Will aggregators be the biggest value capturers in DeFi?

This article is from IOSG Venture.

Aggregators, such as Google, Facebook, Uber, and Airbnb, are the largest value capturers on the internet. When an industry is digitized, a significant portion of its profits flows to the aggregators of that industry online.

DeFi is inherently digital, so will the profits of the financial industry, or currently only the profits limited to DeFi, flow to the aggregators in DeFi, repeating the story that has played out on the internet?

Exploring this question is the reason I wrote this article. It consists of three parts: the first part introduces aggregation theory, much of which is derived from Ben Thompson's series of articles on aggregation theory; the second part discusses the particularities of applying aggregation theory to the DeFi space, which largely stems from the fact that DeFi is based on blockchain rather than the internet. The third part is an observation and reflection on the aggregators in DeFi and platforms that appear to be aggregators.

Aggregation Theory

What is an aggregator, and why can it capture value? Let's take a look at what happened in the accommodation industry after digitization.

In the traditional hotel industry, as shown in the diagram below, accommodation rooms and trust are integrated; users engage with this integration through booking services. However, after digitization, in Airbnb, accommodation rooms and trust are separated; trust is integrated into Airbnb's booking service, and users engage with accommodation rooms through Airbnb.

Image: https://stratechery.com/2015/netflix-and-the-conservation-of-attractive-profits

This change has led to a shift in the value chain. In the Hotel model, hotels (suppliers) that integrate trust are differentiated and scarce, giving them the upper hand in the value chain, and profits flow to them. In the Airbnb model, accommodation rooms (suppliers) are modularized, leading to a large commoditized supply; at this point, the aggregator, which integrates trust and acts as an intermediary between the commodity and users, takes the lead in the value chain, and profits flow to it.

The same thing happened in the taxi industry. In the Uber model, taxis are modularized, and trust, along with scheduling and payment services, is integrated into Uber, changing the value chain and directing profits to the integrator.

Image: https://stratechery.com/2015/netflix-and-the-conservation-of-attractive-profits

From these two examples, it is not difficult to see that when the supply of an industry is digitized and modularized, the business challenge is no longer to maximize limited supply but to manage vast decentralized supply for users. Aggregators are an inevitable and central role in digital markets; they integrate and reassemble the value chain of the industry, thereby capturing value.

But what kind of products are aggregators? Not every product that has aggregation or will have functionality is an aggregator. My personal preferred description of an aggregator is: "Any product that creates customer value on its own is not an aggregator, because ultimately its customer acquisition cost will limit its growth potential."

In general, aggregators need to meet the following three key characteristics. Many products only satisfy one or two of these characteristics, making them platforms rather than aggregators.

• Direct relationship with users
• Zero marginal cost of serving users
• Decreasing cost of acquiring users with scale

The two "spectrums" shown in the diagram below can help us understand aggregators more deeply. The first spectrum is the differentiation of suppliers. The closer to the left side of the spectrum, the less differentiation there is among suppliers, making it more of a typical aggregator; for example, in Facebook, suppliers are almost indistinguishable, "On the internet, nobody knows you're a dog"; conversely, the closer to the right side of the spectrum, the greater the differentiation among suppliers, making it more of a platform rather than an aggregator; for example, in Microsoft, suppliers are specific and integrated with the product.

Image: https://stratechery.com/2018/the-moat-map

The second spectrum is the internalization of network effects. The closer to the left side of the spectrum, the more internalized the network effects are, making it more of a typical aggregator; for example, Facebook, whose product itself is a relationship network, has network effects that directly act on the product, which are fully internalized; conversely, the closer to the right side of the spectrum, the more externalized the network effects are, making it more of a platform rather than an aggregator; for example, Microsoft, which has almost no internalized network effects, has its user numbers not directly related to the utility users can derive from the product.

Image: https://stratechery.com/2018/the-moat-map

By correlating the spectrums with the characteristics of aggregators, we find that the de-differentiation of suppliers leads to a large commoditized supply, which helps aggregators achieve "zero marginal cost of serving users"; the internalization of network effects makes the product utility positively correlated with the number of users, which helps aggregators achieve "decreasing cost of acquiring users with scale."

Aggregators on Blockchain

Whether aggregators are on the internet or on the blockchain, they should meet the basic characteristics of aggregators mentioned in the first part of the article. But how do aggregators based on different networks differ? Currently, three points are observed:

1. Trust, the integration of trust is different. On the internet, trust is integrated into the aggregator, which is a significant reason why aggregators can capture value; but on the blockchain, trust is integrated into the public chain, and this part of the value is captured by the public chain. From the perspective of the value chain, the value chains of the internet and blockchain are entirely different.

2. Data, non-exclusivity of data. On the internet, aggregators exclusively possess various data, including user data, supplier data, operational data, etc., which is a significant reason for their monopolization and subsequent profit extraction; however, on the blockchain, data is open, making it difficult for aggregators to monopolize profits through data walled gardens.

3. Modularization, proactive modularization. The digitization and modularization of traditional industries are generally led by aggregators, with aggregators existing first and then modularized supply. In this model, the interfaces of the modules are provided by the aggregator, and suppliers then connect their products to match the aggregator; consequently, the more suppliers connect, the more users are attracted, and the more users attract more suppliers… This process leads to the difficulty for other similar aggregators to break the monopoly once an aggregator achieves scale, as the product utility of the two will differ significantly.

However, DeFi is inherently digital and modular; it has modularized supply first, followed by demand for aggregators. In this model, the interfaces of the modules are standardized, and besides suppliers actively connecting to the aggregator, the aggregator can also proactively connect suppliers to itself. This means that as long as the aggregator actively connects to supply, the gap in product utility between it and the already dominant aggregator will not be insurmountable.

The internalization of network effects is the primary reason aggregators tend toward monopoly; it hinders the entry of competitors, but on the blockchain, the way aggregators establish advantages through network effects will be significantly weakened, making competition possible.

Based on the above three points, a rough deduction about the value and profit acquisition of aggregators is:

1. Aggregators manage decentralized supply for users and provide a good user experience; aggregators are at the top of the stack, serving as user entry points and traffic gateways. These are values inherent to the aggregators themselves, and aggregators on the blockchain will undoubtedly capture them.

2. DeFi aggregators do not integrate trust, and the part of the value related to trust belongs to the public chain; therefore, in this regard, their value may be lower than that of internet aggregators; trust is integrated into the public chain, allowing suppliers to also be trusted, enabling users to directly obtain goods from suppliers without relying on aggregators, which may also weaken the value of aggregators.

3. The characteristics of aggregators determine that they will tend toward monopoly, and aggregators on the blockchain are no exception; DeFi aggregators will also gain the additional benefits and profits that come with monopoly. However, while aggregators on the internet can exploit upstream and downstream through monopoly, infringing upon their rights and interests, the misdeeds of aggregators on the blockchain can be challenged, making it perhaps more difficult for them to profit through such behavior.

4. Aggregators on the internet, due to strong internalized network effects, mostly tend toward a situation where the strong get stronger, with one or a few monopolists capturing all the value in the value chain at the aggregator level; DeFi aggregators operate in a fully competitive environment, and although there will also be a head effect, there may be a group of leading products that collectively capture the value in the value chain at the aggregator level.

DeFi Aggregators and Platforms

When it comes to aggregators in DeFi, people often think of projects like YFI (yield aggregation) and 1inch (DEX aggregation), but from the logic of traditional internet aggregator theory, they are not aggregators but more like "platforms." Similar to the internet industry, they resemble platforms for ordering organic ingredients or flowers, where the platform delivers ingredients or flowers to users weekly, but the specific products are not chosen by users; instead, the platform selects them from certain suppliers.

Specifically, YFI helps users choose yield strategies, thereby creating customer value on its own; from the spectrum perspective, its suppliers are highly differentiated, the internalization of network effects is weak, and its cost of acquiring users does not decrease with scale, making it closer to a platform than an aggregator. Similarly, 1inch is not a DEX aggregator; it is a DEX platform that helps users choose trading options.

In contrast, projects like Zerion and DeBank, often referred to as asset management platforms, are essentially closer to aggregators. In DeFi, asset management and wallets are the closest to aggregators; they face users and manage assets and goods supply for them. These two may ultimately evolve into similar product forms, with asset management incorporating wallets and wallets engaging in asset management, both evolving in the same direction to better serve users.

The diagram below is a layered representation of DeFi, where aggregators connect upwards to users and downwards to the supply of goods/services. Unlike internet aggregators, DeFi aggregators not only integrate applications but also integrate platforms, which are included in the value chain consolidated by the aggregator.

Another difference is that on the internet, users generally only interact with aggregators to obtain goods; in DeFi, users can obtain goods/services not only through aggregators but also directly from platforms and applications, as mentioned above: trust is integrated on the blockchain, and things on-chain, regardless of which layer they are on in the stack, are trustless.

Although aggregators and platforms may seem somewhat similar and are often confused, they are fundamentally different entities. The former is an open system with modularized suppliers and internalized network effects; the latter is not. The value of aggregators is greater than that of platforms because their scales will not be on the same order of magnitude.

However, this does not deny the value of platforms. In DeFi, three additional factors will increase the value of platforms: first, in DeFi, many users may not know how to choose suppliers, so a platform that can help them select products (providing strategies, routing, etc.) is important, which is different from the internet where users generally know how to choose; second, in DeFi, platforms are situated within the value chain integrated by aggregators, allowing them to benefit from this value chain that has users; third, in DeFi, trust is integrated on the blockchain, making it easier for platforms to be adopted by users and easier to expand.

Conclusion

We did not enter the digital age after the internet appeared; we entered it after aggregators emerged. Aggregators have digitized one field after another, thus digitizing our lives.

Aggregators bring convenience but also a fear of being dominated with no recourse; the monopoly issue of aggregators may be one of the hardest problems to solve in this era, but things on the blockchain inherently possess the ability to resist monopolies, and DeFi is fortunate. While aggregators change lives, they also change industries, as they alter the "open rate" of industries, typically seen in shopping and short videos. Finance is a low open-rate matter for most people, so what changes will a financial aggregator with a good user experience and a rich array of financial products bring to the open rate of finance? DeFi is worth looking forward to.