DAO (Decentralized Autonomous Organization)

Definition: An organization governed by rules encoded in smart contracts and member voting, rather than a traditional hierarchy of executives and a board.

How It Works

Membership and voting power are usually tied to holding a specific governance token, one token often equals one vote, though some DAOs use other weighting schemes
Anyone meeting the membership bar (usually holding some amount of the token) can submit a proposal, spend treasury funds, change a parameter, upgrade a contract
Proposals are discussed publicly, often off-chain on a forum, before a formal on-chain vote opens
Members vote by signing a transaction (or an off-chain signed message later settled on-chain) during a fixed voting window
A proposal needs to clear both a quorum (minimum participation) and an approval threshold (e.g. more than 50%, or a supermajority for sensitive changes) to pass
Approved proposals execute automatically via Smart Contract logic, often after a timelock delay that gives members a last chance to react
Treasury funds are typically held in a shared multisignature wallet or a dedicated treasury contract controlled by the DAO’s voting process, not any single individual
Some DAOs run entirely on-chain, others use off-chain signaling for most decisions and only settle high-stakes actions (spending funds, upgrading contracts) on-chain

Proposal Types

Treasury: spend or allocate DAO-held funds
Parameter change: adjust a protocol setting, e.g. a fee rate or collateral ratio
Contract upgrade: point the protocol at new contract code
Signaling: non-binding, gauges member sentiment before a binding proposal is drafted
Emergency: fast-tracked, shorter voting window, used for urgent fixes like pausing a contract after an exploit is discovered

Voting Mechanisms

Token-weighted: one token equals one vote, simplest and most common, but concentrates power with large holders
Quadratic voting: cost of additional votes on the same proposal grows quadratically, meant to reduce whale dominance
Delegation (liquid democracy): a member assigns their voting power to another address they trust to vote on their behalf
Reputation-based: voting power comes from non-transferable reputation earned through contribution, not purchasable token holdings
Conviction voting: a vote’s weight grows the longer it’s held on a proposal, favoring sustained support over a quick coordinated push

Governance Token Mechanics

Parameter	Purpose
Proposal threshold	Minimum tokens needed to submit a proposal, blocks spam
Quorum	Minimum participation required for a vote to count
Voting period	Fixed window, e.g. 3-7 days, during which votes are accepted
Timelock	Delay between a passed vote and execution, lets members react to a bad outcome
Veto/guardian role	An emergency multisig that can cancel a malicious proposal before it executes

Under the Hood

Governance is only as decentralized as its actual voter turnout and token distribution, a low-quorum vote controlled by a handful of large holders is common in practice.

Worked example: does a proposal pass?

Given: a DAO has 1,000,000 governance tokens in circulation, quorum is set at 10% of supply, approval threshold is a simple majority
Step: 90,000 tokens’ worth of votes are cast, 61,000 in favor, 29,000 against
Step: check quorum, 90,000 / 1,000,000 = 9%, which is below the 10% quorum requirement
Answer: the proposal fails on quorum despite 68% of cast votes being in favor, turnout, not just sentiment, decided the outcome

Worked example: whale concentration

Given: the same DAO, but one wallet holds 120,000 tokens, 12% of total supply
Step: that single wallet alone can meet the 10% quorum bar and, if it votes yes, contributes more to the “yes” total than dozens of smaller holders combined
Answer: a proposal this whale supports can pass even if most small holders never vote, illustrating how token-weighted governance can concentrate effective control in a few large holders

Worked example: quadratic voting cost

Given: under quadratic voting, casting n votes on one proposal costs n² tokens (or voice credits), not n
Step: a whale wanting 10 votes must spend 10² = 100 credits, while ten separate small holders each casting 1 vote spend 1² = 1 credit each, 10 total
Answer: concentrating votes in one account costs far more per vote than spreading the same total across many accounts, which is the mechanism’s intended defense against whale dominance, though it’s still vulnerable to one whale splitting tokens across many wallets

DAO Tooling

Snapshot: off-chain, gasless signaling votes, widely used for temperature checks before a binding on-chain vote
Governor contracts (e.g. OpenZeppelin Governor): on-chain, binding voting and execution logic
Aragon and Tally: interfaces for creating and managing on-chain governance without writing custom contracts
Gnosis Safe: the multisig wallet most DAO treasuries and emergency councils use to hold and move funds
Discourse and Discord: the off-chain forums where most substantive proposal debate actually happens before a vote is ever opened

Why It Matters

Lets a protocol or community manage a shared treasury and make collective decisions transparently, every proposal and vote is publicly visible on-chain
Removes the need to trust a single founder or company with funds and upgrade rights indefinitely
Gives token holders a direct, verifiable mechanism to influence protocol direction instead of relying on a company’s roadmap
Provides an audit trail: every past decision and its vote breakdown is permanently inspectable, unlike closed-door corporate decisions
Enables global, permissionless participation, anyone who holds the token can propose and vote regardless of location or legal status
Automates execution of approved decisions, once a vote passes and clears its timelock, the change happens without anyone needing to manually implement it
Lowers coordination costs for globally distributed contributors who would otherwise need a formal company structure to pool resources and act together

Common Pitfalls

Assuming token-based voting is automatically fair, wealth concentration in a few large token holders can dominate outcomes just like traditional shareholder voting
Underestimating governance attack risk, a DAO’s own voting mechanism can sometimes be exploited, e.g. borrowing a large token balance briefly (a flash loan) just to pass a malicious proposal
Assuming a passed vote is automatically safe to execute, a timelock exists precisely because a majority vote can still approve a proposal with an unnoticed bug or hidden malicious clause
Low voter turnout making quorum the real bottleneck, most token holders don’t vote, so a small, coordinated group can steer outcomes
Treating “decentralized” as a fixed property rather than a spectrum, many DAOs still have an admin key or a small multisig that can override governance in an emergency
Ignoring legal ambiguity, DAO members can carry personal liability in some jurisdictions since the organization often isn’t a recognized legal entity
Underestimating proposal spam or governance fatigue, too many low-stakes votes can cause participation to drop over time, weakening quorum on the proposals that matter
Overlooking front-running risk on public proposals, since votes and their outcomes are visible before execution, actors can position themselves (e.g. trade ahead of a treasury swap) before a passed proposal actually runs

Comparison

	DAO	Traditional corporation	Multisig-only org	Centralized platform
Decision rules	On-chain smart contract, token vote	Bylaws, board, shareholder vote	Fixed set of key holders approve	Executives decide
Transparency	Fully public on-chain	Limited to filings/disclosures	Public transactions, private discussion	Usually opaque
Speed	Slow, voting windows and timelocks	Slow, board/shareholder process	Fast, few signers	Fast
Legal recognition	Often unclear or jurisdiction-specific	Well established	Unclear	Well established
Amendment process	Governance proposal and vote	Legal amendment process	Signer agreement	Internal decision
Liability exposure	Often unclear, sometimes personal	Limited by corporate structure	Depends on signer agreements	Limited by corporate structure

Example

MakerDAO token holders vote on risk parameters for the Dai stablecoin system, such as which collateral types are accepted and their required collateral ratios, with approved changes executing automatically through the protocol’s smart contracts. ConstitutionDAO is a well-known example of the model’s limits: it raised roughly $47 million in ETH from thousands of contributors in days to bid on a physical copy of the U.S. Constitution, lost the auction, and then faced high gas costs and logistical friction refunding everyone.

Referenced by

Dig deeper

Spot an error in this note?