DAO metrics are the quantitative measures used to judge whether a decentralized autonomous organization is healthy: how many members actually vote, how concentrated its voting power is, how quickly it processes decisions, and how large and durable its treasury is. Because a DAO's governance and finances are recorded on a public ledger, these figures are measurable in a way corporate governance rarely is — every vote, delegation, and treasury movement is on-chain. This page is a reference for the metrics that matter and where to read them live; it links out to the deeper concept pillars for each mechanism.

Governance participation metrics

Voter turnout — the share of eligible voting power (or of eligible holders) that participates in a given vote — is the headline number, and it is chronically low. Single-digit-percentage turnout is the norm even in flagship DAOs, and turnout by holder count is usually far lower than turnout by token weight, because a handful of large holders and delegates carry most votes. a16z crypto's DAO research roadmap frames turnout as one of the open problems of on-chain governance, drawing directly on the political-science literature on civic participation.

Related participation figures:

• Quorum attainment — the fraction of proposals that reach the minimum participation threshold required to be valid. Setting quorum too high stalls a low-turnout DAO; too low, and a small coalition can pass binding changes. A frequently-cited cautionary case is a 2021 SushiSwap proposal that would have handed treasury control to a single multisig and passed with roughly 3% of supply voting.
• Proposal throughput — how many proposals a DAO processes over a period, across the proposal lifecycle (forum discussion → temperature check → on-chain vote). Throughput signals whether governance is a living process or a rubber stamp.
• Pass / approval rate — the share of on-chain proposals that succeed. A pass rate near 100% often means real contestation happened off-chain (in temp checks) before anything reached a binding vote, so the on-chain record alone understates disagreement.

Voting-power concentration

Turnout tells you how many people vote; concentration metrics tell you how much it would matter if the rest did. The best-known is the Nakamoto coefficient — the minimum number of entities whose combined voting power exceeds the threshold needed to control an outcome (a simple majority, or the quorum). A small coefficient means a handful of token holders or delegates can dictate results. Empirical work on major token-weighted DAOs has repeatedly found this coalition to be small — on the order of single digits to low tens of addresses for prominent protocols (Analyzing voting power in decentralized governance: Who controls DAOs?, 2024). Because delegations shift daily, the current value should be read from a live tracker rather than a static number.

Complementary concentration measures:

• Delegate breakdown — the share of total voting power held by the top N delegates, and how much of that is self-delegated versus delegated by others. ENS, Uniswap, and Compound publish this on their governance front-ends.
• Gini coefficient — the classic inequality measure applied to the voting-power distribution.
• Voting-bloc entropy — a newer decentralization metric that scores how votes actually cluster into aligned blocs over time, rather than counting token balances at a single instant (DAO Decentralization: Voting-Bloc Entropy, 2023).

Whether a DAO is "decentralized enough" is itself contested; the academic framework When is a DAO Decentralized? (2023) argues it is not a single number but five dimensions — token distribution, infrastructure, governance, escalation, and reputation — measured together.

Treasury and financial metrics

The financial side is covered in depth on DAO treasury management; the metrics that summarise it are treasury size (total value of assets the DAO controls), asset composition (the share held in the DAO's own governance token versus stablecoins and blue-chips — the concentration problem), and runway (how many months of budgeted outflows the liquid, non-native portion covers). DeepDAO aggregates treasury value across thousands of tracked organisations — an order of tens of billions of dollars in total, though the figure swings with token prices, so it is best read live rather than quoted. Pairing runway with contributor-compensation outflows is what turns a treasury balance into a solvency signal.

Where the data lives

No single dashboard is authoritative; the practical answer is to triangulate:

• DeepDAO — the widest cross-DAO index: treasury value, member counts, participation, and governance activity for thousands of organisations.
• Boardroom and Tally (continued as Cactus) — proposal-level and delegate-level data on the DAOs whose governance they host.
• Snapshot — the raw record of off-chain signalling votes and their turnout.
• Dune — community-built SQL dashboards for bespoke, DAO-specific metrics.
• Token Terminal and on-chain analytics — protocol revenue and financial fundamentals behind a treasury.

The limits of DAO metrics

Every metric here is gameable and partial. Turnout can be inflated by a single whale voting on trivial proposals; a Nakamoto coefficient counts addresses, so one entity splitting across many wallets looks decentralised while a Sybil cluster looks like a crowd (see Sybil resistance). Off-chain deliberation — the forum threads and Discord debates where most real governance happens — leaves no on-chain trace, so a healthy DAO can look inert on-chain. And once a number becomes a target (Goodhart's law), participants optimise the metric rather than the health it was meant to proxy. Read metrics as a starting question, not a verdict.

How Caper approaches this

Several numbers a conventional DAO has to monitor, a caper makes structural. Voting weight is not just a token balance to watch for concentration: it is (t·v)/(V·T), where a member's live governance-token holding t is multiplied by their earned, non-transferable vote-token balance v — so a member who has never participated has weight zero, and buying tokens without voting buys no influence (verified against the contract's compute_vote_weight). Turnout and concentration are therefore coupled to participation by construction rather than tracked after the fact. The treasury is the bonding-curve vault, funded as members buy in, so its size is legible from the curve state rather than assembled from multisig balances. And "runway risk" is bounded from the other side: any member can exit for their pro-rata share at any time — the same (t·v)/(V·T) weight — so a stagnating treasury is priced continuously by exits, not discovered in a quarterly report. Caper does not invent new metrics; it removes the gap between the number and the thing it measures.