Revisiting Web3 Aggregation Theory: How Startups Build Market Moats?

Original Title: 《Revisiting Aggregation Theory》

Written by: Joel John

Compiled by: Frank, Foresight News

A year ago, we wrote an article about the theory of aggregators in the Web3 era. In the Web2 era, aggregators benefited from the collapse of distribution costs, bringing together many service providers. Platforms like Amazon, Uber, or TikTok provided services to users through hundreds of service providers (vendors, creators, or drivers) and profited from it, while users benefited from having more choices.

For creators, using aggregator platforms can also scale their influence. For example, I post on Twitter instead of Lens because my audience is more concentrated on Twitter.

In Web3, aggregators primarily rely on the collapse of verification and trust costs. If you use the correct contract address, you don't have to worry about whether the USDC tokens you swap on Uniswap are real USDC tokens. Similarly, on NFT markets like Blur, you don't have to spend resources verifying whether each NFT traded on the platform is authentic because the network bears that cost.

Aggregators in Web3 can more easily check asset prices or find their listing positions by examining on-chain data. Over the past year, most aggregators have focused on integrating on-chain data sets and making them available to users—these data may relate to prices, yields, NFTs, or bridging assets.

The assumption at that time was that monopolies would be established by companies expanding quickly enough in the form of aggregator interfaces. I specifically cited Nansen, Gem, and Zerion as examples. Ironically, in hindsight, my assumption was incorrect, which is what I want to write about today.

Weaponized Tokens

Don't get me wrong. A few months after the first article was published, Gem was acquired by OpenSea, Nansen raised $75 million, and Zerion also raised $12 million last December. So if I look at these matters as an investor, my assumption was correct, as each of these products is a category leader in its own right.

But what interests me in writing this article is that I believe the relative monopolies they would have do not yet exist. Instead, they all face new competition that has emerged over the past year, which is an ideal characteristic of an emerging industry.

So what has happened in the years since? As I wrote in “Royalty Wars,” the relative monopoly of Gem (and OpenSea) was challenged with the launch of Blur in the NFT market. Similarly, the blockchain data platform Arkham Intelligence has combined an exciting user interface, potential token issuance, and clever marketing strategies, including challenging Nansen through referral reward tokens. Zerion may be comfortable, but with the launch of Uniswap's new wallet, it may also see its market share eroded.

Do you see the trend here? As some projects choose to offer tokens to users, those aggregators that have not issued tokens and have grown easily with equity backers are now facing risks. As we gradually move deeper into a bear market, the concept of "community ownership" will become very important, as the limited number of consumers still in the market want to maximize every dollar they spend. Additionally, being rewarded for using the platform instead of paying access fees is also a novel experience.

Thus, on one hand, companies with positive long-term cash flow will see revenue decline, while on the other hand, they will see users flocking to competitors. Is this sustainable? Absolutely not. Here’s how it works:

A project launches a product that implies token issuance, and it’s even better if it’s tied to a referral program. For example, Arkham Intelligence offers tokens to users accessing their platform, and considering the possibility of airdrops, more and more users will spend time on this product.

This is an incredible way to stress-test the product, lower customer acquisition costs, and guide network effects within the product. But the challenge arises in retention; once token rewards are no longer offered, users often turn to other products. Thus, most developers who "imply" they will issue tokens do not know how broad their user base is.

The person in the image below summarizes the philosophical foundation of the average person in the crypto space today, so profound, symbolizing the selfishness that drives our world:

In any case, there has been a historical trend of users abandoning projects that do not issue tokens. This trap comes from the fact that founders (may) believe that users acquired through token incentives are sticky. Under ideal conditions, the graph of token rewards versus product users should look like this:

But the reality is that as token incentives diminish, the initially influxed users almost completely abandon the project. Without the initial incentives that attracted them, they have no reason to continue contributing to the product, a phenomenon that has plagued DeFi and P2E projects over the past two years.

Users who accumulate and hold tokens are the new "community" members, wondering when asset prices will soar enough for them to exit (assuming market participants are rational actors whose behavior aligns with their own interests).

My initial view was to simply integrate the feature sets of multiple products into one interface, using blockchain as the backbone infrastructure to serve as a lasting moat. But this may be wrong, so I wonder why the leading players with relative advantages in Web3 would lose to others.

For instance, Binance overturned Coinbase, which in turn faced competition from FTX; OpenSea faced competitive pressure from Blur; and Axie Infinity developer Sky Mavis may also face pressure from new entrants like Illuvium.

Why do users leave Web3 over time? How can we retain users for the long term?

In Web3, when everyone can publish a version with embedded tokens, what can serve as a moat? I've been pondering this question because we live in a market of narrative shifts, with a new "hot" thing every quarter. This is why the venture capitalists I follow have overnight transformed from remote work experts to experts dealing with geopolitical tensions.

Of course, if you are trading assets, this is effective (by the way, this is the "use case" for most cryptocurrency users). But if you want to build a foundational asset that grows over time (like equity in Google or Apple), then frequent trading is likely a bad idea.

If you ultimately want the time, money, or energy spent to grow without active management, the only way to achieve this is for a product to do two things:

• First and foremost, retain existing users;
• Second, actively expand to avoid competition from other projects eroding your market share;

What should you do? (Note that when people start thinking about moats and retention, it indicates a bear market.)

Competition is a Loser's Game

Part of the explanation for this phenomenon is ranking companies on a spectrum between novelty and convenience. In the early days, primitive tools like NFTs were novel, attracting those willing to try products at any cost.

We were happy to deal with wallet mnemonic phrases and handle fiat deposits because the novelty of using "digital currency" was enough to attract us. If you notice the curiosity users have about "Ordinals," you will realize how patient users can be.

This patience is partly due to the profit factor of early trends, where speculation and profit motives drive users to endure rough user experiences.

On the other end of the spectrum are the highly convenient tools we rely on daily. Amazon is a typical example of an aggregator that has addicted us to convenience. Consumers may benefit from purchasing items from niche stores that are not on Amazon, and the pricing of certain sellers on Amazon may also be incorrect.

But when making decisions, Amazon ensures we don't have to worry about payment methods, delivery times, or customer support. This "saving" of mental labor translates into higher spending on the aggregator (attention or capital).

Many sellers come to Amazon precisely because they understand consumer behavior in the market, which differs from what users see when shopping directly in stores.

Tim Wu's 2018 article summarizes the efforts people are willing to pay for convenience:

We are certainly willing to pay a premium for convenience, even willing to pay higher prices than we realize. For example, in the late 1990s, music distribution technologies like Napster made it free to access music online, and many people used this option. Although music can still be easily obtained for free today, no one really does that anymore. Why? Because the iTunes store introduced in 2003 made purchasing music more convenient than illegally downloading it; convenience trumped free.

Returning to the spectrum I initially mentioned, new technologies often pay users to try them, whereas highly convenient applications can charge users high fees, provided they satisfy users' cravings for convenience.

Today, most consumer-facing applications face the challenge of being in the middle of the spectrum, which I call the "valley of death." They are neither novel enough to entice people to try what they have built, nor convenient enough that users rely solely on them without external support.

Skiff, Coinbase Card, and Mirror excel at this end of the convenience spectrum because they can replace their traditional competitors.

However, take gaming, lending, or identity verification sectors as examples, and you will understand why these themes are currently unable to scale on-chain.

Most applications in the middle position make a fatal mistake by competing with each other. They first increase user acquisition costs and hiring costs through advertising and recruitment, then compete with toxic memes and narratives aimed at opponents. As Peter Thiel said, competition is a loser's game.

When startups begin to compete in small, specialized markets, there are usually no winners. In his words, the only way startups can transition from survival struggles is to have monopoly profits, but how can they achieve that?

New Moats

If companies in Web3 want to move away from tokens as a growth lever, they have only three levers to focus on: cost, use case, and distribution. Some of these situations have occurred in the past, so let me explain them one by one.

Cost

Stablecoins have become a killer use case for cryptocurrencies because they provide a better experience globally than traditional banks. For example, India's UPI innovation may be more cost-effective for domestic payments, but transferring funds between Southeast Asia, Europe, or Africa, or merely transferring balances between U.S. bank accounts, is more sensible using on-chain transfers.

From the user's perspective, the incurred costs are not only the amount spent on transfers but also the time and mental energy allocated to transferring funds. Debit cards serve e-commerce like stablecoins serve remittances: they lower the cognitive costs of making transfers.

When compared to most consumer-facing yield-generating mobile applications—you can offer one or two points of yield calculated in dollars, but when you consider bankruptcy risks, value considerations become irrelevant.

Distribution

If you gather niche users in emerging fields, distribution can become a moat. Think about how Compound and Aave opened up entirely new lending markets. Few people think it’s valuable to spend $100 of ETH to borrow $50, but many overlook that there is a market that has not been served—mainly those crypto millionaires who do not want to sell their assets in a bear market.

So if you think that it’s the people in emerging markets who cannot access credit lines driving DeFi lending volumes, you are mistaken; it’s the crypto millionaires who are using it, a demographic that previously did not have access to banking services. Therefore, focusing on these can make you a "hub" in a niche area, allowing you to concentrate attention on a single feature. Coingecko and Zerion are two companies that excel in this regard.

Given that the marginal cost of guiding users to use new features is almost zero, iterating and adding new revenue sources to the product itself becomes cost-effective. This is why players like WeChat (in Southeast Asia), Careem (in the Middle East), and PayTM (in India) often perform well.

When players like Uniswap launch wallets, they are essentially trying to gather users in one interface, where more features (like their NFT market) can be pushed at a lower cost.

Use Case

Tools like ENS, Tornado Cash, and Skiff have carved out their unique user bases, relying on the product to provide functionalities that traditional alternatives cannot. For example, Facebook will not link your wallet address to your identity, and your bank does not provide the privacy protections offered by Tornado Cash.

These are products with user stickiness, often with no other alternatives that can reach the level of the product. Pioneers in new use cases need time to educate users and help them understand the utility, but they also have the advantage of capturing large new markets.

Just like in the early days of LocalBitcoins, it was the only place for peer-to-peer trading, which helped them gather liquidity in emerging markets like India and kept them ahead until 2016.

It’s difficult to scale by focusing on any of these levers during a bear market; the examples I mentioned above have gone through multiple market cycles. For instance, part of what allowed Axie Infinity to stand out was that the team built for two years before 2020—before the next bull market arrived, the team had built the "muscle" needed to establish community, maintain tokens, and balance the interests of investors (selling tokens) with user interests (earning tokens).

This explains why, when the market declines, developing tools and infrastructure becomes trendy from a venture capital perspective. Because to address the lack of interest from retail users, one can choose to focus on the business-to-business (B2B) side—you build tools for developers, and they are responsible for attracting retail users themselves.

Well-known players like Coinbase recognize this, which is why they release tools like wallet APIs during bear markets.

From Novelty to Convenience

LI.FI is a multi-chain liquidity aggregator that provides SDKs for developers wanting to extend their applications or users to multiple chains.

Suppose Metamask or OpenSea wants to enable developers to allow users to move assets between chains like Polygon and Ethereum. In that case, LI.FI provides a simple SDK to determine the best ways to transfer funds across bridges and DEXs, allowing developers to focus on what they do best.

Imagine aggregators like Li.Fi as Lego blocks that developers can plug into their applications to help users move assets between chains at the lowest cost.

Several players are engaged in the same business, but I use LI.FI as an example because they have genuinely met the conditions I mentioned above, and some of the things they have been doing validate the moats I mentioned earlier:

• LI.FI has started focusing on enterprises rather than retail users. If they capture the long-tail developers building applications that may require cross-chain transfers, they don’t have to worry about directly attracting users;
• Companies using the product can save research and maintenance time. In a bear market, everyone wants to conserve resources as much as possible. Therefore, by default, sales of products like LI.FI become relatively straightforward;
• From the end-user's perspective, the aggregator provides the best cost basis for transfers. Therefore, people want to use products that have already integrated their SDK;
• LI.FI is often the first to integrate new blockchain networks among competing projects in the same space—this places it at the forefront of competitive scarcity;

Finally, their target audience primarily consists of the last participants to leave cryptocurrency. Generally, a year into a bear market, those speculating and trading in the industry are advanced users, and you don’t have to spend much money educating them.

Now, don’t get me wrong; LI.FI is not the only cross-chain bridge aggregator on the market. Although it meets the conditions I mentioned above, such as cost, number of users, and use cases, it’s hard to see any of them building a moat. But I am particularly interested in how they evolve from novel tools to convenient tools.

In the early days, users relied on cross-chain bridges because the transfer process required painfully waiting for exchanges to relay transactions—you had to transfer funds through a centralized platform, undergo security checks, and then hope the funds would arrive, rather than just needing a few clicks.

Of course, DeFi degens are sending billions across chains today, but the average person (like your high school friend) doesn’t care about that.

So how do you survive when the novelty wears off? If you notice how Nansen and LI.FI operate, you can get the answer by observing who they sell their product services to:

LI.FI primarily sells to developers, while yesterday, Nansen launched Query, a tool that allows enterprises and large funds direct access to Nansen data, claiming it is sixty times faster than the closest competitor in querying data. So why are both companies focusing on developers?

This relates to the fact that if a company sells to advanced users, it can simultaneously serve as a tool of novelty and convenience. For example, when developers decide whether to integrate with LI.FI, they typically have a simple calculation in their minds: does the aggregator effectively spend less time and money compared to integrating a single cross-chain bridge?

Similarly, for anyone using Nansen Query, the question is whether the tool saves enough time and effort to justify its cost. If the cost of implementing it internally is lower than paying a third party (like LI.FI), decision-makers are likely to prefer not to build from scratch.

For a company, the fastest way to escape the "valley of death" on the spectrum from convenience to novelty I mentioned in the previous image is to focus on a small number of users willing to pay high fees because your product is now a convenience tool. This provides enough runway for the product to build sufficient user interest and become the preferred convenience tool.

I discussed this framework with Nansen founder Alex, who had a different take, stating that users seek value regardless of market conditions. In a bear market, enterprises and networks are the biggest customers; they need very specific data sets, but these data sets are often not available from third-party vendors.

So adjusting the product to their needs and showing them the value means you can gain more revenue with less competition.

Returning to Basics

When I wrote about aggregation a year ago, I mistook product functionality (using blockchain) for a moat. Since then, a plethora of DeFi yield aggregators have emerged, but most have failed. If competitors can launch products with the same features and better user experiences, or if they issue tokens as Blur did, then simply integrating blockchain may not mean much. In this environment, it’s essential to think about what can truly differentiate a product.

As I write these words, some patterns become very clear:

First, the cost of acquiring users will skyrocket in a bear market because retail users have low interest in it. Unless the product has special novelty or convenience, it is in a strange position;

Second, businesses built for other companies (B2B) may be able to achieve compound growth to survive and then dominate in a bull market, just like FalconX;

Third, poorly designed tokens are temporary moats and long-term liabilities. Few communities accumulate meaningful value for tokens over a sufficiently long period.

When you consider retail markets or niche markets like DeFi, it’s clear that the average person doesn’t care which chain or how decentralized it is; they care about the value they can derive from it. Blockchain can help increase the value that end users can obtain. But founders may often fall into the trap of constructing products and selling services for venture capitalists (or token traders) without genuinely building moats based on cost, convenience, and community.

This may sound a bit cynical, but it’s worth considering.