15 min readProfessional investors are warming up to the idea that Ethereum may host popular applications in the future, however, for many, the investment case for its associated asset, ETH, remains largely unclear. Unlike traditional assets with clear valuation methods, most crypto assets, like ETH, lack well known fundamentals. So when the Ethereum community successfully sparks investor curiosity, articulating their vision of growing adoption, it raises a key question: how does adoption translate into investment potential? What does adoption actually do to ETH’s long-term supply and demand?
Late last year, we took a stab at answering these questions. We examined Ethereum’s mechanics, confronting how protocol rules and user interactions combine to affect ETH’s supply/demand equation. Ignoring analogies and ideological explanations, we focused on measurable factors that directly influence ETH’s value, and the relative importance of each.
The analysis led us to a key insight: the primary driver of ETH’s value is the demand for Ethereum transactions — not staking yield, monetary adoption, financial collateral or whatever else. It hinges on how much users are willing to spend for the services Ethereum provides, very interestingly.
While a major step forward in our opinion, this raised further questions. If transaction fees should be the focus of long-term investors, what use cases on Ethereum drive fee spend? And which categories of use cases will drive future fee spend? How sensitive is each Ethereum service to the cost of transacting?
This report is meant to answer, or bring us closer to answering, these follow-on questions.
We’ll present the results of a new model that categorises and aggregates Ethereum transaction data over time, allowing us to see just how much users are spending and have spent on different services, such as marketplace swaps and stablecoin transfers, as well as the major projects driving such transactions. In other words, we’ll do a full unpacking of Ethereum usage, but in terms of the money users have actually spent to use it.
In the end, we hope to further the conversation on what a rational investment strategy may look like for Ether. This time, by measuring what Ethereum users have been trying to achieve, in hopes that our answers will move us closer to imagining and perhaps even estimating what it will be used for in the future.
Ethereum Usage is Dominated by Application Interactions and Token Transfers
Ethereum usage has evolved significantly since its inception. Initially, Ethereum functioned primarily as a venue for simple asset transfers, a foundational use case that remains important but has since been overshadowed by more complex interactions with applications and infrastructure (Examples include protocol staking, MEV, bridges, oracles and second-layer technologies).
Early on, Ethereum was envisioned as the “World Computer” — a platform capable of executing custom instructions of any kind, and hosting globally accessible, unstoppable applications. To these lofty goals, Ethereum developers have prioritised flexibility and access as protocol properties, leaning on the sophistication of the masses to iterate on the platform design and deploy useful projects.
Consequently, and perhaps not to anyone’s surprise given its global reach, Ethereum has evolved into a playground of speculation. It’s home to a wide array of custom-made digital assets and apps, alongside a growing number of tacked-on infrastructural components that shape the overall user experience. It has attracted substantial user demand; fee spend in H1 2024 totalled nearly $1.5 billion — significant, though notably lower than the $3.5 bn seen at the heights of the H1 2021 bull market.
From what our data shows, by 2018—its third year—Ethereum began to see a broadening utility. Transaction fees, once dominated by simple transfers and managing primitive smart contracts, widened to simple apps, digital identity systems, and business operations (such as on-chain withdrawals). This trend continued with increasingly functional applications, particularly in the financial and gambling sectors, such as Etherdelta, Idex, and Etheroll.
At this stage, users were no longer merely sending tokens; they were engaging with complex, multi-step interactions that leveraged Ethereum’s auto-execution and composability. Transactions were touching many applications in one fell swoop, seamlessly checking off instructions along a path to settlement.
Since 2020, an emerging trend has been the adoption of platform infrastructure, either created by general developers in response to user demand or the Ethereum Foundation itself. These developments have extended Ethereum or re-jigged how it works under the hood. This includes the likes of protocol staking, MEV, bridges, oracles and second-layer technologies, which have all become integral in how the Ethereum machine works, and are indicative of a pursuit towards more advanced usage down the line.
On the surface, Ethereum’s ability to consistently deliver higher utility through more complex transactions is a positive for ETH demand. However, the hard truth is that a very small set of services consistently makes up the majority of Ethereum usage, and these sets largely revolve around speculation or simple value transfer, not necessarily the type of complex “real-world utility” use cases originally envisioned by the developers of the Ethereum Foundation.
Having now looked at the large-scale picture, let’s have a look at each of the main usage categories and which sub-categories and/or specific projects are driving such activity.
Application Interactions are Dominated by Marketplaces, Specifically Uniswap
As Ethereum's ecosystem has grown, the types of applications that users engage with have become wide-ranging. One category, however, has consistently dominated the landscape: digital asset exchanges. In particular, marketplaces where users trade the long-tail of cryptoassets, both collectible and fungible.
In the first two years, Ethereum was a place of experimentation, with developers testing the platform’s capabilities through simple gambling applications like Etherdice and Rouleth. These early apps were significant in that they showcased Ethereum’s potential to handle more than just basic transactions, laying the groundwork for the sophisticated applications that would follow.
Nowadays, on a transaction fee basis, 90+% of usage is with marketplaces. It’s been a volatile shift as shown above — with gaming, phishing scams, and NFTs apps periodically dominating — but on-chain exchanges (often called DEXs), led far and away by Uniswap, have more or less been the dominant use case on Ethereum since summer 2020.
Uniswap alone consistently captures a substantial share of Ethereum transaction fees — 15% in H1 2024. It’s a testament to a core value Ethereum delivers, and its users place, on the ability to speculate freely on digital assets (like ETH, app tokens, and stablecoins).
While NFT exchanges saw a significant surge in the mania of 2021, their relative share of transaction fees has since dramatically declined. In H1 2022, OpenSea commanded nearly half (42%) of all transaction fees spent on applications, totalling over half a billion dollars (US$ 572m). But the collapse in demand has been so immense that more fees were spent using OpenSea in Q1 2022 ($433m) alone than on all NFT marketplaces combined thereafter (Q2 2022 - Q3 2024, $296m).
Other exchanges like 1inch, 0x, and Metamask also contribute meaningfully to financial services as the leading subcategory of application usage, and the emergence of new players like Maestro indicates that competition in this category continues and that there is ongoing demand for user-friendly, liquid, long-tail trading.
While Ethereum supports a wide variety of applications, marketplaces — and Uniswap in particular — are the primary drivers. It’s clear that the ability to trade assets with easy access, transparency, and across the same settlement system is a core component of users’ transactional utility.
Token Transfers are Dominated by Ether and Stablecoins
Another long standing foundational use case on Ethereum are token transfers, and they continue to play a central role in the network’s activity. As Ethereum’s ecosystem has expanded, the types of tokens being transferred have diversified significantly, but Ether (ETH) and stablecoins have emerged as the dominant assets in terms of transaction fee spend.
As the first and, for a time, the only major asset on Ethereum, Ether naturally commanded the majority of fee spend associated with token transfers. However, the introduction and eventual adoption of the ERC-20 standard in 2017 inspired a major shift, with a variety of new tokens minted within the Ethereum ecosystem taking on an increasingly dominant role.
The standard outlined how to easily create customizable tokens, which in turn led to an outbreak of digital assets in an era commonly referred to as the “ICO boom”. As more tokens were created and traded, the landscape of Ethereum assets, and how much users were willing to pay to move them, took on a new life.
Since then, stablecoins in particular have become a significant component of transaction activity. Beginning in mid-2019, Tether (USDT) started gaining traction as a widely-used trading pair and familiar medium of exchange, and the impact of the stablecoin category only grew with the introduction of Circle’s USDC in late 2020. Within Ethereum’s fee landscape, during certain periods, the fees paid for stablecoin transfers have come to rival or even surpass ETH.
While USDT and ETH remain the most significant tokens in terms of transaction fees, the broader digital asset ecosystem has seen the introduction of literally millions of other tokens. Many of these tokens are associated in some manner to specific applications or projects, contributing to the long-tail of token transfer activity on Ethereum.
The NFT renaissance, which strongly leveraged the ERC-721 standard released in 2018, had its largest impact in late 2021 and 2022. However, as the hype waned, so too have transactions associated with moving NFTs, leading to a much less-spirited but still relevant presence in the ecosystem.
On the whole, while Ethereum’s transfer activity has diversified over time, Ether and stablecoins remain the primary drivers of transaction fees. It demonstrates the ongoing importance of ETH as the network’s native asset and the critical role stablecoins play in the broader industry. In our opinion, stablecoins are one of the most likely sustaining, and most intuitive, use cases of crypto platforms that have emerged over the years.
Ethereum Infrastructure Fee Spend Centers around MEV, Bridges, and Layer 2s
The mechanics of how Ethereum operates is substantially different today than 5 years ago: who initiates transactions, how often they’re settled, how fees work, the order of transactions, the supply policy of ETH — all this has changed.
Some changes were planned by Ethereum’s core developing group, others came about naturally through market forces. Take staking, for example. Moving to a system that determines who settles transactions based on how much ETH is locked as “stake” was on the roadmap of the Ethereum foundation prior to the project launching in 2015. However, upon activation in 2020, due to both size requirements and the desire for liquidity, third party providers are the most prominent players in staking, thus settling the majority of Ethereum transactions.
The buildup of infrastructural usage since 2020 is due in part to staking, but to a much larger degree due to layer-two technologies, MEV, and bridges. Layer-twos — which are independently operated systems that periodically settle batches of transactions down to the Ethereum base chain — consistently dominated this category from 2022 to 2023. However, a March protocol-level change to the Ethereum fee market (EIP-4844) reduced the cost of layer-two settlement to virtually nothing (Feb = $41m, June = $1.6m).
Infrastructure-related spend is now more than 50% related to Maximal Extractable Value, or MEV. MEV refers to the value that can be leveraged from the ordering of transactions in the settlement process. Most intuitively, MEV could mean seizing an arbitrage opportunity, where profit comes from the price differences between the same asset of two different exchanges. More cynically, it can be thought of as the opportunity for the validator of transactions to view those transactions before they happen, and replace transactions that will be profitable (for example, one that executes on an arbitrage opportunity) with his/her own transactions.
As Ethereum continues to seek more advanced purposes, and moves to make more complex changes, its infrastructural usage may increase as well. On the other hand, if grassroots infrastructure begins to materially impact general transaction fees, as was the case with layer-twos, changes may work in reverse, where Ethereum’s core developing group moves to intentionally limit such usage.
Ethereum Spending is Down, but persists in Marketplaces, ETH and Long-Tail Asset transfers
Up to this point, we’ve made our way through each main category of Ethereum usage, peering into the relative drivers within each. We’ll now combine it all together to see which specific services are in the highest demand across all segments.
What we take from this table is something that the Ethereum community should be both proud of and concerned about. Ethereum has successfully grown to host a wide range of applications that users are willing to spend over a billions of dollars annually to access; a feat achieved in less than a decade that is worthy of celebration. The hard truth however is that these flows are highly concentrated into a few use cases centered around asset speculation, and the trend of demand for the Ethereum chain is declining.
Given that ETH’s value depends on the demand for Ethereum services, it is of utmost importance to investors that the platform serves lasting use cases, which should ideally become even more valuable with time. While the current trend is discouraging, we are not declaring that Ethereum is doomed. In fact, we find that the Ethereum community is quite open to change and is willing to alter its ecosystem in pursuit of new goals.
It just so happens that a main goal of the previous several years has been scaling via second-layer technologies. And by all metrics, this has been a major success, layer-twos have grown substantially.
However, while the rise of layer-twos has effectively addressed scaling challenges, it has also cannibalised some of the main use cases of Ethereum’s Layer-1 platform, complicating the relationship between ETH's value and the broader Ethereum ecosystem.
In our view, the latest major change, EIP-4844, which strongly incentivised Layer-2s, has worked directly against the economic design benefits of EIP-1559, which tied the value of ETH to its Layer-1 platform demand.
Moving forward, we think the Ethereum community will need to focus on fostering on-chain utility that not only scales but adds meaningful long-term value to users. This is a main research question for the future and something we believe is core to a fundamental ETH investment strategy: what Ethereum services will drive long-term, sustainable user demand?
And while we’ve already tried approaching the question from a conceptual point of view, we still don’t know the full answer to that.
Methodology
Our goal was to identify the motivation behind Ethereum transactions in a data-driven way. We approached this goal by identifying Ethereum addresses, placing them into meaningful categories, and joining this dataset with on-chain transactions. The end output was a dataset that maps all Ethereum transaction demand, denominated in gas, to distinct categories such as decentralised trade, NFTs, etc.
With this output we could then create this report, answering what services users are paying for on Ethereum, and in what magnitudes, as well as build upon these answers to inform what Ethereum might be used for in the future.
The premise we hold is that the nature or use of transactions can be inferred based upon an understanding of the identity of either the sender or receiver of the transactions. In technical terms, the category of the gas spend of a given transaction is either an attribute of the sender or an attribute of the receiver. Given that contract addresses are persistent and available, we found this achievable.
Thus, we built a mapping of addresses to categories across a few different data sources, mostly: Dune Analytics, Artemis, and Etherscan. These data sources identify projects by receiving address, and also categorise them in the case of Artemis. Internally, we then identified the projects of outstanding receiving addresses as needed. Once each address has an identified project, we further mapped each address to a taxonomy of categories.
Once we produced an identifying dataset, we combined publicly available chain data from Google to attain the magnitudes of gas spent, aggregating monthly. We assume however that transactions with under 24k gas spend are ‘Simple ETH Transfers’ and transactions spending to a NULL address are ‘Contract Creations’. We hold these as model assumptions, which we find are correct in the overwhelming majority of cases.
In the end we achieved categorising gas to at least the 85th percentile dating back to 2020 on a monthly basis, 70th percentile dating back to 2016, and 88% of the total gas spent since Ethereum launch.