Crypto wallets are a crucial part of the web3 user experience but struggle to evolve with their growing responsibilities as they were initially not designed with future use cases in mind. Even though their purpose has changed in recent years, the foundational design and architecture has remained more or less the same.
I think that in order to address emerging use cases and user needs, we have to unbundle the wallet on multiple levels.
The wallet started out as a tool for holding your keys and providing you with an interface to manage transactions on the blockchain. However, with the new use cases that have evolved since then, it has had to take on the responsibility of representing your digital identity and carrying it across apps and chains. As a result, wallets as they are built today combine two different functions in one product. Similar to a "bank account", they allow you to manage assets and transactions while also serving as your "passport" to web3.
This „all-in-one“ solution makes sense from a technical perspective (as the wallet still holds your keys), but might be limiting to the user experience as user needs evolve.
As Jay Drain Jr. writes in "The Crypto Wallet: Unlocking Digital Identity“, users might eventually demand wallets that feel more like a „place where they can hang out and interact with their assets“, rather than a „technically complex tool“. In addition, as Derek Walkush puts it in his thread, they might more and more „seek a very specific product experience tailored to their particular needs“, which could subsequently lead to an unbundling of the wallet into verticals, similar to the verticalization of horizontal marketplaces like Craigslist.
At the same time, with the emergence of new use cases and your identifying data getting tied to your keys, the wallet's responsibility of representing our digital identity is becoming increasingly complex, requiring us to develop tools that are better suited for this task.
If these assumptions are correct, moving forward with the „all-in-one“ wallet we use today could present us with a dilemma as we try to meet those potential requirements in the future.
At the root of this problem lies a point Kei Kreutler makes in „Inventories, Not Identities“, that "identity is always inherently relational, defined in relation to environment, authority, and self“. Identity, therefore, is not something static that we carry around with us, but rather a fluid construct that changes depending on the context. This fluidity allows our identity to adapt to new situations and function in different settings, and gives us the freedom to experiment with different ways of self-expression as opposed to defining ourselves in only one way. In this sense, supporting the inherent fluidity of identity is key to maintaining its composability, and thus an essential part of the wallet's responsibility of transporting our identity across the web.
So while on one side the wallet should become a place where people want to spend time, creating a more specific user experience initially leads to a less flexible perception of the identity tied to it, as it will always be defined in relation to that context as well. This limits the number of situations to which it can adapt, making it less composable overall. Conversely, maintaining the fluid nature of identity gets in the way of creating the context needed for the more specific product experience users might be looking for in the future.
I think in order to overcome this dilemma, we have to separate these two areas of responsibility and build solutions that focus on only one of them.
Splitting these two functions would result in two new product groups:
On one hand, we would have wallets focused on creating a place where people can spend time and interact with their assets, and that are tailored towards a specific use case or user persona in order to meet different user needs. These verticalized wallets would not necessarily have to work as standalone solutions, but could also be integrated into other applications (e.g. "in-game" wallets). Gaby Goldberg writes about verticalized wallets in "Stop Calling it a Wallet“ and illustrates it with two examples:
„If this hypothesis is true, I expect to see an emergence of embedded wallets for different verticals or themes, each with their own terminology to onboard new users. Perhaps your digital fashion will be stored in a wardrobe, and you’ll show off your music NFTs by way of a discography.“
On the other hand, we would have wallets solely focused on managing our digital identity rather than also providing us with an interface to interact with our assets, and therefore more likely to ensure identity composability across apps and chains by supporting its fluid nature. From the user's perspective, this horizontal identity wallet might feel like the web3 version of a password manager, a single point from which they can manage the keys that represent their digital identity, and to which their identifying data beyond on-chain transactions and assets could eventually be tied. I think that having all aspects of their digital identity in and around one place like this could help users understand the concept of self-sovereign identity, which requires a complete shift in thinking, as it is the opposite of what we learned in web2.
These two specialized wallet types would be better suited for their respective tasks and could thus better address the various requirements than the combination products we use today.
On a technical level, the wallet is responsible for private key management, managing transactions, as well as providing the user with an interface, leaving very little room for improving the user experience (forcing the wallet to be this „all-in-one“ solution).
Therefore, with the current architecture, unbundling the wallet would lead to two problems:
First, having to handle all of these technical responsibilities at the same time makes creating a wallet a complex task, which might be a barrier for many to build in this space. Second, unbundling the wallet into verticals would require users to manually keep track of their keys and the identities associated with them, as keys and transactions would have to be managed separately within each of these wallets, making the user experience increasingly complex.
Consequently, in order to enable vertical wallets at scale, while simultaneously allowing users to manage their keys conveniently, we might need to approach wallet architecture in a more modular way as well and free the wallet from the burden of being everything at once.
In „Account Coordinator: A Proposal for the Future of Wallets“ Billy Rennekamp proposes a separation of these technical tasks and with that a new kind of wallet, the „Account Coordinator“, responsible for „keeping track of which keys have which capabilities on which networks“ and for managing transactions across apps and chains. While I know a lot has changed since the piece has been written and that there might be other models for how wallet architecture could be approached moving forward, I still wanted to refer to it here as a high-level conceptual model for how wallet architecture could be modularized to illustrate how this could be a more supportive technical foundation for unbundling the product experience of the wallet in the way discussed above.
For one, such a model would allow verticalized wallets to be built at scale, as a separation of the interface from the underlying infrastructure would simplify the process of developing a wallet while simultaneously optimizing it for freedom in design. At the same time, it would lay the foundation for a tool like the "Account Coordinator“, which would allow users to manage their transactions, keys, and the identities associated with them conveniently across the verticalized wallets and apps. This would remove the complexity of having to manage everything manually and create the basis for the identity management wallet proposed above.
Accordingly, a modular approach like this could solve the two main problems the current wallet architecture would face in the unbundling process, and thereby enable the development of these two specialized wallet types.
Closing thoughts
To wrap it up, I think that unbundling the wallet into verticals seems like a logical response to users seeking a more specific experience tailored to their individual needs, as this has been the case with many products in the past when they reached this point (e.g. Craigslist). However, since the current wallet architecture is not able to support this, it is becoming apparent that we might subsequently have to unbundle the core responsibilities of the wallet as well in order to meet user needs in the future and thus unbundle the wallet on multiple levels.
If you have any thoughts on this piece, please get in touch! Any feedback is greatly appreciated :)