In the past 30 days, Ethereum has seen order flow capital exceeding $25 billion, with nearly half originating from proprietary applications. As blockchain space value continues to grow, paving the way for “fat applications,” the privatization of order flow is also on the rise.
How did we get here and where are we headed next? The DeFi summer birthed a multitude of consumer and retail trades, subsequently fostering trading aggregators like 1inch. These aggregators use private order routing to offer users better price execution. Wallets such as MetaMask quickly followed suit, realizing they could profit by providing in-app exchanges for user convenience. This underscores a highly valuable business model for any application that controls end-user attention (and orders).
Over the past two years, we’ve witnessed two other types of participants entering the realm of private order flows: Telegram bots and solver networks. Telegram bots provide users with a convenient way to trade long-tail assets in group chats. As of June, Telegram bots account for approximately 21% of transaction count and 11% of transaction volume, predominantly through private mempools.
On the other hand, solver networks (like Cowswap and UniswapX) have become central venues for high liquidity trading pairs (e.g., stablecoins and ETH/BTC). These networks alter the market structure by outsourcing the task of finding the best trading path to solvers (liquidity providers).
Consequently, trading venues have diverged: user-friendly front ends (including TG Bots, wallet swaps, and Uniswap front ends) cater primarily to smaller, low-value transactions (under $100,000), while aggregators and solver networks are preferred for larger trades.
A closer look reveals that a majority of private order flows originate from front ends (TG bots, wallets, and front ends). With only 15-30% of Ethereum transactions passing through private mempools, the privatization of order flow becomes more apparent, suggesting that a significant portion of private order flow volume is contributed by a minority of trades.
In other words, valuable order flow outweighs the sheer quantity of order flows. Power law distribution of users and order flows leads to an inevitable conclusion: applications will occupy the largest share of total value. This reaffirms the validity of the fat applications theory.
Moving towards fat applications, while Uniswap’s protocol layer is evidently valuable, the more intriguing developments occur at the application layer. Uniswap strives to become a consumer application by enhancing interface, mobile wallets, and aggregation layer functionalities, key components of its vertical stack. For instance, Uniswap Labs’ applications (Uniswap front end, wallets, and aggregator UniswapX) contributed nearly 25% of private order flow amounting to $13 billion in the past 30 days, and close to 40% of total order flow.
In other cryptocurrency domains, applications like Worldcoin dominate nearly 50% of Optimism’s mainnet activity, underscoring the strength of the fat applications theory through building their own app chains. Even top NFT projects like Pudgy Penguins are establishing their chains, leveraging blockchain control to accumulate brand and IP value.
Looking ahead, applications should focus on creating new types of order flows, including generating new assets (such as Pump and memecoins), developing applications for new user scenarios (e.g., Worldcoin, ENS), crafting vertically integrated consumer experiences, and supporting valuable transactions like Farcaster and Frames, Solana Blinks, Telegram, and TG applications, or on-chain gaming.
Final thoughts on fat applications: once the application chain theory becomes industry consensus, fat applications will capture widespread attention. In my view, the bulk of value accumulation will occur at the application layer, where control over users and order flows positions applications in privileged states. These applications may integrate with on-chain protocols and primitives, akin to today’s UniswapX and Uniswap protocol, Warpcast and Farcaster, Worldcoin and Worldchain. Ultimately, these protocols, especially those most integrated on-chain (e.g., MakerDAO), can still accrue significant value. However, given the proximity of applications to users and the off-chain components providing moats for applications, applications might capture more value.
Lastly, I maintain that Layer 1 blockchains (e.g., Bitcoin, Ethereum, Solana) can attain significant value as non-sovereign reserve assets. If given enough time, applications may attempt to build their own L1, akin to how they build L2. Yet, building L2 blockchain space differs vastly from guiding L1 and transforming governance tokens into commodities and collateral assets, hence posing a relatively distant concern.
The core takeaway is that as more applications create and possess valuable order flows, the crypto world will reassess applications, solidifying the trend towards fat applications.
Original article link
