Original Author: JOEL JOHN, SIDDHARTH

Original Translation: Deep Tide TechFlow

An agreement is a set of rules that participants in a system follow. For example, in the military, agreements dictate how people should act. Diplomats have a "protocol" that governs how they interact with each other. Protocol can be seen as a bundle of rules. In the context of machines, especially computers, protocol is defined as rules that dictate how data flows. For example, RSS is a protocol that defines how information about articles is updated in a client. SMTP defines how emails flow to an inbox. Protocol is a bundle of rules specific to a background.

On the other hand, a platform is an operating system, a social network (such as Meta), or hardware (ARM/NVIDIA) that allows a set of protocols to run on it. When you use Outlook (an application) on Windows, you use SMTP (a protocol) to transmit data to Windows (the platform). Currently, there are no natively extended platforms in Web3. Solana's mobile devices may have their own operating systems, which have been fine-tuned to meet industry requirements. Ronin has its app store, allowing the distribution of NFT-supporting games.

However, when you consider the scale of Azure, Facebook, or iOS, you will find that there is no similar scale platform in Web3. (Most likely because we currently do not need them). In 2019, both Samsung and HTC attempted to create hardware-enabled mobile devices, but I think the demand for smartphone devices supporting wallet hardware has decreased with the release of tools like Secure Enclave.

What confuses me is that applications can also be seen as the concept of protocols. Take 0x as an example, is it a protocol or an application? Matcha is an application, while 0x is a protocol that multiple DeFi products can associate with to obtain liquidity. Similarly, OpenSea has Seaport, and their protocol allows various NFT markets to share liquidity. Do you understand the main points?

Because in the early stages, it is difficult for a protocol itself to attract multiple developers, so developers often release an accompanying application to drive activity. If you are an independent application, you are likely to be replaced by other applications. OpenSea lost to Blur in the royalty war. But if you are a protocol with multiple applications built on it, the possibility of being completely replaced is greatly reduced.

Therefore, if you zoom in a little, the strategies of the past few years are relatively simple.

• Launch an application. Drive liquidity by offering token incentives.

• Grow to a certain extent, and now allow third-party applications to leverage your liquidity.

• Release a protocol with governance tokens.

Compound and Uniswap are examples of this strategy. Coincidentally, the core products they release are so powerful that people don't consider building applications on top of the protocol. Products like DeFiSaver, InstaDApp, MetaMask, and Zapper send liquidity to these products. But most user activity happens on the initially released product, the native website of the protocol.

In these cases, the teams build moats in two ways:

• First, by distribution, they become the most reputable brands in the industry.

• Second, through the network effects generated by multiple applications sending liquidity to them.

In other words, in the world of digital assets, applications can evolve into protocols (or platforms). With Roll-ups making it easier for applications to masquerade as L2, we will see more and more applications claiming to be protocols, in the hopes of increasing their valuation.

Utility

During the ICO boom, there was no clear definition of token utility. People generally believed that tokens should not engage in activities that could make them securities, apart from that, there were no other specific regulations. People would try dividends, buybacks and burns (like Binance), as well as governance rights that come with the tokens. The crux of the problem lies in linking economic value to something that can be minted at no cost.

Transaction networks like Bitcoin, Ethereum, and Ripple can claim that conducting transactions requires a small portion of the assets. As the number of transactions increases, the value of the underlying assets (such as ETH, XRP, etc.) also increases. The cost of transacting on Ethereum is equivalent to a low-cost Android device, if someone is also minting while you're trying to make a transfer.

This mindset is helpful for valuing many tokens because their value is based on the revenue generated from transaction fees. Ethereum's EIP 1559 burns a small portion of the token supply with every transaction, making it a deflationary network. This concept is very effective when you are a base layer that derives value from the number of transactions.

However, when you are not a transaction network but an application, requiring users to hold your native asset becomes a hindrance. Imagine if your bank required you to hold their stocks every time you took out a loan. Or if McDonald's servers asked you how many of their stocks you hold before serving you a burger.

Imposing a requirement of bundling actual use cases with the underlying asset leads to terrible results. Exchanges are well aware of this. That is why Binance or FTX (RIP) never required you to hold their tokens to trade. They simply incentivize you to use them by providing discounts with their tokens.

Many tokens known as governance tokens are actually hidden utility tokens. In other words, their utility comes from the idea that they can be used to govern the network itself. There is currently debate about whether DeFi has truly achieved decentralized governance, but the fundamental assumption is that holding an asset helps express opinions on how the product operates.

For many DeFi projects, this means being able to change fee variables, supported assets, and other random features involving finance. In this case, token holders do not receive income from the product, but they "control" the treasury that may generate income. Therefore, when considering fair valuations, multiples based on the number that represents the product creating $100 million in fees are relevant. The P/E ratios of Compound and Aave align with what we see in fintech companies that go public. Markets are driven by narratives in the short term, but over time, they return to rationality.

Markets are narrative machines and occasionally amplify speculation. When this happens, the valuation of a platform is more influenced by the narratives it can drive, not just the fees it generates. Simply put, if a thousand people notice that ten people are using a dApp, the token's valuation may be higher than the fees generated by those ten users.

This is because capital allocators outweigh users due to the liquidity nature of digital assets. Taking Compound as an example, over 212,000 people have tokens stored in wallets outside of exchanges. In the past month, approximately 2,000 people have used Compound for lending. By Web3 standards, this 1% ratio is still a healthy figure.

Ashwath Damodaran refers to this situation as the big market illusion. In a 2019 paper, he explores multiple venture capital firms betting on similar themes, assuming all their bets will eventually be winners. We are currently witnessing this in the field of artificial intelligence.

Billions of dollars flow into multiple companies engaged in the same business, assuming the market is large enough to support them all. Venture capital firms invest capital in the hope that the startups they invest in can stand out and have a large enough market share to justify higher valuations. Given the patterns we often see in startups, many companies end up failing. We see the same trend in the digital asset field.

When a small fraction of users appears, individuals flock to trade a certain asset. Usually, people expect utility to continue to rise and align with valuation. Soon, a new product emerges with a flashy token airdrop. Users start migrating elsewhere, causing the valuation to drop due to insufficient platform usage.

dApps vs Protocols

Now that we have established some basic economic concepts about how protocols and applications make money, let's see which of the two generates more revenue. The chart above does not include Bitcoin and Ethereum as they have a first-mover advantage. It also does not include Solana, in case you were wondering. You will notice that applications like Uniswap and OpenSea generate significantly more revenue than the average protocol. This may contradict the notion that protocols should be more valuable than applications because value flows downward (towards supporting infrastructure).

This is why simply referring to "fat protocols" as a basis for supporting new layer-two solutions is incorrect. Mature applications on Ethereum can generate more revenue than relatively young protocols as dApps tend to earn money by capturing a small fraction of the transactions happening on them. Your fees are proportionate to the amount of capital flowing through the product and your fee percentage. Uniswap and OpenSea are able to earn nearly $2.8 billion because they have high currency velocity (frequency of asset turnover) and executable fee percentages that deliver value to users.

For protocols, forcibly increasing fee percentages as usage grows would undermine network effects unless the use case can justify it. Let me explain. If your livelihood depends on paying transfer fees equivalent to a week's income in Bitcoin on emerging markets, it may be acceptable. But it's not something everyone can afford once fees go up. Bitcoin's immutability and decentralization are characteristics that justify people's willingness to pay a premium.

Bitcoin's fees are justified by the following factors:

• The Lindy effect of the network;

• Its decentralization and immutability.

However, when you introduce stablecoins issued by centralized institutions, the market will reprice the willingness to pay for the protocol. That's why Tron is at the center of stablecoin activity. Here is a quantitative approach. Last week, the average USDC transfer amount on Ethereum was close to $60,000. On Arbitrum, this number dropped to $9,000. And on Polygon, it was only $1,500. Using "average" as a metric can be debated, but the assumption is that as fees decrease, transactions below $100 become possible. The point I'm trying to make is:

• We assume that as the number of transactions and transaction costs increase, the protocol becomes more valuable.

• However, high costs often disrupt network effects where users concentrate on a single network and over time drive them elsewhere.

This is why emerging dApps on new chains have never reached the critical speed of generating enough fees. When you launch a DeFi product on Ethereum today, you are leveraging the network effects of users who have accumulated wealth through ETH, ICO boom, NFT boom, and DeFi boom, as well as the powerful infrastructure that allows people to transact, lend, and borrow. When you build on a new popular Layer 2 solution, you want users to bridge their assets and use your product. It's like doing business in a new country. Of course, you face less competition, but also fewer users.

It's like operating the only Starbucks on Mars. Is it interesting? Maybe. Can you make money? Maybe not.

Community as a Moat

We have been thinking deeply about what the moat is in Web3. Because unlike other industries, most applications in cryptocurrency are known for two characteristics.

• Open source: You allow anyone to copy what you have built;

• Capital mobility: You allow users to take their funds with them at any time.

Despite having these two characteristics, Uniswap, Aave, and Compound have maintained a relative advantage in what they do over the years. Multiple DeFi products have tried to replicate Compound, but this ultimately led to disastrous failure. So what is the moat of these products?

In this industry, the simplest measure of a moat is liquidity. If you are a capital-intensive product, liquidity is the amount of funds available to facilitate product transactions. If you are a consumer application, like a game, liquidity is attention. In both cases, the key factor driving liquidity or capital is community. So in Web3, the only real moat is community. And what keeps early community participants engaged is capital incentives or product utility.

Products like ChatGPT that significantly improve user experience can attract users without providing incentives. Blockchain technology enables applications to occasionally generate similar magic. DeFi crossed this gap in the golden age of AMM and permissionless lending, blooming in June 2020, a time we fondly remember as "DeFi Summer".

A large user group seeking to make quick money through airdrops may appear to be a community. But in reality, it is not. In the long run, this is a "cost" to the network because if you want to maintain the price, buyers of tokens must provide enough liquidity for them. For example, it was revealed yesterday that 93% of the tokens held in the Arkham Intelligence wallet were immediately transferred. Are the members who sold tokens community members or a cost to the network?

If they strategically repurchase, they can become community members. But as long as they don't need tokens to use the platform, they have no incentive to repurchase. They can allocate this money to hundreds of other tokens. DeFi products like Compound and Uniswap have not only token holder communities but also thousands of individuals who leave billions of dollars in liquidity pools of their products.

You can copy their code repositories, but you can't replicate their liquidity pools for a long enough time without establishing a strong community. Capital incentives help retain communities in the long term.

Capital incentives can be in the form of tokens given to users as a reward for performing functions on the network. For example, people who provide Filecoin storage receive tokens for their contributions. Early joining of the network is another way capital incentives accumulate for users. Bitcoin and Ethereum are similar in this regard, as their early adopters have accumulated substantial wealth through early participation and holding.

Users' capital allocation needs go beyond through shared culture. Bored Apes and the many GMs or WAGMI we saw on Twitter during the last bull market are examples of this. Culture helps individuals align their identities and keep them engaged for longer periods of time. Culture cannot be quantified, but the excitement we see around EthCC or Solana's Hackerhouses is an example of it. It provides individuals with a mechanism to build, connect, and conceive without capital investment in the conversation.

The protocol cannot operate solely on the basis of atmosphere; you need people to build on it. Developers are a combination of culture and capital that helps retain users as long-term tools. If we consider the protocol as a country, the practical tools built by developers will keep users (citizens?) on the network for a long time. The protocol can charge fees, just like a highway can charge tolls. But if the fees are too high, they will drive users elsewhere. From this perspective, it is clear that if the use case is consumer-centric, the protocol may not have been designed to make money at all.

The moat in Web3 is formed by users persistently using the network to facilitate economic transactions in applications. Each network has the same set of dApps with different branding but sells the concept of "lower transaction fees" as a unique selling point. We will soon have decentralized ecosystems and user attention will also be dispersed. Indeed, capital will flow into these ecosystems in the short term as users transact, but they will soon become ghost towns like EOS.

In the real world, you cannot duplicate a country. There is no mechanism to expand land area within borders (without violence or economic consistency). That is why people are forced to concentrate in hub cities, which have historically been ports. London, Mumbai, and Hong Kong are all similar in this regard. Concentration of people helps drive network effects within a city. Rents soar, but that also means faster groceries and better services.

In the digital realm, users are forced to concentrate in one ecosystem due to how intellectual property works and the expansion of product suites. Google launching its search engine, Gmail (2004), Android (2005), and Youtube (2006) all contributed to the stickiness of their ecosystem. Users who registered for Gmail inevitably entered Alphabet's other product suite. Apple and Meta also employ similar strategies to concentrate users within their ecosystems.

Apple takes it to the extreme by owning the entire stack from hardware to payments. Concentrating a user base helps achieve economies of scale. I mention this for a reason, and it revolves around developers.

If I were to create a demand hierarchy for early protocols and dApps, it would look like the following diagram. You need developers to:

• Create code;

• Invest capital;

• Attract users.

Without code, we're just spinning our wheels.

Since the early 2000s, venture capital firms that focus on developers have become very prominent. This is because the number of developers is a precise measure of economic activity that can occur on the protocols. Suppose you buy an iPhone because of its camera. It's highly likely that you'll end up purchasing an app that helps you edit images on the device.

So, your initial decision (the camera) drives the secondary purchase (the app). With the emergence of each new app, the ecosystem becomes more powerful because the suite of products available to users expands. The value proposition of purchasing a device is no longer just the camera but the entire open ecosystem available to users.

This kind of change was also evident in the early stages of the internet. People weren't subscribing to the "internet." In their minds, they were subscribing to a webpage that summarized what was happening on the internet. But it was only when people realized they could send emails, and embarrassingly text their crushes at school that the open internet was able to develop.

Now, imagine if there were 20 different internets competing, each with their own listed stocks and app variants with completely different brands and transaction fees. Consumers would be confused, and the internet wouldn't be what it is today. This is where we find ourselves in the Web3 native protocol.

Value Capture

Protocols need a significant amount of user liquidity to support the emerging applications built on top of them. In the era of Rollup upgrades, everyone has the ability to claim to be the next L2 (Layer 2 scaling solution). However, with each new protocol that emerges, we are dispersing the number of users for Web3 native products. There may come a time when users don't know what chain or stack is behind their tools. But we're not there yet.

During this period, it may be a mistake to expect every protocol to capture fees as much as mature dApps. The first-generation blockchain dApps were capital-intensive. The next generation may be attention-intensive. We don't have large-scale Web3 games yet because we prioritize transactions over great gameplay. Blockchain is fundamentally financial infrastructure. It's reasonable to assume that every user wants to transact.

But this line of thinking has somewhat hindered industry growth.

All of this makes me wonder what "value" is. Tokens, such as stocks, game items, and other liquid assets, always have a premium. This premium can fluctuate up and down depending on the hopes or fears of the crowd. Speculation has been a driving force in financial markets for at least eight centuries. I doubt we will change this aspect of human behavior anytime soon.

For founders, the message is clear. You can make money by driving narratives, even if the protocol fees or usage are low. Meme tokens are an extreme version of this situation. Another way is to build a dApp that can generate fees and have reasonable multiples. Aave and Compound have evolved into platforms with these traits. Both approaches require a lot of work.

The best founders we have seen are able to drive narratives and platform adoption, usually disaster ensues when only one aspect is present. Protocols or applications that offer unparalleled core utility may have higher adoption rates due to their stickiness. This is the embodiment of Peter Thiel's belief that "competition is for losers." The more crowded the market segment, the lower the chances of a newcomer achieving higher acceptance or retention of sticky users. All of this is just considering user concentration. What about protocol economics?

Joe Eagan of Anagram presents a good analogy here. The best protocols have a behavior incentive mechanism, much like Amazon. Amazon operated at almost no profit for a long time, but the inherent network effects paid off in the longer term. The most extensive group of sellers on Amazon met the largest group of buyers. In the long run, any "successful" protocol might have similar attributes. Low fees and the broadest range of applications built on top so that users don't need to go elsewhere for their daily functions. Monetization of such a protocol may come from the richness of the data it leaves behind on the blockchain.

This scenario might be launched without a token. It could charge in dollars instead of new native assets. Imagine paying a transaction fee of $0.0001 with USDC. Users would receive a dollar "recharge" to their wallets every 10,000 transactions completed at local stores. But the problem is, most base chains can't do this because their native tokens are essential to their security models, and we're not quite sure how such a protocol would monetize. Such protocols could scale exponentially without requiring users to spend significant funds every time protocol usage increases, as is the case with Ethereum today.

If we believe that blockchain is the infrastructure for transactions, and all actions on the internet will become transactions, we may inevitably need low-cost protocols and fixed costs. Alternatively, we might end up with 50 new L2s, each with extremely high valuations because the market loves novelty. The market can facilitate both, and perhaps that is the beauty of it, practicality combined with speculation.