Thin application: comprehensive upgrade of the "fat protocol" of blockchain value capture theory

Written by: Joel Monegro, co-founder of Blockholder, an investment institution in Placeholder, wrote the famous "Fat Protocol" in 2016. Compilation: Perry Wang

This article is an update of the previous two articles, The Blockchain Application Stack (2014) and Fat Protocols (2016).

When I started thinking about the issue of block application architecture in 2014, I described it as a "functional" hierarchical stack. The first generation of blockchain was designed with "overlay networks" as the bottom layer, which provided many different decentralized services, forming the "shared" protocol and data layer. Above, it is a stand-alone application using these protocols, which redistributes its services to users:

Blockchain Application Stack (2014)

I predict that as encryption technology takes over the network, the architecture will become mainstream for new online services within a decade. It is therefore interesting to review the development of this idea over the past five years.

The most obvious flaw of this idea is that we should build everything on top of the Bitcoin network (at the time Ethereum was not yet available). Now that we have a variety of blockchain networks to choose from, the situation has changed dramatically. We now refer to the "overlay network" as layer 2. Today, the above technical architecture has a better name: "Web3 Stack". But overall, this architecture is valid.

When thinking about this problem two years ago, this model took me to the concept that was hotly debated in the technical world at the time: "Fat protocol". I thought at the time that most of the market value of cryptocurrencies should be captured at the "protocol level", while the value at the network level should be captured at the "application level".

Fat Protocol (2016)

This idea has evolved from the application stack. Most "jobs" and data exist at the protocol layer, and applications tend to provide more limited interface services. In 2014, the "business model" at the protocol level was not obvious. When we invested in budding cryptocurrencies in Union Square Ventures (USV), the potential of the tokens became clearer.

In 2014 and 2016, there were not many real-world examples to watch at the time. It was the "germination period" of the grand blueprint of IT history. But now, we have looked at hundreds of encryption protocols and applications in many markets. The time has advanced to 2020, and after Placeholder's attempts on many aspects of these ideas, now is a good time to redefine old ideas and consolidate what we already know.

Cryptographic service infrastructure

Internet giants tend to keep expanding their platforms and lock users into their own proprietary interfaces to achieve a monopoly of information. In contrast, encrypted networks tend to provide a single service and cannot “own” the corresponding interface because they do not control information. Specialization is beneficial because the more decentralized a network is, the more difficult it is to implement a set of services in one interface, and it will not reproduce the monopoly of giants such as Google, Amazon, Facebook, etc. Therefore, consumer applications in encrypted network / Web 3 are independently built on multiple "composable" protocols, making use of what we call "encryption service architecture"-this is much like microservices, but the difference is caused by Sovereign components.

Cryptographic Services Architecture

In Decentralized Finance (DeFi), people call it "Lego Currency". Think of Zerion (where Placeholder has an investment), Instadapp, and Multis. They all use many of the same protocols, such as Ethereum, Compound, Maker, and Uniswap, to build similar cryptocurrency applications. In this way, they deliver a complete set of financial services-settlement, lending, lending, trading, investment, and more-without having to develop and maintain a complete set of functions, infrastructure and liquidity within the enterprise.

These protocols provide specific services in many excuses, and many applications on the protocol share resources and data, eliminating the risks faced by centralized platforms. Sharing these infrastructures has reduced costs across the board.

Decentralized autonomous organizations (DAOs) and edge applications such as games also present the same mechanism.

Crypto service architecture is undoubtedly a great gospel for startups. Entrepreneurs can quickly launch new applications at low cost by outsourcing many functions to a variety of networks. The cost and resources of adopting a protocol for each application are equal. This is completely different from Internet infrastructure such as Amazon Cloud Services (AWS), for example, the smaller the enterprise, the higher the cost of purchasing AWS services. Many crypto startups stand out because they launched full-featured products before the first real financing. They are the first examples of the new model of "thin applications" that can prove capital efficiency. In contrast, the emerging funds in the traditional Internet world need to be closer to astronomy if they want to compete with established giants digital.

Carry information

Unmanaged is another way to reduce application costs in the crypto space.

The actual Internet giant model relies on creating a data monopoly, because locking users in their own proprietary interfaces is the most powerful way to extract value from user information. These Internet companies are also competing with their collaborating companies, so users have to access many different platforms, and these platforms have obtained user information. With the increasing cost of Internet security and the introduction of new regulatory policies, data has become a burden on companies. Companies that are financially capable and able to bear such costs therefore have a competitive advantage. And resource-constrained startups have to find other business models to compete.

As a user in the encryption field, you carry your own data with you, and no one can control the monopoly. When you sign in to the encryption app by connecting to a wallet, you are sharing a "key" that looks up the information you need on the relevant network. You can share this information with any app, so when you move from one interface to another, your data is with you. You can then control the "private key" (basically a password) required to perform operations such as signing messages or authorizing transactions. As a result, you can effectively keep your data and no one can manipulate it without your consent-unless you entrust the key to the custodian.

For example, in the world of crypto collectibles, artists can use Rare Art to tokenize their work and then sell it on OpenSea. Then someone (like Jake) can buy it and show it in Cryptovoxels's virtual gallery.

A similar example is this: In DeFi, if you use Maker with Zerion and then log in to Instadapp with the same key, you can immediately interact with Maker loans. Because they are built on the same standards and networks, these applications can interoperate by default, and users can move freely between interfaces without losing information or functionality.

The combination of an encrypted service architecture and an unmanaged data model enables startups to compete more effectively with centralized traditional enterprises. This echoes the way open standards drive the IT market cycle. Giving users ownership and control of information can help businesses reduce many costs while meeting the needs of many consumers today. This model does require companies to abandon many traditional means of competition for online services. You lose control, but gain potential efficiency and scale. Adopting this model allows businesses to run at very low cost, and many applications can benefit from each other's success as they share resources at the protocol level.

As a network, thin applications can expand capacity more effectively throughout the market. Every digital artwork on Rare Art indirectly increases the practical value of OpenSea, and activities on Instadapp benefit Zerion (and vice versa). However, what many people don't know is when they can really create long-term business value and defense when everything is open.

Value capture vs. return on investment

The fat protocol believes that the encryption protocol will capture more value than the application interface. A common mistake people make is to confuse the notion of value capture with return on investment. Many have concluded that, although the original text listed the application layer as a requirement for protocol value growth, there is no return on investment in the cryptographic application layer.

It needs to be clear that the less the total value obtained at the application layer does not mean that there are fewer opportunities for excess returns available to the application business. This does not mean that the protocol layer always returns.

Value capture is more related to the total addressable market (or total market size) and other macro factors, while the benefits vary depending on factors such as cost base, growth rate, and ownership concentration. The difference between protocols and applications is how these elements are combined.

Looking at value from a cost perspective is a more precise way of thinking about value distribution. The basic principle is that in the market, cost is an important factor that determines future value, so we can estimate its value structure by studying the cost structure of the market.

In the field of encryption, the network at the protocol layer bears most of the production costs, so more investment is required-this means that more value must be accumulated at the protocol layer to maintain balance, otherwise no investment will occur. Applications have much lower operating costs and require less investment, so they have less demand for market value. However, network ownership and cost structures are much more fragmented than private companies. In general, investing in tokens usually brings you a small portion of value, and this portion must be larger in order to cover your cost of funds.

For example, if the Ethereum network is worth 15 billion U.S. dollars, purchasing 10 million U.S. dollars of Ethereum's native token ETH will get about 0.06% of the market value of the network (based on the current supply), and want to achieve 4 times the return (The value rose to $ 50 million), and the network value of Ethereum needs a net increase of $ 60 billion. At that time, if a seed investment of US $ 1 million (10% of the company's shares) was invested in a successful application business, the company ’s market value “just needs” to increase by US $ 490 million, and the corresponding value of its equity will also reach US $ 50 million—if Considering subsequent additional investments, less than $ 10 million may be required.

But networks and businesses deal with value in different ways. The power to increase the value of a public network from $ 200 to $ 90 billion is very different from the power to increase a company's valuation from $ 10 million to $ 500 million. Every time a token is traded on the open market, its price is determined to be chaotic. In this case, the value of each dollar investment in the network increases and loses much faster than investment in private companies. This complexity adds a lot of leverage to investment in and out. At the same time, business value may be a well-known feature, but private early-stage investments may face higher risks in many unpredictable ways.

Finally, we must consider the combined value of all the underlying protocols used to assess the relevant value. For example, Zerion operates under protocols such as Ethereum, Maker, Compound, and Uniswap. The total value of these networks is much greater than the individual value of Zerion or its competitors. But again, this has little to do with the applications that use these protocols and the return on investment for the enterprise. Encrypted networks may expand to store trillions of dollars in value, but their growth will flatten out. Then, most of the market value can be stored in the protocol layer, and the excess return on investment is transferred to areas that grow faster. However, the current encryption network is far from equilibrium, and we have found high return opportunities at both the protocol layer and the application layer.


Thin applications run at a lower cost because they pass much of the cost on to protocols and users. But its competitors have access to the same production and data resources, so they can replace each other in ways that are not possible on traditional networks. In a way, it is similar to the retail model. The storefront is the "interface" of various products in the retail model, and is distinguished from each other through branding, comprehensive processing and customer experience. However, the field of encryption is not "B2B2C", but P2B2C-Protocol to Business to Consumer, which can be regarded as protocol-enterprise-consumer.

The protocol provides specific services that are bundled at the application layer and distributed to consumers. As in retail, the price is determined by the encrypted network that generates the service (such as the suggested retail price), and fair competition at the application layer makes it difficult for anyone to raise prices unfairly.

This pattern has benefited users a lot and can solve many of the problems we encounter on the web. However, this raises new questions about application-layer defense. When everything is open and competitors can easily replace each other, how do you create long-term business value and defense?

In a crypto economy, application companies must create value beyond the capabilities of the protocol. In many cases, known business models like subscription or transaction fees are a reasonable choice. But as the infrastructure matures and applications become leaner, we need new business models. There are many interesting experiments. For example, Blockstack innovates with "application mining," and NEAR provides its developers with a royalty structure (both reminiscent of Amy James's salutary protocols concept). I am curious about how they will continue to develop, even though I cannot fully accept that the protocol should determine the economics of its application. It will take more space to talk about the scope of the entire experiment, so here I will focus on three general strategies: construction cost moat, vertical integration, and user-miner model.

Establishing a cost moat means that the agreement does not include centralization costs and outreach. The scale of these costs is inherently defensive, as competitors cannot afford the costs of this scale. For example, Coinbase fully taps into two very expensive cryptoeconomic extensions that users are willing to outsource-fiat currency trading and escrow, and profit through classic economies of scale, creating high business value. Nothing new in the sun. The market will not allow Coinbase to increase the cost of cryptocurrency transactions, but they will charge users a fee to cover the large investment required to provide these services. In contrast, leaner applications such as Zerion don't internalize these costs, so they don't charge extra-but as a result they can't use Coinbase's business model or justify charging the same fees. This road works, but it's expensive.

The vertical integration in the crypto economy explores the possibility that explosive applications may gather enough users and eventually become "own suppliers". They can turn themselves into "suppliers" (such as miners) in their integrated agreements and provide services directly to users. We saw this in the old retail model with chain store brands, and it happened again with Amazon promoting its own products to replace competitors' products. Amazon's growth model is by overwhelmingly close to zero profits on the goods sold on its platform, and then using the platform and its perfect demand data to create its own supply chain with unprecedented efficiency. Can crypto applications take a similar approach? What if the application causes the blockchain network to fork later? Will the market allow it? The outside world does not want the application layer to have too much control over the protocol. This is the case in the online world. But this is a possible outcome.

The last user-miner model idea is to use tokens to distribute value and benefits to users. Usually it works by allowing users to mine a certain amount of application-native tokens (as opposed to protocol tokens) and then get benefits like discounts or rewards-but there are many differences. At first glance, it looks similar to customer loyalty / reward systems for very large commodity companies such as airlines and credit card companies. Except for programs that do not provide any benefit. The innovation lies in designing a token model so that users can profit from the growth of the application. It not only brings marginal benefits such as discounts, but also distributes the benefits of application growth to users.

For example, Nexo and Celsius, who provide crypto mortgages, use tokens in this way. Nexo offers discounted rates when you repay interest rates with NEXO tokens. In Celsius, the more CEL you pledge, the more favorable interest rate you can get. If you choose to accept Celsius Celsius to pay interest with CEL, your deposit can also have a better interest rate. Due to the limited supply of NEXO and CEL tokens, as these applications increase in use and more and more people buy and use their tokens, they may appreciate in value. So there are other benefits beyond simple discounts. We've even seen this model in software-as-a-service (SAAS) companies such as Blox, which can offer you monthly fee discounts if you choose to pay with their tokens. How can we further mainstream this model?

What attracts me most is the user-miner model, because it represents true business model innovation. The examples listed earlier are built more like traditional web applications, and they are more centralized and require hosting than thinner applications like Zerion. But I like the key to their pledge model is how they change the user-service relationship. Traditional Internet users are forced to lock because of the centralization of data. Encrypted applications, even if the construction mode is more traditional, do not have the ability to lock users like traditional Internet companies. However, the user-miner model will create a "voluntary opt-in" economic lock-in, and by turning it into a stakeholder, users will benefit from the success of the enterprise. It builds defensiveness through user-ownership relationships rather than user lockdowns. This model has produced a series of fascinating results that I will explore in future works.

Source link: