Author: Eric SJ
Web3 is transitioning from the era of user growth to the era of business model validation. In my previous article, I discussed five proven Web3 business models:
Transaction fee
Stablecoin reserve yield
Funding rate
Sale of block space
Protocol-level service fee
These patterns answer: How do these patterns make money? But there are two even more important questions:
Some income may seem very attractive, but it may not be sustainable in the long term.
Some income may appear slow, but it can be more valuable commercially.
A formula explained: Revenue = User Demand × Usage Scale × Pricing Power × Market Environment
For example, a protocol earning $100 million in a year may represent a genuine business闭环, or it may simply be riding the market cycle—but the issue is the sustainability of this cycle (e.g., past pumps).
The money earned during the casino's peak season and the money earned from renting out infrastructure may both appear as income, but their future expectations differ.
This article breaks down these five proven Web3 business models from the perspectives of revenue drivers and long-term moats.
I. Trading Fee Revenue: Monitor trading volume and user activity
Transaction fees are one of the most straightforward business models in Web3: the logic is simple—transaction revenue = transaction volume × fee rate.
Therefore, the factors affecting income are easy to break down.
Trading volume and market activity exhibit a positive correlation, making this the most obvious variable in the fee model.
During a bull market, asset prices rise, user trading intent strengthens, demand for leverage increases, and trading volume naturally grows, leading to rapid revenue growth for CEXs, DEXs, and Perp DEXs.
However, during a bear market, both trading activity and leverage demand among users decline simultaneously, leading to a noticeable drop in fee income.
This is also why transaction fee-based revenue is the most cyclical.
At the same time, increased trading volume does not necessarily indicate a stronger business model or closed-loop system; what matters more is whether your trading volume stems from genuine user growth or merely short-term incentives attracting temporary traffic.
Taking Hyperliquid @HyperliquidX as an example, its future revenue growth depends not only on the overall size of the perpetual futures market but also on its ability to continue attracting on-chain traders and market makers—the foundation of liquidity.
What exchanges are truly competing for is not the product, but the liquidity network.
Secondly, the fee rate.
The platform cannot indefinitely increase fees, as transaction fees are themselves a competitive tool. As competition intensifies, lowering fees, refunding fees, and increasing user incentives all impact final revenue.
Therefore, for the transaction fee model to grow sustainably, it must simultaneously satisfy: market expansion, increased market share, and stable fee rates.
Early DEXs and current Perp DEXs compete by offering 0 fees to attract capital into their protocols for trading, thereby increasing their market share. However, this raises an important question: Will capital still remain in the protocol once fees return to normal levels?
The OI indicator is a valuable metric. The chart below shows my OI data from last month; currently, there are no significant changes, which to some extent reflects traders' willingness to maintain exposure in a particular asset.
II. Stablecoin Income: Key Factors Are Scale and Interest Rate Environment
The reserve yield model for stablecoins essentially states that income = stablecoin supply × reserve asset yield; thus, these are the two influencing factors.
First, consider scale—it’s the most critical variable. The revenues of USDT and USDC come from the amount of USD assets locked on-chain.
If the supply of stablecoins grows and the reserve size expands, income naturally increases; conversely, if the size decreases, income will also be affected.
The chart below shows Tether's size in the first quarter of 2026, achieving approximately $1.04 billion in net profit at this scale.
Therefore, the core of stablecoin competition is not just about issuing more tokens, but about who can become the on-chain infrastructure for the U.S. dollar. In other words, under today’s regulatory environment, which stablecoin can serve as the primary issuance gateway will determine the depth of its future moat.
The second factor is the interest rate environment.
Stablecoin issuers typically hold U.S. Treasuries, money market funds, and cash equivalents. As a result, their income is highly dependent on risk-free interest rates. In a high-interest-rate environment, reserve yields increase; in a low-interest-rate environment, yields decline.
So even as the size of stablecoins continues to grow, issuers’ revenues may still be affected by interest rate cycles. However, this model has a significant advantage: it lacks extreme volatility, and growth is predictable (though this also means limited upside potential). Additionally, once funds enter the system, they are unlikely to be moved out in the short term.
Large capital also tends to gravitate toward brands that have been proven over time—meaning the longer something has existed, the wider its moat becomes, which is why new stablecoins are increasingly difficult to introduce into the market.
Moreover, this market is gradually opening new growth channels; once a project becomes a traditional entry point onto the chain, it becomes a stable, steady cash cow.
III. Interest Rate Spread Income: Assess Funding Demand and Risk Management
I previously gave two examples of the funding spread model: Aave lending and Ethena funding rate arbitrage.
They essentially profit from the difference between supply and demand for funds.
Using Aave as an example, revenue stems from borrowing demand. During upward market cycles, users' risk appetite increases, leading them to use lending and borrowing to further leverage their positions—this is the source of demand, driving higher utilization rates and consequently boosting protocol revenue. The logic mirrors that of trading fee cycles, both driven by changes in risk appetite.
IV. Block Space Revenue: Primarily driven by on-chain activity
The model for selling block space is also clear: revenue = block demand × gas unit price.
Although the structure is simple, it's worth discussing because this model has certain revenue expectations issues (in my opinion).
Theoretically, the more on-chain users, transactions, and applications there are, the higher the demand for block space, leading to naturally increased revenue—because a highway with no users has no toll value.
However, the gas price per unit is a major drawback; from an industry trend perspective, gas prices have consistently been declining, which is indeed affecting revenue.
With competition between different chains, Ethereum, Solana, L2s, and DA layers are all in direct competition, making gas fees even more intense. Many chains frequently launch zero-gas fee campaigns to attract liquidity and boost on-chain activity.
This involves a balance between increasing demand and decreasing unit prices.
Take Ethereum, for example: two cycles ago, Ethereum's logic was straightforward: limited block space → users compete for transaction ordering → increased demand → higher gas fees → increased network revenue;
But as more blockchains emerge, transaction execution efficiency improves, and more alternative options become available in the market, gas fees have been steadily pushed down—creating a business contradiction:
On one hand: more users and applications require block space;
On the other hand: technological advancements continue to reduce the cost of block space.
This is good for users because transactions are becoming cheaper. But for chains acting as “blockspace suppliers,” revenue per unit has decreased.
This is somewhat similar to the development of internet infrastructure, where bandwidth was initially scarce and expensive; later, as bandwidth continuously expanded, prices steadily declined, and market value came to be concentrated not just with those providing the underlying resources, but more so with those possessing users, ecosystems, and platform capabilities.
So the core issue for the future business model of block space is not just: "Is there demand?"
Instead: Can demand growth offset the decline in unit price?
Five: Protocol-level service fees: Based on usage scale and location
Infrastructure service fees are more like a Web3 version of SaaS—for example, oracles are a classic case.
Its revenue primarily comes from B-end clients: ongoing usage by project teams.
The more projects that use the protocol, the larger the revenue scale, and the higher the migration cost, as the replacement cost after integration is substantial.
However, this comes with a precondition: it must become an industry standard itself—like Chainlink today, which holds more than half of the oracle market, leaving little room for other projects to compete. The moat is extremely wide; even if a cheaper alternative emerges, it’s very difficult to displace existing enterprise-grade integrations.
This type of infrastructure doesn’t sell a one-time product—it sells an ecosystem position. Its long-term value depends on whether more and more projects build around it.
Summary
If you put together five types of business models:
Both trading fees and funding rate models exhibit strong cyclicality, driven fundamentally by on-chain capital's risk appetite.
2. Stablecoin reserve yields and protocol-level service fees both have strong moats, as the root cause lies in the high migration costs on the supply side.
3. The model of selling block space suffers from a continuous decline in unit pricing, requiring consideration of the trade-off between scale and unit price; valuing based solely on revenue is unreasonable (at least for now).