Mandate: Interim Governance Facilitators

The Foundation is viewing this as a formalization of the work I’ve been doing for the last year or so in the calls and the governance ecosystem. The publishing of the mandate will hopefully bootstrap up the process of decentralizing the role among community members in the post-MCD world.



Exciting stuff indeed. You have any idea on how long the foundation would be willing to compensate you (or future GF) instead of letting MakerDAO figure it out?

To be clear, I’m not getting any additional compensation for this other than the satisfaction of furthering the glorious principles of the foundation and open governance in general. Hopefully that will offset the additional targets I’m placing on my back… :roll_eyes:

As far as dates go, I don’t think anyone has ball-parked any numbers. I think the community and the Foundation will have to begin to inventory what actors the system needs, how much compensation is fair, how it will be collected, disbursed, tracked, etc.

It’s going to be a complicated process.


I’m pro you Rich, your the visible heart of this thing. Keep it up!


Regarding point 1, and [quote=“LongForWisdom, post:8, topic:264”]
It would not be overly costly to setup separate forum+onchain polls for the following:

I totally agree. Why agglomerate questions when it’s not that costly to separate them?

Maybe people are happy to activate this “facilitator” role, but are not happy to have Rich Brown as interim for it. (Just saying… I personally think this position is for the moment just a ‘title’ for what he’s already doing, so wouldn’t even spend much time on this).

Disclaimer: Rich is doing a good job. I am happy, as everybody else, to see it formalised in some way in the DAO. This is post about the methodology, not about specific people.


inspired by the excellent post of LongForWisdom I also decided to express my concerns about the recent governance topic: about the Facilitator role and the first interim facilitator, Rich Brown.

First of all, as already pointed out by LongForWisdom, there are multiple questions in this governance topic. I guess more or less everybody would agree that a facilitator role is necessary, or at least useful, or at least not negative. So (I expect) most people would not oppose the activation of such role. So one of the questions is probably universally accepted, but the others need more discussion.

This said:

1) I find it weird that the first role that is voted on in the MakerDAO is that of the facilitator. While important this does not seem the MOST important and the first that should be voted on. I mean, the MakerDAO will require a lot of roles (aka “jobs”) when it’s up and running and independent of the Foundation. Examples:

  • Head of Oracles?
  • Head of Communications?
  • Head of Development?
  • Head of Risk Analysis?
  • Head of Web Services
  • Head of Client Support (?)

Not to mention that, when the DAO becomes very big, we will likely have:

  • Head of Communications > International
  • Head of Communications > USA
  • Head of Communications > Europe
  • and so on.

and then

  • Head of Communications > USA > California
  • Head of Communications > Europe > Germany

and so on…

2) So why start with a job such as the “community facilitator”? Why not try to organise the big picture first, and agree on its structure? Natural questions to be discussed:

  • How many jobs do we want to vote on at the level of the governance (I.e., on-chain)?

I mean, probably we don’t want to vote for each sigle little job (e.g., translator of a couple of documents). We just want to vote on big roles (e.g., Head of translations) which will then deal with a lot of other people “under them” off-chain.

  • How do we plant to recruit people? This time Rich Brown was the only proposed name. I think this is a very bad practice, even if we are still at the very beginning, We should definitely try to have a “transparent application process” with people sending their CV, and other data, e.g., specifying a minimal salary they want.

What if in the MakerDAO chat we currently have, using an anonymous nickname, the head of communication of Harvard University? And what if (S)He was willing to do the “Facilitator Job” for 24 months for free? (very hypothetical, but still…) Would we still vote for Rich Browns? Maybe yes, maybe not.

  • In this light, who is going inside the MakerDAO to evaluate the CVs? Who is going to shortlist the applicants?

  • similarly, what is going to be the procedure to determine the salaries? Are we expected to vote for each salary, of each job? This obviously does not scale. So we need to discuss a hierarchical structure and delegate decisions. This will work only high levels of transparency which, fortunately, should be possible with blockchain.

It seems to me that this structure is what should be designed/discussed/voted-on at the beginning. Not the “facilitator” job + Rich Brown as interim.

3) One could say “OK, but we need to start with somewhere with something…”.

I totally agree. But if we have to do this, let’s do this well, with some conceptual organisation. Let’s do this in an organised manner! Not just by voting on a random role, that is anyway already being well done (and paid for by the Foundation) by Rich Brown.

The facilitator Job, as described by Rich himself during the last call, is supposed to create a discussion link between the community (Reddit, Chat, Forum, other) and the Foundation. Good. This is useful (especially since Rune, Steve etc have not participated to the calls for a long while). But this link can potentially be informal.

Before formalising this job, I’d very much rather discuss and vote on the general “job hierarchy and structure” of the MakerDAO (independent from the Foundation!).

TLDR: in line with my previous post () I feel this vote is again one instance of “let’s play the DAO for a while with some pseudo-random stuff, and buy time”. OK, I want to play. But I’s like to play it properly!


On splitting the proposal

Yep, that was hyperbolic. I get excited sometimes. Apologies.

Feels like common sense is the answer here. It doesn’t need to be subdivided into oblivion, but the four divisions I suggested are logically distinct.

I am 100% against adding complexity without need. I do think there is need to split proposals into logically distinct units, and I’m not even convinced this adds complexity. The Foundation feels similarly, or we would be ratifying all the proposals (Risk, Oracle, Facilitator) in a single giant info-dump.

The disagreement is about where the logical split is and I’m arguing that the logical split is incorrect here. A proposal should include everything required for that logical unit and no more, the role, the person filling the role, the governance cycles and the timeline are logically distinct: they could exist in isolation and still make sense. A more granular split at the level of (as an example) the responsibilities of the facilitator makes less sense: they would lack context in isolation.

I’m not convinced a lot of people feel strongly about this though. So yeah, feel free to ignore me.

On Polling

It’s only unavoidable if you want to run under a consensus decision making system, currently we aren’t. (But I think we should)

I agree, that’s more intuitive, I specifically admitted my alternative was less intuitive (but that I felt it was more efficient)

Polls make discussion and debate more efficient because they allow members of the community to easily see and understand roughly how much support or contention there is around a specific idea.

They are not perfect, but they cost nothing and they allow everyone to better understand the state of consensus.

A couple of examples on how polls could have improved efficiency

  1. You asked many questions around updating signals, which boil down to whether it is realistic to expect people to keep track and update signals in response to changing proposals. If we’d included a poll here we would have some vague idea of the answer. Instead we have an open question and the hope that someone will comment on it (even though comments are more costly and time-consuming than signalling in a poll.) We don’t know if it is contentious, we don’t know if it is clearly one way or another. We don’t know nothing.
  2. It seems fairly obvious to me that this proposal is going to be approved with close to unanimity. This is fantastic, but we could have had ‘evidence’ (as rough as it is) of this information a week ago. If we have the information that everyone is on board with a proposal, then we can skip or cut short the debate and go to an on-chain poll. If this poll had shown that the proposal was contentious, or widely disliked, you could have immediately reacted to this information. Without this, we are left guessing at how much consensus is implied by a certain number or intensity of comments.

On Calendar

I think everyone understands why the timeline is compact. But the goal of launching MCD on time doesn’t feel compatible with giving the community the time to contribute meaningfully to these proposals. There is nothing wrong with this (it’s a trade-off I don’t ultimately disagree with), but I feel like it should be recognised.

My feelings on this (and the reason I brought it up) is that to me, it appears as if this is what is happening, but because I trust that the Foundation is doing what they think is best for the development of MakerDAO, I accept that it is the ‘price of progress’ even if I disagree with it.

The problem is that trust in the Foundation in general was shaken after the purple pill thing, and that some no longer trust the Foundation after that episode. Maintaining trust is always easier than winning it back. If the Foundation wants to win it back (this may not be a goal, in which case fair enough), they need to be above reproach.

This was kind of a metaphysical question, and, on reflection, not a particularly fair one to ask. So apologies for that too.

How can a DAO (decentralized autonomous organization) have roles such as head of this or that? This isn’t a corporation, decentralization and autonomy are conceptually antithetical to traditional corporate structures.

Facilitation of governance is probably the biggest stumbling block for the community right now. We have alot of complexity to contemplate, debate and design. Organizing dynamic collections of ideas and information is a huge deal. Also the risk team mandate was posted only a day later. Facilitation represents the life blood that connects the producing organs of the dao (risk, oracles, general debate, business dev, ect ). Personally I see it as one of the most important aspects for social cohesion.


Depends on what you mean by costly. The one thing we don’t have is time. I probably failed to express that in the initial mandate / calendar. We’ve mapped things out with as much detail as we possibly could to try to align the work governance needs to do to keep pace with the rest of the org, our partners, risk teams, etc. for MCD launch. We don’t want governance to be the blocker if at all possible.

I believe I go into it above, but in the interest of expediting the process as much as possible, everything we need to start bootstrapping this wave of governance is wrapped up in the mandate. The thinking on our end is that we would block out two weeks of debate and in the absence of clear dissent around any of the issues in the doc we will put it to a vote. Why two weeks? Because that is the most time we had in the calendar that would allow the community to ratify the role so we can kick off everything else in the calendar.

I agree that having so much in one proposal is sub-optimal but, in the absence of clear contention about the contents of the mandate, rather than its form, it’s the best way for us to keep velocity up for launch.

In the future we can break things out into as many pieces as the community wants, as we presumably won’t have something like an MCD launch looming in the distance.

1 Like

This first role is the one that paves the way for governance to have the policies and procedures in place to create those other roles.

Very few of those roles have any clear definition, nor does the community have a pressing need for them to get MCD out the door, which is the single goal of the org right now. Not to mention that there is no way for the community to pay for any of those roles yet.

Decentralization is a long path, not a switch we turn on, it is going to take months, if not years, to fully flesh out all the roles required, their management, their mandates, arbitration, compensation, etc, etc. It would take a fully centralized, hierarchical org, 6 - 12 months of significant effort to hire and train that many people with a dedicated team of recruiters. If you add on the layers of complexity of a DAO, then you can start adding multipliers. The community hasn’t even started the debates about what our DAO actually is yet.

The Governance Facilitator is tasked with helping the community to begin to answer those questions. I think you might be putting the cart before the horse here.

This is a model that we will be repeating for roles the Foundation is paying for during the bootstrapping phase. Risk, Governance, Oracle, etc teams and leads will be dedicated resources from the org until such time as the community has organized itself to a place where it can onboard and fund the roles itself.

If someone more qualified than me arrives in the next week or two, has a deeper insight into maker governance, and is willing to work two or three days a week for the next two years for free, you have my solemn promise I will vote for them. :pray:


I will never ignore you! :hugs:

I mention above in another response: I agree, the bundling of everything is non-optimal. It could be interpreted as some sort of bamboozle maybe where we are trying to sneak something in a big wall of text. I don’t think we did though, and it’s all in the original post for people to look for if they don’t believe me.

The reason it’s all in there is because we have a fantastically aggressive timeline to get MCD launched this year. To stay on track we have to ratify a role and a mandate and a proposed calendar of events.

If there is contention about any part of the mandate we can re-evaluate and if there is significant contention, the community can choose to vote no to it and we can go back to the drawing board. The biggest issue I’ve seen yet though is the format of the mandate as opposed to its content.

The next time around (post mcd) we will definitely split things up.

Maybe we are talking about two different things? My belief is that having a poll before people have a chance to talk about the issues at hand potentially invalidates early votes. If people vote “No” because of a typo or a miscommunication, I come back and fix it, how many people will go back and update their votes? That’s why I wanted to have a forum poll 7 days after the initial posting. Which I am now 4 days behind on…

1 Like

The role of Governance Facilitator is highly needed and I disagree with @iammeeoh that other roles are more urgent. Someone needs to be the central point of communication with the community and I think it makes perfect sense to do it this way.

So I completely support this role, the way it is laid out here and @rich.brown for the function but we have to keep in mind it won’t be Rich forever. I have a lot of trust in @rich.brown but I consider this role to be an attack vector because of the (insider) information it will has to deal with and the weight of the persons word in the entire system. Therefor I also think GF’s will need to disclose certain information:

  • Any ties to other projects/investors/…
  • Disclosure of token holdings/control, not only MKR and not an actual number but an indication which can still keep it anonymous … < 100 MKR, < 1000 MKR, < 10000 MKR, …
  • …?

Ratification of this package

I entirely agree with @LongForWisdom that voting for entire packages is far from ideal. I actually don’t have a problem with it in this particular case but I don’t want this to be the default way of going about things. I understand each individual vote, off and on-chain, adds additional work but I don’t feel like governance can be rushed because of it. It’s been mentioned a few times in the calls that a lot of parties need to be aligned to get MCD out the door and governance is a big part of that but the problem is that we, as governance, have absolutely no idea who those parties are and why they are dependent/influential on the timeline.

In regards to voting on packages, this might sound like an exaggerated comparison but I feel like this could quickly end up the way US politics handles bills. A bill gets proposed to handle a certain issue and by the time it gets voted on, 99 other issues, totally irrelevant tot the original issue got tacked on and ratified or denied along with the original issue in a single vote. I really want to avoid such a system.

In short, I will vote for this “package” but I prefer to vote on individual issues whenever possible in the future. For example, I would hate to vote on a single package to add multiple collateral assets. Each asset needs its own vote in my opinion. Not just for clarity about what the vote is about but how else are we supposed to debug a problem with the stability of the system if you change a huge part of the entire system in one single vote? This discussion is for another thread but it’s the best example I have on why I am against package voting.


This is one of the main topics we need to clarify as soon as possible because this single issue will determine if the DAO can sustain itself or not. I assume the Foundation must have some idea of where the money should come from and how much it could be when it started out this project?

1 Like

100% Agree, that system is a mess.

Mitote created a thread with his Thoughts on Compensation. If anyone has ideas about it, I’d recommend they get them down there.

1 Like

While we (as a community) figure out what consensus looks like, how much debate is appropriate, and how and when calendaring happens, there are going to be some bumps in the road.

This first ‘bump’ is entirely on me. I’d originally scheduled the forum poll for this mandate to occur at the beginning of last week. The intention was to wait until we all had a chance to gauge the level of interest and/or contention around the subject, then, once the debate died down, kick off the forum poll to gather signal for the portal poll. I believe it could have been started at the end of last week, but as time is tight, I’d like to begin the forum vote today and have it run until Monday.


In the course of this thread we can see that the amount of likes are very high (you can’t ‘dislike’ but we can make some assumptions on sentiment based on total forum users), the volume of discussion picked up over the tail end of the first week and began seriously in the second, and only began to tail off in the last 5 days or so.

It’s my impression that the community has expressed its concerns and/or support of the broad strokes, with the primary reservation being about the ‘packaging’ of large mandates with schedules, processes, etc. In the future the Foundation will ensure that they are broken up.

In light of this I’d like to kick off the forum poll today and have it run until Monday September 2nd at 3 PM UTC. At which time, if the poll signals assent for the role, a corresponding poll will be entered into the governance portal which will begin at 4PM UTC on the same day and run for 7 days.


Should the Interim Governance Facilitators Mandate be added to the governance portal for formal ratification by MKR holders beginning Monday Sept 2nd 2019 at 4 PM UTC, ending on Monday Sept 9th 2019 at 4 PM UTC?

  • Yes
  • No
  • Abstain

0 voters


I would have expected to agree with this. However some things do go well together. Indeed, somethings need to go together, as in several tactics make a strategy.

For sure Ironrand, it can be necessary to group things, but in this instance I don’t feel that these things are logically grouped in a way that wouldn’t make sense separately.

From Rich’s responses I think it’s fairly clear that they were grouped so that we could save time and hopefully get to MCD faster. This is a valid reason to group them, but it has costs which I dislike (hence my complaints.)

Welcome to the forum!

The formal governance poll is now live:

Agreed. I’m very hesitant to vote in favor of this role without knowing what it will cost, whether there will be additional cost for facilitators who speak different languages, and where all of the money is coming from.

I don’t think the cost is an issue yet since it’s currently covered by the foundation. The eventual cost should be covered by stability fees, like everything else really.


This thread here have some good informations and should be included or summarized with the Gouvernance facilitator thread as a sub-part if not a main subject.

Many questions will remain without answer, but it’s positive to make some changes and give some fresh air.

Having as much feedback as possible with multi-pooling is just right and will help the Governance Facilitator to do his work.