From Our Discord: "Optimal Lifespan of DAOs"

This is a cleaned-up version of a Discord thread I posted and several other community members contributed to between 2021/11/16 and 2021/12/17. The intent with posting this here is to make this material easier to find, reference, and share as appropriate.

Appt Pupil 11/16/2021

Change my mind: DAOs should have short lifespans by design.

Since I haven’t gotten a lot of pushback on my other threads, I take it as a sign I need to speculate more! Here’s my rationale for my opening statement here…

I mentioned in the superpower thread that organizations as we think about them today are optimized around the constraint that they are expensive to start. Thus, every org has a secondary mission of sustainment that causes a lot of misery. Again, expense reporting is exhibit A…

With DAOs breaking this constraint, we need to examine the notion that organizational longevity is a sign of health. News flash: it really isn’t.

Yes, there are companies like Zildjan that have operated for hundreds of years. It seems obvious, though, that most organizations get stiffer and more brittle the older they get. Just like people, although again there are some exceptions (like me, I hope)

Speaking of people, it’s rude to point this out, but so far every critter who was born more than 200 years ago has died. But one of the wonderful things about nature is that one critter’s death is another critter’s sustenance.

The bulk of all living matter consists of just a few elements, commonly referred to as CHNOPS - carbon, hydrogen, nitrogen, oxygen, phosphorus, and sulfur. It’s one of the key enablers of evolution, because unlike most artificial products, living matter can be easily used by other living things to create new living things.

Nature is a good model for how we should think about the DAO ecosystem. When we say wagmi, the “we” is not DAOs - it’s people. DAOs are just the vehicle we use to accomplish our goals, generate new capabilities, hone our skills, and and connect with others.

It’s a tragedy today when a company goes out of business, because it’s hard to get the pieces back together to create value again. But with a DAO, that logic no longer holds. If 4 people vibe really well in a DAO, and the DAO dies, it’s not that hard for those 4 people to form a new DAO. In fact, we’re probably better off if the environment around them was preventing them from achieving their potential!

So here’s a radical proposal - every DAO at its birth should commit to a fixed end date. And not 100 years out - 200 weeks sounds more like it.

Just under 4 years of focus on a clear adventure-vision. We make as much progress as we can, solve hard problems, increase our expertise, and find the people that help us be our best selves. And at the end, we celebrate what we have accomplished, learn from what didn’t go well, distribute the DAO’s resources in a fair way and move on to the next chapter in our adventure.

Furthermore, if the DAO isn’t making much progress, or it can’t attract enough resources to do what it needs to do, the best thing we can do for those involved is to kill it. Free up the resources and people tied up there and let them either join another DAO committed to the adventure-vision, or start one of their own while incorporating what they learned into their plans.

The obvious objection is that if the DAO is doing well, we don’t want to break up the band, so to speak. I’m sympathetic to that idea, yet in that case the folks involved in the DAO can just form a new DAO with the same key players involved (and probably benefit from swapping out some stagnant tools and crufty processes in the process). If that’s too extreme, perhaps we allow the death date to be postponed by a near-consensus of the members.

But the default should be “it’s been a great party, but all good parties end”

The more I look at how to maximize the value of DAOs, the more strongly I feel that the strategic use of DAO death must play a key role. We don’t want to be the oldest flower in a desert - we want to be the ancestor whose legacy is a vibrant field of life in all its wonderful colors.

I’ll let this sit for now, yet would love to hear your thoughts on this notion. Have I gone too far? Let me know!

LBrian 11/16/2021

Given that DAOs are still highly experimental, it’s hard (for me) to speculate about optimal outcomes in the near term. Some DAOs may benefit from a predetermined lifespan, others may attract contributors and capital purely because they are seen as communities where value can be accrued and distributed over a very long time period or ‘forever’. For example, a data analytics DAO that manages and analyzes a repository of globally important historical data stored on Arweave might not benefit from being dissolved or reorganized from the ground up time and time again.

Additionally, the idea of four people splitting off from a DAO to form something else seems reasonable. What about a DAO composed of 15-20 different Orca Pods each with its own working groups and contributors who are pumping out value? Incentives would need to be aligned at an early stage to encourage these people to stick around for the proposed life span of the DAO rather than going somewhere else where they have a bit more security. Ultimately I think it comes down to designing incentives for contributors to work hard to fulfill the vision of the DAO. A quick and simple example of how a short-term DAO might incentivize participants would be to offer some portion of leftover treasury funds for those who stick around for X amount of time. (Lots of issues that may arise in this scenario).

Ultimately fixed end dates may be a useful tool that adds additional functionality to the DAO design space.

Does anyone have any ideas for DAOs might benefit from a long or infinite time horizon?

Appt Pupil 11/17/2021

You raise good points. I can see where the data analytics DAO would benefit from greater continuity. Even in those cases, though, we would still benefit from checking from time to time if the DAO is truly making the best use of the resources and people involved with it, or if it would be more valuable to let them go to a new home.

The mention of incentives is spot-on and the larger DAO example highlights a point I should have been more explicit about. One important reason to have a fixed end date is to allow for a graceful close to the DAO’s work. This would include finding homes for the missions that are valuable to others and distributing the DAO’s resources appropriately.

I challenge the notion that a DAO that does this would provide less security than one that implicitly claims it will live forever. It is more secure because it is making promises that have some hope of being kept, versus “we’ll always be here for you no matter what - until we aren’t”. That said, this certainly needs more thought, so I appreciate you taking time to consider this and provide another useful perspective

LBrian 11/17/2021

100% agree that something like quarterly reporting should be the norm for any DAO short or long term. Short-term DAOs would likely need regular reporting to ensure the DAOs mission is fulfilled in the allotted time.

I also agree that predefined end dates could make the termination of a DAO less messy. Again I see end dates as a tool, a very interesting tool that hasn’t yet seen the light of day (that I am aware of).

Another short term DAO example: Imagine a gaming DAO that pays gamers to complete time-sensitive missions. The entire goal of the DAO would be to acquire a valuable game asset (maybe an NFT) with the goal of selling and dispersing funds from the sale to participants immediately upon retrieval. A predetermined end date may further incentivize players to complete the task as quickly as possible in order to acquire the available treasury funds in addition to the final reward. The organizers of the DAO would also be incentivized to amply spending in the short term in order to achieve the goal as quickly as possible.

Again I could see a longer term DAO working toward this mission as well. Predetermined end dates, in this case, may provide additional incentives for participants who are competing in any highly competitive field. We’ll have to wait and see who wins in the end

One thing I think we will see in the DAO space over the next year is the creation of standards. Things like regular reporting and treasury dispersal at the closing of a DAO will need to be figured out.

ian-nz 11/18/2021

i like this idea. if you have a lot you want to achieve and you know that you have a fixed timeframe in which to achieve it you are going to be more deliberate with your actions, planning and resources.

i like the idea of giving your freshest and best contribution to an adventure-vision to then pass the baton to the next person who is fired up to give their freshest and best.

you are then free to go and scratch another itch and contribute to another adventure-vision you are passionate about growing further and developing in a different capacity to then circle back for another round at the first DAO (if you so wish) with newly-refreshed vision and energy.

a cool idea. i love where this is going.

LBrian 11/18/2021

I like the “adventure vision” framing!

Could these short term DAOs simply be smaller pods within a longer term organization?

Expanding on the example I laid out above:
A large “Treasure Hunt DAO” with a central treasury could spin-off pods that correlate with specific games. Each game-specific pod could have its own treasury funding that is used for incentivizing gamers to do a task within that pod’s specific game environment. These smaller pods could have scheduled end dates rather than the larger DAO having to liquidate its entire treasury, team, and structure.

The possibilities seem endless.

zkchun 11/18/2021

I’m totally yoinking the “adventure-vision” term @Appt Pupil

Appt Pupil 11/19/2021

Your question is one I’ve been pondering a lot, and it’s worth digging into this in more depth.

Let’s use the “Treasure Hunt DAO” as our case study. The question we have to answer is “what is the benefit that we get from everything being in one DAO versus divided up?” A few possibilities come to mind.

One possibility is branding - people come to the Treasure Hunt DAO because they know that’s where the best games can be found. I don’t find this answer compelling, because brand continuity and company continuity differ in traditional business as well. For instance, I just realized the other day that Pringles is no longer a Proctor & Gamble brand - Kellogg’s has owned them for nearly a decade. Did it really matter to me, though? Not really.

In addition, why wouldn’t another DAO with a compelling game just link to all of Treasure Hunt DAO’s games as well as their own? Now they have everything Treasure Hunt has, so they are likely to siphon off some of the top level interest. This dynamic would seem to lead to a situation where the interests of the pods differ from the top level group, so it does not seem sustainable.

Eventually, we’d end up with a situation where one DAO focuses on funding games, another focuses on creating the best marketplace for games, and still another actually creates a compelling game. There doesn’t seem to be much benefit in a composable Web3 environment to having all of these under one roof.

In that scenario, I don’t really care that any of those DAOs are trying to live forever, as they could each transition any brands or other capabilities to a new DAO as needed without impacting the experience of others they work with. Thus, I don’t see where DAO longevity would be driven by this need.

What am I missing here? And I’ll reiterate the question @LBrian raised above because it’s a good one - what are the reasons that a DAO would need a long or infinite time horizon?

ionn 11/19/2021

@Appt Pupil The DAO has long or infinite time horizon goals. Long Now DAO? History/Library DAO? Gold Vault DAO?

The DAO is an organization → organism. Some ecosystems are better suited for longer lived organisms, some for shorter. I guess I’m shifting the burden of proof here, but I would be interested in hearing a compelling reason that all DAOs should be short lived

A DAO to manage a protocol seems like a straightforward example of a DAO I would want to last a long time. At least as long as that protocol lasts. ENS DAO for example, wouldn’t make any sense with a 200 week deadline…

Appt Pupil 11/20/2021

This is a good point @ionn and I will concede that ENS is a good example of a DAO where setting a short term timeline does not appear to have value.

So since we have at least one good example of a situation where a fixed expiration of the DAO does not make sense, we should figure out what factors should be used to determine whether to define a fixed DAO expiration date, and ideally some guidelines on what that duration should be.

Thinking out loud about this, my first thought is to say that it should be defined by the role(s) the DAO is playing in the adventure-vision.

For ENS DAO, they seem to be pursuing an adventure-vision of something like “make Ethereum easy for everyone to use”, and their role in that is to administer ENS.

Since a customer of ENS needs it to work as long as Ethereum is active, and Ethereum is designed to last in perpetuity, that is one reason to have the DAO also designed to operate in perpetuity.

Also, ENS has the characteristics of a utility in the sense that it appears to be the case of a situation where it is better to have everyone using the same one, rather than using competing capabilities.

That said, the scope of the DAO’s role would appear to be best suited to just serving as the system of record for names and the facilitator for changes to the protocol.

Any activities beyond those would not appear to benefit from having a single provider. For instance, building integrations between ENS and other Web3 components would benefit from competition to make them as simple, inexpensive, and scalable as possible. Thus, that would likely cause issues if they are included within ENS DAO’s scope.

So that looks like two factors that can be applied to this question.

  1. Will a customer of at least one of the DAO’s missions need it to operate in perpetuity?
  2. Will the community benefit from the mission being operated as a utility?

If both of these are yes, then setting a fixed expiration date for the DAO doesn’t seem to make sense

That said, all DAOs should have some sort of shutdown and unwinding approach defined. Even ENS DAO needs a way to deal with a situation it is not designed to handle, and there is no good reason not to have a plan on what to do in that circumstance

What other factors should we consider? What flaws are there in this argument?

Appt Pupil 11/20/2021

Why should DAOs be short-lived? Because it is a useful tool for fighting stagnation and bureaucracy. I have seen too many traditional organizations where the principal-agent problem is a crisis.

This relates to an earlier thread I started on organizational missions. Every organization has an unstated mission of perpetuating the organization. My belief is that the the longer the organization’s horizon, the more resources get devoted to perpetuating the org instead of its adventure-vision

Since DAOs reduce the costs of launching and dissolving organizations, there is less value in perpetuating a DAO than a traditional organization. Thus, if we shorten the duration of the DAO, more resources are available towards the adventure-vision.

The lower bound is primarily driven by human needs. People need some level of continuity and certainty to be able to act effectively.

200 weeks was appealing because it is about 4 years in length, and many people are used to that time span due to Gregorian calendar leap years, political election cycles in many countries, and sporting event cycles like the Olympics and football World Cup.

It also can be easily divided into 100 sprints of 2 weeks, which many Agile approaches have found are a useful planning horizon for projects. There may be arguments for a different duration, but so far 200 weeks looks like a good default.

Kathleen 11/22/2021

It’s an interesting thought experiment, thanks for sharing! But I guess for me a DAO is not really comparable with a brand, a company or even an organization for that matter. It’s rather an agile (always improving) framework & tooling for individuals and other DAOs to cooperate with each other. I therefore go with often used description of a borderless ecosystem.

I don’t see any need to kill an ecosystem. If the ecosystem works well (and that’s our task here) it will punish or even kill pods, working groups, other DAOs… which don’t add value. So, I’d say. We don’t need an end date - though we might have working groups with project character and therefore a limited life-span. But we do need to figure out how damaging mechanisms can be avoided. Innovation dilemma, monopolistic dynamics etc.

LBrian 11/23/2021

I agree with your stance here. The whole point of a DAO in my mind is to break up some of the rigidity at the top end of traditional hierarchies. DAOs can create flexible structures of management and productivity where leaders can be phased out or moved to a different department (Pod) by the community. Our goal should not be to mirror the old structures but with defined end dates. It should be to redefine cooperation and how people collaborate within an organization.

Again I think end dates are a tool, just like any other DAO tooling.

Additionally, it’s worth considering the resource drain that would occur when ending a DAO and starting a new one.

Appt Pupil 11/23/2021

I agree that the end of a DAO should not be a destructive event. In our current world, ending an organization has that effect, but it does not have to be that way in a DAO. If we were able to preserve all of the cultural artifacts, relationships, and mechanisms of the DAO and make the transition to a new DAO seamless, would there still be an objection to having them expire?

To use an analogy we have made elsewhere, if the DAO is a virtual container that we can swap out in the same way we can move a program from one physical server to another via virtualization, then a lot of the reasons to try to make DAOs last forever go away.

Another analogy - making a DAO transition like repotting a plant. If the plant thrives, do we care how old the pot is?

The obvious question, of course, is “why mess around with something that is working?” The reason is that it makes it easier for us to move on from things that are working less well to make way for the new. Otherwise, we end up with a landscape that looks like the corporate landscape, where companies whose best days are clearly behind them lumber on because of inertia. If a consensus wants to keep playing on, by all means let’s allow it, but I doubt that we will look at most DAOs that exist now in 4 years and say, “they should keep doing exactly what they have been doing”.

ApePsyko 11/23/2021

In Robert’s rules he suggests nominating leaders. Maybe you could also build an option to ‘dissolve positions or structure’

Gaian 12/17/2021

The main throughline I see in this thread is that being adaptable as an organization is important. Choosing to revisit a decision at a set point in time is invaluable, and while one option at that point is to disband, there are of course other options.

1 Like

I loved encountering the ‘short term’ idea of DAOs when I first arrived. I think for me it was in your Imact DAO document however, and that is a context in which I think it always applies.

The purpose of some DAOs is clearly a lifespan, or even multiple lifespans. Think social DAO, like FWB, service DAOs like RaidGuild, or the DAOs that are running actual blockchains. They definitely do not want to end in 200 weeks.

A service DAO may have completely different members in that time, and hopefully is continually reevaluating how it provides its service. There is no purpose for it be planning its end, however.

Impcat DAOs that are actually successfully having an impact, should be able to measurably see that impact within 200 weeks. And, if it is not as great as they realistically hoped, it suggests that some things do need to change about how it is creating that impact.

High tech, at least software development, has already proven that short deadlines with appropriate chunking of desired outcomes can produce more reliable results than long deadlines and lots of upfront planning. This is probably even more true today during a time of major transition, than 30 years ago when these practices were first being developed.

The challenge is as mammals, we are all creatures of habit, and we know from research that as a group we are very poor at making short term choices for long term benefits, if those short term choices aren’t rewarding in themselves.

So, even with Impact DAOs where I believe this heuristic makes a lot of sense, how do we incentivize members to want to end and restart when appropriate, instead of create the perpetual organization?

People often forget, that a relatively short time span, (11 years I believe,) was how corporations were originally used. This was when their purpose was recognized as a privilege being offered by the State for a public good. Unfortunately, greed and power shifted that perspective on corporations, and now we are essentially controlled by the insanity of the these false corporate persons, and their desire to optimize commerce for asset accumulation vs maintenance of a healthy humanity and living ecosystem.

1 Like

This is exactly the right question. A lot of it comes down to ways to reduce the disruption of a particular DAO stopping and starting. If I can spin up a new DAO that is a better fit for the its current state and hand off the pods that should continue, that would be powerful. Think of it like a crab being able to leave its shell when it has outgrown it for a new shell. If we had “DAO shells” for various types of DAO focuses, sizes, and resources at its disposal that could be easily changed out, it would provide a valuable alternative to having to try to change all the internal mechanisms to accommodate future growth.