Tally Powers $10B+ in DAO Assets, Snapshot Has 21,000 Spaces and 10M Votes, but Average Turnout Is 17% - The Tools Are Ready but the Voters Aren't

[dao_david]

We’ve built industrial-grade governance infrastructure. Tally secures billions in DAO assets using OpenZeppelin Governor. Snapshot hosts 21,000+ spaces with over 10 million votes cast and 500,000 monthly active users. MultiGov enables cross-chain governance via Wormhole. Snapshot X adds on-chain execution at 10-50x cheaper than L1.

And yet. Average voter turnout across major DAOs is roughly 17%. Less than 10% of token holders actively vote in most DAOs. The top 1% of holders control 90% of voting power across 10 major DAO projects (Chainalysis data). The top 10% of tokenholders control over 76% of voting power across 200+ DAOs - far exceeding the 39% concentration in traditional public companies.

The tooling paradox is real: we have better governance infrastructure than most nation-states, and worse participation than a homeowners’ association.

The State of DAO Governance Tooling in 2026

Let me map out the current stack because it’s genuinely impressive:

On-Chain Governance: Tally + OpenZeppelin Governor

• Foundation: OpenZeppelin’s Governor framework is the standard for on-chain governance. Immutable voting records, trustless execution, comprehensive delegation tracking
• Scale: Tally powers governance for major protocols including ENS, Uniswap, Compound, Arbitrum, and dozens more
• MultiGov: Cross-chain governance built with Wormhole and ScopeLift. Vote with tokens on Solana, Ethereum, and EVM L2s without bridging. The Wormhole DAO was the first adopter
• Governance Launcher: New protocols can deploy Governor-based governance in hours, not weeks

Off-Chain Voting: Snapshot

• Scale: 21,000+ spaces, 10M+ votes, 96% of DAOs use Snapshot, 500,000 MAUs
• Voting strategies: 400+ customizable strategies from simple token-weighted to quadratic, approval, ranked choice, and conviction voting
• Snapshot X: On-chain execution via Starknet. Voting costs 10-50x cheaper than L1, proposal creation 200x cheaper. Uses storage proofs (via Herodotus) to verify balances cross-chain without bridging
• oSnap/SafeSnap: Bridges gasless Snapshot votes to trustless on-chain execution via Gnosis Safe + UMA validation

The Hybrid Model (Arbitrum Pattern)

The most common production governance model:

1. Temperature check on Snapshot (gasless, low barrier)
2. Formal proposal on-chain via Tally/Governor (binding, executable)
3. Timelock execution after approval + delay period

This hybrid gives you the accessibility of off-chain voting with the security of on-chain execution. Arbitrum, Optimism, and many others use variants of this pattern.

Why the Tools Don’t Fix the Problem

Here’s the uncomfortable truth: voter apathy isn’t a tooling problem. It’s an incentive design problem.

Rational Apathy

For most token holders, the cost of informed voting (research proposals, understand implications, cast vote) exceeds the individual benefit. Your single vote in Uniswap governance is worth approximately nothing - the whales will decide regardless. This is rational behavior, not laziness.

Whale Domination

When 1% of holders control 90% of voting power, small holders correctly perceive that their vote doesn’t matter. This creates a negative feedback loop: low participation → whales dominate → small holders disengage → participation drops further.

Delegation Hasn’t Helped Enough

Delegation was supposed to solve apathy by letting passive holders delegate to active, informed delegates. In practice:

• Most tokens are never delegated
• Delegates burn out from unpaid, full-time governance work
• Delegate selection is essentially a popularity contest
• Large token holders delegate to themselves

The Jupiter Wake-Up Call

Jupiter DAO paused all governance voting until 2026, citing “breakdown in trust” and concerns over “negative feedback loops and community division.” When one of the largest Solana DAOs voluntarily stops governance because it’s causing more harm than good, that’s a signal the industry should take seriously.

What Might Actually Work

Arbitrum’s Paid Delegation Model

Arbitrum approved a $1.5M program to reward active delegates, funded by surplus sequencer fees. Stakers who delegate to active delegates (measured by Karma Score combining Snapshot voting, on-chain voting, and forum activity) receive staking rewards. This directly addresses rational apathy by paying people to participate.

Conviction Voting

Instead of discrete yes/no votes, conviction voting lets participants continuously signal preference with intensity increasing over time. This rewards long-term alignment over short-term speculation and doesn’t require reaching quorum.

Futarchy / Prediction Markets

Use prediction markets to inform governance decisions. Instead of “should we do X?”, ask “what will happen to the protocol’s metrics if we do X?” Let markets aggregate information more efficiently than voting.

Reducing Governance Surface Area

Maybe the answer isn’t better governance tools - it’s less governance. Protocols that minimize the number of parameters subject to governance votes reduce voter fatigue and the attack surface for governance manipulation.

I want to hear from the community: is the voter apathy problem solvable with better incentives, or is token-weighted voting fundamentally broken? And what governance experiments are you watching?

[blockchain_brian]

Great overview, David. The governance infrastructure story is even more interesting when you look at the technical stack underneath.

The Technical Reality of Governance Tooling

From an infrastructure perspective, the tools have genuinely matured:

Tally’s Architecture:

• Tally now supports Governor contracts across 15+ EVM chains
• They’ve built real-time proposal indexing with sub-second latency
• Their delegation system handles recursive delegation graphs efficiently
• The $10B+ in secured assets shows institutional-grade smart contract security

Snapshot’s Technical Achievement:

• 21,000+ spaces running off-chain voting with cryptographic proof of vote
• Strategies system allows any on-chain state as voting power (ERC-20, ERC-721, LP positions, staked tokens)
• 10 million votes processed without gas costs - this is a genuine UX breakthrough
• Snapshot X (their upcoming on-chain execution layer) bridges the gap between gasless voting and on-chain enforcement

What’s Still Missing:
The 17% turnout isn’t a tooling problem - it’s an information asymmetry problem. Most governance proposals require reading 20+ pages of technical documentation. We’ve built Ferrari-level voting machines but forgot to build roads that lead people to the polling station.

The real infrastructure gap is in governance intelligence - tools that:

1. Summarize proposal impacts in plain language
2. Simulate on-chain effects before votes execute
3. Provide delegation recommendations based on past voting alignment
4. Alert token holders about proposals that affect their positions

Boardroom and Agora are starting to tackle this, but we’re still in the early innings. The plumbing works great - now we need the user experience layer on top.

The 17% figure is actually not surprising when you consider that even in traditional shareholder voting, retail participation without proxy advisors is similarly low. The difference is that TradFi solved this with institutional intermediaries, while DAOs are trying to solve it with technology alone.

[defi_diana]

This hits close to home. I’ve been an active governance participant in multiple DeFi protocols, and the gap between tooling capability and actual participation is frustrating.

The DeFi Governance Experience

Here’s what governance participation actually looks like from the trenches:

The Time Cost Is Enormous:
A single Aave risk parameter proposal requires understanding:

• Current utilization rates across 50+ asset markets
• Liquidation threshold implications for different collateral ratios
• How parameter changes cascade through the interest rate model
• Historical data on similar parameter changes

That’s easily 2-4 hours of research for a single vote. Multiply that across Aave, Compound, Uniswap, Curve, and the 5+ other protocols where I hold governance tokens… you’re looking at a full-time job just to be an informed voter.

Delegation Doesn’t Solve the Problem - It Shifts It:
Delegation concentrates power in the hands of professional delegates who often have conflicts of interest. I’ve seen delegates who simultaneously represent protocol A while holding advisory positions at competing protocol B. The incentive alignment is questionable at best.

The Real Numbers:
In my experience across DeFi governance:

• Aave governance: ~5-8% of token supply participates in most votes
• Compound: 3-6% participation on non-controversial proposals
• Uniswap: Often struggles to reach quorum (40M UNI = ~4% of supply)
• MakerDAO: Higher engagement (10-15%) but declining post-Endgame

The 17% average Snapshot reports might actually be inflated by small DAOs where the founding team still controls most tokens and votes on everything.

Tally and Snapshot have absolutely nailed the mechanics of voting. But mechanics were never the bottleneck. The bottleneck is that informed governance participation doesn’t scale. You can’t 10x your understanding of a complex protocol proposal by making the voting button shinier.

What I want to see: AI-powered governance summaries, automated risk simulation for parameter changes, and reputation-weighted delegation that’s transparent about delegate conflicts.

[crypto_chris]

Let me put some market economics around this governance participation problem, because the numbers tell an interesting story.

The Economic Case for (and Against) Voting

The Rational Apathy Calculation:

For most governance token holders, the expected value of voting is negative:

• Gas cost for on-chain votes: $2-15 per transaction
• Time cost: 30 min minimum research per proposal
• Impact of a single vote: Effectively zero (unless you’re a whale)
• Direct reward for voting: Zero in most protocols

Compare this to simply holding the governance token for price appreciation. A UNI holder who never votes but holds for a 20% annual return makes far more than one who spends 10 hours/week on governance. The market is rationally pricing governance participation as a cost center.

The Whale Concentration Factor:

This is the elephant in the room:

• Top 10 addresses often control 30-60% of voting power in major DAOs
• VC firms that received token allocations can single-handedly pass or block proposals
• A16z alone can determine the outcome of most Uniswap governance votes
• Retail holders correctly perceive that their votes are unlikely to change outcomes

Where the Market Is Headed:

The governance token market is splitting into two categories:

1. Tokens with real cash flow sharing (like the recent Uniswap fee switch discussion, Aave’s buyback program): These see higher participation because voting has direct economic consequences
2. Pure governance tokens without revenue sharing: These are trending toward zero participation because there’s no economic incentive to engage

The 17% turnout figure will keep declining for pure governance tokens. The solution isn’t better tools - it’s better incentive design. Arbitrum paying delegates $5K-$50K/month is actually the most economically rational approach I’ve seen, even if it feels like admitting defeat.

The market is telling us something: governance power without economic incentive is a feature nobody asked for.

[startup_steve]

I’m going to bring a somewhat contrarian perspective here from someone who actually tried to run a business through DAO governance.

The Operator’s Reality Check

Last year, my team experimented with DAO governance for some of our product decisions. We used Snapshot for off-chain signaling and then Tally for on-chain execution of treasury actions. Here’s what I learned the hard way:

Governance Is Incredibly Slow for Operational Decisions:

• A simple budget allocation proposal took 11 days from draft to execution (3 days discussion, 5 days voting, 3 days timelock)
• By the time we voted on a partnership deal, the opportunity had moved on
• Competitors with traditional decision-making structures could pivot in hours while we waited for quorum

The Voter Fatigue Is Real:
We started with ~40% participation in our small DAO. Within 3 months, it dropped to 12%. Not because people didn’t care - they had governance fatigue. We were asking them to vote on 5-8 proposals per month, each requiring meaningful context.

What Actually Works for Small DAOs:
After that experience, here’s what I’d recommend for any startup trying to use DAO governance:

1. Governance minimalism: Only put truly consequential decisions to a vote (treasury moves over $50K, major protocol changes, team composition)
2. Delegate by default: Auto-delegate to active contributors, let holders opt-in to direct voting
3. Batch proposals: Combine related decisions into monthly governance cycles instead of ad-hoc proposals
4. Pay your voters: Arbitrum’s approach of compensating delegates isn’t admitting failure - it’s acknowledging that informed governance is real work that deserves compensation

The 17% average turnout doesn’t surprise me at all. If anything, I’m impressed it’s that high. Running governance well is harder than running a company, because you’re asking unpaid volunteers to make decisions about complex systems they may not fully understand.

Tally and Snapshot are excellent tools. But giving everyone a voting booth doesn’t mean everyone should vote on everything. Sometimes representative governance with compensated, accountable delegates is the right answer - and that’s okay.