So I’ve spent my whole life in web2 startups, growing one from 2 to 700+ and growing many from 1 to graveyard. My jam has always been around the scale part. How do you do it without breakdown down people, losing culture and resist becoming a heartless machine with inaccurate KPIs? This week I did a bit of a dive into scaling challenges DAOs face and what are some ways forward. Warning: lots of snark.

As orientation baseline I think Jessie Walden’s piece of progressive decentralisation from 2020 is a tables stakes when you want to consider the adversarial environment that DAOs operate in. I believe today many DAOs buoyed by the last bull market think they operate in some technical utopia where nothing outside matters and nothing inside besides decentralisation matters. Jessie’s piece it’s well worth your time if you care about long term DAO health: https://a16z.com/2020/01/09/progressive-decentralization-crypto-product-management/. For those that don’t have the time I’ll paraphrase, DAOs should go through three sequential stages of "Product/Market fit", "Community participation" and finally "Community ownership". Attempts to rush and change the order of these will land you in significantly more pain than you already are. It’s seriously good, go read it.

Most DAOs are somewhere between "Product/Market fit" and "Community participation" today. Let’s focus a bit more on the nitty gritty around DAO scaling challenges. Why do you need to care? Because you’re operate in a non-utopian and stochastically complex and adversarial world where the long term survival is heavily dependant how much effort is put into this area. If you want to keep your project at 10 people and take a "que sera sera" approach, sure go ahead just let your experiment be one of the tens of thousands out there. You might luck out, survive and contribute something to the world, I doubt it. Just like how you have limited time in a day to work on this, your DAO, as much as you would wish otherwise, has limited resources and need to be smart about it. If this doesn’t convince you treat it as a cognitive exercise that might distract you away from the burning fires of your project for fifteen minutes.

The following areas are the ones I’ll shine some light on today, these are by no means the only areas but I’m gonna start here and elaborate as I collect more info. Please send me more info for your DAO please 😆.

So you got a DAO and you got early participants who are keen, technically strong, smart and ideologically aligned and very engaged. Everyone is working hard contributing and delivering features that users care about, sweet! Fast forward 12 months, maybe it’s a bear cycle, maybe your token price is down, as you’re no longer the new hip kid or maybe the low hanging fruits are picked and the next class of problems are really hard. Whatever the case your contributors are leaving and what’s worse taking institutional knowledge with them. You think fondly of your third contributor that’s been the lynch pin of your operations, she just told you she’s had enough and moving onto new mushroom based transportation DAO. Man losing her is going to hurt because she was too busy to build and never bothered documenting anything. Some of the best smartest builders I’ve ever worked with were the worst information sharers, not because they can’t do it, because they and their environment incentivised them not too. You snap back to reality and notice tears streaming down your cheeks. That’s okay you think to yourself, you’ve hired new keen ideologically aligned people to join but what are they faced with? Instant un-documented, in-consistent legacy along with randomly hidden institutional history. From the new hire’s perspective: cool new projection, how did we get here? Do I have to read every single improvement proposal and all the comments to know where things stand? (Yes) What design should I use when I build? Who do I talk to about this? How do I catch up on this? Do I read the entire 12 month project history on TG, Discord, Notion… maybe V1 is here, but V2 is there, where the fuck is V3? One big theme with the collaborative and onboarding process in a DAO is that it’s an after-thought usually left to the whims of the natural emergent charismatic leader (think Adam Neumann and his coffee intricacies) or the most vocal intransigent minority, except guess what they leave too. Everything in life has compromises. People love to tout the benefits of "great DAOs and how Web3 enables the unbundling of careers into jobs" the flip side is unless you have an established backbone that is not negotiable you will have a dumpster fire the longer your project lives. The current solution is very Darwinian, the most overloaded project dies by process fragmentation eventually implodes and the everyone moves onto the next shiny thing before that again implodes. The solution here is to treat institutional memory and culture as first class citizens. Even though the new contributors are all different in their choice of baggy clothing, choice of coffees and psychedelics, they need to know for this piece of work, there exists a standard to conform to, the shape of which is dictated by strong culture. Anything less you end up trading short-term gains for long term pain, you know like a junkie…

The above is one side of the coin, here is the flip. Your DAO competes in an dog eat dog world. The contributors are niche of a niche. There’s a world-wide tech. talent shortage, a sliver of the already shrinking market is your crypto talent. Your DAO’s talent is now fought over in one of the most competitive ecosystems ever known to man. Sure it’s a Cambrian explosion, you know what happened during explosions? A lot of fucking things died. You might not like this but you have limited resources and limited runway, I’m sure that huge treasury amassed during the bull market hasn’t taken a 90% haircut just like it did in 2018. Even if you had unlimited runway until the advent of an artificial general intelligence you’re still limited by your contributors and their limitations. Given the dynamic of a hostile environment and limited resources you have have be very careful and very focused on what you’re working on. Oh you don’t? You going to use prediction markets and wisdom of the crowds to spawn an internal Cambrian explosion inside your walled garden of network effects? I don’t know how to tell you this unless, you’re the size of the biggest L1s, you don’t have the scale. Even then I still argue you need strong mechanisms to determine focus and strategy. Strategy is about saying ’no’, sure maybe a nascent and developing path today becomes in hindsight the obvious choice but you don’t know which one of the hundreds of paths before you is that one. So how do you balance the two? In the investment markets you size your bets and you limit how many bets you take on. I see DAO’s today consistently fragmenting their focus and looking at too many proposals and then lament that each of proposals only got three eye balls. It’s not enough to just make the proposal voting system less fatiguing, the DAO has to have a funnel to focus attention on the top most pressing matters. Proposals needs to be sized and weighted appropriately. The way to solve participation isn’t to make it lighter easier to digest, it’s to be laser focused about being laser focused. Sure basic hygiene stuff matters, only involve the appropriate audience for the questions but DAOs needs to focus on the meta of the "Community participation" as much as the matter of it.

There are so many other untapped and interesting areas of DAOs I’ve yet to dive into like what about the OGs of the projects that now get carte blanche about what they work on and how they work? Here’s a nice gem will pause on. Old people like me remember IETF. They’re the peeps that set the standards for the internet, you know that thing all of us are building off, the thing that is fucking well architected and demonstrably scalable. You know what else is amazing about the IETF? Here are their cardinal principles taken from: https://www.ietf.org/about/mission/ * Open process * Technical competence * Volunteer Core * Rough Consensus and running code * Protocol ownership

Now where have I seen those things before? Us humans love to think we’re the smartest things that’s ever existed in all space and all time. Ironically it’s also why we’re cursed to re-invent and re-discover the past, you think two world wars is enough. For christ sakes there’s a movement around learning how to breathe again ("Breath - James Nestor"). Anyway what I’m saying is that this is not new ground, it shares a lot of challenges that very smart people have already thought about and solved. I leave you with RFC 7282 https://datatracker.ietf.org/doc/html/rfc7282 on "rough consensus", it’s fucking brilliant. Why do I keep trying to think when massively wrinkly brained peeps have done it already…