Delegated Voting Procedure: How We Govern

Delegated Voting Procedure : How We Govern

Authors: Samantha Marin, with help and feedback from Alex Clay, Ivan Fartunov, Jessica Smith, and Mario Laul.

This is a procedural document that outlines a possible delegated voting plan for the AN DAO. This document is not intended to replace the charter, but build a basis for large scale decision making in the DAO following the treasury transfer.

Who are we?

We are the Aragon Network DAO, a DAO that builds products for the next generation of DAOs. We’re a product DAO that ships industry-leading, open-source, blockchain-enabled technology.

  • DAO Members: ANT holders.
  • DAO Contributors: Any ANT holder who is part of a guild, subDAO, or working group in the AN DAO. DAO Contributors may have governance rights in their respective guild/subDAO/working group.
  • DAO Core Contributors: Any ANT holder part of a guild/subDAO/working group who works in a full-time capacity for the AN DAO and/or is in a leadership position. DAO Core Contributors may have governance rights that go beyond their fellow contributors, such as veto power or more votes.
  • Delegates: Individuals who vote on matters for the entire DAO, using tokens that have been delegated to them. Delegates don’t have automatic special power in guilds/subDAOs/working groups. If they are a core contributor in Aragon, they must state conflicts of interest on proposals that directly impact them, and can choose to abstain from votes in which they have those conflicts.

How do we vote?

In this structure, the Aragon Network DAO will use delegated voting to ratify decisions. ANT holders can delegate their ANT tokens to one or more delegates. There is no token allocation cap to be a delegate, but there is a cap in order to be paid to be a delegate.

Each on-chain vote will be live for 7 days, with a 3 day timelock following the results.

How do we select delegates?

The 11 paid delegates will be chosen every 12 months by an ANT vote, where ANT holders allocate certain percentages of their tokens to the various delegates. The 11 prospective delegates who receive the most percentage of the ANT votes will become the paid delegates for the following 12 months.

The two aspects of the application process are the written application and the community AMAs (sync or async).

Delegates can run again after their previous 12-month term, and token holders may choose to delegate to the same person as before.

Delegates can be either individuals or small groups/organizations.

  • Written application: These statements will be stored in a public repository. Please be short and concise.
    • ETH address
    • ENS name, so delegators can easily identify you
    • Discord username
    • Twitter handle
    • Reasons for wanting to be a delegate
    • Web3 interests (choose 5 or fewer)
      • Accessibility, Bridges, Cryptography, DAOs, Data and analytics, DeFi, Developer tools, Economics, Events, Environment, Gaming, Governance, Identity, Infrastructure, Legal, Messaging, Music, NFTs, Oracles, Privacy, Security, Self-management, Social impact, Visual art, Wallets, Writing, Zero knowledge.
    • Qualifications and experience
    • Technical experience
    • In your words, what is Aragon?
  • Community AMA: Delegates will need to answer community questions, whether in written format or a video AMA.
    • Here, the community members will ask the prospective delegates questions. This will help ANT holders determine where to delegate their tokens.
    • The documents and/or session recordings will be made public.

There will be a two-week application period, where prospective delegates can submit their applications. The vote will begin two weeks before the previous delegate period expires, and will end one week before the period expires. On the expiration date of the previous delegation period, all tokens will be returned to the original holders, who can then choose to re-delegate to the next round of delegates.

Timeline example:

March 1 (day 1): Delegate application period opens.

March 2-13 (days 2-13): Written applications and AMAs are held.

March 14 (day 14): Delegate application period closes at midnight UTC.

March 15 (day 15): Delegate vote is held—ANT holders use the written applications and AMAs in their decision making process.

March 22 (day 22): Delegate vote concludes. 11 paid delegates are selected.

12 month term begins March 22.

Rules governing delegates:

  • Token Minimum: All paid delegates must have at least 0.1% of the total circulating ANT share to be able to vote.
  • Token Maximum: No delegate may receive more than 3% of the total circulating ANT share. Delegates must close off to additional delegation once they reach the 3% threshold. Any delegate who goes beyond 3% will receive a one-week warning. If they still have 3%+ after that week, their wallet will be barred from voting.
  • Term Limit: Paid delegates have a 12 month term limit. Once their limit is reached, tokens will be returned to holders, who will need to re-delegate. Token holders can re-delegate to the same delegate if they choose. Delegates must announce one month before their term limit expires or personal resignation, so token holders can re-delegate.
  • Compensation: Delegates will be modestly compensated a base rate of 3,000 USDC per month (flag for the community, should it be more to attract more technically proficient delegates?). All delegates will be compensated the same, no matter the number of tokens delegated to them.
    • The only way to raise delegate compensation is with a 1-of-1 vote from the DAO members, not a token-weighted vote where delegates control votes.
    • In addition to the 3,000 USDC, delegates are compensated on a sliding scale based on the percentage of tokens allocated to them.The structure could look something like this:
      • 0.10% to 0.30% = an additional 750 USDC per month
      • 0.31% to 0.50% = an additional 1,000 USDC per month
      • 0.51% to 0.70% = an additional 1,250 USDC per month
      • 0.71% to 0.90% = an additional 1,500 USDC per month
      • 0.91% to 1.20% = an additional 1,750 USDC per month
      • 1.21% to 2.20% = and additional 2,000 USDC per month
      • 2.21% to 2.99% = an additional 2,250 USDC per month
    • The minimum a delegate earns monthly is 3,000 USDC. The maximum is 5,250 USDC.
    • Delegate payment is income. Do your own research as to how your jurisdiction taxes income.
  • Signaling: Delegates must post on Twitter, Discord, or Forum how they plan to vote 48 hours before the on-chain vote begins. This allows delegators to move their tokens if they do not agree with the voting stance. The Snapshot vote (described below) is also an important part of the signaling process.
  • Strikes: Delegates will be subject to a “three strike” system. Strikes include missing on-chain votes and voting differently than their signal. If they reach three strikes, their wallet will be barred from voting. Discourse and Snapshot signaling votes are highly encouraged, but will not be punished if missed.

Delegates should….

  • Vote in the way they feel best benefits the Aragon Network as a whole, not a small group.
  • Operate “open by default,” meaning they share their ideas and opinions with the greater ecosystem.
  • Make an announcement one month prior to stepping down (if mid-term) or if they don’t plan to run for the next term limit. This allows ample time for new delegates to prepare statements.
  • Put forth their own proposals and ideas for the DAO, and shepherd them through completion.
  • Check on workstreams to have full transparency of how those workstreams are running.
  • Take their job seriously, putting time into reading and understanding each proposal.
  • Call upon subject matter experts if they need help understanding the proposal.
  • Represent the Aragon ecosystem, both in our internal channels and externally, such as through conferences, Twitter, articles, podcasts, and more.
  • Operate with professionalism.

Delegates should not……

  • Exceed their 12-month term limit without a re-delegation period.
  • Exceed the 3% maximum token threshold.
  • Bribe delegators for votes.
  • Collude with other delegates for malicious reasons.
  • Actively campaign for tokens.
  • Raise their compensation via a delegated vote. Rather, that must be done via 1-1 voting.

How do we prepare and pass proposals?

Proposals, or Aragon Governance Proposals (AGPs), are fully-formed ideas that are ready for the DAO’s feedback. Proposals are polished, thoughtful, concise, and free of jargon.

Proposals include:

  • Title: clear and concise, not vague or confusing.
  • Short description: think of this as a TLDR or abstract.
  • Longer description: the meat of the proposal. Include the scope of the project here, too.
  • Technical specification (if necessary): for proposals that include code or necessary legal jargon, add those in here. Think of this as the “expert deep dive.” Most proposals won’t need this—it should all be in the longer description.
  • Metrics or Key Performance Indicators: if the proposal passes and you begin working on it, what metrics or KPIs will you use to show progress?
  • Team description: who will be working on and shepherding this project? Describe them and their relevant experience.
  • Next steps if passed: describe what your first few weeks will look like if this proposal passes.

All proposals should be aligned with Aragon’s manifesto and its mission, vision, and values.

Three-step proposal passing process:

  1. Proposal is posted to Discourse for 1-person-1-vote signaling, with a poll with the following options: “Yes, move to Snapshot” and “No.” Discourse quorum and pass rates are still being determined—up for debate on the forum.
  • Discussion and voting period: Minimum 7 days, maximum 14. The proposer is responsible for moving the proposal to Snapshot and not letting proposals linger on Discourse for more than 14 days.
    *The proposer may post a second draft of the proposal, if comments warrant it. The draft should be live for 7-14 days with a poll.
  1. Proposal is posted to Snapshot for token-weighted signaling. Snapshot quorum and pass rates are still being determined and up for debate on the forum.
  • Voting period: 7 days.
  • The proposer must go to the original Forum post on Discourse, edit their post with “This post moved to Snapshot on [DATE]” or “This post did not pass and was deprecated on [DATE].” This is so external parties can see what happened to proposals.
  1. Proposal is posted to the chosen voting platform for token-weighted, on-chain confirmation. On-chain vote quorum and pass rates are still up for debate.

Once the proposal has passed the on-chain vote, it is up to the proposer to shepherd and ship the idea. The proposer may gather a team to assist them, but they must be responsible for executing the proposal.

How do we fork?

Any ANT holder can fork the Aragon Network DAO at any time. To fork the DAO, an ANT holder may post a proposal to Discourse outlining their planned fork of the DAO. This proposal does not need to be voted on, because the proposer may simply create their own fork by themselves. The fork proposal is simply a way to signal to other interested parties that a fork is occurring.

A forked DAO….

  • Has its own token, separate from ANT.
  • Operates in a separate Discord, Discourse, Snapshot, and voting platform.
  • Exists because of irreparable differences between the original DAO and the forked version.

How do we try new governance models?

All governance experiments must be run in a small group prior to putting forward a proposal to the DAO. No governance experiment will be used on the wider DAO level in an untested state. Guilds and subDAOs should write their own governance charters and use that as a testing ground for new methods, then take those learnings to the larger DAO. An example of a guild running its own governance experiment and sharing it with the larger DAO: Writers Guild of BanklessDAO.

Once a governance model has been proved in a small group, that group may propose trying the model on the DAO-wide scale. Metrics and data on how the governance model went are necessary to include in the proposal, and members of that group should be able to speak on how the model can be used at a wider scale.

What do you think of this procedure? I’d love to hear feedback in the comments and the poll below.

Leave a signalling vote here, so I know where to take this draft next:

  • Yes with no edits/minor edits
  • Yes with heavy edits
  • No, another structure entirely

0 voters


As per the proposal that was voted on and passed, I recommend making it explicit that it’s a fully liquid system in the sense that any ANT holder can either vote directly, delegate their voting power to one or more delegates, or un-/re-delegate at any time to vote on their own behalf.

EDIT: The main benefit of this is that it encourages token holders to delegate while not disabling broad participation.


This is in conflict with the following sentence under “How do we vote?”: “There is no token allocation cap to be a delegate, but there is a cap in order to be paid to be a delegate.”

1 Like

Hey @Samantha, I must say that this is a fairly well-written proposal. Kudos to you for putting this together. Based on this post, I have the following comments:

1. Delegate Voter Term: I believe that 12 month period for elected voters in a DAO setup is not good for the functioning of a young and dynamic DAO. Firstly, electing 11 trustworthy candidates for a year is hard when the average number of full-time contributors within the DAO is quite less. Such votes tend to be based on visibility and popularity. This is not wrong but in a young DAO, this might be detrimental as many candidates might have not spent an adequate amount of time with ANT holders or the DAO contributors.

Suggestion: The term should be for 6 months and every 3 months 50% of the delegates should be re-elected. For example, if we elect 6 delegates in March, then the other 5 should be elected in June. This will ensure that the decision-making by delegates is always dynamic based on the contributor sentiment. I am aware this system might seem complex at first, but this will ensure that the DAO is not controlled by a group for a long period of time.

2. Circulating Supply Limits: I believe the Aragon Network has been facing the issue of low ANT holder participation for a while. The figures 0.1% and 3% are very high. Most votes we see today on the VOICE app, don’t even exceed 1% of the total ANT supply in terms of voter participation. We have not tested this and don’t know really how much ANT will actually be delegated. We are also unaware of how much of the circulating supply is held by centralised exchanges. Hence, having a 3% upper limit might lead to a concentration of power in the hand of 1 or 2 members within the DAO. This will be quite worrisome. The only way to mitigate this is to set a quorum of at least 20% of ANT supply. Being in this space for this long, I am aware that this is unrealistic.

Suggestion: We should take an average of the ANT supply used in the past votes to determine the upper and lower limit. Also, the quorum should be such that no delegate controls more than 15-20% of the entire Vote.

3. Conflict of Interest: Delegates might be full-time or part-time workers within the AN DAO working with guilds or might have independent proposals. Hence, the decisions they take as delegates might allow them to determine their own compensation for other projects or even influence other delegates. Also, even if they abstain from voting on their own proposals, voting delegates might have to please the abstaining delegates as they will vote on each other’s proposals in the future.

Suggestion: No person should be allowed to hold any other role other than the role of a delegate within the AN DAO (during the term of being elected).

4. Term Limit: As suggested before, 12 month terms are not good. Similarly, having the same delegate for more than 2 consecutive terms can lead to complacency and votes based on popularity. Also, delegates might get attached to the power they hold over the DAO if they exercise for long without checks. A constant room for flexibility within the delegates is very important. Otherwise, rigidity can lead to the destruction of the very fabric of the DAO ecosystem.

Suggestion: A delegate can only be elected for two consecutive 6-month terms. After the term limit is reached, such a delegate can only be re-elected after 6 months.

5. Robust Delegate Off-Boarding Mechanism: This is extremely important to ensure that delegates are not mistreating contributors or misusing their powers.

Suggestion: During the experimentation phase, a proper system for offboarding delegates should also be designed.

6. Liability Issues: Recently, many lawsuits are being filed in various jurisdictions for misuse of funds or losses generated by DAOs. In the scenario where all transactions are approved by the 11 delegates, they may also be held liable (both criminal and civil) by third parties or contributors. This is quite worrisome for delegates if they are exposed to litigation. This situation becomes worse when the proposal is made by an anonymous party and the accused cannot be identified.

Suggestion: This requires further research.

7. Experimentation: As stated in your post, I think this is key for a successful launch of the delegate voting system.

Lastly, I believe that it is fundamental for a DAO to remain in a space where a hierarchy doesn’t exist. It is probable that having delegates (even if elected) might lead to the creation of a system where the delegates feel that the entire DAO is essentially reporting to them. I would try my best to avoid that.

Also, agree with @mlphresearch: I recommend making it explicit that it’s a fully liquid system in the sense that any ANT holder can either vote directly, delegate their voting power to one or more delegates, or un-/re-delegate at any time to vote on their own behalf.

Thanks @Samantha for getting the discussion going. I hope something good comes out of this.


This is interesting. I was originally leaning towards 12 months because (1) un-/redelegation is always an option throughout the period, and (2) the DAO gets used to having to pay closer attention to this matter always at the same time of the year. But I think you’re making an important point which is that, at the beginning, it might be beneficial to encourage more frequent DAO-wide assessment/reconsideration.

I think it’s important to remember that, in all delegative and representative systems, there is a very strong tendency for incumbents to get re-elected. This makes sense because (1) people are biased towards supporting/trusting familiar names, and (2) people rightly associate competence with prior experience. Until practice proves otherwise, I think it makes sense to assume that this also applies to DAOs. The potential negative effect of this tendency is that newcomers - in part regardless of their competence or the quality of their ideas - are automatically at a disadvantage. This is one of the reasons why it makes sense not to apply a minimum ANT threshold to becoming a delegate because it would make it more difficult for up and coming community members to permissionlessly build a delegate track record and thereby attract more delegations.



My ùnderstanding is the delegates would be in place ahead of S3 to assess funding proposals as of December am I mistaken?

This was my understanding from people that have higher context of the AA transition than I, people also listed as advisors here.

@AlexClay @Fartunova @jessicasmith can you guys clarify please.

I recognise March is an example timeline but it is rather misleading communication if in fact we are electing delegates in late October.

This is a very comprehensive and extentsive proposal to the point of being an entire Charter. And although it is proposed not to replace the Charter, having predefined the entire governance process clearly it intends to. Are we going for another 150+ lost in tangential conversations forum thread as per the transfer of funds?

More practically this raises many many governance alterations so how would you propose formulating the vote for this?


As I wrote in another thread, I personally think the Charter/Constitution is not the ideal place for operational (incl. governance-related) details, which are better addressed in other (version-controlled) parts of project documentation. As I see it, the purpose of the Charter/Constitution should be to define the core structure/principles (in this case the subDAO/guild system and both direct and delegated ANT voting) and, more importantly, to inspire and guide not through step-by-step instructions but through vision, values, and the general ethos/spirit behind the core principles - things that shouldn’t require changing as the DAO learns and evolves. In that sense, the Charter/Constitution can be thought of as a preamble to all other governance documentation - something that animates everything else without ever feeling like a straightjacket.

Of course, I also realize that this is partly a matter of semantics or, more specifically, naming and thereby structuring different parts of governance-related documentation. Anyway, my point here is that it’s naive to think that the DAO won’t need to tweak various proceducal details over time, so it makes sense to organize documentation accordingly.

EDIT: Before writing a new Charter/Constitution, I recommend reading this:


This was simply an example time line, but can see the confusion. If the November transfer is to take place they will need to be in place in advance of S3. So around October time.


Apologies for cross-posting (especially as this is not a thread on the Charter/Constitution per se), but since it did come up: before writing a new Charter/Constitution, I recommend reading this:

First of all, awesome job @Samantha and team!

There are some points that require refinement (like aligning the vote thresholds to the VOICE attendance in practice), but overall this proposal makes sense.

I am 100% pro this way of governing, as I love the “get things done mentality”.
In DAOs people are tempted to keep discussion instead of realizing the DAOs vision and via this way the DAO is tempted to make decisions more efficient.

I think it is best to start with a rough process (like outlined above), and make it relative easy to adjust it while doing it. Put only the most relevant points “in stone”, and lets improving the process along.

1 Like

Hi Samantha and team,

Really great job on this! And I agree with @mlphresearch feedback.

I have 1 important question: have we included feedback from external parties who are major delegates in large DAOs already? I would love to send this for feedback from Lefteris (Optimism, ENS, Gitcoin), Griff (similar), Simona? and a few others? Maybe Juan from Maker?

I don’t want to include them if you have already gotten some external feedback from existing delegates elsewhere. So let me know and i’d be happy to reach out.



@Samantha thank you for leading this towards a well-thought-through proposal and @mlphresearch for thoughtful and helpful comments.

Josh is a friend. We work together on DAO Star One. Depending on how the new charter process goes, I can pull him in for a feedback session.

This can be recalculated monthly to accommodate for the comments @lion917 raised.
So in October 2022 we will have the initial (up to) 11 compensated delegates. However, if a delegate loses their delegated power and falls below the 0.1% threshold or there are 11 other delegates with more voting power, that individual/team will no longer be compensated.

The 12 months is the time frame at which all tokens get automatically undelegated to avoid entrenchment - this is also an important technical requirement for the new system cc: @sembrestels

I’m not 100% sure how I feel about it, but it seems somewhat reasonable. There may be a cap on the additional compensated work delegates can complete with the DAO, either as a full/part-time contributor, as bounty recipient, or through external advisory/ research/ development and other structures.

Two additional comments in this context - this should be limited to the compensated delegates only, and a delegate’s compensation should be increased if this rule is introduced. Given the power and responsibility, delegates yield a compensation of $36k per year and is unlikely to attract dedicated top talent without ulterior motives. The number of compensated delegates should be decreased to lower the financial burden on the network to maybe 5 or 7.
[P.S.] For the sake of transparency here, I do plan to apply for a delegate role so I might be biased.

@lion917 - I get the concern about the 3% and think it is viable. The bottom honestly seems very fair, considering that this lower limit is for the person/team to qualify for compensation as a delegate. Anyone can vote regardless of how much voting power is delegated to them.

@lee0007 - indeed, the timeline in there is just an indication. Realistically ANT holders should be able to start delegating voting power in early October if we are to adhere to the timeline as per the will of the community as expressed by the vote on the original proposal.

Many changes must happen to the charter (if we assume the current one as a starting point). This proposal does not change that need substantially, as it adheres to a proposal that passed with ~3% of the network support. The Charter in its current form passed with <0.2%…it is very clear which of the two should be the legitimate mental anchor if we should have any. Also, I 100% agree with this:

1 Like

These are important points. Question is: what are the time commitment and knowledge/skills required to perform delegate duties, and whether it makes sense to concentrate the delegate compensation budget among a small group of delegates vs. distribute it more broadly? I’m worried that this question is too difficult to answer with precision ex ante, or that the answer may change (perhaps even repeatedly) as the DAO finds its rhythm under the new system. It’s also worth remembering that, when it comes to organization building, it’s generally easier to add than remove.

How much is 0.1% of ANT circulating?

You raise a great point @Anthony.Leuts. We should ideally take feedback from the contributors of DAOs where this system already exists. We will get some insightful feedback.

1 Like

It’s ~38k ANT or about $63k at current prices, as per coingecko reported circulating supply

1 Like

I agree to answer this is difficult. One line of thinking:
The delegates are directing substantial financial resources owned by the network. For the lack of a better word, they are the fiduciaries of the treasury with a responsibility to the network to allocate resources in the best possible way that would enable the network to pursue its objective sustainably (e.g., build useful DAO infrastructure).

To achieve this, the delegates should be able to formulate and deploy some sort of network-level strategy which dictates resource allocation trade-offs. It might be best for that strategy to be unified across delegates, but potentially different delegates will pursue different network-level strategies (getting consensus is tough). Some might decide strategic trade-offs are not a thing and assess proposals on ad-hoc basis until the budget for the year is spent.

To execute their chosen strategy, delegates will have to fund or defund specific work streams (e.g. building smart-contract infrastructure) and initiatives (hackathons, integrations with other projects) that aim to incentivize network stakeholders to contribute useful work to the network.

This requires delegates to assess and plan what type of contributions are needed by the network when, whether the initiatives and workstreams put forward for funding are meeting these needs, whether the teams putting forward those proposals can ship, and whether the results meet deliverables. Forum politics will sadly be an unavoidable aspect of the role.

Ideally, we want delegates to be experienced in making decisions on a strategic level while having a sufficient understanding of practical implications and interdependencies. It should be a mix of backgrounds.

TLDR: It sounds like a demanding role in expertise and time. Parts of it can, over time, be spun out into different guilds - i.e. a guild that reviews independently deliverables sounds like a good idea but the incentive structure there can get wanky

1 Like

Samantha, love the thoughtful attention to detail here!

I think your vision for the future and “end state” for delegation and the governance culture surrounding it is roughly aligned with what I envision and with those who I have spoken to in detail about this, at least in spirit. I do think we can take a less prescriptive path to get there though.

To clarify, when I say less prescriptive, I am not referring to any notion of power or politics, but rather just to increase flexibility and reduce logistic or bureaucratic overhead. Many of the recommendations here that are being proposed as something to codify, could probably be emergent outcomes of a more lean governance system, via smaller proposals that are informed by each previous governance iteration. @lee0007 's observation about this potentially getting complicated into a web of 150+ tangential interrelated conversations sounds likely, and even we were to get a vote through, we could end up with something really tough to work with or understand.

Antifragility via negativa is how I’m thinking about this, or perhaps just with a more agile mindset that can respond to lean governance changes more iteratively.

Hope that makes sense! Can definitely chat more fluidly about this soon if what I’m saying isn’t clear or seems silly.


This raises something that I’ve seen cause issues in other DAOs: the lack of clarity/process when it comes to developing strategies and using these strategies as a reference point when assessing proposals/plans/budgets. It’s difficult and expensive to develop and execute plans without strategies, and it’s difficult and expensive to strategize without a shared vision/mission, so there’s clearly a sequence of steps implied here.

My personal view on this is that developing strategies requires input from different corners of the DAO and should not be centralized with the delegates. Even though the DAO can pursue multiple strategies at once, it will inevitably be forced to prioritize some over others when it comes to allocating resources through ANT votes. Perhaps the initial role of delegates (especially those that receive compensation) should be to publicly scrutinize proposals and provide a sanity check based on whatever vision/mission and strategies the DAO has adopted? Any ANT holder, regardless of their role in the Aragon ecosystem, can always vote with their tokens whichever way they want. But, when it comes to delegates (who are likely to accumulate decisive levels of voting power), I think their Code of Conduct should be very clear about commingling of roles / conflicts of interest.


I tend to agree that the first iteration should focus on getting the basics right and not be overly prescriptive about the rest. A simple role description and a general code of conduct for all delegates (public commitment recommended), and a slightly more detailed role description and a code of conduct for paid delegates (public commitment required) should be enough to get going. Both can be regularly revisited to integrate learnings from practice.