From Mempool to Intent Pool: The Future of Frontrunning

Mempools are broken markets. Public transaction ordering creates a negative-sum game where searchers and MEV bots extract value from users through frontrunning and sandwich attacks, degrading network performance for everyone.

Intent-based architectures invert the model. Users submit desired outcomes (intents) instead of specific transactions, shifting competition from the public chain to off-chain solvers and fillers like those in UniswapX and CowSwap.

The value capture flips. In a mempool, value leaks to adversarial bots. In an intent pool, competition between solvers returns value to users as better execution prices and guaranteed results.

Evidence: On Ethereum, proposer-builder separation (PBS) via MEV-Boost already privatizes block construction. Intent-centric systems like Across and Anoma extend this privatization to the user transaction layer itself.

The mempool is obsolete. Its public, first-come-first-served design is a security vulnerability that creates a multi-billion dollar MEV industry for searchers and validators, extracting value directly from users.

Intent-based architectures invert the model. Users submit declarative goals (e.g., 'swap X for Y at best price') to a private intent pool, not executable transactions. Solvers like UniswapX and CowSwap compete privately to fulfill them.

This shifts power to users. The competitive solving market internalizes MEV, returning value as better execution. Protocols like Across and SUAVE demonstrate that private order flow and auction-based settlement are superior.

Evidence: Over 70% of swaps on CowSwap receive price improvement via its batch auction solver competition, a direct transfer of value from what would be MEV back to the user.

Mempools are public attack surfaces. Every pending transaction broadcasts its intent, creating a zero-sum game where searchers and validators extract value via frontrunning, sandwich attacks, and arbitrage.

The MEV supply chain is now institutionalized. Firms like Flashbots and Jito Labs professionalize extraction, offering tools like Flashbots Protect and Jito's bundles to route around public chaos, but this centralizes power.

The user experience is adversarial. Retail traders on Uniswap or Curve consistently receive worse prices than the quoted rate, with losses estimated in the hundreds of millions annually to MEV.

Evidence: Over $1.2B in MEV was extracted from Ethereum and its L2s in 2023, with sandwich attacks alone accounting for ~$240M, per EigenPhi data.

The mempool is obsolete for high-value transactions. It functions as a public, unprotected broadcast channel, creating a zero-sum extraction game where searchers and builders profit at user expense. This model is unsustainable for mainstream adoption.

Intent-based architectures invert the paradigm. Users declare a desired outcome (e.g., 'swap X for Y at best rate'), not a specific transaction path. This moves competition from the public mempool to a private solver network where entities like CowSwap solvers compete on fulfillment quality.

This shifts MEV from extraction to service. In an intent pool, solvers internalize MEV (like arbitrage or batch liquidity) to subsidize better prices for users. Protocols like UniswapX and Across demonstrate this, using fill-or-kill intents to guarantee execution without frontrunning risk.

The new supply chain is solver-centric. The critical infrastructure shifts from block builders to intent aggregation, routing, and solver reputation. This creates a professionalized MEV service layer, with projects like Anoma and SUAVE building the foundational infrastructure for this new market structure.

The centralization fear is misplaced. Critics assume the intent-solving layer becomes a single point of failure. This confuses the role of a centralized sequencer with a decentralized marketplace.

Solvers compete in a permissionless auction. Unlike a single L2 sequencer, the solver network is open. Projects like UniswapX and CowSwap demonstrate this model creates redundancy and reduces extractable value.

Trust shifts from operators to economic security. Users trust the cryptoeconomic design, not a specific entity. Solvers post bonds and face slashing for bad execution, enforced by protocols like Across and SUAVE.

Evidence: Ankr's RaaS data shows fragmentation. The rollup-as-a-service market is already decentralizing execution. Intent solvers will fragment further, creating a more resilient system than today's L2 oligopoly.

Public mempools are dead. Their function as a broadcast mechanism for vanilla transactions will be replaced by private order flow submission directly to builders via intent-based protocols like UniswapX and CowSwap. This shift is a direct response to the unsustainable economic leakage of public transaction data.

The new battleground is solver competition. Execution quality, not gas price, becomes the primary metric. This creates a zero-sum game for solvers competing on price improvement, where the best aggregated liquidity and routing logic (via 1inch, Across, Socket) wins the user's order.

Cross-chain intents are the killer app. Native intents abstract away chain-specific complexities, enabling atomic cross-chain swaps without user-side bridging. Protocols like LayerZero and Circle's CCTP will become standard infrastructure for intent solvers, not end-users.

Evidence: UniswapX already processes over $10B in volume by routing orders through a private network of fillers, demonstrating the economic viability of the intent-based model over direct AMM swaps.