Understanding Decentralization & Prioritizing an Operations Team

Decentralization is a more nuanced concept than the popular meme would have us believe. It is not as simple as decentralization = good.

There has been a lot of debate about decentralization on the Temporary Transition to Foundation Model thread. Mostly the arguments against are a version of “decentralized is better, giving power to the council is centralized and bad.” I totally sympathize with the concern, it’s a good one, but also believe it to be based on misunderstanding of how decentralized systems actually work, and that this could slow down our critical priority of establishing an operations team.

To start, I want to point out that yearn is more centralized than it appears.

“Heretic!! Isn’t YFI the most decentralized DeFi project ever? Doesn’t that mean we’re decentralized?”

Yes and no. Yes, it was an amazing, historic distribution and the ownership of YFI is more decentralized than other governance tokens – but that is just one piece of a much more complex entity. YFI ownership is well decentralized, but the yearn organization is not — and that’s ok. It’s a process. And nothing is completely decentralized or centralized, it’s a gradient.

Voting power is less decentralized than YFI ownership. For example, a group of the largest holders vote together as @yfi_whale (thread)

There is nothing wrong with doing this. They are acting in their self-interest with the sovereignty available in this medium. Their group is a type of informal and behind-the-scenes centralization. This kind of informal centralization is happening all over the yearn ecosystem.

Decision Making
The actual machinery of our collective intelligence decision making functionality is mostly obscured and generally centralized.

yfi_whale is making decisions on how to vote in private messaging groups. Andre and a small group of core contributors are making decisions on how to spend their efforts in private messaging groups. Special interest groups formed by whoever is motivated to work on specific areas are working in public on discord, making decisions, and plain taking initiative.

The bigger decisions (like to approve a new website design, or vote weighting) will happen on-chain but far more decisions are happening all the time in various small groups that have massive impact on yearn. There wasn’t a vote about making yinsure.finance, I assume Andre decided to do it on his own (as he should). There wasn’t a vote about which frontend design concepts to move forward on and who’s in charge, there was just discussion in discord and lots of hard work and self-managed initiative.

This kind of decision making is glorious and essential. If we tried to make every decision in the most decentralized way we could, via on-chain YIPs, it would be impossible, and it actually wouldn’t be as decentralized as you’d hope.

On-chain voting is our only formalized and transparent decision making system — and that’s actually a centralization of our decision-making ability. It’s a major weak point.

Functional Decentralization
The informal and hidden centralization happening now is part of a natural process. It is up to this forum as the only formal decision-making body within yearn to help empower these small groups to make decisions for yearn in a transparent way so that they can be effective, accountable, and accessible for community members to contribute to. This is not ceding power to leadership groups, I do not advocate for that. I advocate for empowering special interest groups to take action according to their best judgement on our behalf and with our support and oversight.

We need a diversity of decision-making capabilities and groups to thrive. Some may take the form of DAOs, others will be more informal, and others springing ad hoc from the natural hierarchies and teams that emerge in the communications channels.

We will not survive as a “fully decentralized” organism with just on-chain voting. That isn’t even possible. If we don’t trust our community and grant them formal decision making powers (like the multisig council) more and more small centralized groups will form in the shadow to pre-process our decisions for us and we’ll have decentralization theater, not decentralization.

Prioritizing our Next Steps
I believe our critical priority right now is to approve an operations team to make budgets and spend money.

While we could present budgets here in the forum and debate them and vote on-chain that’s slow and timid. Do we have such little faith in the quality of our people that we can’t assign any substantial responsibility outside this DAO?

The risks of not being able to spend money and move forward efficiently is an order of magnitude bigger than the risk of mismanaging our funds. We have to evolve, grow, learn, and try things.

So here’s what we have to do:

  1. Update the membership of the multisig (thread)
  2. Give the multisig limited powers to make and spend budget as an operations special interest group not as a leadership team or foundation (thread)

Everything else is secondary to this. Without functional ops we can’t hire devs or pay people for the work they are already doing, and we can’t move at the speed this industry demands.

I don’t want to open a can of worms, but I do want to point out that there are other options than having the multisig function as the ops team. They already have the critical job of managing the YFI governance contract.

  • We could give that power to the yDAO (bias: I am a member) which is already established to hold and spend money.
  • Or we could create a new ops team and give them the power.

All that being said, I think that if we can update the membership of the multisig in a reasonable timeframe so it’s composed of committed and active members (it’s a full-time job) that’s the easiest solution.


Bravo ! I really appreciate your clear and well thought contributions ! Same to @Substreight !


Well put, I agree with everything in this post.

1 Like

Great post @tracheopteryx

If we really wanted to be an extremely effective decentralized organization, we would model the principles of the Pirate Party:

3 components

  • Leadership
  • Framework
  • Swarm

Key ideas which are not intuitive

  • Nobody tells anybody what to do, you lead by inspiring others by example, not by instruction
  • Leadership sets vision, not message
  • Leadership sets exciting, achievable goals which the Swarm can achieve independently of the Leadership note: maybe run with simplified OKRs
  • Swarm is open, transparent, inclusive, where everybody can see almost everything all the time

It’s all about trust, transparency, speed and scalability.

  • No regrets, no blame games, no backseat drivers to those who are in the arena taking action, no hindsight. Mistakes are going to happen, we learn and we move on.

  • Goals have to be specific, achievable, inclusive for YFI governance and epic.

  • If any 3 YFI holders support an idea, it can be done. They take accountability and authority to accomplish the idea. Asking permission only gives someone else accountability for your actions.

  • 3 basic group sizes: 7, 30, 150. Forums used to communicate at all levels.

  • Meetings minimized to sync and health check only, maximum 1 hr.

  • Most planning and work to be done in the swarm, in forums, discord or in person.

  • Voting creates opposition and should only be done in rare / critical circumstances e.g. resource allocation or irreversible decisions. For a great example read https://fs.blog/2018/04/reversible-irreversible-decisions/ regarding reversible and irreversible decisions:

If a decision is reversible, we can make it fast and without perfect information. If a decision is irreversible, we had better slow down the decision-making process and ensure that we consider ample information and understand the problem as thoroughly as we can.

For example, my new website design which can create immediate improvement in experience for users as soon as PR is conducted, is a reversible decision being treated as an irreversible one.

  • If something is disliked, do something better. Don’t instruct anyone on what they MUST or SHOULD NOT do.
  • Be a do-ocracy. Allow many different things to be tried.
  • If voting cannot be avoided e.g. resource allocation, we try to do a consensus circle (in a swarm) or entrust this decision to consensus among the ops team.
  • This must be FUN. People do stuff because they want to do it. This is why, while producing excellent thoughtful work that helps achieve Yearn goals, I choose to insert memes and jokes.
  • We are here to create an inspiring example for what is possible for the world, right?

As Vitalik said https://twitter.com/VitalikButerin/status/1294186507465420802
“I think the real chad is the friends you make along the way.”

Feedback, thoughts always welcome and appreciated.


Beautiful. I know Andre created YFI to offload some of the decision-making from him to the community– but I think for this to keep moving forward, we need to place much of the everyday decision-making back in the hands of Andre and those he trusts– and realistically, most of yearn’s decisions have been and continue to be made in this small circle anyway (with lots of feedback from the community).

We are already nowhere close to the system Compound has setup, where they vote on literal code changes to the protocol– we vote on “ideas” and then hope that Andre et al. will implement them properly. We’ve continued to trust them this whole time– so now let’s back up that trust with the compensation they deserve to keep this moving.


It’s super easy to advocate for full decentralisation. It’s popular and sounds nice. It’s just not dynamic enough at the moment.

I agree 100% with the priorities you’ve stated. Changing multisig members is crucial, the process is really slow today. As well as giving the new members more power, so they are able to allocate funds. There’s people working on keeping all the ecosystem running, not only Andre, and they all deserve recognition.

We have to make sure Yearn stays lean and dynamic. It’s the only way forward.

100% behind you here. I would be for giving the ydao more power to help out, but I think votes there take 3 days so that would be just as long to approve money to send out. But I’m sure the ydao can change stuff like that. That being said if the multisig is up for all the tasks that’s fine too. I do expect them to sign trx in a timely manner though, that goes for whoever votes for the funds.

1 Like

As I said before, I really appreciate both @substreight’s and your @tracheopteryx ideas.

I think we can all agree on following statements :

  • Yearn is a great
  • We love and use Yearn products and are committed with Yearn’s survival and development
  • We pursue Yearn’s long term growth and success
  • We embrace openness and transparency
  • We want to support and nurture Andre’s brilliance, developments and ideas

If we agree on those then we should delineate this agreement in some form of foundational letter.

First step should inmediately accept, vote and implement @Substreight’s Operational Team.

We should also embrace @uhmpeps key ideas, as guideline to act as a decentralized group.
But I want to take it a step further.

Let me outline an organizational structure idea, where there exist initially 5 bodies :

1. YFI HOLDERS, which are Yearn owners. It is a large, decentralized and heterogenous group.

This larger community votes via YIP on following core issues:

  • Operations Team (MultiSig) integration and membership
  • extraordinary (250k plus) expenses
  • Protocol changes
  • Governance and Economics proposals from yDAO

2. yDao

IMO should work on a strategic level !

Work out a YFI Constitution or Foundational Letter with :

  • Vision: Long term growth and success
  • Objectives : Strong, Secure Protocols & Strong Brand
  • YFI Governance and YFI Economics rules
  • Voting and Decision making processes
  • Conflicts and dispute resolution
  • Ownership and management of assets (incl. intellectual property)

Some very smart and committed people already integrate current yDao and does amazing jobs (FAQ, Communication, web development, documentation, etc.) (@saltyfacu , @dark, @substreight, @devops199fan , @milkyklim and many others). I think that’s the place were this should thrive !

3. Operations Team

6/9 MulitSig - with deeper involvement and constant participation :

  • Treasury Operation
  • 250k rolling Budget Allocation
  • Assess deliverables

Some sort of rotation rule should be mandatory: Say 2 Participants rotate every 6 months (one who is less active, one who is oldest in the position).

4. Special Interest Groups

They should be community driven roles with :

  • self-established objectives known for all the ecosystem with absolute accordance to long term vision and YFI constitution.
  • Key results driven (@uhmpeps’s OKR as stated above)
  • Periodic Accountability
  • Mainly grant based work

a. Vault Strategies

  • rebalancing analysis & execution
  • an independent multisig ?

b. Risk Mitigation

  • Assess, evaluate and mitigate current and future risks.
  • Defend our system against whale takeover/dominance (as said by @zeframlou)
  • Security Retainers, Audits and Bug Bounty programs
  • Protocol Insurance
  • Embrace Andres work, ideas and involvement, but lay grounds for an truly autonomous long term existence.
  • Legal & IP
  • surely tech work will be on dev’s side, but I can see a deep community concern and involvement that will advance these issues.

c. Marketing
Certainly will be more than one SI Group imo
@uhmpeps and many others doing great stuff here !

  • I. User aquisition & AUM growth
  • II. FrontEnd & UI
  • III. Brand + Community (Twitter, Youtube, AMA, TG, Discord, Medium, etc.)
  • IV. Non Tech Support

5. Development Team

Last but not least: “@andre.cronje’s department”

There undoubtedly should be some permanent roles and hirings, but we also should embrace community grant based involvement.

We should definitely and clearly coordinate this with Andre. What does he want to do ? Where is he heading ? Does even Andre want to lead and coordinate dev work ? Is he willing to delegate dev work ?

Development Functions :

  • Core / Protocol
  • New Strategies
  • New Vaults (snx, eth, sbtc, lend, etc etc)
  • New Products (like yinsurance)
  • Tech support
  • Opex - Infrastructure, Gas, tools, etc

I said initially 5 bodies, because sometime in the near future an yDAO and Operations Team merging seems logical next step.

Do you think this makes sense ?
Anyway, I’m happy to add my thoughts and pls do give feedback on these !


Absolutely love this, feels like my brain got an upgrade reading your post.
Would like to hear thoughts from the folks you mentioned.