It’s Teams All the Way Down: Use SubDAOs to Progressively Decentralize

A PDF version of this research piece can be found here.


Phil Jackson organized the Jordan-era Chicago Bulls around the triangle offense. Tony Tsieh reorganized Zappos around holacracy, a more flat hierarchy with fewer managers. NASA’s Apollo program was a collaborative, multi-year effort with dozens of teams across military agencies, universities, engineering, manufacturing, and science divisions. If you’ve been a part of teams before, you know that there are size and complexity limits that can accelerate– or hinder a DAO. And no two teams (nor DAOs) are exactly alike.

It’s common to think about DAOs as the next evolution to the hierarchical corporate structures that have dominated industry and Silicon Valley alike. Team structures within a DAO are no different. Corporate hierarchies homogenize as companies grow. But DAOs aren’t relegated to the same fate. Finding the right team structure– and evolving to fit the current needs and ambitions of a DAO are critical to balancing resiliency and innovation. There is no one-model-fits-all.

The most important asset in a DAO is its people. How best to structure the team affects the organization’s agility, resiliency, success, and even survival. We want to walk you through some of the models out there, assess their strengths, and ultimately help you and other high-growth DAOs better understand these choices and experiment with the model that works best for them. We believe that onchain startups must govern with speed if they want to win and that means not thinking about decentralization as a tradeoff to velocity.

Always has been.
Always has been.

Model #1: Parent DAOs & Sub-Teams

We’re going to start with guilds, sub-teams and subDAOs. No matter what you call it, these groups work as an extension of the parent DAO (we call them Sub-Safes or SubDAOs, meaning teams that can independently operate and hold their own treasury, but are accountable via clawbacks or checks and balances to the parent team/DAO).

The parent DAO funds these teams, but each of them manage their own treasury, set their own governance model (which may or may not look like the parent or other subDAOs) and execute transactions or other onchain tasks on behalf of the parent DAO. The decision-making methodology for these subDAOs may match the parent (e.g., using the same erc-20 token as their parent), or they may be completely different (where an EOA or NFT may define the roles, titles, or leaders) within a team.

Regardless of how sub-teams are organized, there’s typically one common divisive aspect– can the parent DAO enjoy veto power? This is good to safeguard against bad acting teams who do not invest their treasury/budget to suit the DAO’s interests.

However, this “check” can also slow agility. So it can be a boon or a bane here. The more oversight and clawback ability the central DAO has over the subDAO, the less autonomy and decentralization the sub-team has. To succeed, DAOs need to bring an experimentation mindset; and make sure the tools they use allow them to tweak and change things when a model does not work for the subDAO (and/or the parent)!

Case Study: MakerDAO’s Endgame

Perhaps no DAO is as famous (or infamous) as MakerDAO. After ratifying the Constitution last year, the team has worked to solidify its bet that the next stage of growth will come from subDAOs. Nicknamed Endgame, the proposal splits the DAO into six subDAOs– each will be self-governing and manage their tokens within the MAkerDAO ecosystem.

An overview of how the Maker subDAOs will organize. 
An overview of how the Maker subDAOs will organize. 

Indeed, the growth core unit (an early sub-team at Maker) led the business development that drove DAI ubiquity via hundreds of DAI markets and dapp integrations. In addition, the DAO will establish 3 teams, including a voting committee, delegate group, and conserver group. Not every major stakeholder agreed.

How the Maker SubDAOs fit into the bigger picture.
How the Maker SubDAOs fit into the bigger picture.

The team has an ambitious roadmap to further solidify DAI’s place as a keystone stablecoin in DeFi infrastructure– the industry will be watching closely how these governance choices influence growth. Learn more about their subDAOs here.

Model #2: The Council

Some DAOs elect a council that manages and votes on behalf of the stakeholders. Sometimes, these councils are elected by token or NFT holders. Other times the council may be established by the initial founders– sometimes in a decentralized way, sometimes not. The council usually holds centralized veto power; a mechanism designed to defend against governance attacks. However, a common thread between most councils is that this power only be used in dire, existential situations where inaction could result in the DAO’s demise.

Case Study: NounsDAO

You know NounsDAO. Loveable noggles. A DAO that auctions off one noun per day. NounsDAO has a six-member board that includes four founders. This board or council enjoys veto power over any proposal it determines “introduces non-trivial legal or existential risks to Nouns DAO or the Nouns Foundation.” These councils serve as a more-centralized backstop to onchain decisions that could threaten the viability of the DAO as a going concern. Last year, the DAO forked. You can learn more here and see the governance history. Today, the treasury stands at 5.6K ETH or $18.4M USD. The DAO continues its mission, with regular votes on projects to support.

One way teams can balance the power of the council is to institute term limits and empower the DAO to vote for new members. This gives a chance for the community to elect council members that are expected to be good stewards of the DAO for the time allotted. Decent DAO and Fractal have benefited from past Nouns research. We’re big fans of Nouns and eagerly await the cool art and cultural projects they fund next.

Model #3: Bicameral / Tricameral Parliment

Checks and balances. Separation of powers. Trust and verify. Trias politica. The idea of state governments having direct say over certain aspects of others is not new. In the search for models beyond 1 token : 1 vote, the systems that are currently implemented in many sovereign nations may serve as the next viable model for sovereign decentralized organizations. DAOs are not corporations, indeed.

Case Study Optimism: Two Houses, Dual Governance

Optimism (OP) is an optimistic rollup L2 scaling solution. Between its partnership with Coinbase and 25% market share, the team has a solid foundation and bright prospects. In its two-house model, Optimism has a Token House that focuses on project incentives, protocol upgrades, and use of treasury funds. Votes are completed via token holders and delegates. This group acts like a typical 1 token: 1 vote DAO. The Citizens’ House, however, is the experimental component of OP’s governance model. This group is focused on retroactive public goods funding. Core contributors get a soulbound token. The protocol’s vision for how Impact=Profit manifests itself in governance.

Optimism has made clear this model will evolve as the DAO experiments with governance, starting with the bicameral model. The team believes this balance of one house focused on a traditional, plutocratic model, coupled with a non-plutocratic model will serve to balance short-term incentives (e.g. price appreciation) with long-term (revenue, strengthening of public goods). Learn more here.

We’re actively researching this methodology to consider making it available in Fractal. Would your team use it? Let us know.

Mix and Match to Find What Works

No single model will suit an entire DAO, and often, sub-teams will have to experiment with different models to learn what works and what doesn’t. One governance model may work for the parent DAO while a completely different model may work for one or more subDAOs. Teams need to consider the goals of the team, how they operate, what influence they have on the broader organization, among many factors. Or, you may find that a combination of approaches, unique to each team/sub-team, is a better fit for your organization. The key is to experiment, test and learn, evolve, and be ready to accept that combinations of models may win out.

Case Study Panther Protocol and the “LaunchDAO:” Vote-Only Tokens

We’ve been researching several ZK projects recently. In most cases, such as Panther Protocol’s governance model, councils are designed to serve as an initial way to bootstrap governance from a centralized initial launch to a more decentralized model. The Panther team took cues from Gro DAO, namely to create an onchain vote before distributing tokens using a voting-only token that becomes redundant when the actual governance/utility token is launched following approval to form/create the DAO and associated tokenomics model for the protocol. To do this, the team utilized SafeSnap, a tool that combines a Gnosis Safe module and Snapshot to execute voting in the early days of a protocol.

How SafeSnap works: take off-chain voting and solidify it onchain. 
How SafeSnap works: take off-chain voting and solidify it onchain. 

Case Study: ETH Lizards, a Combo of Models

We’re big fans of ETH Lizards. Gaming guild, investment DAO, memelords. The team has a council of five NFT holders. This team oversees investment opportunities and manages the treasury/vault. The councilmembers are entitled to revenue as a payment for deals that they source during their tenure. The council is voted in by the broader ETH Lizards NFTholder community.

ETH Lizards’ DAO/SubDAO structure.
ETH Lizards’ DAO/SubDAO structure.

In addition, ETH Lizards utilize a tri-cameral sphere-of-influence model, where subDAOs focus on a particular strategic goal and maintain a council above them. For example, the council above is a part of the Portfolio group for identifying promising, early-stage web3 gaming partner projects to partner with and invest in.  However the team also could deploy Community and Partnership subDAOs which have completely different remits, decision making and treasury management.

Case Study: Hats Protocol, Infinite Diversity, Infinite Combinations

Modularity is important from an interoperability perspective, but also at the protocol layer. We’re bullish on Hats because the protocol empowers DAOs to create their own unique organizational structure with as much customization as needed (or as needs change).

From the team’s docs: Hats are programmable, revocable, and legible roles, which can be collectively controlled by the wearer of the "Top Hat", which could be an individual or a group, such as a DAO. Hat-based roles can be flexibly imbued with responsibilities, authorities, accountabilities, context, and more.

Here’s a look at how Haberdasher Labs and the Hats protoDAO use their own protocol to manage their teams’ access and governance:

In Conclusion: The Key is to Experiment

There is one constant: rarely will a team land on the optimal governance model at the first go. DAOs that want to build resiliency while they progressively decentralize must experiment. And to experiment, it’s critical that you consider how changes to the governance model(s) are committed onchain.

The largest DAOs in the industry may have the time and engineers to customize these changes, but for most, having an app where these changes can be proposed and made without requiring its own dev effort is critical to agility– learn what works (and what doesn’t) fast. Experiment, discuss what works and what doesn’t, and iterate.

*                              Shutter DAO on Fractal following the DAO Blueprint.*

We built Fractal to account for the fact that teams may choose a model and then find they need to take another direction. If and when that time comes, Fractal gives you the ability to propose changes to your own subDAO and execute them– in the app, without writing custom smart contract code, all built on top of the security of your Safe multisig. We believe in one mantra: Govern at startup speed.

For more information on setting up your subDAO, visit and read about how to get started here.

About Decent DAO and Fractal

Fractal is part of Decent DAO, an EVM-compatible platform that enables tokenized smart contract protocols to progressively decentralize governance and operations. Decent helps teams go from a multisig or Snapshot to fully modular, onchain governance by offering composable tools, integrations, and workflows that support decentralization without compromising on execution.

Decent DAO is supported by the Decent Foundation and backed by investors and partners including BlockTower Capital, GSR, Cumberland DRW, and 1kx.

Learn more at, on X @decentdao, and Farcaster /decentdao

Subscribe to Decent DAO
Receive the latest updates directly to your inbox.
Mint this entry as an NFT to add it to your collection.
This entry has been permanently stored onchain and signed by its creator.