Onchain Atlas

DAO Governance

Token or membership holders propose and vote on decisions — spending treasury funds, changing parameters, upgrading contracts — instead of a company's executives deciding alone.

Also called: onchain governance · token voting · decentralized governance

What it is

DAO (Decentralized Autonomous Organization) governance is the umbrella term for the systems a protocol or community uses to make collective decisions on-chain or semi-on-chain: who can propose changes, who can vote on them, how votes are weighted, and how passed proposals get executed. It's less a single mechanism than a category — most DAOs combine several of the more specific mechanisms in this Atlas (token voting, delegation, timelocks, treasuries, sometimes ragequit or slashing) into one governance stack.

How it works

  1. A DAO issues a governance token or membership unit that represents voting power, distributed via sale, airdrop, or contribution.
  2. Any holder meeting a proposal threshold (a minimum token balance, or DAO membership) can submit a proposal — text-only ("temperature check") or binding (a transaction the DAO will execute if it passes).
  3. A voting period opens, typically days to weeks, during which token holders cast votes for/against/abstain, often weighted by tokens held or delegated.
  4. Many DAOs let holders delegate their voting power to representatives who vote on their behalf, since most holders don't actively track every proposal.
  5. If quorum (minimum participation) and approval thresholds are met, the proposal passes.
  6. A timelock delay often follows before execution, giving members a window to notice a malicious or unpopular passed proposal and exit or organize opposition.
  7. After the timelock, the proposal executes automatically via a smart contract (releasing treasury funds, upgrading a contract, changing a parameter) — no human intermediary is needed to carry it out.

Why designers use it

  • Distributes control over a protocol's treasury and parameters to its actual stakeholders rather than concentrating it in a founding team, improving legitimacy and credible neutrality.
  • Makes decisions transparent and auditable — every proposal, vote, and execution is recorded on-chain.
  • Allows a protocol to keep evolving (upgrades, treasury allocation, partnerships) after the founding team steps back or moves on.
  • Aligns incentives between token holders and the protocol's long-term health, since holders who vote badly bear the consequences through their own token's value.

Failure modes

  • Voter apathy: turnout is often extremely low, so a small, motivated minority (or a single whale) can pass proposals that the broader token base would have opposed had they voted.
  • Plutocracy: because voting power is proportional to tokens held, wealth concentration directly becomes decision-making concentration, often the opposite of the "decentralized" branding.
  • Governance attacks: an attacker can borrow or buy enough tokens briefly (flash-loan or otherwise) to pass a malicious proposal, then exit before consequences land, unless timelocks and snapshot mechanics prevent it.
  • Proposal spam and voter fatigue lead to rubber-stamping — voters stop reading and just follow delegate recommendations or vote "yes" on everything from familiar-looking proposers.
  • Legal and practical ambiguity: "the DAO decided" doesn't map cleanly onto real-world liability, contracts, or enforcement, creating unresolved questions about who is actually accountable when something goes wrong.

What to check before using it

  • What's the actual quorum and participation history — is governance realistically decentralized, or does a small group of insiders control outcomes?
  • Is there a timelock between a vote passing and execution, long enough for the community to react to a malicious or buggy proposal?
  • Can voting power be acquired temporarily (flash loans, exchange-held tokens, borrowed governance tokens) to swing a vote, and is that blocked via snapshotting?
  • Is delegation available and are delegates' voting histories visible, so passive holders can still exercise meaningful oversight?
  • What is actually outside governance's reach (e.g., core protocol security parameters, emergency pause powers) and who controls those instead?

Experiments that used it · 4

Shown oldest first, so you can watch the design evolve.

WHALE
Pseudonymous NFT mega-collector WhaleShark issued a 10M-supply ERC-20 social token whose value was underwritten by 'The Vault,' a publicly audited multi-million-dollar NFT collection governed by a token-holder DAO.
2020 partial success
Fei Protocol
A decentralized, ETH-backed stablecoin (FEI) that pioneered 'direct incentives' (mint rewards / burn penalties on DEX trades) and Protocol Controlled Value, but launched into a peg crisis and eventually wound itself down after a merger with Rari Capital and an $80M exploit.
2021 failed
Mango Markets
Solana's flagship cross-margin lending and perpetuals DEX, drained of ~$110M via oracle manipulation of its own MNGO token in 2022, then wound down in 2025 citing regulatory constraints.
2021 exploited
Marinade Finance
Solana's first liquid staking protocol, which issued mSOL against algorithmically delegated stake and later evolved into a validator stake-auction marketplace with slashable performance bonds.
2021 ongoing