Published on Apr 17 2023