The "number of voters" shown on governance voting pages can be an illusion. If you only judge by total count, the apparent vibrancy is often just a few individuals' enormous delegated votes being counted repeatedly.
Here are 3 steps to distinguish delegated votes from real participants and understand the true state of governance.
Step 1: Check the "Number of Voters" Metric – Addresses vs. Entities
On governance platforms like Snapshot and on-chain voting systems, the "number of voters" shown by default counts the number of wallet addresses that voted, not "real people." An address could represent a genuine active participant, or it could represent a "delegate" proxy who has gathered voting power from thousands of tokens.
What to do: Open the voting page and look at the detailed results of the proposal.
How:
Check the voting results page, which usually displays "total voters" and "total voting weight" (the total number of tokens that participated in the vote).
Compare the two numbers: If the token amount corresponding to the total voting weight is far larger than what the actual number of voting addresses could reasonably hold, it indicates a large portion of votes came from delegated proxies.
How to know it's done: You now understand that the "voter count" data you're seeing does not yet distinguish between "addresses that voted directly" and "delegated voting power" – they are two different things.
Common pitfall: Taking the "Unique Voters" number on a proposal page as the real number of community participants. Data on Snapshot shows that the average number of voting addresses for most crypto projects is below 500; only Sushi consistently has over 1,000.
Step 2: Examine Delegation Relationships – Separate Delegated Votes from the Total
To tell "delegated votes" apart from "direct votes," you must look into the system's delegation records.
What to do: View "voting delegation records" on a block explorer or governance dashboard.
How:
Scenario A (Using Snapshot): Snapshot doesn't show delegation relationships directly, but you can check the project's governance page for a "Delegates" list, which shows the main proxy delegates and how many votes they have received.
Scenario B (Using on-chain governance like Compound Governor Bravo): On Etherscan, locate the DAO's governance contract, call the
getVotes(address)function to query voting power for different addresses, or directly inspectdelegatesevent logs. Research on protocols like MakerDAO indicates that token holders can discern delegates' behavior and reward them: delegates who vote correctly tend to see a larger increase in delegation share the following month.
How to know it's done: You now have, for the top few high-weight voters in this proposal, their respective "own vote" amounts and "delegated vote" amounts.
Step 3: Recalculate Effective Participation – The Real Number After Stripping Out Delegation
Only after you strip out delegated votes can you see how many people are truly making active decisions.
What to do: Use delegation data to back-calculate the number of unique addresses that voted directly.
How:
Formula:
Directly voting addresses = Total voting addresses - Addresses that only voted via delegated proxyIf a community has 500 total voting addresses, and 300 of them each have less than 1 vote weight, but 200,000 votes come from just 2 large addresses and those 200,000 votes account for 70% of the final result – then the appearance of 500 participants is essentially a result swayed by 2 proxy agents.
How to know it's done: You've calculated the real number of direct participants and you know the share of total voting weight coming from delegated votes.
Caution: Concentration of delegation is common. Research shows that in Nouns DAO, delegates consistently hold about 68%–78% of the voting power, while the share from ordinary members voting directly is far lower. Therefore, when you see a "large number of voters," you must be alert that this is not active community participation but rather the result of power concentrated in a few delegates.
