Aggregation Theory in DeFi
This week at DeFi Summit in London, I presented my thoughts on using Ben Thompson’s Aggregation Theory as a model for understanding DeFi. It’s something that has been top of mind for us at Topo Finance, as we aggregate functionality of underlying protocols to create our interest rate optimization product, but it’s clear the DeFi ecosystem as a whole is thinking hard about these types of frameworks. In every presentation I attended, someone asked the presenting business, “How do you plan to monetize?” This is a subset of the the question Aggregation Theory answers: Where and how does value accrue?
Traditional and Digital Value Chains
Aggregation Theory came about as a way to explain and understand how companies like Google, Facebook, Netflix, Uber, and AirBnB came to disrupt their traditional counterparts and establish near-monopolies in their industries. It ultimately boils down to the value chain, which for a consumer market consists of suppliers, distributors, and consumers, and how that value chain changes with the dawn of the internet.
In the pre-Internet world, supply and distribution were integrated –– think editorial content+newspapers, video content+broadcast capabilities, taxis+dispatch. In all cases, distribution is a really hard and expensive problem to solve. This leads to scarce supply, as you’re not going to create, say, editorial content unless you have a distribution channel lined up. Naturally, companies with distribution channels integrate backwards into supply, and their main challenge is maximizing and protecting the value of their supply.
The core disruption of the Internet is that distribution costs go to zero. Instead of having to print editorial content on thousands of pieces of paper, you can now just type it up and post it on the Internet. The effect this has is the modularization of supply: a huge abundance of commoditized supply existing in little chunks all across the web. This is the sign of true digital disruption: when an industry’s supply is digitized and modularized, incumbents should start sweating. (This also means that for all of Fintech’s innovation, they are not true disruptors).
Incumbent business models are so threatened because the hardest problem for a business has shifted from “How do I protect and maximize my limited supply?” to “How do I manage this huge amount of scattered supply for a customer?”
This shift in the problem creates a shift in value chain emphasis, in Ben Thompson’s words:
This has fundamentally changed the plane of competition: no longer do distributors compete based upon exclusive supplier relationships, with consumers/users an afterthought. Instead, suppliers can be commoditized leaving consumers/users as a first order priority.
Aggregation Theory Network Effects + Drivers
Supply management becomes the key question for aggregators. With the overabundance of supply that comes with digital goods, users find the greatest value in curation and discovery of supply. Google search results and Netflix recommendations provide curated options, while Uber and Lyft automatically assign the best option for the user –– but both approaches rely on wrangling a massive supply into a convenient user experience. With consumers as the highest priority, UX becomes the start of the aggregator flywheel.
The aggregator’s virtuous cycle is fairly intuitive: a better UX in managing supply leads to more users, which incentivizes suppliers to join the aggregator’s platform, which then increases the user experience and the cycle continues. This leads to a winner-take-all scenario, in which one or two aggregators attract all users and suppliers.
As mentioned, this is fairly intuitive, but the underlying drivers showcase the true characteristics of aggregators. UX is critical because aggregators have a direct relationship with users. These users are added at zero marginal cost, allowing the platform to scale. The platform is necessarily a demand-driven network, where suppliers are granted access to users when they join.
DeFi: The Modularization of Capital
It’s worth understanding in detail how aggregators are formed and how they find success, because things get a lot more complicated in the open ecosystem of DeFi. At a macro level, however, the process is identical. I find looking at AirBnB’s example particularly relevant to the DeFi industry.
We can see that hotels traditionally have integrated property and trust –– you see a big building with the Hilton logo and know you can spend a safe night there. Reservations are handled on a hotel-by-hotel basis and are therefore modularized, with each individual hotel having a relationship with the guests. AirBnB, on the other hand, modularizes property in people’s homes, and integrates trust and reservations directly in the core experience of their app, held in the user’s hands.
Ben Thompson here makes a statement highly relevant to the DeFi space:
The commoditization of trust is far more injurious to hotels than you might think…In the pre-Airbnb days travelers — and sublessors — justifiably prioritized trust above all else. In other words, the implication of Airbnb building a platform of trust is not that a homestay is now more trustworthy than a hotel; rather, it’s that the trust advantage of a hotel has been neutralized, allowing homestays to compete on new vectors, including convenience, cost, and environmental factors.
This is of course the opportunity presented in DeFi. As we modularize capital (Be Your Own Bank), we integrate trust and financial services directly into the functionality of the ecosystem.
This is where DeFi gets interesting, however, because financial services are not like reservations. They are a diverse set of functions, which all need to be accessible to the same users. So the DeFi offering actually looks more like the below, where individual financial services aggregate users and suppliers, and then are aggregated themselves.
Aggregating Aggregators: Composability and DeFi Value Chains
Unlike traditional aggregators, which exist in closed systems and data silos, DeFi aggregators exist in an open ecosystem and are inherently modular themselves (#moneylegos). This means that we have capital aggregators, services aggregators, and interface aggregators.
Capital aggregators, like Compound, Set, b0x, dYdX, are the underlying protocols that provide functionality of a financial service. As they provide a better UX, earn more users, and attract more suppliers, they are naturally aggregated by a higher level. These higher level aggregators then route more users and suppliers to them, meaning the key to their success lies in their ease of integration and security rather than their UX. This leads me to believe there will be no winner-takes-all at this level, particularly as service aggregators like Topo route capital according to an algorithm as opposed to brand notoriety or user interface. If an Uber/Lyft competitor wanted to challenge those apps’ dominance, they’d need a super unique value prop to win users and suppliers. If a platform wanted to challenge Compound, all they would need to do is provide better rates, easy integration, high security, and get plugged into a platform like Topo to access necessary liquidity.
One layer above capital aggregators, we have service aggregators, which provide interoperability among protocols while adding additional functionality. This is where Topo, Totle, Staked, Dex.Ag, Idle Finance, and MetaMoneyMarket sit. A core belief here is that modular supply is inherently messy, and when we’re talking about completely new market structures and incentive mechanisms around people’s finances, the ability to provide users with a simple way to know they are allocating their capital efficiently is extremely powerful. Unlike capital aggregators, user experience at this level is paramount, as service aggregators seek to abstract complexity and automate decision-making, while opting to both own the user and allow integration from above.
At the top of the chain, there are the interface aggregators (wallets and asset managers), which provide the user with a single touchpoint to access underlying functionality. Interface aggregators are focused on owning the user, a mandate of traditional aggregators and the starting point for the aggregation flywheel. That flywheel must start with a differentiated UX, however, and it’s unclear if that’s possible in an open system in which anyone can plug in and copy functionality. We see this in the offerings of Zerion, Argent, DeFiSaver, and Instadapp, all which are essentially abstracting the same underlying smart contracts to provide the same functionality. This is not to say that those services are not useful, because they are critical and high-quality products, it is merely a question as to whether that strategy will produce significant network effects.
It should be obvious that this is an oversimplified depiction of the current ecosystem, and that we’re already seeing things shift relating to service and interface aggregators. Both groups are attacking UX in a slightly different way, and they will converge in their approach over time.
More Questions than Answers?
One of the clear conclusions that arises from the above aggregated aggregator models is that an open ecosystem operates VERY differently from a closed system. We see that all the aggregators above have some of the hallmarks of traditional aggregators (zero marginal cost for users, demand-driven network, increasing UX quality), even as they lack other characteristics (necessity of direct relationship with users, clear winner-take-all outcomes).
Ultimately, this means that Aggregation Theory is not an ideal framework for evaluating value accrual for platforms built on top of a DeFi network. It raises more questions than it answers. Is liquidity rather than users king? What is the impact of forking an existing protocol? What happens when a more conservative wallet provider with a huge user base enters the space?
Zooming Out: Ethereum as Aggregator
While I feel that my foray into evaluating DeFi businesses through the lens of Aggregation Theory has not provided clear answers, it does serve as an effective framework for DeFi as a whole.
While only time will tell how the platforms on Ethereum accrue value, we already see Ethereum well into the winner-take-all effects of aggregation. It’s core UX has brought users and suppliers to its network in a way that no other blockchain has seen, and those building on top are only bringing better UX and fellow builders and users.
It’s nothing new to say that Ethereum is going to win the DeFi games, but Aggregation Theory provides more weight to that assertion than simply “Eth has more devs.” It’s the user experience those devs create which accelerates the aggregator’s virtuous cycle.
This is obviously not a comprehensive analysis of the topic — in creating the talk I cut ideas from my brain and in writing this piece I cut ideas from the talk. In conversations around this topic, many people have brought up additional questions and theories of how this all plays out, which is exactly the point! Please comment or reach out with holes to poke or ideas to contribute :)
If you’re involved or interested in the DeFi space, Topo Finance is releasing a closed beta of our interest rate optimization product in the next couple weeks, if you’d like access join our Telegram, email me at email@example.com, or find me on twitter @trentofelmore. We’re on a mission to create decentralized and automated wealth management tools, we’d love you along for the ride.
We will be launching Topo’s open beta at Devcon in Osaka, if you’re going to be there, we’d love to have you at our launch party!
Follow Topo on Twitter here.