Original author: Decentralised.Co
Original translation: Block unicorn
This article was inspired by a series of conversations with Ganesh Swami and covers seasonality of revenue, evolution of business models, and whether token buybacks are the best use of protocol capital. This is a follow-up to my previous article on the stagnant state of cryptocurrencies.
Private capital markets such as venture capital oscillate between excess liquidity and scarcity. When these assets become liquid and outside capital pours in, market frenzy drives prices up. Think of a newly launched IPO or token offering. Newfound liquidity allows investors to take on more risk, which in turn drives the birth of a new generation of companies. When asset prices rise, investors seek to move funds into early applications in the hope of earning higher returns than benchmarks such as ETH and SOL. This is a feature, not a bug.
Cryptocurrency liquidity follows the cyclical cycle of Bitcoin halving. Historically, market rebounds occur within six months of Bitcoin halving. In 2024, ETF inflows and Saylors purchases became supply absorbers for Bitcoin. Saylor alone spent $22.1 billion buying Bitcoin last year. But Bitcoins price surge last year did not translate into a rebound in the long tail of small altcoins.
We are witnessing a time when capital allocators are liquidity-strapped and attention is spread across thousands of assets, while founders who have been working hard on their tokens for years are struggling to find meaning. Why would anyone bother building a real application when launching a meme asset can generate more financial returns? In previous cycles, L2 tokens enjoyed a premium driven by exchange listings and venture capital backing because of perceived value. But as more and more players flood the market, this perception (and its valuation premium) is disappearing.
As a result, L2s have lower token valuations, limiting their ability to subsidize smaller products through grants or token-based revenue. This valuation surplus in turn forces founders to ask an old question that plagues all economic activity — where does the revenue come from?
So transactional
The above chart does a great job of explaining how crypto revenue typically works. For most products, the ideal state is what AAVE and Uniswap are like. Blame it on the Lindy Effect or early-stage advantage — both products have maintained fees for years. Uniswap can even add front-end fees and generate revenue. This shows how well it is defined by consumer preferences. Uniswap is to decentralized exchanges what Google is to search.
In contrast, friendtech and OpenSea have seasonal revenue. During the NFT summer, the market cycle lasted two quarters, while social finance speculation lasted only two months. Speculative revenue from a product makes sense if the scale of the revenue is large enough and consistent with the product intent. Many meme trading platforms have joined the club of over $100 million in fees. This number is the size of what most founders can expect in a best-case scenario, either through tokens or acquisitions. But for most founders, this kind of success is rare. They are not building consumer applications; they are focused on infrastructure, where the revenue dynamics are different.
Between 2018 and 2021, VCs heavily funded developer tools in the hope that developers would attract a large number of users. But by 2024, two major changes have occurred in the ecosystem. First, smart contracts enable unlimited scaling with limited human intervention. Uniswap or OpenSea do not need to scale teams in proportion to trading volume. Second, LLM and AI advances have reduced the need to invest in cryptocurrency developer tools. Therefore, as a category, it is at a moment of reckoning.
In Web2, API-based subscription models worked because there were so many users online. However, Web3 is a smaller niche market with few applications scaling to millions of users. Our advantage is a high revenue per user metric. The average user of cryptocurrency tends to spend more money at a higher frequency because blockchains enable you to do that — they make it possible for money to move around. Therefore, over the next 18 months, most businesses will have to redesign their business models to capture revenue directly from users in the form of transaction fees.
This isn’t a new concept. Stripe initially charged per API call, Shopify charged a flat fee for subscriptions, and both have since moved to charging a percentage of revenue. For infrastructure vendors, this model translates fairly directly to Web3. They will cannibalize the market with a race to the bottom on the API side — perhaps even offering their product for free until a certain volume of transactions, and then negotiating revenue share after that. This is the ideal scenario.
What would this look like in practice? One example is Polymarket. Currently, the UMA protocols token is used for dispute resolution, and the token is tied to the dispute. The greater the number of markets, the higher the probability of a dispute. This drives demand for the UMA token. In a trading model, the required margin could be a fraction of the total bet amount, such as 0.10%. For example, a $1 billion bet on the outcome of the presidential election would generate $1 million in revenue for UMA. In a hypothetical scenario, UMA could use this revenue to buy and destroy their tokens. This approach has its benefits and challenges, as we will soon see.
Another player doing this is MetaMask. The transaction volume processed through its embedded exchange function is about $36 billion. The exchange revenue alone is over $300 million. A similar model applies to staking providers like Luganode, where the fees are based on the amount of assets staked.
But in a market where the cost of API calls is decreasing, why would a developer choose one infrastructure provider over another? Why one oracle over another if revenue sharing is required? The answer lies in network effects. Data providers that support multiple blockchains, provide unparalleled data granularity, and can index new chains faster will become the first choice for new products. The same logic applies to transaction categories such as intent or gasless exchange facilitators. The more chains supported, the lower the margins, the faster the speed, and the higher the likelihood of attracting new products because this marginal efficiency helps retain users.
Destroy all
The shift to tying token value to protocol revenue is not new. In recent weeks, several teams have announced mechanisms to buy back or burn tokens in proportion to revenue. Notable among them are SkyEcosystem, Ronin Network, Jito SOL, Kaito AI, and Gearbox Protocol. Token buybacks are similar to stock buybacks in the U.S. stock market - essentially a way to return value to shareholders (or in this case, token holders) without violating securities laws. In 2024, the U.S. market alone will spend about $790 billion on stock buybacks, compared to $170 billion in 2000. Whether these trends will continue remains to be seen, but we are seeing a clear bifurcation in the market between tokens that have cash flow and are willing to invest in their own value, and tokens that have neither.
For most early stage protocols or dApps, using revenue to buy back their own tokens is probably not the best use of capital. One way to do something like this is to allocate enough capital to offset the dilution from newly issued tokens. This is how the founder of Kaito recently explained its approach to token buybacks. Kaito is a centralized company that uses tokens to incentivize its user base. The company receives centralized cash flow from its enterprise clients. They use a portion of that cash flow to perform buybacks through market makers. The amount purchased is twice the amount of newly issued tokens, so in effect the network becomes deflationary.
Ronin takes a different approach. The blockchain adjusts fees based on the number of transactions per block. During peak usage, a portion of the network fees goes into the Ronin treasury. This is a way to control the supply of an asset without necessarily buying back the tokens themselves. In both cases, the founders designed mechanisms to tie value to the economic activity of the network.
In future posts, we will dive deeper into the impact these actions have on the price and on-chain behavior of the tokens participating in such activities. But what is clear at this point is this - as valuations are suppressed and the amount of venture capital flowing into crypto decreases, more teams will have to compete for the marginal funds that flow into our ecosystem. Since blockchains are essentially funding rails, most teams will move to a model that charges fees proportional to transaction volume. When this happens, if teams are tokenized, they will have an incentive to issue a buyback and burn model. Teams that do this well will be winners in the liquidity market.
Of course, one day, all this talk of price, earnings, and revenue will become irrelevant. We will once again spend money on dog pictures and buy monkey NFTs. But if I look at the current state of the market, most founders who are worried about survival have already started a discussion around revenue and burn.
