One of the most fun and terrifying aspects of DeFi is the constant innovation, there are new and crazy projects forming constantly. Low barriers to entry mean experimentation has never been easier, enabling global permissionless innovation in finance. Over the last few years, I’ve been in the privileged position to speak to almost one hundred teams raising their first rounds of funding. Out of these conversations, I’ve found many founders still struggle to identify the optimal way of raising capital for DeFi and crypto projects. These days I advocate a fairly novel form of capital formation which I will outline in this post. You could call this approach “DAO first”, and it’s been used to generate powerful communities numerous times now - some examples are Barnbridge, dHEDGE, Illuvium, Spool, Thales, and Lyra.

Most of the teams I speak with are raising initial capital and looking for angel investors, but I also talk to teams who’ve been around for a while and have raised a few rounds of capital. These rounds come in many forms, but invariably they create friction for the team looking to move towards a more decentralised form of governance and raise further funding from their communities.

A few months ago, I spoke to a team that had just closed a pre-seed round and was preparing for their seed round. I walked them through my preferred approach, they loved it, but unfortunately, they raised their previous round with a convoluted combination of a SAFT plus some exotic equity warrants 🤯. Yes, they were in SF, obviously.

We immediately started discussing how they could potentially modify the structure to discharge the obligation it had created in a way their existing investors would accept. Luckily I’m an expert in unwinding idiotic decisions! Expertise I have acquired through many years of dealing with myself. Eventually, we came up with a reasonable plan for them to take to their lawyers, but before we ended the call the founder said, “I wish I had come across this model before I had done the pre-seed”. I realised that even though I have articulated it hundreds of times to many people in the space, I had never taken the time to put it down in writing. So I promised I would publish a blog post about it. Now many months later I am getting around to doing so…

If you ask ten different DeFi investors what the best mechanism is for raising capital for a protocol, you will get ten different answers depending on how they came to be investing in DeFi. Most of the DeFi founders and DeFi funds are pretty open to DAOs at this stage, whereas some of the OG crypto funds are less so, but ultimately it comes down to investor protections. SAFE notes, SAFTs, warrants and all the other weird and wonderful mechanisms that are used to raise capital are mainly there to provide investors with a sense of security. My argument is that in DeFi you are getting a false sense of security in exchange for hampering the future growth of the protocol. In the current environment, there is no reason when launching a new protocol to trade off its future autonomy and decentralised governance to provide boomer investors with a security blanket!

The advice I give to many founders is to spin up a Gnosis multisig or an Aragon DAO, send me the address and I will contribute funds. I don’t want a false sense of security, I either believe the project has a chance of success or I don’t. I, therefore, optimise for ensuring the protocol has the best foundation possible, not for holding a worthless PDF with some investor rights on it. My reasoning for this is that by creating a very light DAO framework immediately the foundation is laid for handing governance over to a community much sooner and more seamlessly.

DAOs are becoming much more common and investors are getting comfortable with them, but they are launched in a way that is different from what people expect from an investable company. You have the pre-DAO period before the DAO is instantiated where all you have are a bunch of meatspace agreements, some explicit and some tacit. These agreements dictate how control of the DAO will be distributed based on who has contributed what. The instant the DAO contract is deployed and the governance distribution is crystalised, token holders have control of decisions. This can be strange for someone who is used to investing in an incorporated entity and reading the documents to understand the cap table and who your counterparties are. As a seed investor in a DAO, you are committing to invest in something that doesn’t yet exist, it is not until tokens are distributed that the DAO actually exists.

DAOs also forgo reliance on external courts of law. The DAO controls itself from the inputs of DAO members alone, it is by definition self-contained and need not rely on any external arbiter to determine control. Of course, there will always be the necessity of meta-governance, but the DAO handles this through stakeholder inputs controlled by its own internal rules. The critical thing to realise here is that the state transition from non-binding meatspace agreements to “DAO” must be legitimate. If the DAO is deployed with rules that break the previously agreed approach it will not have legitimacy, and will likely need to be redeployed or could fail completely. There have been examples such as Curve where an “external party” took it upon themselves to deploy a DAO based on what appeared to be the community consensus rules and the community accepted it as legitimate.

The key outcome you want from a DAO formation is that it be sufficiently distributed from the moment it is instantiated, that it is acting as a coordination mechanism rather than a control mechanism. A DAO should transmute community preferences into tangible outcomes. While this ensures the project is protected from external capture, it also creates risk for all parties because people contributing capital have minimal to no recourse outside the DAO itself and people contributing ideation and effort have no direct control over “their” idea. But if the DAO is launched in a legitimate fashion you have a platform where all parties can come together to coordinate the next phase in the project. If the DAO structure and token allocation properly aligns incentives, the launch of the DAO should cement this coordination rather than unravel it. It also forces all stakeholders to cooperate to ensure the best outcome. One of the dangers of launching a DAO is that a new project is unlikely to have a strong community and this could lead to a scenario where very few people are taking the initiative to help progress the project, but this is offset in the nascent stages as you have a small group of people all coordinating to achieve an outcome and communication is fairly easy. As the project scales communication will become harder and this will necessitate slowing down and being more explicit and deliberate about communication. The tradeoff is that the project will have gathered many more resources by this time, in terms of asset value and human engagement from the community.

I wrote recently about another aspect of DAOs that remains challenging, how to have leaders with legitimacy to lead the people delivering the protocol. The key is that humans must deliver the code and other outcomes, but those same people cannot control the protocol. For a protocol to be sufficiently decentralised and resistant to capture the token holders must have control. This can either be directly (ask Maker how that is going) or indirectly through delegated democracy. But there is a very fine line here and the broader DeFi community is still navigating these issues.

So now that we have addressed some of the high-level issues, let’s get a little more explicit about the reasons for going DAO first. Every new project starts with an idea, someone(s) must have that idea. Once you have a concept you can start to discuss it, this is the first glimmer of a new community. If people find the idea compelling they may agree to commit resources to it. As the project gathers momentum it becomes self-reinforcing. It demands more attention, which requires resources. This is the point where most founders seek outside capital and start fucking things up. So let’s go from here.

A new project that wants to form a DAO doesn’t necessarily need external capital at the time of formation, it could start by distributing tokens to the initial contributors and retaining within the DAO a percentage of the tokens for later capital raises. But it can also inject capital at genesis to de-risk execution and generate momentum. Regardless of which path is taken, it is critical that tokens are distributed and control is relinquished. By relinquishing control the initial contributors to the DAO are signalling to their community that they intend for the project to be controlled by the community, not by a few people in privileged positions.

Now there is a practical consideration here, how do you balance the contributions of capital against the contributions of labour and other inputs? The answer is either through open and transparent negotiation or not at all. All contributors must discuss the division of governance power within the DAO and agree in principle to execute a contract that enacts that on-chain. Some of you are probably saying this puts the non-capital contributors at risk. Because they are only contributing labour and not capital, are they not at risk of some kind of regulatory interference? I believe the answer is no in most jurisdictions on the basis that once the DAO is instantiated none of those non-capital contributors control the DAO. This requires careful coordination, but in most cases, a minority of tokens is allocated to labour contributions so if properly managed with governance power vesting there should be no single entity that controls the DAO.

Now we have a contract on Ethereum that is hopefully the legitimate expression of the contributors in terms of distribution of governance rights, we must capitalise the DAO and further distribute the tokens. There are many ways to go about this but one of the most common is to take tokens allocated to capital raising and distribute them via an LBP, the only issue with this is you need an initial amount of capital to put into the pool, many teams will do an OTC round before the LBP to inject a minimum amount of capital into the DAO. There are many other approaches, some teams have used auctions, others have used Gnosis or Uniswap. But regardless of which approach is optimal for your DAO, you will almost certainly need funds to enable the DAO to operate.

One of the big questions ahead of capital coming into the DAO is what the optimal distribution of tokens is. This changes significantly at different times depending on the overall sentiment in the crypto market. But the following ranges and the considerations around each are fairly good starting points:

Treasury: 10-30% Yield Farming and other liquidity incentives: 20-50% Contributors: 10-25% Early Investors: 5-15%

Treasury: One of the challenges with tokens vs equity is that it is very difficult to issue more tokens once you have set the monetary policy. It is not impossible, but ideally, you should aim to retain sufficient tokens in the treasury to cover future capital requirements. A range of 10-30% is typical these days, on the lower end of 10% often teams will carve out a separate pool for seed capital in the DAO which may account for another 5-10%.

Yield Farming: Yield farming when done well creates very powerful network effects for a new project and will help to attract new community members. Retaining a significant portion of the token supply for yield farming and incentives creates maximal flexibility for the project and also ensure that tokens will be more widely distributed. A low percentage for yield farming and community distribution should ideally be offset with a large portion for the treasury and early liquidity bootstrapping.

Contributors: This portion can include only non-financial contributions or can include pre-seed and seed capital. This should be very carefully managed to ensure tokens are not overly concentrated and risk triggering arguments of control by a select few individuals.