Sports betting accounts for the vast majority of prediction market volume, upwards of 90% on some major platforms. The success of sports as the primary prediction market category is because the sports betting industry is deeply flawed. And while there’s clear opportunities outside of sports betting, the first killer submarket of prediction markets is changing the rigged nature of online sportsbooks and casinos that exploit users.

Ask any sports bettor and they explicitly know the game is stacked against them because, well, the house always wins.

The middlemen who collectively control the sports betting industry – “the house” – always win because they set the odds such that they always have a statistical advantage. Traditional platforms (e.g. DraftKings) actively ban users for winning too frequently. And users who aren’t banned are subject to the sportsbook changing odds at will and taking extractive “vigs” (fees on sports betting platforms).

All of this effectively amounts to a bad market structure with a heavy tax on participation.

Prediction markets create a better market structure for sports betting, by offering:

1. A peer-to-peer marketplace with global liquidity

2. Better odds

3. A market without exploitative vigs (e.g. fees)

Today, we’re excited to announce Pantera is leading the Series B investment in Novig, a sports prediction market that offers the best experience for consumers.

Novig is delivering a new peer-to-peer exchange with better consumer protections: people wager on sports and have higher probabilities of winning because they’re not betting against the house.

This exchange market structure results in better odds, where Novig users actually have 5% better odds compared to traditional platforms. Better odds translate to more opportunities for traders, which shows in the numbers – 23% of Novig users are profitable compared to 2% on traditional platforms. And, since Novig’s market is peer-to-peer, Novig users can even set their own odds for certain markets.

The sports betting market structure is reshaping in real time because of new prediction markets like Novig, which expand the total addressable user base for sports bettors by tapping into global liquidity and providing better consumer protections.

We’re thrilled to back Novig’s founders Jacob Fortinsky and Kelechi Ukah who founded the company with a core thesis: prediction markets offer a better experience for sports markets. The entire Novig team is a collection of sports and markets obsessed builders and bettors who are focused on creating the best sports prediction market for consumers.

Even the origin of the company’s name points back to the mission: removing the “vigorish” ’– the platforms cut to ensure guaranteed profits – and offering users a ‘no vig’ experience. And so, unlike traditional sports books that ban profitable bettors and profit on ensuring consumers lose, Novig aligns its growth with that of its users, scaling with the success of their customers.

The “house” always wins because it owns the rules. Novig’s betting that they can change the rules, replacing the house with a global peer to peer market. We’re making the same bet.

For sports prediction market fans, download the Novig app today or listen to the Inside Prediction Markets Podcast by Novig. And if you’re obsessed with markets, sports, and want to challenge the house, Novig is hiring

Subscribe now

Information presented is for informational purposes only and does not intend to make an offer or solicitation for the sale or purchase of any securities. All investments involve risk and unless otherwise stated, are not guaranteed. There can be no assurance that this investment will provide a return for the investing fund. Be sure to consult with a tax professional before implementing any investment strategy. Investment Advisory Services offered through Pantera, a Registered Investment Advisor with the U.S. Securities & Exchange Commission. Registration does not imply a certain level of skill or training. Past performance is not indicative of future results.

“Capital markets are the story we tell ourselves about the future—and the collective bet we place on it.”

— Anonymous

The original primitive of Ethereum was permissionless capital formation and launching tokens in the form of initial coin offerings (ICOs).

Over the past decade that primitive – permissionless capital formation – has evolved multiple times over. ICOs were effectively permissionless crowdsales, but lacked programmability and robust liquidity management. The most recent iteration of capital formation has been around launchpads like Pump which pioneered a simple mechanism to launch tokens and bootstrap liquidity that has led to a resurgence of permissionless capital formation.

Subscribe now

Capital markets are iterative systems, constantly evolving towards efficiency. The next phase of capital markets is the inevitable shift towards being internet-native and programmable.

Today, we’re excited to announce our Seed Round investment in Whetstone, the makers of Doppler Protocol, a custom capital markets protocol for asset issuance and trading. Several projects, including Zora, Paragraph, Cooprecords, and Ohara among others have leveraged Doppler Protocol to create tokens for new asset classes around content and AI generated applications.

Within its first several months, Doppler has launched 5.5 million tokens, creating an estimated over $1.5 billion in value with assets that have done over $1 billion in trading volume. Doppler provides capital markets infrastructure for companies and token launchers to re-invigorate Ethereum’s founding use case – onchain IPOs.

The Path Towards Onchain IPOs

The traditional IPO process has long been criticized for its high costs, inequitable access, and misaligned incentives that result in predative underpricing and inaccurate price discovery. Direct listings and SPACs emerged as avenues to re-direct capital formation to avoid the issues of IPOs, but to limited success and not without introducing their own disadvantages.

Crypto’s capital formation and listing process has similarly faced issues with capital fundraises often disconnected to liquidity creation (e.g. token generation events/listings).

Initial exchange offerings and exchange listings provided projects distribution and access to broader capital, but have devolved into a game of backroom deals and pay to play structuring that has driven poor incentive alignment and subpar outcomes for all those involved.

Pump’s ICO – which raised $1.3B in total with $600m of it onchain in under 12 minutes – offers a glimpse into a better system, one where onchain IPOs are not only possible, but advantaged over traditional IPOs.

While exchanges were able to purchase tokens on behalf of their users, 75% of participants went through Pump’s auction page versus a centralized exchange. But, it wasn’t without its drawbacks. Of the ~24,000 KYC participants, only 42% of wallets were actually able to purchase tokens, in part due to some users being unable to get purchase orders filled in time. A future where more IPOs are onchain requires more customization and programmability around capital formation and token issuance.

Doppler enables projects to launch tokens with more functionality, including customizing vesting terms, controlling participant allocations, preventing market maker sniping, and improving price discovery. Even more notably, tokens launched through Doppler have the ability to monetize their token’s core asset – the trading fees. By tokenizing the LP position of a token and assigning it to the token’s treasury, the token LP becomes a revenue generating asset for the project. Instead of exchanges benefiting exclusively from token appreciation, tokens can monetize their own volatility.

Whetstone, and Doppler, are founded by Austin Adams. Prior to Whetstone, Austin was a core protocol researcher at Uniswap Labs and helped design Uniswap V4. Before joining Uniswap, Austin worked as a researcher at the Federal Reserve where he focused on global capital markets.

We’re incredibly excited to back Austin and the Whetstone team. Austin has deep expertise in mechanism design and building financial markets protocols. The Whetstone team is a collection of markets obsessed builders and operators, focused on innovating on the structure of internet capital markets.

Doppler is a story on the future of custom capital markets.

If you’re a team looking to launch onchain, book a 15 minute discovery call to learn more about how Doppler can help.

Subscribe now

Information presented is for informational purposes only and does not intend to make an offer or solicitation for the sale or purchase of any securities. All investments involve risk and unless otherwise stated, are not guaranteed. There can be no assurance that this investment will provide a return for the investing fund. Be sure to consult with a tax professional before implementing any investment strategy. Investment Advisory Services offered through Pantera, a Registered Investment Advisor with the U.S. Securities & Exchange Commission. Registration does not imply a certain level of skill or training. Past performance is not indicative of future results.

Hey everyone – I wanted to share that I’ve started a podcast. There are many crypto podcasts out there, but this one is mine.

I’ll still continue writing and sharing my thoughts through this newsletter, but if you want to hear some of my less filtered ideas, I’ll be sharing some of that as I chat with guests on the podcast.

There’s three podcasts that have dropped so far, dropping the links below for your viewing pleasure. If you have a minute, please like, subscribe, or give a rating on Spotify.

Episode 1 with Paul Veradittakit and Franklin Bi

• Paul Veradittakit is a Managing Partner at Pantera Capital

• Franklin Bi is a General Partner at Pantera Capital

Watch on YouTube

Listen on Spotify

Episode 2 with Alana Levin and Hootie Rashidifard

• Hootie Rashidifard GP and Founder Hash3, an early stage crypto fund.

• Alana Levin a former colleague and investor at Variant, another early stage crypto venture fund.

Watch on YouTube

Listen on Spotify

Episode 3 with TulipKing and Brian Breslow

• Mikey Kremer AKA Tulip King – Head of Strategy at CounterpartyTV, arguably the most popular entertainment financial channel in crypto.

• Brian Breslow an investor at No Limit Holdings, a crypto venture fund.

Watch on YouTube

Listen on Spotify

For anyone that’s ever bet on an NFL game, you immediately realize that it feels like a levered form of entertainment.

The addition of financial gain or loss, amplifies the entertainment experience. This same experience, often amplified by volatility, is what has propelled prediction markets, memecoins, and the trend towards broader financialization of everything.

But, today, markets and entertainment are mostly unbundled. That is to say, where you bet and where you get entertainment are decoupled.

Discovery is distinct from trading. People see what they want to buy on Twitter (I’m still not calling it X), see a friend share a memecoin in a group chat, or hear about an asset on a podcast and then open their brokerage account to slam the “market buy” button, slippage and fees be damned.

The next great bundling will be of media and markets. Entertainment and finance. Livestreams and leverage. Culture and prediction markets.

As more assets move onchain and internet capital markets mature, interfaces and front-ends will have a broader asset universe and more financial products that they can offer to their users.

Twitter could easily be a venue where $TICKERs have integrated purchases options or embedded sports betting within social networks powered by underlying prediction markets.

It’s worth noting that in some instances the orderbook is the entertainment – watching an asset go up or down elicits emotional reactions that when coupled with social experiences (comments, follows, socials, etc.) breed higher entertainment value.

Internet capital markets will bundle the moment of attention and the moment of monetization. The feed merges with the terminal. Culture can produce live events and shows that are interwoven into the trade. Media and markets, once separated, will now forever be intertwined.

Subscribe now

The first employee hired at eBay was for someone to open checks.

eBay was so successful out the gate that it needed a full time employee just to help open and process how much money the company was making. Name a better first-hire, I’ll wait.

Subscribe now

In the beginning, eBay acted as the middleman, processing checks that people sent in exchange for goods. Naturally, an online payments system emerged for the platform, with PayPal – outcompeting even eBay’s own subsidiary, payment service product Billpoint – eventually getting acquired for $1.5 billion by eBay. Chef’s kiss.

And since the internet kickstarted the online payments arms race, the physics of finance have demanded that payments get faster and cheaper. Each new payment innovation – whether ACH, SSL checkout, QR codes, point of sale solution, APIs-first product, mobile checkout, or RTP networks, stablecoins – is driving towards the same end state. Faster and cheaper payments.

However, while the physics of finance pushes the payments system towards efficiency, it has also introduced its own entropy, creating a patchwork of payments networks each with another layer of integration complexity without solving fundamental issues around settlement speed, fragmentation, or risk management. Enter stablecoins and Coinflow.

Today, we’re excited to lead the Series A investment in Coinflow, the payments infrastructure for modern money.

Coinflow provides a unified payment stack with instant settlement and fraud protection, solving issues that legacy providers haven’t addressed.

One of the causes of this patchwork payment system is that global payment rails and networks have acted as walled gardens, fragmented because of jurisdictional boundaries. For example, the U.S. is unlikely to integrate with Brazil’s payment infrastructure, PIX, and vice versa because of national interests. Stablecoins promise an interoperability layer across nations and networks, while also offering a remarkably cheaper and faster transaction experience than legacy payment infrastructure.

Today, Coinflow offers a one-stop-shop payments solution for marketplaces, fintechs, remittance providers, gaming platforms, and ecommerce merchants worldwide. Their product offering consists of fiat and crypto payment options for end consumers, stablecoin payout options for merchants, and an in-house fraud detection model and chargeback protection product for merchants.

The founding story of Coinflow is not dissimilar from that of PayPal, which realized that accepting money over the internet was hard but that it shouldn’t be. While building his previous company, Cofounder and CEO,Daniel Lev, realized that accepting stablecoin payments was hard because money rails aren’t connected, but should be. Since Coinflow’s seed round in 2024, Coinflow has achieved 23x revenue growth, expanded payment coverage to 170+ countries, and reached a multi-billion-dollar annual transaction volume run rate.

The future of money is programmable and over the next decade stablecoins and blockchain rails will rebuild and connect global payments infrastructure. The physics of finance demands cheaper and faster transactions. And modern money – stablecoins – demand a modern payment solution, just as the internet demanded a new payment stack. Coinflow provides the payment infrastructure for modern money.

If you’re a merchant looking to elevate your payments technology, Coinflow is your ideal growth payments partner —book a demo today!

Share

Subscribe now

Around 25% of Solana transactions fail to execute. Translation failure rates exceed 40% during times of peak congestion.

Solana’s future as the ecosystem for Internet Capital Markets requires the ability for applications to guarantee transaction execution for users.

Today, we’re excited to announce our Seed Round investment in Raiku, a marketplace for blockspace on Solana, in partnership with Jump Crypto, Lightspeed Faction, Reciprocal Ventures, Initial Ventures and Hashkey Capital. Raiku already has partnerships with some of the most trusted validators in crypto, including Figment, Kiln, Staking Facilities, Blockdaemon, Chorus One, H2O Nodes, among others.

Although Solana will continually strive to be faster, its high demand for transactions will likely continue to outstrip its available supply of blockspace. Raiku emerges as a solution to offer a more reliable transaction experience for applications and users.

Raiku offers a two-sided execution marketplace and blockspace market that aligns incentives for developers, validators, and institutions.

Raiku consists of a few core components

• Coordination engine – power slot auctions, transaction scheduling, and guarantees transaction inclusion

• Slot marketplace – The marketplace for Solana blockspace

• Raiku sidecar – Native client integration with Solana validators

• Global accounts – Support for external execution environments and shared state

By creating a dedicated coordination network for transactions, Raiku can enable validators to offer pre-confirmations, or the ability to pre-book blockspace. With new transaction types, validators gain new non-MEV revenue mechanisms and the ability to earn more for high priority transactions. In exchange, institutions, applications, and users will enjoy a better transaction experience that is more reliable and consistent.

Solana’s vision for Internet capital markets demand speed and transaction certainty. Raiku offers high-performance block building, using Solana’s core validators to ensure fast, predictable transactions with guaranteed block inclusion.

We’re incredibly excited to back Robin and the Raiku team as they re-imagine the future of blockspace market structure on Solana. If you’re interested in working with Raiku, reach out to build with Raiku or become a validator.

Share

Share

As crypto moves towards a world of real-world applications, interacting with offchain, real world data is critical. As a result, applications today are cobbled together by external middleware that connect blockchains and offchain services (i.e. oracles, RPC nodes, indexers, etc).

Smart contracts can’t see outside data, which is why devs have to pay to ingest offchain data. More importantly, smart contracts, while powerful, aren’t able to write or read to APIs which limits some of their functionality.

Rialo reimagines the approach by building a vertically integrated full-stack solution that can natively access web2 APIs.

But why does vertical integration of offchain middleware and the native access to web2 APIs matter?

Integrating offchain middleware into the blockchain layer provides two core benefits:

• Reduces the tax on developers from a UX and cost perspective

• Expands the market of applications that can be built on top of blockchains

Vertical Integration of Offchain Middleware: Better For Builders

For any application built on blockchains, paying just two offchain middleware vendors can destroy a significant percentage of an app’s margin. This directly makes it harder for developers to build and pushes costs upstream to users.

Each hop down this stack extracts its own margin for its service. Modern applications depend on a variety of services – oracles, indexers, schedulers, bridges, relayers, block builders – each with its own uptime target, incentive model, and rent. These service providers are the new normal for developers and aren’t going away, but they can get cheaper and easier to integrate into applications.

If we look at the established middleware products we can see how extractive this squeeze is. To take an example, for an onchain lending protocol, the additional cost of recurrent liquidation checks are:

• 300% extra for scheduling

• 200% extra for the data

• $4k/month for indexing for fast results

That is to say, if the protocol charges $1 for gas, the app pays an additional $6 for offchain service provision. This is before even carrying out the liquidation! Lenders like Aave pay around 5% of the loan to liquidation service providers. Keeping to Aave as an example, the mean loan is ≈ $100 k ( $9B total borrows / 90k monthly borrowers) and so a 5 % liquidation cost adds an additional ~$5,000 per liquidation.

It isn’t always this expensive, some of these functions are periodically offered by newer, less mature service providers who subsidize costs to gain market share. But this is only a temporary effect: we don’t pretend that Uber or Doordash will always cost $5 just because they price that way when they launch in a new city.

When functions like scheduling and native web‑calls are bundled directly into the base protocol — Rialo’s approach — the stacked middleware mark‑ups disappear and the developer experience improves.

Expanding the Universe of Integrated Applications

As blockspace continues to commoditize, the differentiators become less about performance (TPS) and more about developer experience, distribution, and application specific capabilities.

While an ideal world is one where all activity happens natively onchain, the real world is messy and bringing more assets and value propositions onchain requires access to third-party APIs and offchain data. Blockchain networks have historically pushed these decisions towards middleware solutions rather than internalizing how their networks interact with offchain data and third-party APIs.

Without a seamless interface to real‑world data and services, even the fastest blockchains will leave apps feeling disconnected and higher friction for users.

With Rialo’s integrated approach, a new class of applications becomes both possible and seamless to implement:

• Onchain apps based on web2 API data: Apps with Zapper-like triggers (e.g. email action to transaction) that combine offchain data sources with onchain actions.

• AI orchestration: AI agents that can transact efficiently with better indexing and access real-time information

• DeFi oracle integrated apps: Oracle integrated you can make liquidation logic programmable at the app level. No need for the liquidation bot tax

• Enterprise-grade automations: enable transactions triggered by CRM or email events.

• RWA-enabled applications: apps that leverage real-time data from traditional financial markets (e.g. interest rates,

• Blockchains with embedded social data: Web2 API integrations that unlock social data for next-gen crypto applications.

Investing in Rialo: The Vertically Integrated Blockchain

At Pantera, we’re excited to lead the $20 million funding round for Subzero Labs as they build Rialo. The Rialo team is especially situated to build an application specific blockchain. Cofounders Ade Adepoju and Lu Zhang were both early engineers at Mysten Labs who contributed to the Sui network. Beyond helping build Sui, Zhang built Diem, Meta’s crypto initiative, and large-scale AI/ML infrastructure at Meta and Google. Meanwhile, Adepoju developed distributed systems at Netflix and advanced microchips at AMD.

With years of deep involvement in multiple layer-1 protocols and decentralized applications, Adepoju and Zhang have arrived at a vertically integrated approach that abstracts middleware and lays the foundation for new types of real world applications.

Special thanks to Matt Stephenson and Lauren Stephanian for comments and insights in this essay.

Subscribe now

As interest rates come down, DeFi yields start to become more attractive. Increased volatility brings more users, yield, and leverage. Combined with more sustainable yields from RWAs, it suddenly becomes much easier to build consumer-grade crypto financial apps.

When we merge these macro trends with innovations around chain abstraction, smart accounts/wallets, and the general shift to mobile, there’s a clear opportunity to build consumer-grade DeFi experiences.

Some of the most successful crypto financial apps over the past few years have been at the intersection of improved UX and speculation.

• Trading bots (e.g., Telegram) – offer users trading within their messaging and social experience

• Better crypto wallets (e.g., Phantom) – improve the existing wallet experience and offer a better experience across multiple chains

• New terminals, portfolio trackers, and discovery layers (e.g., Photon, Azura, Dexscreener, etc.) – offer advance features for power users and give users access to DeFi in a CeFi like interface

• Robinhood for memecoins (e.g., Vector, Moonshot, Hype, etc.) – to date crypto has mostly favored desktop, but mobile-first experiences will dominate future trading apps

• Token launchpads - (e.g., Pump, Virtuals, etc.) – offer access to permissionless token creation for anyone, regardless of technical capabilities.

As more consumer DeFi applications launch, they’ll look like fintech apps with the standard UX that users have come to appreciate, but they’ll aggregate and provide an opinionated experience of DeFi protocols on the backend. These apps will be opinionated about the discovery experience, the products offered (e.g., types of yield), appeal to power users (e.g., offer convenient features like multi-collateral leverage), and generally abstract the complexities of interacting onchain.

The RWA Flywheel: Endogenous vs Exogenous Growth

High interest rates since 2022 have supported a massive influx in real-world assets (RWAs) onchain. But now, the transition from offchain finance to onchain finance is accelerating as large asset managers like BlackRock realize that issuing RWAs onchain brings meaningful benefits including: programmable financial assets, lower cost structure for issuing and maintaining assets, and greater asset accessibility. These benefits, like stablecoins, are a 10x improvement over the current financial landscape.

According to RWA.xzy and DefiLlama, RWAs account for 21-22% of assets on Ethereum. These RWAs are mostly in the form of grade-A, American Eagle-backed, U.S. Treasury bills. The growth has largely been driven by high rates that make it easy for investors to be long the Fed over DeFi. And while the macro winds are shifting to make T-bills less appealing, the Trojan horse of onchain asset tokenization has entered the walls of Wall St, opening the floodgates for more RWAs to come onchain.

As more traditional assets move onchain, this will kick off a compounding flywheel effect, slowly merging and replacing legacy financial rails with DeFi protocols.

Why does this matter? The growth of crypto comes down to exogenous capital vs endogenous capital.

Most of DeFi is endogenous – largely circular within the DeFi ecosystem – and capable of growing on its own. However, it has historically been quite reflexive: it goes up, it goes down, and then back around. But over time, new primitives have steadily expanded the DeFi pie.

Onchain lending via Maker, Compound, and Aave expanded the use of crypto-native collateral as leverage.

Decentralized exchanges, in particular AMMs, expanded the universe of tradable tokens and jumpstarted onchain liquidity. But DeFi can only grow its own market to a certain extent. Although endogenous capital (e.g., the speculation of onchain assets) has driven the crypto markets to a robust asset class, exogenous capital – capital that exists outside of the onchain economy – is necessary for the next wave of DeFi growth.

RWAs represent a massive amount of latent exogenous capital. RWAs – commodities, stocks, private credit, FX, etc. – present the greatest opportunity to expand DeFi beyond the circular siphoning of capital from the pockets of retail to the loins of traders. Just as the stablecoin market has required growth via more exogenous uses beyond onchain financial speculation, so too will other DeFi activities (e.g., trading, lending, etc.).

The future for DeFi is for all financial activity to move onto blockchains. DeFi will continue to see two parallel expansions: the similar endogenous expansion via more onchain native activity, and exogenous expansion from real-world assets moving onchain.

The Platformization of DeFi

“Platforms are powerful because they facilitate a relationship between third-party suppliers and end users." – Ben Thompson

Crypto protocols are about to have their platform moment.

DeFi applications are all moving towards the same business model, evolving from standalone application-protocols to fully-fledged platform-protocols.

But how exactly are these DeFi apps becoming platforms? Today, most DeFi protocols are relatively rigid, offering a one-size-fits-all service for applications that want to interact with these protocols.

In many instances, apps simply pay the protocol for its core asset, such as liquidity, as a standard user would, rather than having the ability to build differentiated experiences or programmed logic directly within the protocol.

Most platforms start this way, solving a core problem for a single use case. Stripe initially provided payment APIs that allowed individual businesses, like online stores, to accept payments on their websites, but it only worked for single businesses. Once Stripe launched Stripe Connect, it enabled businesses to process payments on behalf of multiple sellers or service providers, making Stripe the platform that it is today. It later offered better ways for developers to build more integrations, expanding its network effects. Similarly, DeFi platforms like Uniswap are now moving from standalone apps (e.g., DEX) that facilitate swapping to building a DeFi platform that enables any developer or app to create their own DEX on top of Uniswap’s liquidity.

The key enabler of DeFi’s platform transition is a shift in business model, with the evolution of singleton liquidity primitives.

Singleton liquidity primitives – Uniswap, Morpho, Fluid – aggregate liquidity for a DeFi protocol allowing access by two modularized pieces of a value chain (e.g., liquidity providers and applications/users). The liquidity provider experience becomes more streamlined, allocating capital to a single protocol rather than protocol of differentiated pools or isolated vaults. For applications, they can now rent liquidity from the DeFi platform rather than simply aggregating the core service (e.g., DEX, lending, etc.).

Some examples of the emerging DeFi platform protocols:

Uniswap V4 is pushing towards a singleton liquidity model whereby applications (e.g., hooks) can rent liquidity from Uniswap's V4 protocol rather than simply routing liquidity through the protocol as with Uniswap V2 and V3.

Morpho has moved towards a similar platform model where MorphoBlue acts as the core liquidity primitive layer that can be permissionlessly accessed by vaults created through MetaMorpho, a protocol built on top of the liquidity primitive MorphoBlue.

Similarly, Instadapp’s Fluid protocol creates a shared liquidity layer that's leveraged by its lending and DEX protocols.

While there are differences between these platforms, the common throughline is that emerging DeFi platforms share a similar model, building singleton liquidity contract layers with more modular protocols on top that enable greater application flexibility and customization.

The evolution of DeFi protocols from standalone applications to fully-fledged platforms marks a maturation of the onchain economy. By adopting singleton liquidity primitives and modular architectures, protocols like Uniswap, Morpho, and Fluid (formerly Instadapp) are unlocking new levels of flexibility and innovation. This transition mirrors how traditional platforms, like Stripe, empowered third-party developers to build on top of core services, driving greater network effects and value creation.

As DeFi enters its platform era, the ability to facilitate customizable, composable financial applications will become a defining feature, expanding the market of existing DeFi protocols and enabling a new wave of apps to build on top of these DeFi platforms.

Stablecoins are a trillion dollar opportunity.

That is not hyperbole.

While crypto is often thought of for its volatility, tokens, and liquidity-profile, the other side of the crypto barbell that more silently carries the banner for crypto adoption is stablecoins. For folks that are new here, these are cryptodollars that are pegged 1:1 with their underlying fiat using either algorithms (less popular) or reserves (more popular) to maintain the peg.

Stablecoins have gone from 3% of blockchain transactions in 2020 to now consistently representing over 50% of blockchain transactions. Stablecoins are crypto’s killer value proposition and unlike a lot of crypto are non-speculative in nature.

In a short period of time, stablecoins have showcased their ability to be one of the transformative innovations within crypto. And 2024 has been a breakout moment for stablecoins, transacting over ~$5 trillion in adjusted volume, over $1 billion transactions, across nearly 200 million accounts.

Stablecoins witnessed impressive growth during crypto’s last bull cycle, but this time around, stablecoins are being used beyond the DeFi ecosystem. Over the past few years, stablecoins have demonstrated the core value proposition – seamless cross-border payments, initially through access to U.S. dollars. Correspondingly, the geographies that see the greatest stablecoin growth are in emerging market corridors where access to dollars is in high demand.

Stablecoins offer a 10x value proposition to the traditional payment rails across both B2C payments (e.g. remittances) as well as in B2B cross-border transactions.

Cryptocurrencies have long promised a solution for the trillion dollar cross-border payment market. In 2024 there will be ~$40 trillion cross-border B2B payments made via traditional payment rails (excluding wholesale B2B payments) (Juniper Research). Within the consumer payment market, global remittances account for hundreds of billions in annual revenues. And now, stablecoins offer the means by which to make global, cross-border remittance payments on crypto rails a reality.

Amidst this more rapid adoption of stablecoins among both B2C and B2B payments, the supply of stablecoins onchain and transaction volume are reaching all-time highs.

The Trifecta: Better. Faster. Cheaper.

There’s an old adage in business – It’s rare that a product comes along and is able to offer something that is better, faster, and cheaper. Oftentimes, the product can be two of those things, but not all three. Stablecoins offer a better, faster, and cheaper way to move money around the world.

Stablecoins offer a 10x value proposition to traditional dollars for both businesses and consumers.

Better: Stablecoins offer a more accessible product, one that’s available 24/7, 365 days a year. They can be transmitted globally, across borders with ease and offer a programmability that makes stablecoins a superior product to fiat.

Faster: Stablecoins are unquestionably faster, settling instantly versus T-minus 2 or T-minus 1 days to settle.

Source: BVNK

Cheaper: Stablecoins are cheaper to issue, transfer, and maintain than fiat. In 2023, Stripe facilitated over $1 trillion in payments volume with a fee structure that starts at 2.9% with an added 30 cent charge for domestic card-based transactions. On high throughput blockchains like Solana or Ethereum L2’s like Base, average stablecoin payments cost less than a cent.

The Emerging Stablecoin Stack

Although the stablecoin stack is ever-evolving, there are emerging layers:

• Merchant Layer – Applications and interfaces that originate retail or business transactions

• Stablecoin Orchestration – Providers that offer last mile on and off ramp, virtual accounts, cross-border stablecoin transfers, or stablecoin to fiat conversion

• FX and Liquidity – Providers that offer cross-border stablecoin interchange to other USD-pegged stablecoins, fiat, or regional stablecoins.

• Stablecoin Issuance – Companies or protocols that offer white-label stablecoins or first-party stablecoins with differentiated features

Similar to how crypto exchanges emerged in every corner of the world to cater to local participants, we anticipate a variety of crypto crossborder applications and processors to emerge as they cater to specific stablecoin corridors.

Just like in traditional finance and payments, building moats at each part of the stack is important to expand business opportunities beyond the initial value proposition. We have given some thought to which moats are defensible and can expand through time for each layer:

• Merchant Layer – Moats are made by owning a user’s or business’s stablecoin flow. This presents the opportunity to upsell other services, sell that users flow, and own the customer experience end to end. The Robinhood of stabledollars will emerge following a similar playbook.

• Stablecoin Orchestration – Licenses! Who will have the most reliable, global coverage at the cheapest rates. Will it be developer friendly? Look at the Stripe x Bridge acquisition to where and how moats form here.

• FX and Liquidity – Liquidity begets liquidity, flows beget value accrual. Any player that can source proprietary liquidity and price it effectively will outperform newer entrants without it. This is why some large exchanges service most of the stable flows in certain major corridors today. We also believe that a transition from OTC style FX to exchange style FX to onchain style FX will facilitate faster payment and transactions at this layer.

• Stablecoin Issuance – Issuance will commoditize over time and lead towards the inevitable launch of dozens of large brand stablecoins (e.g. PYUSD). As other layers of the stack grow (i.e., Merchant, Orchestration, and Liquidity) we expect these to be well positioned to launch their own stablecoins whether to capture yield, build their branded stablecoin, or build proprietary stablecoin liquidity and flows.

These layers merge over time as the layers of the stack get bundled. The merchant layer is in the best position to aggregate other layers of the stack to provide additional value to end users, increase margins, and create additional revenue streams. They will have the power to choose which FX to do themselves, which on/offramp rails to own or rent, and which issuers to use.

Moreover, we anticipate that stablecoin issuance will become increasingly common for large fintechs and ecommerce providers that facilitate significant flows. The next generation of neobanks and fintechs will be defined by stablecoins. As recently as this month we have heard interest in large card networks like Visa, banks like JPM, and asset managers like Blackrock to explore stablecoin projects of their own.

Looking Ahead: The Next Decade of Digital Dollars

The tokenization of dollars is still in its infancy.

Even as stablecoin MAUs reach ATHs, we believe adoption will continue as hundreds of millions of people interact with stablecoins over the next decade.​

Importantly, stablecoin users continue to rise even amidst the volatility of exchanges volume. From bull to bear, stablecoins prevail and expand their digital reach.

As crypto rebuilds the financial system from the ground up, stablecoins exist in parallel, integrating into the traditional financial payments network.

While large players like Stripe, Visa, and Paypal have all entered the stablecoin market, we see abundant opportunity for new stablecoin focused protocols and companies.

Some ideas we’ve been excited about:

• Stablecoin Neobank – The advent of mobile enabled Neobanks to accrue significant value. Crypto Neobanks will not only provide best in class payment rails, but they’ll enable the next generation of consumer financial apps that aggregate payments, trading, yield, lending, and other core financial services.

• Onchain FX – while today most stablecoins are USD-pegged, we expect more currencies to come onchain, enabling an onchain FX layer to develop. More immediately, with the abundance of USD-pegged yield bearing stablecoins offering varying yields and value propositions, we anticipate a need for a FX layer for these initial USD-pegged stablecoins.

• Telegram payment rails – Telegram offers a native wallet for payments, but we also see a unique opportunity to leverage new interfaces like TG mini-apps to build a new payment layer on top of Telegram.

• Remittances on Crypto Rails – Remitly. Wise. Intermex. Ria. MoneyGram. Western Union. All remittance companies, each with hundreds of millions to billions in annual revenues. Remittance companies take fixed fees which are meaningful on low dollar amounts (e.g. $6 on a $60 dollar transaction) or high fees 30-100bps per transaction. Stablecoins reduce the cost to send money globally and make the process seamless. Money. “The remittance margin is the stablecoin opportunity.” – Jeff “Stables” Besos

• Global Venmo – Building P2P rails to bring Venmo like capabilities onto a global scale. Remittance is typically a one sided flow, this would be serving the social commerce use case in a more two sided flow.

• Stablecoin powered treasury management and operations – As the fintech landscape expanded from payments with PayPal, it created multi-billion dollar opportunities across wealth management, personal finance, payroll, business spend and expense management, Neobanks, financial accounting and reporting, lending/mortgages, among others. Similarly, stablecoins offer an opportunity to rebuild many of these tedious processes with better rails, powered by stablecoins. In the near term, treasury management and operations deal with complex operations which makes the value proposition for stablecoins ripe for disruption.

Closing Thoughts

Stablecoins represent a trillion dollar opportunity. We’re looking to back the founders and visionaries who see what stablecoins can be, unburned by what the financial system has been.

If you are a founder building stablecoins, a current operator at a traditional fintech or payments company interested in working with stablecoins, or anyone with views on the space – please don’t hesitate to reach out to ryan.barney@panteracapital.com and mason@panteracapital.com.

In addition, Pantera is looking to host stablecoin focused events in SF and New York over the next year. Spots are limited but if you are interested please sign up here.

Subscribe now

As a consequence, one of the most common and successful business models in crypto is the exchange model. If you’ve been in crypto for some time, this isn’t necessarily surprising, but I think it's often under-appreciated when evaluating crypto companies and protocols.

Exchanges can appear in a variety of instances but generally create a venue for transactions for some asset or service, and a mechanism to process those transactions.

Nowadays, I try to ask the question – how could this protocol or business become an exchange? For certain businesses it’s quite clear, but for others, it requires some imagination. 

Subscribe now

One general rule of thumb is – the closer proximity of the app to the transaction, the more likely that a budding app can blossom into an exchange. This is intuitive for applications that perform transactions themselves (e.g. swaps) but is also true of applications that sit in privilege positions of the stack (e.g. are in the flow of value) or that start as SaaS products but can grow into platforms with robust marketplaces. 

Exchanges are well-positioned to develop in scenarios were: 

• New assets emerge onchain

• Apps control distribution and can introduce transactional behavior 

• New services emerge that impact valuable onchain state or are somehow connected to transactions 

• Crypto games control their own asset issuance and have open economies

• Developer platforms can introduce service marketplaces or auction houses for transactions

When new assets emerge onchain

This is probably the most obvious instance in which an exchange is ripe to emerge. Coinbase created a marketplace for bitcoin. Early entrants like Etherdelta showcased the market demand for trading long tail tokens, although second mover Uniswap emerged as the dominant venue. The list goes on for both newly created assets NFTs (i.e., Superrare, Opensea) and for derivative assets like stablecoins (e.g. Curve) and perps (i.e., Deribit, dydx, etc.)

To be sure, not all exchanges succeed or have similar moats. Some new assets such as ERC 1155s or Bitcoin ordinals have been best served by incumbent exchanges (e.g. Magic Eden) rather than newcomers. Exchanges, based on new assets, are typically most defensible when they have a combination of liquidity network effects, control end user distribution, or are the privileged issuer of the underlying assets that they trade. 

Apps that control distribution 

Generally speaking, controlling end user distribution enables an application to level up into an exchange. Phantom and MetaMask are wallets, but the advent of MetaMask Swaps turned all wallets into exchanges. Telegram bots are exchanges, but located in a more convenient place for users – their DMs. Social apps like Farcaster, Lens, and Unlonely offer built in transactions for exchanging assets – memecoins, NFTs, points, etc. – at the point of distribution. Similarly, Solana blinks now offer a model for any app to implement in-app transactions and become exchanges. 

In the future, I expect more distribution focused apps to offer in-app transactions leveraging primitives similar to Solana blinks, Farcater frames, or Lens open actions. Historically, web2 apps were forced to direct users to transact on other platforms, but now blockchains can offer transactions at the point of distribution, meaning all apps can be marketplaces. 

Marketplaces for new services 

Service providers that impact valuable onchain state or are somehow connected to transactions, have the ability to become exchanges. For instance, oracles bring external data (e.g. stock pricing data) onchain, which can impact the onchain price of other assets, trigger liquidations, or result in arbitrages that generate instances of oracle extractable value (OEV). Oracle providers like Pyth step into an exchange model by creating auctions that bundle transactions with oracle updates and let people pay for priority in back running transactions. Bridging and cross chain interoperability services naturally act as exchanges, by similarly controlling pricing of assets 

More emergent are ZK proof marketplaces and services that must be submitted on-chain. As a result, ZK proof marketplaces (e.g. Gevolut) and aggregators (e.g. Nebra) will have to compete with transactions and other proof generators for scarce blockspace. While the proof service landscape is emerging, the leading providers will benefit from economies of scale (more proofs → more aggregation → cheaper proving – more proofs) which will potentially allow these service marketplaces to become valuable exchanges.  

Crypto games

One notable difference between web2 gaming and web3 is the shift towards open economies with transferable assets. Since crypto games like Axie often control asset issuance and velocity (e.g. turnover) they can verticalize their own in-game exchanges or otherwise act as asset brokers. This naturally provides web3 games the opportunity to introduce an exchange business. 

Developer platforms as marketplaces or auction houses 

Developer platforms are a little less obvious, but benefit from economies of scale and possess privileged positions in the stack that put them in close proximity to transactions.

Rollup as a service providers (RaaS), such as Conduit and Caldera also share the ability to insert themselves into the value flow by creating their own sharded sequencers that become a focal point for transaction ordering for RaaS supported chains. While shared sequencers can certainly order blocks themselves, they can alternatively auction transactions and capture MEV. 

Another core piece of developer infrastructure are wallet as a service providers (i.e., Dynamic, Crossmint, Privy) which have the ability to connect user funds across apps providing a sticky user and developer experience. With this lock-in, WaaS providers offer in-app transactions, swaps for on and off ramp, and possibly offer similar wallet swap plug-in features. 

As the crypto economy grows and all assets move onchain, we’ll undoubtedly see many more exchanges emerge. 

Exchanges are everywhere, some in plain sight, some waiting to be uncovered. 

If you’re building a new type of exchange, my DMs are always open. 

Thanks to Franklin Bi and Zaid Fattah for feedback on this essay.

I’m excited about the opportunity to work alongside Paul, Dan, Franklin, Lauren, Ryan, Nihal and the rest of the Pantera team to continue supporting world class founders. 

Why Pantera 

As the first dedicated crypto venture fund, Pantera has been backing founders for over a decade, investing and supporting the top founders in crypto. The Pantera team has made significant steps to drive the industry forward towards mainstream adoption and I’m excited to do my part in helping Pantera shape the future of crypto. And perhaps most notably, Pantera has assembled a world-class team and track record of backing founders building out the future of web3.

Prior to joining Pantera I was an investor at Variant, an early stage crypto fund. Before investing, I spent time as a Sr. Research Analyst at Messari as well as on the operational side as a marketer for ConsenSys and its portfolio companies. 

Throughout my career, from marketer to analyst to investor, I’ve always focused on the application layer of the crypto stack. Because of inherent social coordination principles of crypto networks, I firmly believe that applications in crypto will shape the nature of how people interact, which is why my initial focus at Pantera will revolve around investing in the next generation of crypto apps that onboard tens of millions of users. Still, I’m always open to connecting with individuals building critical infrastructure that will serve the crypto apps of the future. 

If you’re a builder, please don’t hesitate to reach out to me at mason@panteracapital.com.

Marketplaces and networks that launch with a token from the jump (or before establishing sufficient organic demand), must find PMF in a shortened window or they’ll otherwise spend their proverbial tokenized bullets from the growth gun. 

My friend and fellow investor, Tina calls this the “hot start problem” where the presence of a token limits the window of time a startup has to find PMF and gain enough organic traction such that the startup can retain users/liquidity as token rewards diminish.

Apps that launch with a points system, also encounter the hot start problem as there’s now an implicit expectation of tokens from users. 

I quite like the framing of the “hot start problem” since a core differentiator of crypto compared to web2 is the ability to leverage tokens – financial incentives – as a tool to bootstrap new networks.

This strategy has proven effective, especially from the likes of DeFi protocols such as MakerDAO, DyDx, Lido, GMX, among others. Token bootstrapping has also shown to be effective for other crypto networks from DePIN (e.g. Helium) to infrastructure (e.g. L1s) and certain middleware (e.g. oracles). However, networks that opt into the hot start problem by blitzscaling with tokens face several trade offs including, obfuscating organic traction/PMF, pre-maturely spending the proverbial bullets from the growth gun, and adding friction for accomplishing operational tasks because of DAO governance (e.g. fundraising, governance decisions, etc.) 

Opting Into The Hot Start Problem

The hot start problem is favorable to the cold start problem in two scenarios: 

• Startups competing in red ocean markets (markets with a high degree of competition and known demand) 

• Products and networks with passive-supply side participation 

Red Ocean Markets

The core disadvantage of the hot start problem is the inability to ascertain organic demand, but that problem is lessened when building in a category with robust product market fit. It is in this instance that the second movers, by launching a token early, have the potential to successfully compete with earlier market entrants. Defi offers the most examples of second-movers overcoming the hot-start problem, effectively leveraging tokens to bootstrap new protocols. While Bitmex and Perpetual Protocol were the first centralized and decentralized exchanges to offer perps, later entrants like GMX and dYdX leveraged tokens to quickly grow liquidity and become perp category leaders. Newer DeFi protocols within lending like Morpho and Spark have successfully bootstrapped billions in TVL compared to first movers like Compound, although first mover, Aave (formerly ETHlend) still reigns supreme. Nowadays, when there’s clear signs of demand for new protocols, tokens (and points) are a default of the liquidity bootstrapping gameplan. For instance, liquid staking protocols have aggressively leveraged points and tokens in order to grow liquidity in a highly competitive market. 

Elsewhere in crypto consumer land, Blur showcased the playbook for competing in red ocean markets with its market-defining points system and token launch that catapulted Blur into a dominant Ethereum NFT trading venue by trading volume. 

Passive vs Active Supply-side Participation 

The hot start problem is easier to overcome in passively supplied networks compared to actively-supplied networks. The brief history of tokenomics shows that tokens are useful at bootstrapping networks when there’s passive jobs to be done – staking, providing liquidity, listing assets (e.g. NFTs), or set-it and forget-it hardware (e.g. DePIN). 

Conversely, although tokens have also succeeded at jumpstarting active networks like Axie, Braintrust, Prime, YGG, and Stepn, the premature presence of tokens often obfuscates true product-market-fit. As a result, the hot start problem is significantly more challenging to overcome in active networks compared to passive networks. 

The lesson here is not that tokens are ineffective in active networks, but instead that apps and marketplaces that launch token incentives for active jobs to be done – usage, gameplay, gig work, services, etc – must take extra steps to ensure that token rewards are going towards organic usage and driving important metrics like engagement and retention. For instance, data labeling network Sapien gamifies the task of labeling and has users stake points in order to earn more points. In this instance, passively staking while performing some action has the potential to act as a loss-aversion mechanism, ensuring higher quality data labeling from participants. 

Speculation: Feature vs Bug

Speculation is a double edged sword. It can be a bug if integrated too early into a product’s lifecycle, but when done strategically, it can also be a powerful feature and growth tool for garnering user attention. 

Instead of tackling the cold start problem, startups that choose to launch a token before gaining organic traction opt for the hot start problem. They accept the trade-off of leveraging tokens as an external incentive to draw user attention while wagering on their ability to discover or create organic product utility amidst increased speculative noise.

The privatization of order flow is only going to grow as the value of blockspace commodities and paves the way for fat applications. 

Source: Orderflow.art

But how did we get here? And where are we going? 

The short answer for how we got here is, foodcoins. The slightly longer version is that Defi summer catalyzed an abundance of prosumer and retail trading, which subsequently led to the birth of transaction aggregators like 1inch that offered users better price execution through private order routing. Wallets (e.g. MetaMask) quickly followed suit, realizing that they could monetize user convenience by adding in-app swaps, proving that an extremely valuable business model existed for any application that controlled the end user attention (and orders). 

Over the past two years, we’ve seen two additional categorical players enter the private order flow landscape – Telegram bots and solver networks. Telegram bots have kept in line with MetaMask’s “convenience fee” offering users an easy way to trade long-tail shitcoins assets from the comfort of their group chats. As of July, Telegram bots account for ~17% of transactions and 6% of trading volume, the majority of it going through private mempools.

Source: Dune

On the other, fat-head side of the market, solver networks (i.e., Cowswap, UniswapX) also emerged as core venues for trading highly liquid pairs in significant size (e.g. stablecoins and ETH/BTC). Solver networks changed the order flow market structure by outsourcing the job of finding the optimal route for a given trade to a competitive market of solvers (market makers).

As a result, there’s a budding bifurcation in trading venues, whereby convenient front-ends including, TG Bots, wallets swaps, and Uniswap’s front-end are predominantly used for longer tail, lower value (sub $100k) transactions while aggregators and solver networks are the go-to venue for larger transactions, usually involving stablecoins and majors (ETH/BTC). 

Upon more granular filtering you’ll notice that most of the private orderflow comes from aggregators (e.g. 1inch) and front-ends (TG bots, wallets, and front-ends). 

The privatization of order flow is even more pronounced when we consider that, on transactions basis, only 30% of Ethereum transactions go through private mempools, meaning that the minority of transactions contribute to a significant percentage of private order flow volume. 

Source: Dune

Put another way, valuable order flow is more important than order flow quantity. The power law of users and order flow leads to an inevitable conclusion – applications will accrue the greatest proportion of overall value. Said another way, the fat app thesis is alive and well.

Towards Fat Apps

Uniswap’s protocol is clearly valuable, but the more interesting story is happening at the application layer as Uniswap has made efforts to become a consumer application – verticalizing key components of their stack – from expanding the capabilities of their interface, mobile wallet, and aggregation layer. For instance, Uniswap Labs’s applications – Uniswap’s front-end, wallet, and aggregator UniswapX – originate ~16% of the $8 billion in private order flow volume over the last 30 days and nearly ~18% of total order flow (private and public).

Elsewhere in the crypto landscape, apps like Worldcoin account for nearly 50% of Optimism mainnet activity which has pushed them towards launching their own appchain, further highlighting the fat app thesis and the power of controlling demand (e.g. users and transactions).

Even top NFT projects like Pudgy Penguins that possess strong brands are building their own chains, with Luca, the CEO and cheif Pengu explaining, that controlling the blockspace upon which your distribution is built is advantageous for value accrual to the Pudgy brand and IP.

Looking forward, apps should look towards creating new types of order flow whether through the creation of new assets (e.g. Pump and memecoins), by building applications that create new user utility such as identity (i.e.. Worldcoin, ENS), or by crafting better consumer experiences that are vertically integrated and support valuable transactions such as Farcaster and frames, Solana Blinks, Telegram and TG apps, or onchain games. 

Final Thoughts on Fat Apps

It’s worth noting that the fat app thesis has been top of mind for many in crypto since the end of the last cycle once the appchain thesis cemented as part of the consensus viewpoint.

My current view on how the Fat App Thesis plays out is that we’ll see the majority of value accrue to the application layer of the stack, where the control of users and orderflow keeps apps in a privileged position. These applications will likely be coupled to onchain protocols and primitives, similar to those of today including UniswapX and Uniswap Protocol, Warpcast and Farcaster, Worldcoin and Worldchain. Ultimately, these protocols, especially those maximally onchain (e.g. MakerDAO) can still accrue significant value, but the apps will likely capture more value given their proximity to the users and offchain components that give apps more defensible moats.

Finally, I still believe that there’s still a path for Layer 1 blockchains (e.g. Bitcoin, Ethereum, Solana) to capture significant value as non-sovereign reserve assets where the underlying assets (e.g. ETH) accrue immense value. It’s possible that given enough time, apps will try to build their own L1s just as they’ve built their own L2s, but spinning up commodity L2 blockspace is very different than bootstraping an L1 and transforming a governance token into a commodity and collateral asset, so that’s probably a debate for a distant future.

The core takeaway is that as more applications create and own valuable order flow, the crypto world will re-rate applications as people come to the inevitable conclusion – fat apps are inevitable. 

But crypto doesn’t only change how attention is valued—it also promises to change who the value of attention accrues to. 

In 2016, Tim Wu introduced the term “attention merchants” to refer to the way publishers and, later, platforms have exploited users’ attention for profit. Crypto creates a pathway for users to become their own attention merchants, recapturing the value of our attention through ownership of attention assets. 

The most notable instance of this trend is within SocialFi, where users are able to own the flow of attention to assets such as memecoins, influencer access keys, content, etc. By creating pathways for users to directly engage with attention-based assets, SocialFi platforms challenge the traditional power dynamics of the attention economy, turning users from passive consumers to active participants—the new attention merchants.

The SocialFi Frontier 

SocialFi is shaping up to be a defining category within web3. Crypto social networks like Farcaster are gaining steam, with over 75k DAUs. Telegram bots, which couple group messaging and trading, facilitate billions in trading volume. And information markets are now moving toward financializing social graphs such as Twitter (i.e., Trends.market, Fantasy.top) and Farcaster (i.e., Swaye, Perl, Arrina). 

While not all social platforms will have financial incentives attached, SocialFi represents one evolution of social from indirectly valuing social capital to t more efficiently valuing social- and attention-based assets. As a socio-economic technology, crypto enables social applications to add other financialized elements (e.g. asset trading) or natively integrate financial primitives into the application layer (e.g. Friendtech bonding curves). The SocialFi trend is being driven by consumer desire to own and trade in attention assets. Users are choosing to spend time in applications that allow them to earn based on their attention, or play financial games to enhance the social-entertainment experience.

For instance, there’s Fantasy, a fantasy sports trading card game and information market built atop X’s (formerly Twitter) social graph. Fantasy allows creators to earn money from their social media presence, while enabling players to earn rewards based on their intuition and knowledge of certain social accounts. Elsewhere, new social networks like Friendtech, Unlonley, and Sanko allow creators to directly monetize their social interactions through chat access passes. This benefits users who purchase access passes early, rewarding them for allocating their attention to undervalued creators and groups. 

The core benefit across both new information markets and social networks is that creators and users are now the attention merchants, owning the attention assets within these applications and monetizing the attention through app usage. 

Many applications have responded to the user desire to embed commerce and finance inside within their social experiences:  

• Messaging → in-message transactions and trading

• Gaming → ownable assets and in-game economies based on real-moneyTM

• Social → ownable social graphs, channels, content, and platforms 

• Memes → scenecoins and derivative memetic assets 

• Information markets → new markets for social-based entertainment, influencers, and social capital

• Exchanges → new protocols for issuing social- and attention-based assets  

Over the past year, the SocialFi ecosystem has grown rapidly, with an explosion of companies across attention asset exchanges (e.g. memecoin protocols), PvP (player vs player) social games, new forms of information markets, and financialized social networks. The driver behind this expansion is the maturation of crypto infrastructure across two vectors, scalability and usability, which has supported new types of consumer experiences (e.g. mobile PWAs), cheaper transactions (e.g. L2s), and faster app iteration cycles through improved developer tooling (i.e., account abstraction and wallet as a service tooling). 

Social Networks

Social networks can broadly be categorized into two subgroups with their respective creator monetization models: parasocial and bi-directional. 

Parasocial networks are platforms where creators and fans have a one-sided relationship. The one-sided relationship is often paired with a direct monetization model such as subscriptions (i.e., Substack, OnlyFans, Patreon) or through a direct ad revenue cut to the creators (i.e.,YouTube, TikTok).

On the other side of the spectrum are bi-directional networks where creators and fans have a dual-sided relationship (i.e., Twitter, Reddit, Facebook, Snapchat). Bi-directional social networks allow users to monetize distribution that encourage reach rather than restrict it, as is the case with token-gated access (e.g. influencer gated chats). Web2 bi-directional networks like Twitter and Linkedin have historically made it more challenging for creators to directly monetize their reach. Instead, creators have had to for strategies such as affiliate programs, directing users to other monetization sites (e.g. Twitter → Substack), or promotional campaigns. 

By re-imagining users as the new attention merchants, SocialFi offers a variety of new monetization options for both categories of social networks. Parasocial networks offer creators the ability to further monetize the top percentiles of their audience through tokenizing content, influencer access, ephemeral perks (e.g. time-limited rewards), or social status. Parasocial networks Drakula and Friendtech have tokenized content and creators, respectively, enabling the top creators to earn revenues from trading volumes. Sofamon showcases one example of a token model where individuals can slowly purchase an aesthetic good (e.g. avatar clothing) until they own a whole item, which is then wearable. 

Web3 social networks offer new options for monetization. One example is username and namespace monetization, which can generate revenues for valuable namespaces that scale to millions of users. Elsewhere, bi-directional social networks can make better use of in-app transactions. This can manifest as in-social network marketplaces, channel storefronts, or in-app games. 

The key difference with web3 bi-direcitonal networks versus web2 social networks is that the new attention merchants—users and creators—will be able to better earn from their activity. For instance, imagine if Reddit subreddits moderators could own their channels, earn revenue based on ads they showcased, or earn a piece on transactions that go through their channels because of the communities that they curate. 

PvP Social Games

As consumer infrastructure has matured, it has unlocked a new frontier of PvP (player vs player) social games. Most notably, a wave of Survivor-esque competitions—including Crypto The Game, Blessed Burgers, among others—have emerged that offer new digitally native and highly social game experiences for users to earn valuable prize pools. Other applications, like Rug.fun or PvPWorld, provide game theory strategy games where users can coordinate with others in order to win prizes. Contrast this with web2, where most mobile games monetize attention through traditional advertising, or offer users a way to pay to play the game (e.g. users don’t have to wait for cool down periods). Game developers now have new business models where social games resemble content, with developers releasing multiple ephemeral applications that offer shortened play cycles where users can earn significant rewards from participation before moving onto the next game. 

New types of social games should optimize for: multiple winners, which enables greater participation; easy-to-play games, where average users feel like they have high perceived odds of winning; and social interactions, which further enhance the virality of these games. These proposed game dynamics are more incentive aligned than web3 games, which have historically bent towards pay-to-win type games or farm-first vs fun-first games. 

New Marketplaces and Exchanges

Crypto dominant use cases revolve market creation, specifically the issuance of new asset classes, bringing existing assets onchain, or expanding access to digitally native assets. 

• Information markets – Information markets like Polymarket have the potential to create more efficient political markets and support the creation of new types of event markets based on real world events, culture, and business. 

• Attention exchanges – Launchpads like Pump and Ape.store let users create new assets (e.g. memecoins) based on one quality:attention. Elsewhere, Sofaman is tokenizing status and culture by enabling users to create a telegram-based digital avatar that has branded clothing for sale on a bonding curve.

• Telegram bots – Telegram bots bring marketplaces and social-financial games into the messaging experience and offer a more convenient experience for users 

• Points and pre-tokens – Points have been an effective incentive strategy for teams to test user behaviors and experiment with dynamic incentives. Points marketplaces like Michi and WhalesMarket and pre-token marketplaces like Aevo can help create more efficient markets for tokens. 

Several sub-trends are spurring the creation of new marketplaces and exchanges. First, an increase in the verticalization of social and financial platforms is driving these apps to issue new types of assets. Second, growing user ownership around onchain activity through earning points, tipping, and tokens is increasing the surface area of assets for users to interact with, encouraging the creation of new trading venues. And finally, users are now interacting with assets, such as memecoins, where they feel a greater sense of perceived autonomy. Similar to real world cultural assets such as sneakers or music, users feel a sense of perceived control over the popularity and potential appreciation of these cultural based assets, since the fundamental metric that gives the assets value (a users’ attention) is controlled by the end consumer. 

Building for the New Attention Merchants

There’s a paradigm shift happening in social, where the dynamic between users, creators, and attention is being redefined. At the heart ofthese trends is the transition from users and creators as the supply and demand side of the attention economy to being able to become the merchants of their own attention. 

It’s admittedly hard to design new financial or social primitives, let alone primitives that blend the best of both into a unified experience. The early social-financial tools, toys, and games that become the next era of SocialFi networks and applications will be those that speedrun experimentation, testing new consumer behaviors and capitalize on the emergent behaviors and revealed preferences of consumers. 

But not all supply is the same.

There’s active supply, meaning it must engage continually in the marketplace. And then there’s passive supply, which requires initial onboarding but minimal ongoing maintenance; it’s set-it-and-forget-it.

Subscribe now

Tokenized marketplaces with active supply are much stickier and often have greater defensibility once they get to scale because the compounding nature of demand offers better economics for suppliers. Whoever reaches the greatest demand liquidity first wins because they can provide higher earnings on the supply side, which further spins the flywheel.

Conversely, passively supplied marketplaces are better able to rapidly scale supply growth without an equivalent level of marketplace demand—but their stickiness is harder to secure. Builders can take advantage of these dynamics while bootstrapping a tokenized marketplace, but only if they understand the core tradeoffs of each type.

Active vs. Passive Marketplaces

To understand why active-supply marketplaces tend toward defensibility but are harder to scale—and why passive marketplaces trend the opposite way—it’s essential to look at their general characteristics, each one existing on a spectrum.

Human Labor vs. Resources 

Active forms of supply are nearly synonymous with human labor. As of yet, people can’t sit back and passively rent out their brain power like storage space. For example, Braintrust, a decentralized job network, requires constant supplies of human talent to fulfill the real-time demands of employers. 

On the other end of the spectrum are resources, such as hardware, NFTs, and capital. These represent passive supplies. For instance, DIMO, a data-sharing network for cars, requires users to buy and connect a DIMO hardware device. After this one-time setup cost, the device continuously streams vehicle data to the DIMO network with little input from the user.

High Opportunity Costs vs. Sunk Costs

In active marketplaces, supply-side participants choose the marketplace with the best combination of income/yield and token-appreciation potential. Axie Infinity, which popularized play to earn, competed with other marketplaces in which users could spend their labor to earn yield. Without strong organic demand, active marketplaces must perpetually compete against all the other ways users could spend their time earning tokens.

Passive marketplaces, however, have upfront or sunk costs in the form of depreciating supply-side assets. As a result, supply-side operators will passively contribute their physical assets to a marketplace so long as it’s profitable to do so. For instance, GPU owners have an incentive to supply their compute power to GPU marketplaces to make the most of their depreciating assets. Even in the absence of demand equilibrium, passively supplied marketplaces can continue to support a high quantity of supply using token incentives. 

Quality-Dependent Supply vs. Quality-Independent Supply

It’s far easier to scale a marketplace when you have a clear view of supply quality. Passive-supply marketplaces, with their physical supplies, are better suited to this than active-supply marketplaces. That’s because their supply often comes with quantitative qualifiers that make scaling quality easier. For instance, GPUs have different quantitative classifications (e.g. A100 vs RTX 4090s) that also track quality. 

This is seldom the case with active-supply marketplaces, which are dealing with high variations in human capabilities. Gig-based platforms such as Braintrust or Nosh are only as good as their workers, but the demand-side participants will have differing opinions on the quality of those workers.

Token Design Implications

So how should builders go about bootstrapping tokenized marketplaces and getting them to scale? What implications do a marketplace’s supply characteristics have on token design?

Actively Supplied Marketplaces

There are several key token design implications for actively supplied marketplaces:

• Scale token incentives with demand growth 

• Incentivize supply-side loyalty, quality, or reliability

• Create dynamic incentives to prevent side-switching 

In passive marketplaces, supply can wait for demand to catch up (e.g. Filecoin). Not so in active marketplaces, where humans are dealing with high opportunity costs. Thus, builders must prioritize demand-side growth to remain competitive. However, tokens can help bootstrap initial demand so that supply-side participants want to join the marketplace. 

One strategy for scaling active marketplaces is to dynamically scale supply-side incentives so that token distribution closely mirrors growth. A related mechanism is to onboard supply in a permissioned manner, such that a stable or threshold earning rate (e.g. yield) is consistently offered to supply-side participants, which in turn keeps supply-side labor engaged and reliable. 

Regardless, this limitation of actively supplied marketplaces actually makes them stickier as they grow—they are able to offer more consistent earnings as demand increases. From an incentive design perspective, such tokenized marketplaces should focus on providing continual rewards that keep users active on the platform. Moreover, they should dynamically adjust these rewards to incentivize users who provide consistent supply over users who churn. 

But while token incentives are valuable for bootstrapping supply and demand, some innovation is likely required at the servicing, verification, and reputation layers to scale supply quality, an important characteristic of actively supplied marketplaces. 

Here, tokenized marketplaces must take lessons from traditional managed marketplaces. For instance, the RealReal and StockX provide verification services to ensure the physical supply is legitimate. Similarly, Braintrust acts as a middleman and provides a quality-assurance layer as part of its marketplace offering while leveraging tokens to help onboard supply (e.g. referrals) and vet peer-to-peer talent. 

Active-supply marketplaces that can leverage the scaling effects of tokens can do even better. By using a network of stake-based intermediaries or a token-incentivized verification and curation layer, they can enhance the quality assurance process and generate more efficient marketplaces.

What about those tokenized marketplaces where supply and demand come from the same users, such as with NFT marketplaces or play-to-earn games like Axie and Stepn? 

Side-switching marketplaces will need to be more dynamic when altering token incentives because they are most likely to find that the speculative flywheel of a token can obfuscate organic demand. Marketplaces where side-switching is prevalent can help alleviate the reflexive nature of growth by incorporating vesting and/or other specific lockups into token rewards, incentivizing long-term participation. Actively supplied marketplaces should incentivize greater supply-side diversity to bring on more prosumer and professional supply over retail supply, which can be more fickle. 

Passively Supplied Marketplaces

There are also important token design implications for builders of passively supplied marketplaces:

• Scale supply-side quantity aggressively to achieve commercially viable scale 

• Build greater defensibility through demand-side products (e.g. SDKs, APIs) or supply-side lock-in via proprietary hardware 

• Incentivize supply-side loyalty, quality, or reliability

Passively supplied marketplaces typically require a certain supply threshold to be met before the marketplace is commercially viable (i.e. generates robust demand), so builders should initially focus on supply-side growth. Moreover, this supply is typically measured in quantity, not quality. For instance, data collection networks like DIMO, Hivemapper, and Wynd, need significant data before the aggregated data or services built on top of them become valuable. 

Because all passively supplied marketplaces are easier to scale, a new entrant isn’t going to secure demand merely by gathering sufficient liquidity. Instead, it will typically need to compete on product by offering an orchestration/onboarding layer or building SaaS components (e.g. SDKs and APIs) that help demand access the marketplace’s supply. GPU marketplaces like IO.net offer aggregation services that make accessing GPUs easier for end-compute users. Similarly, DIMO has built a marketplace that enables DIMO token owners to purchase services for their cars. 

Another way to make passively supplied marketplaces more defensible is by transitioning from commoditized supply to proprietary supply. Wireless marketplaces like Helium and XNET are leveraging proprietary supply to build their telecom infrastructure. 

Finally, given the high sunk costs of passively supplied marketplaces, supply will often continue servicing a network so long as rewards outweigh operating costs. When sunk costs are high and opportunity costs are low (e.g. Blackbird restaurants accepting FLY tokens), supply will be more retentive since the supply-side participant has an inherent incentive to service the marketplace. Conversely, when high sunk costs are coupled with high opportunity costs (e.g. GPU owners servicing compute marketplaces), demand or token rewards could be a deciding factor for where supply chooses to allocate resources. 

Share Unsyndicated by Mason Nystrom

As usage of crypto social platforms and financial games increases, the way they are constructed will continue to evolve. We can expect to see a trend towards increasing verticalization, with projects seeking to offer a more seamless and comprehensive experience for users that enables the creation of new consumer behaviors and attention or social-based assets. While not all web3 social experiences are financial, the blockchain rails that underpin crypto consumer applications allows for new token-incentivized behaviors and digitally native assets to be imbued into the social experience. 

The existing SocialFi stack is comprised of four core layers: 

• Discovery - where people discover what to buy

• Execution - where assets are bought and sold

• Liquidity - where assets reside and are pooled

• Asset issuance - where assets are created

Today, the stack is fairly fragmented, with the user discovery and social experience decoupled from execution (e.g. trading), liquidity, and asset issuance. But as the SocialFi frontier expands, applications will continue to verticalize attention and markets in an effort to control user social experiences and attention asset liquidity. 

SocialFi app builders will need to own multiple layers of the SocialFi stack in order to build protocol defensibility. Attention asset trading (e.g. execution) and issuance are commoditized layers of the stack – token issuance is increasingly easy, and execution can be added anywhere that owns eyeballs. Owning either the discovery or liquidity layer will be of increasing importance, since these are defensible layers of the SocialFi stack that exhibit strong network effects. 

Within SocialFi, most applications are choosing between two approaches towards verticalization:

• Exchange-first approach: Build an exchange or marketplace where users trade attention assets (e.g. Robinhood for memes) and evolve into a social/discovery platform from there

• Social/discovery-first approach: Build a social platform to own discovery and then layer in financial elements and primitives with consumers/attention merchants as a key stakeholder in the platform

Exchange-First

Any social network or discovery platform faces significant headwinds: bootstrapping new social graphs, spurring new consumer behaviors, and retaining user engagement in an era of a hyper-competitive attention market. Given these obstacles, the exchange-first approach is arguably easier to jumpstart, because users’ appetite for speculation helps to surmount those headwinds. However, this approach faces more competition because of the ease with which exchanges can be bootstrapped compared to social networks and the benefits that social networks retain once a threshold density has been reached. 

From the exchange-first approach, deep verticalization of the SocialFi stack has proven effective because exchange-first applications have attention-trading capabilities built in. For instance, Friendtech has emerged as one of the most vertically integrated SocialFi apps, with control over the entire stack. The app serves as the focal point for key discovery, exclusive execution, and leverages a native financial primitive––a bonding curve––that issues assets with utility specific to the Friendtech app.

Newer SocialFi protocols have also vertically integrated the SocialFi stack. For example, memecoin issuance and discovery platforms like Pump and Ape Store, allow users to easily deploy a memecoin on a bonding curve. This enables users to easily purchase tokens directly from the bonding curve, as opposed to having to wait for someone to seed liquidity into a DEX or liquidity pool. While some Pump initialized memecoin execution and discovery can occur on other platforms—Dexcreener and Twitter, respectively— protocols like Pump still provides a unique social discovery experience and execution platform for its newly launched tokens. 

Discovery-First

Historically, the discovery-first SocialFi approach has seen success through the likes of Twitter, Farcaster, and Telegram, and also through market terminals like Dexscreener and Coingecko. Many of these applications have tried to move down the stack to offer execution (e.g. trading) of tokens, but have yet to fully lean into offering custom, proprietary trading experiences.

The exception is Telegram, which has successfully integrated the social-financial experience. Still, Telegram’s UX is limited, and while degens have opted for it for convenience, there is room in the market for more Robinhood-esque experiences that offer seamless trading UX, easy onboarding, and retail friendly features like commission-free trading. Moreover, new primitives like Farcaster frames and casts and Lens open actions further enable new types of financial transactions within these discovery-first networks. 

Final Thoughts: Be Opinionated 

Builders can create compelling social-financial games and networks by taking a unique approach to how app monetization and financialization impact their apps. The exchange-first approach is easier since it doesn’t necessarily rely on creating new consumer behaviors, people already want to trade attention. But, a discovery-first approach where a platform controls eyes is historically more defensible than one that only controls orders (e.g. transactions). The primary goal of the discovery or social-first approach is to iterate quickly, testing new consumer behaviors and social-financial dynamics until users showcase their revealed preferences that have the potential to grow into large social networks. I’m confident that the most successful apps will have opinionated, vertically-integrated designs that create liquid marketplaces for new types of assets or that otherwise catalyze new consumer behaviors.

Subscribe now

Whether it's a domain extension, social media username, or phone number (e.g. 1-800-FLOWERS), unique digital identifiers have been coveted since the issuance of the first web domains, which converted IP addresses into human-readable names and eventually gained value via their searchability with the continued adoption of the internet. Surprisingly, the rise of social networks made the market for namespaces less liquid, since you can’t easily buy someone’s handle on web2 social platforms. Twitter even banned the sale of handles, which is ironic for a company in dire need of more sustainable cash flows.

Although the .eth namespace has been around for years, the integration of a cross-chain interoperability protocol (CCIP) has further spurred the creation of new namespaces by connecting ENS subdomains with traditional, offchain domain names. For example, in August Coinbase leveraged ENS and CCIP to start offering its users blockchain usernames under the DNS domain cb.id (masongnystrom.cb.id is both a DNS address and a web3 wallet). 

With the rise of namespace issuers, crypto consumer applications will have the opportunity to create their own native namespaces—but this time around, we’ll be able to own social usernames or have governance over the root namespaces (e.g. ENS). With that capability will come lucrative opportunities for the top crypto consumer applications and their early users. 

To date, blockchain projects that wanted to create onchain namespaces have opted for one of two paths: 1) launch their namespace using ENS subdomains, or 2) launch their own app-native namespace protocol that serves the native platform. 

Deciding whether to build your own namespace 

Social graphs like ENS can be leveraged to bootstrap an identity layer for a new application. Coinbase, a platform where the user’s public-facing identity isn’t fundamental to the product experience, has opted to use ENS subdomains to generate user identities quickly while also enjoying benefits such as DNS compatibility and integrations with prominent distribution layers such as MetaMask. Farcaster has similarly decided to leverage ENS because it possesses a base of high-value users who want to display their existing identity (e.g. .eth address) within Farcaster’s social graph.

On the other hand, projects like Lens and Yat have opted to issue usernames separate from the ENS ecosystem by creating their own namespace protocols. These unique namespaces carry similar value propositions to ENS subdomains but have the distinct benefits of greater protocol flexibility, autonomy from a third-party issuer, and the ability to earn fees from namespace issuance and annual renewal fees.

And charging for namespace renewal in social networks serves a purpose beyond generating revenue: it prevents squatting on valuable namespaces by speculators, which is good for a social network as it matures and onboards more users. Moreover, crypto social apps or protocols like Lens will use usernames from their unique namespace as an incentive to join the network, offering firstcomers their preferred namespace with the potential to monetize the namespaces at a later point in time once some network density has been achieved. 

However, platforms that build their own namespace protocols forgo the benefits that come with building on top of ENS, most notably a substantial ecosystem of integrations and an existing user base of people who already possess a .eth address and speculate on ENS-related subdomains. 

Onchain namespaces offer a viable business model and incentive mechanism for new types of crypto social networks. Most applications will be best served by leveraging ENS subdomains as their namespace and issuing those usernames to their users for free. Platforms that have socially relevant namespaces or that have sufficient discovery and distribution (e.g. social networks) will have the greatest chance of bootstrapping their own namespace protocols and charging renewal fees for usernames in the long run. 

If you’re building new social applications leveraging unique namespaces, I want to hear from you: mason@variant.fund. 

Share Unsyndicated by Mason Nystrom

+++

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Variant. While taken from sources believed to be reliable, Variant has not independently verified such information. Variant makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. This post reflects the current opinions of the authors and is not made on behalf of Variant or its Clients and does not necessarily reflect the opinions of Variant, its General Partners, its affiliates, advisors or individuals associated with Variant. The opinions reflected herein are subject to change without being updated.

You need not look very far to see evidence of this trend. Searchers deploy bots, like Jaredfromsubway.eth, to take advantage of a human user’s desire for convenience by front-running their DEX trades. Banana Gun and Maestro, which allow human users to make bot-enabled trades from the convenience of Telegram, are consistently some of the most frequent gas guzzlers on Ethereum. And now, on new ephemeral social apps like Friendtech, bots are entering the fray after the app finds initial human user adoption, and can inadvertently further bootstrap the speculative flywheel. 

All this to say that bots, whether profit-driven (e.g. MEV bots) or consumer-driven (e.g. Telegram botkits), are increasingly often the prioritized users on blockchains. 

While bots in crypto so far have been fairly rudimentary, bots outside of crypto have started to evolve into robust AI agents thanks to the rise in large language models (LLMs), with the end goal of autonomously handling complex tasks and making their own well-informed decisions. 

Building these AI agents on cryptonative rails yields several important enhancements: 

• Native payment rails: AI agents can exist outside of crypto, but if we want AI agents to perform complex actions they’ll need access to capital. Crypto rails present a meaningful improvement for giving AI agents access to capital over having them obtain access to bank accounts or payment processors (e.g. Stripe), or deal with the vast majority of other inefficiencies that exist in our offchain world.

• AI agent wallet ownership: AI agents connected to wallets will have the ability to own assets (e.g. NFTs, yield, etc.), giving agents the digital property rights that are inherent to all crypto assets. This is especially important for agent-to-agent transactions.

• Verifiable, deterministic actions: AI agents will be most effective when actions are provable (where they can ensure that some action was completed). Onchain transactions are deterministic in nature—they either happened or didn’t—which means AI agents will be able to more accurately complete tasks onchain than offchain. 

Of course, there are limitations to onchain AI agents.

One limitation is that AI agents will need to perform logic offchain to be performant. This means onchain AI agents will host their logic/computation offchain to optimize efficiency, but agent decisions will be executed onchain, allowing for verifiable actions. Importantly, AI agents can also use zkML providers like Modulus to ensure their offchain data inputs are verified. 

Another key limitation of AI agents is that they are only as useful as the tools they are given. For example, if you ask an agent to give a summary of a real-time news event, the agent needs to have a web scraper in its toolkit to comb the internet to perform the given task. Do you need the agent to save the response as a PDF, add a filesystem to the toolkit. Want the agent to copy trade your favorite Crypto Twitter influencer? The agent needs access to a wallet and key signing permissions over said wallet.

Looking at the current landscape on the spectrum of deterministic to non-deterministic, most crypto AI agents perform deterministic tasks. That is to say, humans program the parameters of the tasks and how the task (e.g. a token swap) is accomplished. 

Crypto AI agents have evolved from the early days of keeper bots—which are still utilized across DeFi and oracle apps—to today’s more sophisticated agents that leverage LLMs, including autonomous artists like Botto; AI agents that can bank themselves using Syndicate’s transaction cloud; and early AI agent service marketplaces like Autonolas. 

There are already a variety of exciting applications at the bleeding edge:

• AI agent-enabled “smart wallets”: Dawn leverages DawnAI to provide an AI agent that can assist users with sending transactions, executing tradings, and other real-time onchain insights (e.g. trending NFTs). 

• Crypto gaming agents: Parallel Alpha’s newest game Colony aims to create AI characters that can own a wallet and transact with each other. 

• Enhanced toolkits for AI agents: AI agents are only as good as their toolkits, and interacting with blockchains is currently a nascent area. Crypto AI agents need wallets, ways to fund them, permissioning capabilities, integrated AI models, and the ability to interact with other agents. More concretely, Gnosis has showcased what this early infra looks like with its AI mechs, which wrap AI scripts with a smart contract such that anyone—including another bot—can call the smart contract to perform the agent action (e.g. make a bet on a prediction market) while also being able to pay the agent.

• Augmented AI traders: DeFi super apps that provide elevated actions for traders and speculators including: DCA into a position if conditions are met; executing trades when gas prices drop below a certain price; monitoring new meme token contracts; and determining order routing without the user having to know where to onboard, etc. 

• Enabling the long tail of AI agents: While large applications like ChatGPT are good for some general chat purposes, AI agents will need to be fine-tuned for numerous industries, subjects, and niches. Marketplaces like Bittensor create the incentives for “miners” to train models for specifics tasks (e.g. image generation, pre-training, predictive modeling) around target industries (e.g. crypto, biotech, academia). While Bittensor is nascent, developers are already using Bittensor to build applications/agents on top of the long tail of open source LLMs. 

• NPC consumer app agents: Non-playable characters are common in games like MMORPGs, but less common in multiplayer consumer apps. However, the financialized nature of crypto consumer apps makes AI agents excellent participants to introduce new types of game mechanics. Open AI infrastructure company Ritual recently released Frenrug, an LLM-based agent that operates inside of Friend.tech that will execute transactions (buy or sell keys) based on user messages. Friend.tech users can attempt to convince the agent to buy their keys, sell others’ keys, or attempt to get the Frenrug agent to use its funds in other creative ways. 

As more applications and protocols leverage AI agents, humans will use them as a conduit for accessing the crypto economy. And while AI agents look like toys today, in the future they will augment daily consumer experiences, become key stakeholders in protocols, and create entire economies between themselves. 

AI agents are in their infancy, but these first-class citizens of onchain economies have only just begun to showcase their potential. If you’re testing the boundaries of how AI agents augment onchain experiences, please reach out on twitter.

Share Unsyndicated by Mason Nystrom

Special thanks to Tom Waite and Sami Kassab for conversations and feedback on the ideas within the essay. Thanks to Dan Roberts for edits and making the essay sound buttery. 

The value of these networks is clear: they offer a better solution for powering consumable resources, from computing to energy to data. In turn, these resources are directly consumed by companies or, more often, leveraged in their own products and services. For instance, decentralized networks like Hivemapper sell data directly to transportation companies like Uber, which in turn use the imagery data to improve their product offerings. Similarly, Livepeer enables livestreaming applications to tap into its marketplace for video transcoding services but is also integrated by companies like Bonfire, which lets creators easily spin up their own livestreams.

To better evaluate the potential of these networks, we need a better way of categorizing them. The current sector name that has caught on is DePIN (coined in 2022 by the folks at my former employer Messari), but I suggest introducing further granularity and splitting decentralized infrastructure networks into two groups:

• Decentralized physical infrastructure networks (DePINs): cryptonetworks with consumable, non-fungible resources that leverage incentives to deploy location-dependent hardware devices

• Decentralized resource networks (DeRENs): cryptonetworks that leverage incentives to build marketplaces and increase the supply of existing or idle consumable, fungible resources that rely on location-independent hardware 

DePINs differ from DeRENs across three core dimensions:

• Resource fungibility

• Hardware location deployment

• Resource creation

Resource fungibility

The most meaningful of the above differences is the fungibility of the consumable resource. 

In resource networks, the consumable resource is fungible because the network’s hardware assets are typically interchangeable. For example, the computational resources provided by networks like Akash or Render are highly fungible—the processing power of one GPU is the same as any other GPU with the same specifications and capacity. With the exception of highly specialized activities like high-frequency trading, users typically do not care where their hardware is deployed geographically, so long as the network latency is acceptable compared to centralized architectures. 

Conversely, DePINs leverage non-fungible or semi-fungible resources. In such cases, the consumable assets are not easily interchangeable and hardware is often unique to a particular network. For instance, Hivemapper’s dashcam maps a specific location, which produces data unique to that location and place in time. Further, an imaging network like Spexigon cannot contribute its aerial image data to the Hivemapper network; the asset of each network is image mapping data, and it’s exclusive to that network.

There are, of course, assets in the middle of the spectrum. Energy, for instance, is semi-fungible because it can be used for a variety of purposes but its utility is limited by the distance of transmission.

Hardware location and resource creation

Hardware location and resource creation are closely related; the deployment of application-specific and location-dependent hardware often coincides with building a proprietary resource.

On this count, DePINs face more challenges building both the supply and demand sides of a market. The supply side requires location-dependent hardware setups, while generating demand depends on there being enough scale on the supply side to make the network valuable to consumers. 

It is easier for resource networks to bootstrap supply since the idle supply can come from anywhere and often doesn’t require the creation of new hardware and infrastructure. But resource networks with fungible assets also face greater competition since the cost of switching from one resource network to another is lower.

Building moats in DePINs and DeRENs

Crypto-based resource networks must still compete with their web2 counterparts like AWS and Google. While DePINs and DeRENs can leverage tokens to subsidize initial resource costs, the most successful networks won’t compete on price alone, but instead will unlock new demand or expand the market in unique ways. 

Arweave, for instance, didn’t compete on file storage pricing. It offered new functionality and convenience through permanent storage and eventually found traction storing NFT metadata. In the DePIN category, mobility networks like DIMO aggregate previously siloed data that power a new wave of applications, from battery intelligence and energy management to better vehicle commerce. 

Another successful strategy is to vertically integrate and generate demand by building the initial product that leverages the infrastructure or resource network. Render couples its GPU-rendering capabilities with its Octane software, which drives usage of the underlying compute resource network.

If you’re building new resource networks or physical infrastructure networks that create demand for new users and applications, I want to hear from you: mason@variant.fund. 

Thank you to Jesse Walden, Sami Kassab, Dustin Teander, and Alana Levin for feedback, comments, and conversations that helped develop many of these ideas.

According to VanEck data, Reignmakers has generated $52 million in revenue for DraftKings, with $72 million in secondary sales YTD. And DraftKings isn’t the only successful NFT fantasy game.

Some of the top NFT collections by secondary sales are in the same sports category:

• Sorare 

• Mythos (NFL Rivals)

• NFL All Day (Dapper Labs)

• NBA Top Shot (Dapper Labs)

Fantasy sports contests have obvious product market fit in web2, but NFTs further expand the capabilities through digital ownership, allowing users to earn more money speculating on their passion for sports. Ownership is a zero-to-one innovation. Ownership creates new game dynamics, helps build psychological attachment for the product, and ultimately creates better outcomes for users. 

Let's examine some core lessons from fantasy sports NFTs:

Create post-mint utility: NFT collectibles have value, but the unlock is in the nature of programmable assets—NFTs that can be used for new consumer experiences. Solve the post-mint problem and you’re one step closer to finding product-market fit. 

Leverage freemium to whale tier pricing models: DraftKings onboards new users with free (non-transferable) NFTs. Fantasy sports apps also tier their leagues by pricing and prizes, similar to how casinos alter bet minimums at different blackjack tables—it’s all the same game, but users decide their risk/reward profile. This enables DraftKings to price consumers all the way from the lowest paying fan to the largest whale. 

Abstract crypto: DraftKings is using NFTs (sorry: “digital collectibles”) and ERC-20 tokens (sorry: “crafting” tokens), but DraftKings abstracts all of the crypto complexity to focus on the user experience. Custodial wallets. Deposits via bank accounts. Gasless experiences but using Polygon for cheap fees on the backend. All these decisions compound to a better product experience. 

Build “convenience-walled gardens”: users will always pay for convenience. In the context of NFTs, that means users will buy NFTs where they can consume them. DraftKings marketplace is a custodial platform and takes a hefty rake (10%) on secondary sales, but devoted users don’t seem to mind. Platforms have started using custodial wallets to provide “convenience-walled gardens” that remove crypto in exchange for added product stickiness. 

Lean into the hobbification of finance: People are willing to invest time and capital into their passions. Crypto apps can be very successful when they combine deep seeded passions (e.g. culture) with financial elements. NFTs are one manifestation of bringing culture onchain so that people can speculate on it. Prediction markets are another example of a category that allows people to take advantage of knowledge they have about their interests (e.g. politics). 

It’s no secret that NFTs have struggled during the bear market, but I’m still extremely excited that NFTs—or whatever companies choose to call them—will continue to be used to build new crypto consumer apps and improve existing experiences (i.e., fantasy sports, entertainment, commerce, etc.) 

The future is non-fungible and collectible.