The Future Of Network Effects: Tokenization and the End of Extraction

KJ Erickson
Public Market
Published in
10 min readJul 17, 2018


How businesses built around network effects are about to undergo a radical transformation that will impact us all

In the first of this two part series, I argue that the network effects of tomorrow will often be built around decentralized tokenized ecosystems in which there is no distinction between network participants and network owners. This transformation will occur because tokenized ecosystems can capture the value of the network for participants without the economic rent extraction (fees and data) that is characteristic of centralized network effect platforms — resulting in better outcomes for everyone. The combination of reduced costs, higher value capture by participants, and the powerful growth incentives that well-designed tokenized networks provide create the conditions necessary to supplant some of today’s most powerful companies.


Network effects are at the heart of the most valuable companies built in the last quarter century. Because network effect businesses provide more value to participants the larger they grow, they can be immensely hard to get off the ground.

The ones that do reach critical mass tend to careen towards natural monopoly, creating competitive moats that allow the network owner to extract huge fees — understanding that even if network participants are frustrated, they have few alternatives.

Most insidiously, if the network owner is a private company with a fiduciary duty to its shareholders, charging monopoly rents is not optional: the company’s responsibility to shareholders means it will always look to maximize profits.

This is the ‘Extraction Imperative’. As the business scales, the negative impacts of the Extraction Imperative on ecosystem participants start to counteract the positive impacts of the network effects — until the business is no longer positively impacting the very network of individuals and entities that drive its success.

This piece examines the nature of network effects, the good and the bad they can create, and how the maturation of blockchain technology can radically transform the nature of the businesses built around them.

Background: What’s in a Network Effect?

The simplest definition of a network effect business is a product or service that gets better for everyone as its network of users or providers grows, such as a social network like Facebook.

There are actually many types of network effects. In “The Network Effects Manual”, venture firm NFX outlines 13 different categories, including those we’re used to: “Personal” (Facebook), “2-Sided Marketplace” (Amazon), “2-Sided Platform” (iOS), and “Data” (Google).


In each case, the value of the network grows as more nodes are added. In the case of 2-Sided Marketplaces like eBay or Amazon, the more sellers there are, the more value there is for buyers in the network. In turn, the more buyers there are in the network, the more sellers are attracted to join. These positive feedback loops can lead to rapid scaling once the flywheel begins to turn.

‘Market Power’ and the Extraction Imperative

Network effects are good for users then, right? The unfortunate answer when dealing with private, centralized network owners is “At first, yes. Over time, not so much.”

As the network is scaling up, network effects make the services that harness them better able to deliver value. The bigger LinkedIn is, the wider array of potentially-relevant professional contacts we can stay in touch with. The more sellers Amazon has, the broader the array of products we can buy conveniently from a single digital destination.

Unfortunately, network effects are so powerful that they drive network owners towards natural monopolies. In other words, as the network grows, both the value of participating and the cost of switching are sufficiently high that the entire group of potential participants consolidates around a single network.

The downsides start to occur when these powerful network effects are owned by centralized private companies. As network effects push industries towards monopoly, the owner of the network has the ability to assert what MIT Cryptoeconomics Lab economist Cathy Barrera called “Market Power.”

“Market power arises when users or customers have few comparable alternative options for sources of the good or service being provided. This gives the seller the ability to raise prices, or in the case of some internet giants to charge transaction fees, compile and sell user data, all as a condition for giving users access to the platform.”

Network effects give private companies market power. Market power gives private companies pricing power. Pricing power allows companies to extract more in rents from their users than is economically optimal — leading to the classic deadweight loss scenarios we all remember from Econ 101.


What is the Extraction Imperative?

All for-profit companies have a fiduciary duty to serve their shareholders’ best interests. In the US, shareholder rights are so strong that shareholders can sue corporate managers for not delivering on their duties.

Thus when a network effect business is housed in a private corporation, that business has an incentive to maximize shareholder returns by extracting as much profit as possible. We call this the Extraction Imperative. It lies at the heart of the problematic nature of building network effects within private companies.

In his now-classic essay “Why Decentralization Matters,” Chris Dixon of Andreessen Horowitz captures the results of the Extraction Imperative with the following S-Curves:


These S-curves show how in the early days of a business that is building network effects, its incentives are aligned with its users and the business’s actions are accretive to its network participants. As the network grows, incentives diverge and the business’s practices becomes more extractive and/or competitive. As Chris Dixon puts it, “the relationships with network participants move from positive-sum to zero-sum.”

Enter Tokenized Networks: Network Effects Without the Extraction Imperative

Up until now, it has simply been assumed that this Extraction Imperative was an unavoidable consequence of network effects businesses — the intrinsic downside to the value they provide. Under the private corporation model — the dominate organization model we’ve had for the past century or more — that assumption holds true.

But a well-designed decentralized tokenized network can change the game.

As we saw above, “Market Power” describes a scenario in which there is a breakdown in the interests between network participants and network owners. The ability to exert market power rests on two factors: 1) that network owners and network participants have different incentives; and 2) that network participants don’t have quality, realistic alternatives.

Let’s examine what happens to each of these factors in a tokenized ecosystem:

1. Tokenized Ecosystems Align Incentives

When a network is owned by a private, centralized company, economic value ultimately flows away from network participants and towards an exogenous group of shareholders who hold equity in the corporation. In a network in which there is no single entity that owns and controls the network, this dynamic does not hold true.

When designed properly, decentralized, open source, tokenized cryptoasset networks solve the problem of incentive alignment between network creators and network participants. When the software is public, and the organizing body is a nonprofit rather than a for-profit, the Extraction Imperative is eliminated — because there are literally no longer shareholders with a claim on cash flows. The value is held in the network itself, represented by token ownership. Everyone who participates in the activity of the network, using and accruing tokens in the process, is effectively a network “owner.” Because there is no division between owners and participants, there is no point at which their incentives diverge.

2. Tokenized Ecosystems Ensure Competitive Pressure & Realistic Alternatives

The second condition for market power — the lack of a realistic alternative — is also an artifact of privatized centralization that doesn’t exist in the open source context. Centralized companies hold proprietary IP that creates a competitive moat and insulates the company from competitive pressure that keeps prices low and the need to please network participants high.

Properly designed tokenized networks, on the other hand, are run via open source software developed via nonprofit foundations without shareholder interests to serve.

Open source blockchain-based networks come with the ability of the community to dissent from leadership decisions by forking the code into a new version of the blockchain that doesn’t reflect the contentious policy. In the wake of the DAO hack for example, the Ethereum community disagreed on whether or not to effectively overwrite the chain to reverse the hack, leading to the fork that would end up producing Ethereum Classic.

In short, a tokenized network can be designed in such a way that it does not satisfy either of the conditions — divergence of incentives between owners and participants or lack of meaningful alternatives for participants — required for a central organization to exert market power.

The Double Network Effects of Tokenized Platforms

What this means, practically, is that properly designed tokenized networks that are able to achieve similar network scale to their centralized peers will provide substantially more value to the actual users of the network.

Not only do the network participants get to reap the rewards that previously went to the private network owner (for example, the $34B that Amazon makes in commissions and fees from independent sellers), but they also get to recapture the substantial deadweight loss that results from monopolized pricing. Ultimately, this means that the networks will grow bigger, quicker because all the value accrues to the participants themselves.

This is revolutionary: tokenized networks are able to provide all the benefits of network effects without the costs to the network participants that are unavoidable under private, centralized network ownership. Importantly, these cost savings can be more than financial: they can also come in the form of increased privacy, ownership of data, and more favorable network policies.

In another 2017 essay “Crypto Tokens: A Breakthrough In Open Network Design” Chris Dixon described how tokenization can actually increase the speed with which network effects take hold:

“Token networks…align network participants to work together toward a common goal — the growth of the network and the appreciation of the token…Moreover, well-designed token networks include an efficient mechanism to incentivize network participants to overcome the bootstrap problem that bedevils traditional network development.”


The term “double network effect” was coined by Nick Soman to refer to the additional network effect that tokens can create by providing an “efficient mechanism to incentivize network participants.”

In short, if the first network effect is the impact of new participants on the utility of the network in terms of whatever it’s supposed to provide, the second or “double” network effect is the ability to deploy tokens to incentivize users to build the network faster.

Tokenization can improve the speed of network growth through two mechanisms:

  1. Network ownership: In well- designed blockchain networks, the users of the network are also holders of the tokens. Therefore, they have a vested interest in helping that network grow so long as they believe that network growth will help the value of their token holdings.
  2. Token incentivization of pro-growth actions: Token rewards can be deployed to incentivize the behaviors that the network needs to grow. For example, they can be used to power referral programs, or to reward actions like social sharing. Importantly however, because of the larger dimensions of token economies, many believe that tokenized growth incentives are likely to increase both the percentage of network participants who participate and the average number of growth actions per participant. The reason? As Soman puts it:

“When tokens can be traded instantly for cash, the current cash value of a token arguably represents its floor. It can also be spent within applications, or held for potential future value, creating a fear of missing out that may make signup even more enticing.”

The combination of the above factors — the alignment of incentives creating more value to be captured, and the way that tokenization can incentivize users to be growth agents — has profound implications for how quickly a network can grow.


Network effects define the businesses that most shape our daily lives. While we have come to expect that the fees and data they demand are simply the cost we pay for the benefits they offer, decentralized blockchain-based networks offer an alternative. In this alternative, the market owners are the market participants themselves, and tokenization not only keeps incentives aligned, but actually adds a new layer of economic incentive to network growth.

Together, these factors suggest that network effects businesses of the future will be organized not as centralized for-profit corporations built on extractionary business models, but rather as decentralized token-based economies with incentive alignment between network owners and participants.

What’s more, it’s somewhere between “not-impossible” and “likely” that the twin benefits of rent-free pricing and economic upside for contributing to growth will fuel the expansion of these new networks to sizes that eventually overtake some of today’s network effect stalwarts.


Enjoy this post? In part 2, we’ll look at the vulnerabilities of existing centralized network effects businesses with an eye to this question:

Which categories of centralized network effects businesses are most ripe for disruption?

Click the follow button to ensure you don’t miss out.