👋🏻Welcome back, especially to my newest subscribers after my last piece Is Writing Over?. If you like reading my in-depth essays half as much as I enjoy writing them, please subscribe with your wallet by clicking below:

If you enjoy reading my writing as much as I enjoy creating it, please subscribe by clicking the button below:

Last year, fueled by 2021’s saccharine crypto utopianism and a dog-eared copy of the Sovereign Individual, Twitter users envisioned themselves waving goodbye to corporate overlords to be full-time DAO contributors.

I was one of them.

To me (and many others), it seemed inevitable that swarms of contributors would converge to contribute to DAOs. Powered by ✨Vibes✨, DAOs would accomplish ambitious objectives and make web2 incumbents look like an Ancien Régime.

Blockchain technology would enable thousands of internet strangers to collaborate on and ship projects without centralized companies.

But that just kinda, sorta, didn’t happen.

Operationally, DAOs feel more like the stinky ol’ companies they wanted to replace. A core team receives salaries and freelancers collect occasional payments on clearly defined tasks, assigned by the core team. Payments are often in crypto (but not always).

DAOs haven’t ushered in a new era of work. That doesn’t mean DAOs have vanished or that they’re idle. Many have crashed and burned. Some continue to hold votes, create initiatives and ship projects.

But we’ve stopped seeing DAOs as a wunderwaffe and realized it’s just an org structure. Right now, DAOs feel a bit like when a dad buys a new power tool and then figures out what it’ll be useful for.

In their current form, Most DAOs are startups with voting. A board of (often unqualified) directors vote on usually unanimous issues. Contributors wade through proposal demands that make it feel like you’re wrangling an academic grants committee.

But I’m not a DAO pessimist. I still believe DAOs can create meaningful change for the better. They can still align internet strangers to collaborate on worthwhile work. And they can proliferate decentralization and become awesome enterprises. Just not the way they’re trying right now.

Let’s start from first principles to discover how DAOs can accomplish this. In this essay, we’ll discover why organizations exist, examine what’s actually new about DAOs and outline how they might accomplish things traditional businesses can’t.

Before we start, let’s define what a DAO is for the purposes of this essay.

Most blogs and media outlets define DAOs as basically „It’s a company but you can vote“.

More specific definitions might mention Discord servers, ERC-20 tokens or Snapshot votes. But tools or activities don’t capture DAOs. Let’s get on the same page by defining each letter of the acronym.

• Organization: An organization exists to facilitate an activity. While a community is simply a group of people with shared characteristics (as I explained in my essay on NFT communities), an organization coordinates activities

• Autonomous: „Autonomous“ does not mean „automated“. In this context, autonomy is about independence. It means that DAOs are not subsidiaries of state- or privately-owned enterprises, which may include a degree of automation.

• Decentralized: What’s decentralized and what’s not is debated. What matters is that blockchains are immutable, which means DAOs can’t be shut down by any one person or entity, while tokens ensure that there’s some distribution of governance rights.

These descriptions aren’t complete, but sufficient for this essay. Now let’s discover the first principles of what DAOs could be good for—and how to make them the forces of change they can be.

We’ll start with a painfully obvious-sounding question that’s actually not-so-obvious. Why do we have organizations at all?

Just like understanding cryptocurrencies must start with „what is money?“, understanding DAOs must start with „What is an organization?“.

The foundational text on that question is an essay by Ronald H. Case. His 1937 article „The Nature of The Firm“ answers the following question:

In other words, if the market is great at allocating resources, why isn’t everyone a freelancer? You could buy labor from everyone you work with and they buy yours. Instead of a company paying everyone from a central pool of money, everyone pays each other for labor from their own bank accounts.

Ultimate decentralization.

But Coase (and anyone who thinks about this for more than 2 minutes) argues that wouldn’t work. You’d spend most of your day invoicing, paying bills, hiring, firing, negotiating…

The ‘work of doing work’ skyrockets until you’re not doing your real work.

This eats into the labor you can sell and costs you money. The more people you add, the higher the share of organization versus money-earning labor.

You might reach some local optimum where you’re still making money while trading freelance labor with everyone else, but you’d probably work with a handful of people at best.

That’s because network effects scale exponentially. Each person added introduces more connections to the networks.

Humans created organizations to scale collaboration by minimizing transaction costs.

In most organizations, you don’t have to negotiate paperwork with each coworker. You don’t even have to talk to most coworkers!

By simplifying, standardizing and codifying interactions and transactions, organizations increase how many people can collaborate, which allows more complex chains of activities to emerge, which creates more intricate products and services.

‘Work of doing work’ costs can be considered transaction costs:

If you worked in growth marketing for a SaaS platform, you might have a weekly meeting. With marketing aiming at increasing sales, this meeting exists to create a customer => company transaction.

But the meeting only coordinates the labor involved in that transaction, it does nothing to facilitate the transaction itself. You get paid by your company for participating in the meeting (labor that does not lead to transactions). While meetings don’t show up on invoices, customers pay for your internal coordination indirectly: If your company was more efficient, you could increase scale or lower headcount, both of which would lead to lower prices.

The marketing meeting is a transaction cost. So are internal correspondence, negotiations, credit card fees, broker commissions and so on.

(For the purposes of this article, we’ll examine the transaction costs related to collaboration and organization, not monetary ones like fees and commissions)

Organizations still have limits dictated by transaction costs. At some point, the marginal cost of adding another person to an organization—due to coordination, politics, bureaucracy and deadweight—aka transaction costs—exceeds the marginal benefit.

The stereotypical workday at a large corporation is a swamp of meetings, forms and approvals and very little doing what you were hired to do. These transaction costs are often a feature, not a bug: As startups become corporations, they go from upside-maximizing to downside-minimizing. This prevents major failures, but also stunts innovation.

Organizational innovation expands the boundaries of human collaboration.

And because each relationship creates transaction costs, productivity declines. While a basic organization decreases transaction costs compared to the dystopia of forced freelancing, adding too many people causes it to devolve into a bureaucracy where nobody does actual work. The internet has lowered transaction costs associated with freelancing, but only tiny organizations are viable as freelance collectives—a design agency might work, but good luck manufacturing the iPhone.

That’s because networks scale exponentially. While 2 people create 1 connection, 4 people create 6. When you reach 25 people, there are 300 possible interconnections. Chances are that you’ve worked in a functional organization with more than 25 people without maintaining 299 lines of communication.

Just like organizations were an innovation that lowered marginal transaction costs, you can innovate on the organization to lower them further.

Structure and process innovations like Taylorism invented our modern conception of management. The aim of management is to reduce transaction costs by reducing the interconnections between people.

Because managers report to leadership and others report to managers, hundreds of possible connections are reduced to two. This reduction in transaction costs enabled larger organizations, which enabled mass manufacturing.

Further innovations that lowered transaction costs, both technological and organizational. From project management tools to management methodologies, we keep finding ways to reduce transaction costs.

Reduction in transaction costs via technology or management innovation enables also breeds innovation.

In his 2008 book „Here Comes Everybody: The Power Of Organizing Without Organizations“, Clay Shirky explains how there’s always a „Coasean floor“: A set of activities which would have value to people, but where transaction costs are too high to make it worth doing.

Shirky describes an example of an obscure New Jersey parade of which you could find hundreds of photos online. In the early 2000s (when the book was written), this was astounding. I’ll paraphrase his argument:

Many events were not worth covering in the mainstream. And if they were worth covering, then you’d see the few photos the newspaper’s photographer took, not hundreds of pictures from hundreds of people. Pre-internet, transaction costs for producing a collection of every picture taken at an event were too high to be worth it.

Imagine doing all of the following:

• Run newspaper classifieds to make people aware you’re looking for pictures from the event.

• Convincing people to print their pictures and mail them to you at their own cost.

• Sorting through the images to make sure they’re all from the event

• Arranging the pictures in a layout for print.

• Working with a company that prints them and get them to print samples.

• Pay for a print run of the collection

• Find a way to sell the collection

There are easier ways to waste your time and money.

Today, transaction costs of collecting photos from any event are effectively zero: You add „#burningman2022“ to your Instagram post. Instagram does the rest. We take for granted that we can find 122,031 pictures from Burning Man 2018 in a few seconds, but it’s quite an accomplishment!

There is value to this, but pre-internet transaction costs made it not worth doing. It was buried under the Coasean floor. A more recent example is the creator economy:

There have always been thousands of people interested in medieval history, vinyls of Japanese jazz music and Stone Age construction techniques. But finding, reaching and distributing media to them just came with such high transaction costs that it wasn’t worth doing, which made creating in these spaces not economically viable, which meant nobody did it.

Transaction cost reduction lifts products and services above the Coasean floor and makes them worth doing.

ConstitutionDAO is an example of this. Without smart contracts and ERC-20 tokens, transaction costs would’ve been too high to coordinate thousands of people to build trust in an entity and send their money to it.

But few DAOs do genuinely new things. Most DAOs do stuff companies have always done, but with higher transaction costs. When everything requires proposals and votes, DAOs do the same things companies do, but they’re slower in executing them.

Just because you have the option to call a vote doesn’t mean you constantly have to do it!

Look at most Snapshot votes and you see that much of it is LARPing democracy. Out of the last 10 FWB votes, only one ended with the winning option having less than 95% of the total vote, with none being close at all. The same is true for most DAOs: It’s rare to see a vote with less than 85% on the winning side. The sentiment should’ve been obvious.

For DAOs to proliferate, the future can’t be in doing stuff we did before crypto, but with endless proposals, snapshot votes and a token. In other words: We need to figure out which things DAOs could unearth from the Coasean floor.

What might those be?

Before Wikipedia angered high school teachers all over the world, there was nupedia. Nupedia was supposed to port the encyclopedia to the internet. Jimmy Wales wanted subject matter experts to contribute knowledge and have it reviewed by an editor before it goes live. Same process as a physical encyclopedia, just without having to print anything.

nupedia eventually failed. But the online encyclopedia wasn’t dead. It just needed the exact opposite approach!

Nupedia was reborn as the decentralized Wikipedia. Editing tools were available for anyone to contribute as little or as much as they wanted. Transaction costs were (and still are) basically zero! You didn’t have to ask for permission, write a proposal, attend meetings or take a seminar.

The user base exploded—and so did Wikipedia’s content, mostly created by amateurs who contributed as little as a comma or as much as thousands of words.

Ranked #7 on the internet’s most visited websites, Wikipedia’s neighbors on #6 and #8 (Baidu and Yandex) have 45,000 and 10,000 employees, respectively. Wikipedia has around 280—and legions of volunteer editors who are compensated with digital “barnstars”. This is even more impressive when you consider that Wikipedia isn’t some indie alternative to Encyclopedia Britannica, but the leading product in its market.

Wikipedia is the most impressive, but there are more projects where pseudonymous contributors build impressive things together:

• Video game modding communities often create “mods” that could be sold as standalone games.

• Fan fiction series often spawn multiple novels’ worth of writing.

• Open source software forms the foundations of much for-profit software.

Wikipedia, modding, fan fiction and open source all fulfill some part of the promise of DAOs: Decentralized creating with ephemeral teams of highly motivated contributors who don’t need to know each other.

These contributors receive Github ‘stars’, forum ‘likes’ and Reddit ‘upvotes’.

And none of them ever touch a blockchain.

Many of these proto-DAOs eschew commercialization. Many gaming mods could sell as standalone games. And while there’s probably some idealism around creating public goods, I reckon it also comes from tacit knowledge that introducing money would skyrocket transaction costs, plunge the project into the Coasean floor and kill it.

Imagine trying to publish a book of fan fiction by 10 contributors. You’d have to find out everyone’s real name, figure out payments to each member’s home country, maybe even a business entity, negotiate proceed splits… you may as well just write your own book.

This is a perfect example of the Coasean Floor: There’s value to publishing the book (the content is popular), but transaction costs make it such a pain that it’s not worth doing.

Explicitly commercial projects have higher transaction costs. People write fan fiction in their free time without expecting money because the activity itself is the reward. But nobody comes home after a long day of work and designs toothpaste packaging for toothpaste. There are two reasons for this:

1. Utilitarian activities are usually less fun than creative pursuits.

2. It feels wrong when others commercialize work we do for free.

That’s because the more we consider something work (as opposed to pleasure), the more we want compensation. And compensation causes negotiations, agreements, invoices, contracts—transaction costs. And as transaction costs rise, this work falls below the Coasean floor and doesn’t get done.

Contribution to a DAO may not feel as work-ey designing toothpaste packaging for a company does, but it’s rarely as fun as writing fan fiction.

That’s why DAOs do need blockchains (and, more specifically, tokens) to create commercially successful offerings. Tokens also align incentives and provide a standardized unit of measurement.

The proverbial handshake deal might have low transaction costs, but only works in high-trust environments. A handshake deal works when somebody’s reputation is at stake. But there’s nothing at stake when MonkeyBoy69 can run with millions and reincarnate themselves as iHaveYourNose91. This makes it impossible to trust anonymous strangers online.

A (fungible or non-fungible) token gate means everyone sends costly signals, which improves trust and thus lowers transaction costs. The other part is the efficiency of smart contracts.

Instead of agreeing on a bunch of things every time you contribute to a DAO, you interact with a smart contract where terms are pre-defined, movement of assets is transparent and enforcement doesn’t require lawyers and courts—as long as you don’t break the law.

The jurisdiction of smart contracts ends where the traditional financial system begins. But ERC-20s and smart contracts drastically lower transaction costs, which lets DAOs enable new types of work previously below the Coasean floor.

DAOs have a unique opportunity: They could get the best work of highly skilled professionals.

Organizations have always evolved. We take limited liability companies, corporations and conglomerates for granted. But no law of the universe requires them to exist. As Venkatesh Rao notes:

Corporations are a recent invention, and instances that had the ability to transform the world in magical ways did not really exist till the [East India Company] was born. Businesses of course, have been around for a while. […]

What was new was the idea of a publicly traded joint-stock corporation, an entity with rights similar to those of states and individuals, with limited liability and significant autonomy (even in its earliest days, when corporations were formed for defined periods of time by royal charter).

Organizations change to help humans solve the challenges the world presents. The present iteration of work in organizations is a skeuomorph of the industrial age. Productivity soared when factory workers turned screws for 8 hours per day, so office workers should hit keys for 8 hours per day, right?

Knowledge work has different economics.

Factory work has a linear relationship between screws turned and cars produced. Extra screw-turning means the factory worker creates extra value. But if you’re a copywriter, the additional value plummets between spending one hour writing 10 headline ideas and spending ten hours writing 100 headline ideas.

The same is true for lines of code, marketing graphics and all other types of knowledge work.

Labor and value were linear in the industrial era, but they’re exponential in the knowledge era.

Skilled knowledge workers create most of their value in a few hours per week. A well-known study found that most office workers spend about 3 hours per day on value-producing work, even if they spend 8 hours or more at the office. If power law dynamics apply to those ~15 weekly hours, about 3 hours generate the majority of the someone’s impact.

To fill the 8 hours mandated by industrialism, offices make up meetings to socialize and play status games, but most of those could be eliminated with zero loss in productivity.

DAOs can arbitrage this inefficiency because for contributors, the high-impact hours don’t come in a package deal with a cubicle, endless meetings and tuna-microwaving colleagues. On the DAO’s part, the few high-impact hours don’t come in a package deal with the low-impact hours.

So if DAOs enable skilled tech workers to contribute their best work with radically lower transaction costs, both sides win.

This scenario allocates resources (capital and labor) more efficiently. DAOs only buy the high-impact hours, contributors only spend the high-impact hours.

Web3 detractors and other heathens might now say we’ve come full circle and invented freelancing, which isn’t exactly new.

As we saw in Coase’s argument against the freelance dystopia, freelancing increases transaction costs compared to employment. From reading/writing applications to negotiating prices and handling invoicing, freelancing has high transaction costs:

Many skilled tech workers would be happy to make money on the side for a few hours of work. But freelancing comes with invoicing, outreach, etc., which turns the few hours into many.

Transaction costs throw this potentially fruitful collaboration below the Coasean floor.

Powered by ERC-20s and smart contracts, DAOs can turn the high-transaction-cost freelancing economy into a low-transaction-cost contributor economy where contributors focus on creating value (instead of proposals).

This requires DAOs to innovate beyond proposal-writing and constant votes, which are pure transaction costs. The goal should be to lower the transaction costs of DAO contribution to let new ways of collaboration emerge.

But for that, we’ll have to change a few things. Right now, contributing to most DAOs has giant transaction costs: You have to buy some tokens, write a proposal (often multiple pages), hope your proposal goes to a vote, then market your proposal so people vote for it. Only then do you get to contribute.

Instead, see how easy our role model Wikipedia makes it:

While Wikipedia has gotten less anarchist over the years, you can still contribute by clicking edit. Your edit could transform the page, add a paragraph or merely correct a typo.

On the other hand, DAOs require you to write multiple pages worth of proposals, do Q&A and then change a paragraph somewhere. Assuming there is a Coasean floor of people who have something to contribute but won’t go through the bureaucracy of proposal-vote-implementation, we need something simpler.

Minimum viable contribution is the smallest unit of activity a DAO recognizes and rewards. Like a software MVP, it’s neither perfect nor complete. But it’s a starting point that has some merit.

In Work is a Software Product, I argued:

We access work as a software product. This is not only true conceptually, it’s true literally. I do 90+% of my work in:

• Google Docs

• Google Meet

• Google Mail

• Notion

• Slack

(You could replace Notion with Google Sheets and Slack with Google Chat and all my work runs through Google).

Even if you don’t work in marketing, you probably do most of your work using less than 10 pieces of software. It’s obvious enough that we access work through software. But work also behaves like software. [...]

Your outlook changes when you stop thinking about your team as labor units to arrange and start thinking about them as software users. This improves things for everybody:

1. UX design is a user-focused process that aims to help people reach their goals. Make it easy for your team to do their best work—and they’ll do their best work.

2. Great UX scales. Software keeps working after you stop, so you’re creating assets that help your team do their best work, even when you’re not actively managing.

Few companies nail this. If work is software, most companies have atrocious UX.

This should also be true for DAOs. Let’s set aside the lens that views DAOs as cities, countries or organisms and see them as a software product. Good UX simplifies and clarifies things for users.

There’s a concept every self-help book regurgitates: To build a flossing habit, floss one tooth. And as you’re flossing one tooth, you’ll start flossing all your teeth. The principle: Start with something tiny and your brain won’t dread it as much. This tiny action contributes little to the goal, but it removes every excuse to start.

Looking at DAO contributions through a product lens reveals a shocking picture. Before doing anything fun (contributing) you have to sift through previous proposals to figure out how to write one, hope you can find some documentation or a template, then you have to write it and pass a vote… just for a chance to get to do the thing you want to do.

To lower transaction costs, we need to remove these barriers and reward permissionless contribution as small as “flossing one tooth”—writing one line of code, adding one line to a piece of content, etc.

DAOs tend to see contributions at the project scale. Contributors need to own and manage the project they proposed. And a lengthier process makes sense for consequential projects. But MVC looks at contributions at the task scale.

MVC itself may not do much, but it’s intended to show people how easy it is to contribute and receive a (tiny) reward. This could lead to bigger contributions in the long run, especially if DAOs invent intermediate steps between MVC and proposal-worthy projects.

Allowing any token holder to contribute may sound chaotic. But that’s the point of the Coasean floor: We need to unearth what nobody is doing, but DAOs enable. And it might look like thousands of people contributing something tiny (or more). This will require organizational innovation (the same way industrialization required the invention of management) and likely new technologies like token-gated collaborative tools.

DAOs can’t keep being startups with the operations of academia. MVC is one way to lower transaction costs to simplify contribution and fix DAOs and discover what decentralized value creation might look like.

That’s it for this week! If you enjoyed this essay, please click below to subscribe for more in-depth writing that’s actually worth your time:

If you liked this essay, I’d also appreciate a retweet on the tweet below to make sure more people see this piece:

Finally, if you especially loved this essay, feel free to collect it as an NFT to show some appreciation. Collectors will also get first access if I launch anything token-gated (but no promises).