ERC-4337 is 1 Year Old Today - The Good, The Bad, The UglySummary

From lukasschor.eth

Exactly one year ago, the first production-ready ERC-4337 entry point contract was deployed. How far have we come since then? What does the data say? Let’s look into the good, the bad and the ugly. 👇

The Good #1: Numbers go up!

We’ve seen an incredible surge in userOps since a year ago. Up to over 2M userOps in a single month. And an accumulated 12.5M userOps in the last year. Over 3M accounts were deployed with ERC-4337 support. This is significant! 🚀

Monthly UserOps split by network

The Good #2: Awareness

We’ve seen an incredible increase of interest for Account Abstraction across the board. The @4337Mafia TG group reached 1850 members. Lots of talks at conferences, bounties at hackathons, media coverage etc. AA has become mainstream! 🎉

Beyond 4337 event in Tokyo 2023

The Bad #1: Shallow adoption

Looking under the hood, the picture is a bit more sobering. Just a few apps make up most of the adoption, primarily to enable free NFT airdrops or similar usecases. On all ERC-4337 accounts there’s only around $1M in assets.

https://www.bundlebear.com/apps/all

AUM of ERC-4337 account implementation with largest amount of accounts deployed. Second largest account implementation has $130k in assets.The retention shows a similar picture of the adoption not being that sustainable yet. Most ERC-4337 are used once (e.g. to claim an airdrop) and then never again. Even CEXs launched their smart wallets, but struggled to make them sticky.

Onchain retention of ERC-4337 smart wallets.

Large CEX's smart wallet (name blacked out as I don't want to finger-point) saw some interest during the launch but adoption has since completely plummeted.

The Bad #2: Bundler Economics

So far, bundler are not generating massive amounts of revenue onchain. While this may change, after gas costs, the entire bundler market generated only around $200k in net revenue over the last year. Granted, there is the general expectation that the adoption of ERC-4337 will grow drastically, increasing the market size for bundler. But at least for now, we are far off from having a profitable bundler ecosystem.

https://dune.com/niftytable/account-abstraction

The Ugly #1: Fragmentation

With various account implementations emerging, there starts to be more fragmentation. Users are not able to us the account created with smart wallet A in smart wallet B. Some efforts for better standardisation are ongoing, but turn out to become political themselves (ERC-6900 vs. ERC-7579 debate). At least to me it is unclear to what extend we can retain the "portability" properties of web3 accounts as we transition to smart accounts. And there might be some efforts to take advantage of this to create platform lock-in.

https://docs.zerodev.app/blog/why-7579-over-6900

The Ugly #2: Entry point contract upgrades

Within a year we already saw the third entry point contract being introduced with version 0.7.0. Each time, user technically have to upgrade to the new entry point to ensure lasting compatibility with bundlers. This process touches the core configuration of the account and is therefore quite security critical. If we want to achieve mainstream adoption of ERC-4337 smart accounts we'll have to get to a point where these upgrades are required much less frequently or even have long-term-support versions of the entry point contract.

Summary

A year after the first production-ready deployment of the ERC-4337 entry point contract, the landscape shows both, progress and challenges. User operations have surged, and awareness around Account Abstraction has significantly increased. However, issues like limited adoption depth, low retention, and fragmentation, present hurdles to be overcome. So while the tech has matured and proven scalable, the product-market-fit is yet to be validated fully.