Why Account Abstraction Infrastructure is the Real Web3 Bottleneck

The problem is infrastructure, not cryptography. ERC-4337 solved the core protocol-level challenge, but the user experience is still broken by a lack of interoperable, reliable services for paymasters, bundlers, and key management.

AA shifts complexity from users to service providers. This creates a new infrastructure moat for platforms like Stackup, Biconomy, and Candide that manage gas sponsorship, transaction bundling, and social recovery.

Fragmentation kills composability. A wallet using Pimlico's paymaster cannot seamlessly interact with a dApp relying on Alchemy's, creating walled gardens that contradict Web3's open ethos.

Evidence: Despite ERC-4337's launch, AA wallets represent less than 1% of on-chain activity, with infrastructure complexity being the primary adoption barrier.

User experience is the bottleneck. Scaling solutions like Arbitrum and Optimism have reduced gas costs, but the fundamental wallet and transaction model remains hostile. Users still manage seed phrases, pay for failed transactions, and cannot batch actions.

Smart contract wallets are the substrate. Protocols like Safe and Argent prove the demand for programmable accounts, but they operate as isolated islands. The ERC-4337 standard provides the missing universal infrastructure, enabling a shared mempool and paymaster network.

The real scaling is cognitive. The next 100M users will not tolerate the current friction. Bundlers and paymasters abstract away gas and sponsorship, enabling use-cases like social recovery and session keys that L2s alone cannot provide.

Evidence: Safe's $100B+ in assets and the rapid integration of ERC-4337 by Starknet, Polygon, and Base demonstrate that the market prioritizes flexible account infrastructure over pure TPS gains for adoption.

Smart accounts are not interoperable. A Biconomy or Safe smart wallet on Polygon cannot natively sign a transaction for an Optimism dApp, forcing users back to seed phrases and breaking the abstraction promise.

Paymaster services are fragmented. Gas sponsorship from Etherspot or Pimlico is a siloed feature, not a network primitive, creating vendor lock-in and preventing a universal 'gasless' standard.

Bundler infrastructure is nascent. The ERC-4337 bundler network, run by Stackup or Alchemy, lacks the decentralization and liveness guarantees of base-layer blockchains, introducing a new centralization vector.

Evidence: Over 5 million ERC-4337 UserOperations have been processed, yet daily active smart accounts remain a fraction of Externally Owned Account (EOA) wallets, proving adoption is gated by infrastructure, not demand.

Account abstraction adoption stalls without robust infrastructure. Smart accounts require secure key management, gas sponsorship, and batched transactions, which are non-trivial to build in-house.

Vertical integration creates walled gardens. Projects like zkSync's native account abstraction or StarkWare's smart accounts offer seamless UX but lock users into a single L2 ecosystem, limiting interoperability.

Modular stacks enable permissionless innovation. Standards like ERC-4337 and infrastructure from Stackup, Biconomy, and ZeroDev let developers assemble best-in-class components for bundlers, paymasters, and signers.

The winner-takes-most dynamic is premature. The infrastructure layer is fragmented; no single bundler or paymaster network has achieved dominance, creating a strategic opening for new entrants.

Evidence: ERC-4337's UserOperation mempool remains underutilized, with most AA activity still flowing through vertically integrated L2s, highlighting the modular stack's deployment challenge.