Fat Protocol or Fat App Discussion 2.0



It is better to compare blockchain protocols with ISPs or platform businesses such as Uber/Apple Store rather than “HTTP”, “FTP”, or “SMTP”, and a blockchain protocol can be viewed as a collectively owned and operated “business” with no majority shareholder;

There is no right answer for investing in “protocol” or “application” either in the web age or the blockchain age: a “protocol” can capture huge value as what Taobao does or lose money like Uber or Zillow;

A blockchain protocol can capture more value than an application on it only if the protocol builds a high market barrier through technology or network effect and charges less than the value created. However, it is unlikely that any open-source protocol can maintain such high barrier, and an oligopoly with low margin is expected to appear as what happened in the telecommunication industry;

Since a blockchain protocol is not owned by any entity, a freemium model used by many platforms is not applicable since nobody owns the “traffic” and can design products to monetize the “traffic”. Therefore, the potential value of a protocol is limited even if the protocol dominates.

For people who have not heard about “fat protocol”, please refer to the original article from Union Square Ventures and this article from Lightspeed Venture Partners.

1. What is comparable to blockchain protocols?

Many people compare blockchain protocols with “HTTP”, “FTP”, or “SMTP”. Technically, it is right to do the comparison since all of these are protocols, a defined set of rules and regulations that determine how data is transmitted.

However, from an investment perspective, the comparison between blockchain protocols and web protocols does not tell us much. Instead, as an infrastructure of the blockchain, the blockchain protocols should be compared with internet infrastructure operators and platform businesses.

As investors, we can view the blockchain protocols as “companies” owned and operated by many people distributed in the world with no major shareholder. The mining fees are the money that users pay to the protocol operators for the blockchain infrastructure as what users pay for internet access and e-commerce infrastructure.

Let’s compare the blockchain protocols with the Internet Service Providers (ISPs) and platform businesses as follows:

2. Why are the ISPs and platform businesses comparable with blockchain protocols?

First, product-wise, all the businesses are serving a huge audience. These companies may not be serving all human beings, but they usually do not have a very specific targeted customer segment.

Second, technology-wise, all the businesses in the three categories are providing a basic layer where third parties can offer differentiated services/products on it. It is important to notice that all these businesses sacrifice the customizability in exchange for the massive audience while leaving the differentiation to third parties.

Third, business-wise, all the businesses are enabling other new third-party businesses and are bound closely with the third parties. There are only two results for these businesses and the third-party businesses: win-win or lose-lose.

3. What matters in the end?

In the fat protocol thesis, there is a famous diagram showing the value captured by the protocol layer and the value captured by the application layer. However, the writer did not clearly state how “value” was defined — revenue or profit? Another question is: for investors, the total value matters or the average matters?

Fat protocol diagram (Source)

From my perspective, it is clear that profit matters more than revenue.Revenue does not necessarily reflect the value created by a company: Amazon has $220.96 billion revenue in 2018 with a 4.03% net margin ($8.90 billion), while Google has $129.87 billion revenue with a 14.45% net margin ($18.77 billion). The market tells us that Google creates more than double value for its users compared to Amazon with 58.77% of Amazon’s revenue. Therefore, from a profitability perspective, an investor should consider Google (profit capturer) over Amazon (revenue capturer).

On the other side, the total captured value by all companies does not draw the full picture, while it is the average value captured by an individual company that matters most. For example, Taobao/Tmall, the monopoly e-commerce marketplace in China, generated a GMV of $768 billion (sales of the merchants on the platform) in FY 2018 with $28.148 billion revenue itself. However, the $768 billion sales are generated by about 10 million merchants. Therefore, the average annual revenue of a merchant on Taobao is about $76,800, which is almost zero comparing to Taobao’s revenue. In terms of profit, Taobao/Tmall also has a margin of about 53%, while it is safe to say the merchants usually have a single digit margin. Therefore, an investor should invest in Taobao (the “protocol”) instead of the merchants (“applications”) on it.

On the other side: according to IBIS World, the global ISPs generated about $563.674 billion revenue in 2018 with the biggest five players estimated to account for less than 40.0% of industry revenue. The industry margin is estimated to be 13.4%. According to CIA, there are about 12,034 ISPs in the world. Therefore, on average, a small-to-medium-sized ISP will have about $28.11 million annual revenue with $3.76 million profit. Therefore, an investor should invest in Netflix (the “application”) instead of a small-to-medium-sized ISP (the “protocol”).

Therefore, if we think from the profitability of an individual company, there is no right answer for whether we should invest in a “protocol” company (including ISPs, platforms) or an “application” company even in the current web stage.

4. What else can the comparison tell us?

(Source 12345678)

Through the table above, we can see that in many cases the platform businesses (“protocols” for small businesses) are very attractive given the large transaction volume and a relatively high margin. However, it is not always true since the result depends on whether the platform is able to build a high market barrier. This logic holds for the Web “protocols”, and I believe it will also hold for the blockchain protocols.

5. Under what circumstances will a blockchain protocol become attractive?

Based on what we discussed above, it is reasonable to believe that the following requirements need to be met for a blockchain protocol to be more attractive than the applications: (1) high market barrier; (2) high value created.

For the market barrier, it should be easy to understand: if the “business” (remember a protocol can be viewed as a collectively owned and operated business) is highly profitable — people are willing to pay for it, there should be many “companies” who want to get a slice of the “market”. If that happens, the margin will be reduced, many players will then leave the market, and the margin will remain low with a limited number of players. If a protocol wants to build the barrier, it can leverage either the network effect (first-mover advantage) or avant-garde technologies.

For the second prerequisite — high value created, it decides people’s willingness to pay (WTP). Blockchain, as a new type of database or infrastructure, is competing with the existing web. How much do people want to pay for the “decentralization”? How much do people want to pay for “privacy”? If it cost people more to reduce information leakage than the potential damage it might have, who will pay for that? Ask the Chinese people, they will tell you “freedom” is not priceless as long as the government maintains a high GDP growth.

Let’s assume a workable blockchain protocol has been invented (most of the protocols today still have many serious problems that limit their adoption), we use a decision tree to evaluate under what circumstances a blockchain protocol is an attractive investment opportunity.

Through the decision tree, we can verify our conclusions. Only in two scenarios will a protocol capture more value than an “application” built upon the protocol, as what happened to Taobao, Facebook, and Apple Store:

  • Scenario 1: it is difficult to replicate the protocol due to its technology or network effect, the protocol charges less than the value it created/consumers perceived, and application developers must have all activities on the chain(otherwise the app developers can still use the traditional infrastructure).
  • Scenario 2: it is easy to replicate the protocol, application developers agree to use the protocol and charges consumers for the mining fee, and nobody creates a similar protocol with lower fees to compete, which is very unlikely to happen.

Therefore, it becomes clear that if a protocol is very technically defensive — hard to copy — and creates a high value to the users, then we can expect it to become the Taobao or Facebook of the blockchain. If it is not defensive and we are not investing purely based on unreasonable speculation, then we might expect it to become another Zillow or Didi Chuxing, and we should wait for better deals. A reasonable speculation example can be that: Uber can be attractive even with a negative margin due to the company’s potentially different business model in the autonomous driving age.

Another thing we can learn from the comparison is that: many platforms offer the infrastructure for free but monetize the traffic through other products, while this model is not applicable to the blockchain protocols. For example, WeChat do not charge the users, while Tencent makes a lot of money from its games with most users coming from the WeChat platform. However, since the blockchain protocols are collectively owned and operated by millions of individuals in the world, the monetization of blockchain protocols is limited: nobody can exclusively own the huge traffic and design a product to monetize the traffic, and charging a usage-based fee seems to be the only monetization method.

On the other side, given the open-source nature of blockchain protocols, it is highly possible that an oligopoly or monopoly will appear in the end with a relatively low “margin”. This is a scenario similar to what happened to Uber/Zillow/Didi Chuxing or the ISPs or the telecommunication industry: there is almost no secret in offering the commodities, and people go to capture more value by moving onto the “basic layer” — Internet or phone or TV — and offering differentiated services/products.

To conclude: first, protocols does not necessarily have a brighter future than applications; second, a protocol can “succeed” only if it can build a high market barrier and offer a high value to the users; third, protocols are expected to have a low margin — “mining fee” — and cannot monetize the users/traffic through other means due to its decentralization nature.

This research project is completed through a collaboration between Fan Wen at Yale SOM and Josh Liggett at OurCrowd.

Disclaimer: The article is not, and should not be regarded as “investment advice” or as a “recommendation” regarding a course of action, including without limitation as those terms are used in any applicable law or regulation.


This site uses Akismet to reduce spam. Learn how your comment data is processed.

  • HI Fan and Josh, As always, I try to keep track of your work – always enjoy reading, and I always find something to relate to Geeq. Since we last exchanged comments, Geeq has transformed from a pure protocol project to a public blockchain infrastructure as-a-service project, which is what we realized we had specified. In terms of your last two paragraphs, our business model is exactly what you propose: very slim margins on tremendously high volumes. How, you might ask? Good question, as most platforms rely on network effects (which, of course, would benefit Geeq). However, we’re cheap enough to create two-sided markets for very low cost transactions. Machine-to-machine markets, IoT, micropayments. That ought to be a good start.
    Also, our emphasis is on low barriers to entry for users, rather than erecting high barriers (which defeats the purpose of decentralization, in my view). And our revenue model is different.
    Always glad to get back in touch with you two. Geeq is moving along nicely, thought you’d like to know.

hide comments

This is a unique website which will require a more modern browser to work!

Please upgrade today!