After the approval of AGP-1, the Aragon Governance Proposal process, by ANT holders in November 2018, the Aragon community began preparing for the first official vote under the new system. Proposals were readied and submitted, reviewed by the Aragon Association (AA) Board of Directors, and finally voted on by ANT holders from January  24-26, 2019.

This post will review what went well and what could be improved in Aragon Network Vote #1 (ANV-01), as informed by both my own perspective as one of the AGP Editors and based on results from a brief survey that was released for feedback at the same time as the final results blog post. As such the commentary will be a mix of my own opinions and feedback received from others.

You can find the results from the feedback survey here.

I’ll note that, as you can see in the spreadsheet, there wasn’t much data to go off of based on the survey results. Like the vote itself, turnout for the survey was quite low. There were only seven responses to the survey, and four of those that responded did not vote. This is itself something that we need to work to improve. However we did receive feedback via other channels as well, including Aragon Chat, the Aragon Forum, reddit, private messages, etc. Thank you everyone who gave us feedback!

What went well

There was only one positive comment in the survey responses from someone who voted, which said that voting was “highly positive, the feeling is great”. This is nice to know!

From my own perspective, I will say there are two things in particular that went well:

  1. There were more proposals submitted than I expected. In total, twelve AGPs were submitted on time. And the quality of these proposals was pretty high. Nothing I would consider spammy or scammy.
  2. As far as I can tell, the voting tech worked flawlessly. We didn’t receive any technical support requests during the voting period, so it seems the instructions we published were informative and the tech worked as expected. Big kudos to the dev team for building a great product, and to the AA for opening the votes on time.

A third positive outcome was for the proposals that were approved. Thank you ANT holders for approving funding for Aragon One, and congrats to all other authors of the proposals that were approved!

What could be improved

Based on the feedback in the survey, comments we received through other channels, and my own observations, there’s clearly a lot that can continue to be improved with the AGP process.

Communication

We have a lot going on in the Aragon project. Between the Flock teams and the AA we publish at least one blog post per week on average, we have a packed social media content calendar, and around the time of this last vote, we also had AraCon. So it’s easy to see how communications about the vote could have been lost in the noise.

One survey respondent said they did not even know the vote was happening. Another comment sent to me privately mentioned that they knew the vote was happening, but did not receive the instructions in time to get their tokens off an exchange to vote. And given that several AGPs were submitted last minute, one even submitted too late, we can also do more to communicate with AGP authors and make sure that they are getting their proposals in with plenty of time for public review and comment before the proposals are finalized and reviewed by the AA.

Related to the proposals that were submitted last minute, there was also some confusion about how to participate in the Flock program and how it relates to the AGP process. The Flock program managers and AGP Editors could work together to better communicate how a Flock proposal moves through Flock first and then the AGP process.

Drafting and reviewing proposals

One piece of feedback that I received during the public review stage was related to the bidding process for AGPs. If multiple teams are bidding to perform a similar service for the network, then it may be strategically advantageous for them to disclose the bidding price at the last moment before proposals must be finalized to prevent the bid from being undercut by another team. But doing so reduces the quality of the public review, since reviewers lack important price information to judge the proposal and give feedback that may improve the pricing offered. So there’s this unresolved tension around the bidding process when multiple teams are competing that we haven’t come up with a good solution for yet. If anyone has any ideas, please share!

Another aspect of the review that could be improved is punctuality in turning in the AA review votes. The last of the AA review votes weren’t turned in until a full day and a half after the review was scheduled to end. This delayed the publishing of the “final details” blog post, including instructions for voting, which had an impact on voter turnout (as noted in the communications section above), and also left proposal authors waiting for an answer longer than expected. Two of them reached out to me to ask what the status was, and while there were some “pre-commitments” on GitHub it’s preferable to have an official answer in a timely manner. As with most situations like this, there were reasons for the delay, but more effort can be made and advance planning done to ensure that the review is completed on time.

Voting

There were a few suggestions for how to improve the voting process itself that came through the survey responses, including:

  • Make it more obvious that voting is happening when it is, by putting the message everywhere someone looked on Aragon communication channels that the vote was happening. I think we did a pretty good job of this but welcome specific suggestions for how we can improve.
  • Make it easier to vote. A couple of comments mentioned that voting with one transaction per proposal was annoying, so it’d be nice if there was a way to bundle all votes into one transaction. There was also a suggestion to have a “splash page” to vote on with direct hardware wallet integration so voters could just plug in and vote all in the same place, no extra software necessary and no jumping between different Voting apps.
  • There was some confusion about why the proposals had different support requirements. The reason for this was explained in the “final details” post about the vote, but given this respondent’s other answers it’s likely they did not see/read that post, indicating that we could have done more to increase visibility of this important post and make sure that voters know they should read it before voting.
  • Make the link between the proposals in the voting app and the proposals on GitHub cleaner, and have an official “one-pager” with a summary of each proposal so people can quickly see what they’re voting on. Right now we use hashes and GitHub links in the Voting app to establish a secure link between what’s in the app and what’s on GitHub, but it’s done in a hacky kind of way and could definitely be cleaned up and made more “native” to the app.

Many of these comments also cover feedback I received via other channels, and I’m sure there are other things that could be done to improve voting (some of which may be addressed with our work-in-progress v2 voting infrastructure). I don’t have anything else to add here in terms of critique. But I do want to again say that while there is obvious room for improvement, what we do have works amazingly well and the product team deserves massive props for what has been achieved so far with Aragon’s on-chain governance tools.

Vote analysis

There were some good suggestions in the survey results for insights that we might try to gain by analyzing data from the votes, such as how long the voters had been holding their tokens for, which proposals had the most activity and if there’s a way to connect the timing of votes to activity off-chain (e.g. promotion on social media), participation rates, and determining how many of the tokens that didn’t vote are held in just a few wallets (e.g. exchanges). We haven’t completed a deep analysis like this yet but perhaps it’s something we can fund using the new AGP-10 Community Funding DAO.

What we do already know from looking at the blockchain is that we have a lot of room for improvement with voter turnout. While the Aragon Network Token contract showed around 20,000 unique addresses holding ANT at the time of the vote, only around 60 unique addresses participated. All together these 60 addresses controlled at most 7.85% of the ANT supply. However there are reasons to be optimistic that participation rates will improve in future votes thanks to improvements to both the voting technology and how we communicate with voters.

Next steps

While we aren’t going to be able to build the perfect governance system, toolchain, and communications plan overnight, we can use this feedback to inform a plan that improves on all these areas incrementally over time.

As the communications team lead at Aragon One I have already begun incorporating much of the comms-related feedback into a new Aragon Network Vote communications plan that the comms team will use to plan its communications around future votes. I’m also working on making the Aragon Governance page on the Aragon Wiki the “single source of truth” for information about past, current, and future votes so people can go there and find all the information they need to vote confidently. We’ll still use the blog and other well-known community channels to publish important announcements but we’ll start aggregating links to these announcements on the wiki page whenever it’s relevant.

The Aragon Association has reviewed this post before publishing and agrees that they can do more to smooth out the review process. And the product team has been aware of many of the tech improvements that could be made, such as delegated voting and making voting less expensive, and is also taking inspiration from some of the new suggestions to improve the voting experience. Expect to see these improvements roll out bit by bit in future Aragon releases and research blog posts.

Thank you

Big big thanks again to everyone who participated in the survey and sent in feedback through other channels. It’s already making an impact and we’re excited to try out some new ideas during the next Aragon Network Vote. And many thanks again to everyone who participated in the last vote. Get your proposals and tokens ready, because the next vote is just a couple of months away, currently scheduled for April 25-27th!