*This blog is a transcription of the Pheasant blog here.

The Rise and Fall of the Fat Protocol Thesis

A Perspective by Tomo, CEO of Pheasant Network

In 2016, the “Fat Protocol Thesis” emerged as a provocative idea in the blockchain space. Proposed by a prominent crypto investor, the thesis argued that most of the value in decentralized networks like blockchains would accrue at the protocol layer rather than the application layer. This view represented a radical departure from traditional internet value distribution and has shaped how investors evaluate blockchain projects since.

Fast forward to 2024. Eight years have passed since the first publication of this thesis—a significant period in the fast-moving world of crypto. During this time, many revolutionary ideas have either evolved or been reevaluated. So far, this thesis has largely held true: more value is currently captured at the protocol layer. However, I believe its validity may now be diminishing.

So, if the Fat Protocol era is winding down, are we on the verge of a ‘Fat Apps’ era? I believe so, though we’re not quite there yet. I think the transition is gradual, and today we find ourselves in an intermediate, preparational phase.

What is the Fat Protocol Thesis?

Before unpacking my own thoughts on the “fat app” trend, let's first revisit the now-controversial Fat Protocol Thesis.

The Fat Protocol Thesis proposes that, in the blockchain ecosystem, most of the value accrues at the protocol layer rather than at the application layer. This is the reverse of how value is distributed in the traditional internet, where applications like Facebook or Google capture the lion’s share of value, while the underlying protocols remain thin, largely commoditized, and generate little value themselves.

Two core principles underpin the fat protocol-thin application dynamic: the low entry cost for applications and the presence of native protocol tokens.

First, developing a blockchain protocol is significantly more complex and resource-intensive than building an application on top of it. This is because the foundational layers are open, shared, and composable, providing a solid infrastructure for apps to build upon without reinventing the wheel.

Those familiar with DeFi may recognize this as the “money Lego” concept. Uniswap, for example, didn’t need to create its own data storage mechanism or invent new currencies; it simply built its decentralized exchange on top of Ethereum (Layer 1), integrating existing assets like DAI, USDC, and others.

This dynamic is a marked departure from Web2. While Google relies on fundamental protocols like HTTP, its business model of monopolizing users and profits naturally led it to build everything else from scratch to create services like Gmail—everything from data storage to user authentication and server infrastructure.

Because blockchain applications can build on these pre-existing protocols, they don’t need to develop their own infrastructure from scratch, drastically lowering their development costs. This lower entry cost means applications capture less value—they require fewer resources to create and maintain. In contrast, protocols, which demand much more capital and effort, naturally capture more value to justify their higher costs.

The second factor driving this “fat protocol” phenomenon is the presence of native tokens. According to the theory, protocol tokens create a unique value cycle not seen in Web2. These tokens attract early investors, developers, and services to the protocol. As the ecosystem around the protocol grows, so does demand for the token, driving up its value. This, in turn, attracts more products and services to the protocol, further increasing token value. Early investors hold onto their tokens, anticipating even greater returns as the protocol gains popularity.

This creates a dynamic where, as the author observed, “the market cap of the protocol always grows faster than the combined value of the applications built on top, since the success of the application layer drives further speculation at the protocol layer.”

A Common Misconception

The unique token-driven dynamic, paired with the low barrier to entry for apps, forms the foundation of this radical concept. However, one critical distinction must be made: value capture is not the same as investment return. In a follow-up post in 2020, the same author clarified that while the protocol layer in crypto captures more value, this doesn’t always lead to substantial investment returns.

To put this into perspective: Imagine you invest $1 million in a Layer 1 protocol with a $1 billion market cap. If the network doubles in size, your investment also doubles, giving you a $2 million return. Now, consider investing that same $1 million in an app with a $10 million market cap. If that app grows 10x in the same time frame, your return would be 10x, or $10 million. So, while the protocol layer may capture more overall value, the application layer could deliver much higher investment returns. This is why value capture shouldn’t be the sole indicator for evaluating investments.

The Limitations of the Fat Protocol Model

As I mentioned earlier, while the Fat Protocol Thesis was influential at the time, I believe it’s now outdated. It was certainly prescient and made sense given the state of the crypto space back then, but the landscape has evolved so drastically over the last eight years that the factors the thesis relied on no longer hold up as a strong foundation.

So, what exactly has changed, and why has this once-groundbreaking idea lost relevance?

One major shift is the maturation of blockchain protocols, which now support far more applications, users, and transactions than they did in 2016. When the thesis was first released, Ethereum, the first smart contract blockchain, had only been public for a year. Although the network had a market cap of around $1 billion, there were few successful applications on it, making it reasonable to assume that most of the value would accrue at the protocol layer—because, at the time, it did.

Since then, both infrastructure and protocols have matured significantly. They now support a wide array of applications, users, and transactions. The rise of DeFi and NFTs exemplifies this. This flourishing decentralized app ecosystem challenges the original assumption that protocols would always capture the lion’s share of value.

Secondly and more importantly, market sentiment has shifted. In 2016, most development efforts focused on building and refining core blockchain infrastructure—scalability, security, and consensus mechanisms were the priorities. Innovation and excitement were concentrated at the protocol layer.

As protocols matured, however, the focus changed. The conversation moved from “protocols are the future” to “what’s the point of protocols without application value?” As the ecosystem evolved, it became increasingly clear that infrastructure alone wasn’t enough to drive adoption. Users interact with applications, not the underlying protocols. As decentralized apps gained prominence, attention naturally shifted to user-facing solutions. This shift suggests that the initial emphasis on protocol dominance was likely overstated, based on the early stages of blockchain development.

Is the Fat App Era Upon Us?

With these shifts in mind, let’s return to our original question: If the Fat Protocol Thesis no longer represents the future, are we on the verge of the Fat App era? In many ways, yes. That said, I don’t expect this shift to fully materialize for another 2-3 years, and there are two key nuances to consider.

First, it’s the collective app layer that becomes fatter, not individual apps. As the number of decentralized applications (dApps) grows, along with their users and transactions, the app layer will collectively capture far more value than the protocol layer. However, I don’t foresee the rise of a few dominant apps monopolizing the crypto space in the same way that Google or Amazon dominate Web2.

Blockchain’s open and decentralized nature doesn’t allow apps to lock in users and their data, which is how big tech companies achieve monopolies in Web2. Instead, we’re more likely to see a diverse ecosystem of applications—a “fat layer” of many apps. Some will offer specialized functions, like Uniswap’s exchange features or OpenSea’s NFT marketplace, while others will integrate multiple services to offer users a smoother, more comprehensive experience.

Second, while the protocol layer may seem to be thinning relative to the expanding app layer, it’s still growing in absolute terms as the definition of ‘protocol’ has evolved. When the Fat Protocol Thesis was first introduced, ‘protocol’  primarily referred to Layer 1 blockchains like Bitcoin and Ethereum. These were the dominant platforms with large user bases, and their ecosystems were relatively simple.

Today, the landscape is far more diverse, with a myriad of smart contract platforms available to developers—such as Solana, Avalanche, and Sui—due to lower entry barriers and the proliferation of tools, resources, and documentation. Additionally, ‘protocol’ now encompasses Layer 2 and even Layer 3 networks, which didn’t exist in 2016. For Ethereum in particular, there are dozens of these networks. So, while the protocol layer’s relative share of value capture may decrease, it’s still “fattening up” in absolute terms as Layer 1 chains expand and new layers are built on top.

A crucial point I’d like to highlight here is the lack of unity within the protocol layer itself, causing significant fragmentation. There are many Layer 1 blockchains, but they often lack interoperability, creating isolated ecosystems. Even within platforms like Ethereum, which have multiple layers (Layer 2 and Layer 3), there are barriers to seamless connection both vertically (between layers within the same network) and horizontally (across networks of the same layer).

Where Are We Headed From Here?

The notion of fragmentation and lack of interoperability is crucial because these are key reasons why I believe the ‘Fat Protocol to Fat App’ transition hasn’t occurred yet. The protocol layer’s fragmentation significantly limits the app layer’s ability to capture its full value, as users, liquidity, and development resources remain siloed. This fragmentation also increases development costs and results in poor user experiences, both of which stifle innovation and reduce the value captured by individual applications.

Given this scenario, I believe the next 1-2 years in the blockchain space will need to serve as a preparatory period for the Fat App era. While the Fat Protocol phase has brought us essential infrastructure services, we still lack the middleware necessary to bridge the gap between the infrastructure and application layers. This is where I anticipate increased focus and rapid development. Middleware services—such as Chain Abstraction technologies and Bridge-as-a-Service (BaaS) frameworks—will be critical in enabling exponential value capture at the application layer, setting the stage for a true Fat App era.