This is a transcription of the talk that Annika Lewis gave at ETH Denver on March 3rd, 2023: Making Decisions in a DAO: Gitcoin Lessons Learned

I have been with the project Gitcoin for about a year and a half now, previously serving as a workstream lead and presently as a steward with the project. My perspective comes from being a full-time contributor and an external steward at Gitcoin. The other day, I told my mom about this talk, and she asked me to describe it. So I walked her through what I was going to discuss, and at the end, she played it back to me, saying, “Okay, so it’s kind of like making decisions when nobody’s in charge.” And I was like, “Yeah, thanks, Mom.” So, that’s how we settled on the alternate title, “Making Decisions When Nobody’s in Charge.”

Everyone familiar with DAOs or crypto Twitter knows that DAO governance and decision-making has become an absolute meme. Countless memes about DAO governance highlight how messy and difficult it is to make decisions.

It’s a challenging topic, and today we will discuss DAO governance, specifically in a Gitcoin context. There are three parts to this talk. First, I’ll tell you about Gitcoin for anyone unfamiliar with it. Then, I’ll discuss how Gitcoin is governed - what the governance system is like and how they make decisions in theory. Finally, I’ll discuss lessons learned using practical examples from the Gitcoin DAO in the past year.

Gitcoin, if you’re unfamiliar, is the preeminent funder of public goods within the Ethereum ecosystem. Gitcoin and all its programs, including the grants program, have distributed over $50 million to early-stage projects in the Ethereum ecosystem and beyond. In addition to the Gitcoin grants program, Gitcoin is the creator of Allo Protocol and the newly launched Grants Stack, which coincidentally launched yesterday at ETH Denver. So, if you’re a builder creating a community or a grants program and want software to help you, check out Gitcoin’s Grants Stack and the Allo Protocol.

If you’re trying to fund early-stage projects, Gitcoin is the place to be.

When I talk to people about Gitcoin and people in the ecosystem generally, they often say, “Wow, Gitcoin is at the forefront in terms of DAOs, and there’s so much scale in their decision-making process.” Gitcoin, ENS, Optimism, and many other friends are far along regarding DAO governance and decision-making. However, as I see it, we are just in the earliest early days. So even though Gitcoin has several processes, there’s still so much for Gitcoin to learn, along with all of the other DAOs that are leading at the forefront. We are genuinely in day one of this space, and these are just the very early days of DAO decision-making.

So, how does Gitcoin’s governance system work, who are the necessary parties, and how are decisions made at the DAO?

In a nutshell, Gitcoin has two pillars to its governance: the external side of things and the internal side of things. On the external side, Gitcoin has a governance token called GTC, which operates on a one-token-one-vote governance decision-making system. All GTC holders have a say in decisions made at the DAO and can vote. Anyone who has GTC in their wallet or has GTC delegated to them can participate in decision-making regarding how the DAO is run. A subset of GTC holders can volunteer to be signed on to be stewards for the DAO.

The stewards of the DAO commit more time and involvement to governance and decision-making beyond just being token holders. Stewards participate in monthly calls and provide counsel and oversight. Furthermore, a smaller group called the Steward Council, consisting of 10-15 people elected by the DAO every six months, offers more dedicated hands-on strategic oversight, similar to a board of directors.

Moving on to the internal side, Gitcoin currently operates using a workstream structure with five workstreams (such as public goods, funding, marketing, fraud, etc.). Each workstream functions as a standalone unit within the DAO, requesting its quarterly budget and operating independently. One or two workstream leads represent each workstream to the DAO and the broader community.

The workstream leads form a group called CSDO (Cross Stream DAO Operations), which can be loosely considered the DAO’s executive team. CSDO handles more minor decisions that don’t necessarily involve the external stewards and ensures work is carried out collectively across workstreams.

What I learned during my time at Gitcoin:

The first and foremost lesson I have learned is to be very explicit about the governance surface area. By that, I mean being extremely clear about what is governed by a specific process versus what decisions are made through other mechanisms. I have a memorable example from the past year that illustrates this point.

A “sort by” feature on many websites allows users to sort items by various criteria, such as recommendations or price. At Gitcoin, we had a similar feature within our grants program, showing grants in rank order. One day, we made a small product change that slightly tweaked how the grants were displayed on the UI. One of our engaged community stewards noticed this and posted it on Twitter. The product team confirmed that the change had been made, and the steward asked why it hadn’t gone through governance and whether the community should have been involved in the decision-making process. This sparked an interesting debate about the governance surface area and which types of decisions should go through governance and which shouldn’t.

For me, this was a valuable lesson in the importance of being transparent and upfront about the exact governance surface area. Otherwise, there can be tension and a need for more clarity in decision-making processes. It’s impractical for every decision to go through an extensive governance process. Still, this example highlighted an opportunity to clarify the governance surface area and change its implementation.

The next lesson that I learned is don’t ever let governance bite you in the ass.

I’ll use the previous example of that small product change to illustrate how over-governance might be challenging in a DAO. In the Gitcoin governance process with the GTC token voting, proposals need to be on the Governance Forum for at least five days and then on Snapshot for a vote for five days. So, if you have your five days on the Governance Forum and your five days on Snapshot, you’re looking at a 10-day decision-making cycle, in addition to all the time engineers have spent working on that product feature. This process is excellent for significant decisions requiring much debate, dialogue, and discussion. Still, it may not make sense for more minor decisions where it could be a form of over-governance.

Chase Chapman, an absolute expert in DAOs, had an insightful tweet about this, discussing how we often talk about centralization risk but not decentralization risk. This occurs when we’re making small decisions involving too many people, leading to inflexibility in responding to a changing environment. Especially when building software, this can unnecessarily stifle innovation. All of this ties back to the point about governance surface area, clarifying it, and being very clear about when to delegate decision-making to expert teams versus having a full-on community process.

The last lesson I’ll share, which is more of a feel-good one, is to iterate. We have learned much from a governance standpoint by testing, changing, and repeating the process. I believe that’s what we all need to do in Web3 as we explore these early days of governance.

I’ll share a few successes I’ve seen at Gitcoin regarding how it has iterated on governance in the past couple of years. The first is the evolution of steward governance. As mentioned earlier, community stewards are similar to delegates in other DAOs and play a crucial role. Gitcoin has done an excellent job of changing that process over the last couple of years since its implementation. Scott, one of the founders of Gitcoin, launched the steward governance process almost two years ago. The steward council was implemented about a year ago, and now they’re on version 2 of the steward council. There’s a constant desire to iterate and work with the stewards to decide how governance should change. I think that’s admirable and leads to excellent outcomes in innovating, experimenting, and making necessary changes.