or, The Death of Communities via a Thousand Incentives
Before getting into VC a year ago, I spent most of my 20s as a builder in crypto, primarily trying to understand the role that token incentives could play in the successful achievement of community goals. The first product I built and launched was The Bounties Network, which was one of the first Ethereum protocols to grow >$1m TVL. We worked with a lot of your favorite early tokens/DAOs, and the goal was to build an open marketplace (ie hyperstructure) which could power a plethora of vertical marketplaces via a shared layer of liquid tasks. We eventually sunsetted this effort for a number of reasons, including our discovery that despite open marketplaces now being possible thanks to smart contracts, there were still a number of market forces which would make such a system extremely difficult to establish and scale, due to the business realities that each of the participating marketplaces would face. Suffice it to say that I have a deep experience with how token incentives can be leveraged within crypto communities.
The next thing I went off to go build was during the 2020 DAO rush — a platform called Myco which allowed groups of people with aligned goals and values to come together and work in a unified space and get rewarded their contributions with a token that could be used for revenue sharing purposes vs being traded on the open market. We tried a number of different pathways in search of a repeatable pattern or blueprint which any community could pick up and run with, regardless of their individual aims. This project also eventually failed (for different reasons), but the primary one was that the hypothesis we were testing ended up being completely false, and that’s the subject of this post. We thought that the biggest factor limiting community success was the lack of predictable structures – in their incentives, in their internal processes, and in the spaces they inhabit. We thought that if we could productize those elements and lower the barriers for launching/sustaining communities, we could help enable a new generation of groups to create value for themselves and their stakeholders, in whichever direction they chose.
One of the tragedies of startups is that people shy away from sharing the hypotheses which they tested and ended up being wrong. My hope is that by more vocally sharing some of what we learned, I can save some of you the years which I spent walking down a dead end road. Nonetheless, this post will be full of conjectures and beliefs proven on small datasets, and so I encourage anyone who disagrees with me to try and succeed where I failed.
The first and most obvious fiction that I want to dispel is the idea that extrinsic incentives can be used to make a community stronger. They can’t and they won’t; incentives are the kiss of death for any real community.
A community is an organic group of humans who are aligned on a given mission and values, working together to achieve that goal. Whether it’s a local exercise group, a book club, an open mic, or any other permutation of organic community, the one thing they have in common is that they’re built a strong foundation consisting of highly aligned members who are there because they are intrinsically motivated to be there. Any community you’ve ever seen succeed got there because the people who started it/built it up truly just cared about the core mission, and nothing else.
While there were already some different papers out there which covered the idea of “motivation crowding”, I believed they were mostly wrong. I thought that if the extrinsic reward was a token which couldn’t be traded openly (like an extremely efficient form of sweat equity), we could get around the very real challenges which organizations face when introducing extrinsic monetary incentives (which “crowd out” the native intrinsic motivating factors). I thought that if we simply productized communities in the right way, that we could figure out a repeatable solution to this very complex problem. It ended up not being possible.
Without spending too much time summarizing the key insights of the above papers, the crux of the problem boils down to the fact that extrinsic incentives, if you do need to introduce them, are extremely tricky to administer over repeated cycles, and there are a number of key factors which need to be kept in mind in order for the crowding-out effect to not occur.
The incentives can’t be predictable — rewards cannot be expected, and instead need to arrive only by surprise, creating delight that doesn’t enforce future expectations of rewards from the same (or new) members.
The incentives can’t be exclusive — rewards which are only feasibly available to a select few members will position members against each other and turn it into a competition (ie the top X% of contributors get $Y). Instead they can only be milestone based, which eventually becomes unsustainable for communities as they scale (especially because those milestones need to feel attainable for all members, regardless of how early they are in their journey into the community).
The incentives can’t become part of the culture — rewards can’t be discussed, their value on the open market shouldn’t ever be talked about, and in general the culture can’t ever become associated with the rewards in any way.
The incentives can’t be very large — rewards which are meaningful enough to attract new members will inevitably attract un-aligned individuals, who don’t share the native intrinsic incentives. In a world where many folks are increasingly looking for new sources of income (and because of global income inequality), this is extremely difficult to achieve in practice.
The incentives can’t overshadow social recognition — rewards can never be perceived to be more valuable than the recognition that members would otherwise natively get from other community members for completing tasks that contribute to their shared goal.
As you can imagine, using tradable tokens as the incentive mechanism (vs pure cash) makes all of these problems harder to solve, not easier. Token incentive systems require predictability in their incentive delivery in order to avoid being captured by a small number of “trusted administrators”. Token economies which seek to grow and scale cannot possibly provide non-exclusive rewards, or rewards which are small enough to avoid it becoming a pull for joining the community. Token communities which don’t have the token as part of the culture are oxymoronic. And perhaps most importantly, the social recognition conferred for great contributions will always be less loud than the volume of conversation happening around the tokens/incentives themselves.
As a young man in my 20s, I thought that problems like these were a green-field opportunity for me to explore, a beast I could try to tame. I thought, if only I could try to out-smart greed, we could come up with a pattern for bootstrapping abundance. I was totally fucking wrong!
All of the literature on motivational crowding concludes that in theory it’s possible to introduce extrinsic incentives, but that it’s tricky and should be done with great care. Here I will state plainly that if you’re building a community and intend to use a token to try and introduce extrinsic incentives, you will likely fail, and your community will eventually die (albeit perhaps extremely slowly, sustained by the energy of token holders who are incentivized to believe a dream which cannot ever be true). The intrinsic motivations of community members which originally made your community great and defined it will be crowded out by new motivations and people, who are there for entirely different reasons. Not only have all of my personal attempts in this area failed, I’ve also seen a plethora of tokenized communities eventually succumb to the same forces I witnessed. Once the crowding out occurs, you can almost never go back.
Going back to my original statement, I believe that the very thing which separates a community from other types of organizations is that their members are there for intrinsic purposes. This to me is at the core of what makes something a community — the intrinsic motivation is part of the definition, not an afterthought or a downstream effect. In the absence of intrinsic motivation you have something else… like a company, a non profit, an ecosystem, an economy, a machine, or any number of other structures of human organization. But without intrinsic motivation, I would argue that by definition you do not have a community.
Many amazing communities exist around or alongside these other types of human organizations — like a fan group which organically forms around a musician, or a collective of people who all wear the same brand and make it a core part of their identity. While in their late stages these kinds of communities eventually flounder (because people want to be part of the community for reasons other than simply liking the thing itself), in the early days these groups attract truly like-minded individuals who are all there because they simply care enough to be there. For anyone who’s ever had the pleasure of being a part of an early movement like this, you can attest first hand just how magical they are, how special the relationships formed end up being, and how impactful they can be to the life arcs of their members. Real communities are serendipity engines, and are the closest thing that most people will get to witnessing real magic.
I believe that for any real community, their primary goal should be to defend themselves against status/incentive seekers who join the group for reasons other than the original intrinsic motivations which the community is founded upon. There are many ways this can be accomplished, but the best one I’ve found is for the community to intentionally hide itself, becoming increasingly difficult for non-aligned individuals to find it, creating mechanisms for people to self-filter out via their non-adherence to the core values, until the only people left knocking on the community’s door are the ones which have self-selected to be there and have proven their own intentions. This notion stands in direct opposition to any of the ways that one might think of or treat a traditional product, which is propagated to a maximum number of willing users. Great communities can’t and shouldn’t scale infinitely, unless the dissemination of the ideology at their core is the driver of that scale, and nothing else.
While this post might seem to stand in direct opposition to my original belief in tokens as extremely powerful tools (and products!), I want to try and reconcile my beliefs into what can ideally be a somewhat coherent path for success in building tokenized ecosystems.
Firstly, and perhaps most importantly, I believe that tokens and communities should exist at opposite ends of a given ecosystem — the people there for the money and the people there because they care should be kept distinct, and the latter should actively work to distance itself from the former in any way possible. While it’s possible that some individuals may move from being intrinsically motivated to relying on extrinsic org incentives, these positional switches should be shared explicitly, and those individuals should be prepared to leave behind the benefits they got from being part of the community vs the organization.
I believe that, if a community can properly defend itself against extrinsically motivated individuals, it (and ONLY it) should be allowed to make critical administrative decisions about the org itself, essentially functioning as an independent volunteer-based board of directors. Those decisions can then flow down to the organization (which is charged with executing those decisions), but they should not be made by the org itself, and the org should remain perfectly open to being overpowered by the community at any moment.
I believe that tokens are still an extremely powerful tool for incentivizing meaningful contributions to decentralized organizations, especially from a large network of contributors vs a select few which have been curated into the org. Tokens can be transmitted as efficiently as cash, but still provide major economic upside (assuming they aren’t sold) when the total value of the economy increases. As long as the utility of the economy is tied to the token itself (ie people getting value from the ecosystem need to buy in and hold/burn the token), then number will go up for everyone who holds it, without the need for actually distributing revenue shares or dividends to individual holders. I believe that this is part of what tokens are most useful for, and what will lead to them inevitably becoming the substrate for a new wave of 21st century economic activity. They aren’t equity, and they aren’t cash; they’re something much better. I think we’re still in the very early innings of fully understanding how tokens can be leveraged to create successful and repeatable economic patterns.
However, of the many things tokens are useful for, “rewarding communities” is not one of them.
As you can tell, the Ethereum community (as well as most other tokenized communities) breaks many of the rules outlined here, and this is why I believe that the Ethereum “community” is now a shell of its former self. What we have before us is the late stages of a once thriving intrinsically motivated community which has now succumbed to the pressures of greed and status, both of which have become the primary motivators of many well-meaning individuals, even some whom were originally part of the intrinsically-motivated subset. This crowding-out effect isn’t their fault; it’s a natural outcome of communities which confer external status and economic benefit based on contributions. Crowding out naturally occurs in any community where the members all hold the same asset, and the value of that asset skyrockets rapidly. Our ecosystem is on the cutting edge of this socioeconomic research, and I don’t fault any individuals for their roles in getting us this far in this experiment.
However, I can no longer sit by and pretend that what we have today is anything close to an intrinsically motivated community. It’s a web of incentives, most of which are impossible to track despite their roots being tied to a ledger that’s perfectly public. Even if we could track them, it’s not clear that by simply being aware of an individual’s incentives you can actually solve this crowding out problem, although it does probably help a ton. I personally believe that the Ethereum Foundation should be overseen by people who are intrinsically motivated by the core mission, and should eject anyone who is not. That doesn’t mean that extrinsically motivated contributors don’t have a welcome home in the Ethereum ecosystem; in fact their membership is both necessary and implied by the economy we’ve created. But we should stop pretending that everyone who engages with Ethereum is part of its “community” — this defies the very definition of the word.
As always, I want to publish this post with a caveat that… I might be wrong! While I haven’t personally ever seen the productization of community work, I always welcome anyone to try where I have failed, and to do their own research within this space. Communities are beautiful and fascinating creatures, and even in the worst case I promise you’ll probably have a great time, and hopefully make some life long friends along the way :)
Thanks to Zain Bacchus for encouraging me to publish this post, and for his feedback.