Over the last eight weeks, I led a minigrants program through ArbitrumDAO and Plurality Labs called “Our Biggest Minigrants Yet.” Each week, we accepted grants applications on JokeRace for a predetermined category, like DeFi or sustainability. The top four projects with the most votes each week were awarded 2,500 ARB, for a total of 10,000 ARB in grants distributed each week and 80,000 ARB distributed over the course of the eight weeks.
Below I’ve included a program recap, detailing performance over the eight weeks, as well as some learnings for future program managers of similar governance programs.
“Our Biggest Minigrants Yet” Program Objectives
This program was initially launched under Plurality Labs’ second strategic priority as outlined in their program overview here:
“Achieve governance optimization by identifying and iteratively improving key capabilities to increase DAO performance and accountability.”
Governance in large organizations like ArbitrumDAO can be disorganized and ineffective without structure and process. Moreover, properly incentivizing community members can be challenging. The goal of this program was to develop and implement a weekly governance process through experimentation in order to fund worthwhile grants seamlessly and that would ultimately lead to better DAO performance and accountability.
If successful, the impact I expected to see this program have on the Arbitrum ecosystem was a more engaged and higher quality community. The reason I decided to use JokeRace as the governance platform for this program was because I wanted to experiment with various ways to make governance fun and properly incentivize participation, and JokeRace had the most versatile and features-aligned platform to achieve these goals.
“Our Biggest Minigrants Yet” Program Recap
Prior to the start of the eight week program, we ran two JokeRace contests to determine the voting council for the minigrants program, as well as the grant categories for each of the eight weeks:
We accepted everyone who nominated themselves for a spot on the voting council and invited the ArbitrumDAO Citizens Retrofunding Round 1 recipients to fill in the remaining spots for a total of around 50 members on the voting council.
Once our voting council members were finalized, we ran a grants categories contest on JokeRace that was only open to voting council members to determine the eight categories for our eight weeks of grant giving. We took the top result and slightly modified it based on discussions held in the private voting council Telegram chat to come up with the following eight categories:
-
Empowering web3 communities (January 15-19, 2024)
-
Play to earn reimagined (January 22-27, 2024)
-
The future of identity in web3 (January 29 - February 2, 2024)
-
Scaling DeFi for the masses (February 5-9, 2024)
-
Bridging the gap to TradFi (February 12-16, 2024)
-
Sustainability in the blockchain age (February 19-23, 2024)
-
Unleashing the power of DAOs (February 26 - March 1, 2024)
-
The next frontier of NFTs (March 4-8, 2024)
You can click on any of the links above to view all applicants and final results from each round.
Outside of our first week, during which we received 43 applications, the number of applicants in each round oscillated over the eight weeks, with the lowest week receiving nine applications (week 6) and the highest week receiving 28 applications (week 3).
The category we started the program with in week 1 was the broadest: empowering web3 communities. Almost any project could be said to fit within this category, which is why I believe we saw the highest number of applicants in round 1.
The categories where we saw the lowest number of applicants were sustainability in the blockchain age (week 6) and play to earn, reimagined (week 2). This could tell us a number of things - that Arbitrum isn’t the most conducive L2 to building in these categories, that the Arbitrum ecosystem hasn’t made as much of an effort to attract projects in these categories, or that these categories are simply more difficult to build in.
The percentage of votes participating over the eight weeks varied from 22% to 39%, with the highest voter participation in week 4 and the lowest voter participation in week 7. We exceed our voter participation goal of 20% in all eight weeks of the program.
One possible reason for the low voter turnout in week 7 was because week 7 took place during ETH Denver, and many voting council members may have been attending conference events and offline during the voting period that week.
Overall, the gradual decline in votes from week 4 to week 7 could be attributed to a number of factors - that eight weeks was too long to retain momentum for a program, or that morale declined after an issue with an applicant in week 5.
“Our Biggest Minigrants Yet” Program Learnings
In an attempt to keep this retro short enough that people will actually read it, I’ve attempted to summarize my observations from this experience into five key learnings that I believe may be useful for other program managers and governance leaders to know.
Learning #1: Compensate council members to eliminate any perception of impropriety
At the start of the program, the council members had a discussion around whether we - as council members - were eligible to apply for grants during one or more of the eight rounds. We wanted to give council members a fair opportunity to benefit from the grants program if they’re building meaningful projects that would otherwise qualify for a grant, but we didn’t want there to be any perceptions of impropriety.
After days of discussion around this topic, we decided the fairest model would be to allow council members to apply for grants but disqualify them from voting for their own project. They would still be eligible (and encouraged) to vote for other projects that applied in the same round they applied in.
However, in Round 5, this became a huge point of contention when one applicant believed it to be unfair that a council member had received a grant in Round 4 and again in Round 5 with a different project. While the council member did not actually violate any rules, multiple members of the ArbitrumDAO community believed it unfair (or at least a bad look) to allow council members to apply for grants in this program.
Upon further reflection, I believe the best way to structure a grants program with a voting council is to set a predetermined compensation model upfront and disallow any voting council member from participating in the grants rounds in order to avoid any appearance of impropriety with the public, which may result in diminished trust in the council.
I believe council members should be compensated in some way for their time, as they all committed a lot of time over the eight weeks to reading through applications, conducting their own additional research on the projects, and analyzing which project(s) they thought were most deserving of funding.
From our Round 5 Retro:
I’ll share more on the other bullet points in Learning #2 below.
Learning #2: Establish community mods and rules from the start
This program was primarily set up and led by me, with input from our ~50 council members. We did not assign any other explicit roles during the program, nor did we establish community guidelines as we did not have a Telegram or Discord community to manage at the start of the program. Our public Telegram chat was established during Round 4, about midway through the eight week program.
During Round 5, an applicant joined the public Telegram chat and started spamming the chat with unsubstantiated accusations and even taking it a step further to harassing other applicants and council members in their DMs and publicly on Twitter. Some council members felt immediately that this person should be booted from the chat, with some even deciding to leave the chat themselves so as not to have to be subject to the person’s offensive messages.
I encouraged everyone to stop engaging with this person but let the person remain in the Telegram group for a few days before removing them from the chat. In the spirit of decentralization, I did not feel that I should have the power to censor and remove people from the chat, but several council members believed I should have removed this person much sooner than I did.
In retrospect, I believe we should have established a community channel and community guidelines from the start. This could mean assigning a council member to be the community moderator and compensating them for this role, or it could mean establishing community guidelines in writing from the start so that should any conflicts arise, there would be no questions about how to address and resolve the conflict.
Having a designated community moderator and/or written community guidelines can prevent conflicts from escalating and group morale from taking a hit.
Learning #3: Set project parameters to reduce risk of funding scammers
Scammers have gotten more and more sophisticated over time, and it can be difficult nowadays for even a well trained group of crypto native individuals to spot a scam project, especially when the goal is to fund early stage projects that haven’t received much funding yet.
While running this program, I learned that there are clever ways to not only make a fake project seem legitimate but to also pass the KYC process, for example by using someone else’s information.
In response to some questions that arose about a project in one of the earlier rounds, I put together this project evaluation guide to help our council members do the best job they could to ensure we were funding well deserving, legitimate projects rather than scams.
At the end of the day, there is no way to guarantee with 100% certainty every single time that a project is not a scam, but creating guidelines like this can help everyone DYOR to the best of their abilities and reduce the risk of inadvertently funding a scam.
Learning #4: Reduce the governance time period to create more urgency to act
When it comes to getting people to participate in governance, whether through voting or submitting proposals, the one consistent trend I’ve observed over the past two years of running a variety of governance experiments through Rehash, JokeRace, and ArbitrumDAO is that most people will wait until the last minute when there is a sense of urgency to act.
Whether you leave your proposal open for voting for one month, one week, or one day, most people will act within the final 24 hours. Same goes for submitting proposals or ideas for contests.
Throughout the eight weeks of this program, we consistently saw the number of applications double during the final 12 hours, despite applications being open for 72 hours each week and announced weeks in advance.
If you’re struggling to increase governance participation in your organization or program, the solution is almost certainly not to increase the time period for participation. On the contrary, you might find success with decreasing the time period for participation and more closely investigating your marketing tactics or other aspects of your program.
Learning #5: Reduce intermediaries as much as possible to reduce friction and increase performance
The final learning I’ll include here is mostly targeted toward program managers or DAO leaders to best set themselves up for success.
I believe the most successful program managers are those who can own the program from start to finish, and the most conducive environments for success are those that support the goals of the program manager from an organizational and administrative standpoint but are otherwise non-prescriptive.
As a program manager, you’ll need access to marketing tools, like social accounts, and a clear understanding of processes that extend beyond running the grant, like gathering KYC information and liaising with the larger organization that’s providing the funding. You’ll be best set up for success if you can clearly communicate and obtain the resources you need before starting the program in order to reduce the need for intermediaries during the program, which can slow things down and create roadblocks.
If you made it this far, thank you and I hope you found some value here. I’d love to hear your thoughts and feedback on this - feel free to share them with me on Twitter @ddwchen or Farcaster @treegirl, and don’t forget to subscribe here to get updates from me in the future.