GovAlpha Quarterly Review Q2 2021

Source: GovAlpha Quarterly Review Q2 2021 - Google Docs

Executive Summary

Successes

  • MIPs Portal - Success
    MIPs Portal is up and running, functionality is flexible and extensible.

  • New Monthly Governance Cycle - Success
    Easier to engage with, understand, and requires less effort from voters.

  • Expansion and Redundancy - Qualified Success
    We have another Governance Facilitator and we have more contributors.

Failures

  • SourceCred - Failure
    We were not able to make any noteworthy improvements to SourceCred-in-Maker last quarter.

  • Maker Operational Manual - Qualified Failure
    We never got around to launching this as a thing, but we did produce lots of valuable documentation on parameters and modules in the Maker Protocol.

  • Governance Participation - Ongoing Problem
    Continues to be difficult to motivate large holders to vote. Delegation should mitigate this, but problems will be ongoing.

  • Leadership Burnout - Ongoing Problem
    Lots of responsibilities and stress levels can run high, plan to mitigate with multiple strategies going forward, but will require constant attention.

Spending

  • Under-budget
    GovAlpha came in significantly under-budget for the quarter.
    Spent 156,962.22 DAI out of 240,000 DAI.
    This is a result of being over-conservative given a lack of experience in budgeting.

Focuses for Next Quarter

  • Delegation
    Launch and ensure that there are multiple quality delegates for MKR Holders to choose from.

  • Education
    Focus on education about governance in MakerDAO. Make sure governance is informed as possible.

  • Quality Assurance
    Ensure that our processes are working effectively and resulting in reasoned decisions and intentional outcomes.

Conclusions

  • Overall - Qualified Success
    Progress has been made in the most key areas.
    The biggest issues have been identified and are being worked on.
    Some non-critical projects were deprioritized to ensure that we hit the most critical areas.

Analysis

Last Quarter’s Focuses

Expansion and Redundancy

“A key focus for this quarter will be to expand the number of Facilitators of GovAlpha from one to three.” -
LFW, Soul of Optimism

Evaluation: Qualified Success

As most are aware, we don’t have three governance facilitators at this time. However, we do now have two, which is an improvement from the start of the quarter. While it would have been good to end with three, I don’t believe that it would have done anyone any favors to attempt to force the issue, so on the whole I am satisfied with the result here.

Expansion outside the governance facilitator role, though not explicitly called out in the previous quarter’s budget, has been a success. Over the quarter we were able to onboard both Pablo and Elihu into the core unit as contributors, in addition to helping Payton into the facilitator role.

The additional resource has been of great help and I do believe that it has meaningfully increased redundancy within GovAlpha, and for the MakerDAO governance processes.

Additionally, we have recently made a push to onboard a second wave of contributors. Although it is very early in the process, several individuals have expressed interest and initial conversations have seemed promising. I’m confident that this will further increase GovAlpha’s bandwidth and ability to effectively fulfill its mandate.

In retrospect, it would have been preferable to make this second onboarding push earlier, and in the future, we will try to ensure that we give onboarding the priority it deserves, even if it means pushing back other initiatives, given that it should increase our capacity in the longer term.

SourceCred

"Functionality has become available that will allow us to weight Cred generated by ‘likes’ in threads within a particular category or tag” - LFW, Idealistic Rube

Evaluation: Failure

Unfortunately, in the area of SourceCred, we have not made the sort of progress we were hoping for at the start of March.

While the development of SourceCred within Maker has continued in a limited fashion, it’s fair to say that there have not been any appreciable improvement benefiting the DAO in the last quarter. There are several reasons for this.

#1 Lack of bandwidth

The primary reason for this outcome is that neither I nor later Payton had sufficient bandwidth to manage SourceCred work the way we probably should have been doing. I don’t view this as a failure precisely, as bandwidth was being spent on more critical tasks, but it was unfortunate that we were not able to give this area more attention.

#2 Technical Complexity

In addition to bandwidth issues, the technical complexities of getting the SourceCred algorithm to do what we wanted it to do were larger than anyone anticipated. In combination with the difficulty in verifying the outcome (comparing cred scores under old versus new versions), this has eaten up a lot of the time spent on improvements.

#3 Organizational Changes

In the last quarter, both SourceCred and GovAlpha underwent moderately disruptive organizational changes. On our side, we moved from being funded by the Maker Foundation to being funded by the DAO. On the SourceCred side, there was also some restructuring of responsibilities that invited delays.

On the other hand, while not achieving the stated goals or any visible improvements to SourceCred-in-Maker over the last quarter, there were a couple of less visible improvements that are worth mentioning.

#1 SourceCred Outreach

The SourceCred outreach program was an initiative proposed by Elihu to contact active members of the forum and social media and inform them personally about SourceCred. While I was initially skeptical, this program does appear to have attracted several community members to sign-up to SourceCred, many of which continue to contribute actively in the community.

#2 SourceCred Technical Proposal

In addition to our more modest feature requests for SourceCred, we have managed to finalize a proposal for technical work that will bring SourceCred more directly into the control of Maker Governance. This proposal will move the SourceCred distribution mechanism fully on-chain and allow cred data to be provided by an off-chain oracle and contested on-chain. Once work on this proposal has been completed, this will make SourceCred less trustful and more directly controllable by MakerDAO.

On balance, the failure to make significant progress on SourceCred largely comes down to a lack of bandwidth. Unfortunately, this may not be something easy to rectify in the future. While useful and a net benefit to governance, actively improving SourceCred is not very high on the priority list of GovAlpha compared to other initiatives, leaving it vulnerable to the same de-prioritization if there is limited bandwidth available.

The solution here is to expand the available bandwidth for the team, which should come as a result of the ongoing onboarding effort.

Fortunately, SourceCred continues to provide value in its current form, and maintenance is much less bandwidth-intensive than improvement, so I am confident that SourceCred continues to have a place within MakerDAO.

MIPs Portal

“Up to this point, Governance has not been able to view, search or discover MIPs in an attractive and accessible way. This should change in the coming months.” - LFW, Surprisingly Accurate.

Evaluation: Success

The MIPs Portal is something I would describe as an unqualified success over the past quarter. While there is still work to be completed and numerous improvements to be made, its utility and potential are already extremely clear.

Core, extensible functionality is already present and the combination of tags and custom filters will allow the MIP Editors to manage the content of the portal efficiently and effectively without requiring additional development time.

While MIPs remain an area that much of the community is either unfamiliar or reluctant to engage with, we feel that this portal is a very positive step towards making MIPs as accessible as possible.

Development on the portal was initially very rapid, and dSpot deserves a lot of credit here for their great work. In the last month and a half development has slowed due to dSpot’s other commitments and organizational growing pains, however, they retain our confidence, and development work on additional features and polish has begun again in July.

Maker Operational Manual

“The goal of this project is to create and maintain a set of documents that will allow MKR Holders to make informed choices when voting or signaling within the Maker Governance processes.” - LFW, He Tries.

Evaluation: Qualified Failure

Unfortunately, GovAlpha was unable to launch the ‘Maker Operational Manual’ this quarter despite efforts to the contrary, leaving us to classify this focus as a qualified failure.

Despite the overall failure, there were a couple of successes in this area.

First, GovAlpha did manage to create parameter documentation for some of the most-changed parameters in the Maker Protocol. These documents are linked in every appropriate poll, executive vote, and signal request, and I believe that they provide a useful educational resource for members of Maker Governance when they vote in one of our governance processes.

Second, we also managed to produce parameter documentation for many of the most important parameters introduced in the Liquidations 2.0 upgrade just prior to the first votes on the upgrade. We believe this documentation meaningfully increased the availability of information on these parameters.

The failure to launch the Maker Operational Manual as a platform for this documentation in a reasonable timescale is the main failure here. As with some of our other focuses, the issue here was one of bandwidth and prioritization - specifically the bandwidth of leadership and decision-making within GovAlpha.

I don’t believe that there is a huge amount of work involved in getting this up and running, but it’s been blocked on decision-making bandwidth and lack of a ‘project owner’ to keep the project moving as consistently as would have been ideal.

As bandwidth and priorities shift over the coming quarter, we will attempt to rectify these issues and launch the documentation platform.

Other Points

New Monthly Governance Cycle

Evaluation: Successfully mitigates some issues.

This quarter we replaced the Monthly Governance Cycle defined in MIP3 with a new one defined in MIP51.

The new Monthly Governance Cycle has lots of advantages, but some drawbacks.

+ Requires less participation from MKR Holders
+ Does not require out-of-schedule executive votes.
+ Has a longer participation window.
+/- Allows for more changes to be made with less effort.
- Breaks link between outcome and implementation (can’t implement technical changes as a result of a poll.)

This change should help to solve many of the problems that were caused by a lack of participation under the old monthly governance cycle.

Lack of Governance Participation

Evaluation: Ongoing problem.

We have struggled to meet the desired level of participation in both governance polls and executive votes over the last quarter. It has become necessary to ‘chase down votes’ over the last few months in a way that was never required previously.

This situation came to light due to the absence of a large MKR Holder that had previously voted very diligently. This had helped to mask the non-participation of other larger MKR Holders.

In the very short term, there is probably not an easy fix available for this issue. The short-medium term response has been to work on delegation as a priority. Several larger MKR Holders have indicated that delegation is ‘the solution’ that will allow them to participate more reliably (through delegates.)

We are hopeful that delegation will prove to be the silver bullet that resolves the first-order problems caused by a lack of participation (i.e. difficulty operating the governance processes reliably.) However, lack of participation will remain an issue to be addressed in the future to ensure effective and decentralized governance of the Maker Protocol into the future.

Leadership Burnout

Evaluation: Ongoing problem

Over the last quarter, it’s become clear that the environment and responsibilities faced by facilitators of GovAlpha can result in a build-up of stress that negatively impacts productivity and job satisfaction. This issue is exacerbated by the difficulty in taking vacation time away from MakerDAO.

This problem will continue to exist and require constant monitoring, but several strategies should mitigate its impact and ensure that those affected can remain productive and satisfied.

  • Expansion of GovAlpha such that responsibilities can be delegated effectively.
  • Ensuring that vacation time is taken when needed in volumes that mirror industry standards.
  • Ensuring that personal management resources are available for those affected.

We hope to implement these strategies in the coming quarter.

Forum Management + Moderation

Evaluation: Ongoing problem

Over the last month or so, it’s become clear that the MakerDAO official forum needs a more active hand in management and moderation. The volume of posts has increased as the DAO has expanded, and this has come with more contention over the direction taken in various areas.

GovAlpha has continued to work on the MakerDAO forum to mitigate some of these issues to mixed success. So far we’ve attempted the following:

  • Making greater use of categories and subcategories.
  • Using the pinning functionality more liberally to temporarily highlight key threads.
  • Making heavy use of Group and Title functionality to identify members of the DAO and their respective areas of expertise and associated core units.

There are still many avenues available to us here, and we’ll be exploring them over time. Some of our current ideas include:

  • Integrating and organizing tags more effectively.
  • Investigating and integrating forum plugins to improve functionality.
  • Expanding moderation team, and having moderators screen posts in certain categories more actively (for example signal requests, MIPs, etc)

Spending

Source: GOV-001 - Quarterly Domain Budget - April 2021

Actuals versus Budget

Spending Breakdown

Spending Type Budgeted Actual Spending
Core Unit Wages 86,000.00 65,021.14
SourceCred 81,000.00 75,346.00
MIPs Portal 20,000.00 12,574.93
Status UI Support 0.00 4,000.00
Admin 500.00 20.15
Contingency Item Budgeted Usage
Status UI Support - 4,000.00
Contingency 52,500.00 -

Spending Conclusions

  • Significantly under-budget (better than the alternative, probably!)
  • Overestimated contingency requirement.
  • Took on Status UI Support even though not initially budgeted for.

Overall the budget layout for GovAlpha in the first quarter was too conservative, this reflects my relative inexperience managing budgets and is a result of erring too far on the side of caution.

This resulted in an opportunity cost to the DAO to the tune of the over-budgeted amount. I consider the downside to this outcome to be limited at the current stage, given that the DAO is already accumulating significant opportunity costs for the capital it is not actively utilizing. While this outcome is not optimal, it has not inflicted any real cost to the DAO at this stage.

The next quarter’s budget has already been approved, I suspect it will have the same issues. When writing future budgets GovAlpha will be more conservative with its estimates.

The unspent funds will be returned to governance on request or if it appears that the return would meaningfully benefit the DAO in some way. Returning the funds to the surplus buffer is non-trivial so absent a clear need or request to do so, we will deduct the difference from the next budget transfer to GovAlpha.

Next Quarter’s Focuses

These are laid out in more detail in the last quarter’s budget, which can be found here.

  • Delegation
    Launch and ensure that there are multiple quality delegates for MKR Holders to choose from.

  • Education
    Focus on education about governance in MakerDAO. Make sure governance is informed as possible.

  • Quality Assurance
    Ensure that our processes are working effectively and resulting in reasoned decisions and intentional outcomes.

Conclusions and Final Words

Overall, the first quarter of GovAlpha’s operation while funded from the DAO has been a qualified success. Progress has been made in several key areas, and existing processes and responsibilities have been discharged successfully in a large majority of cases.

Many of the issues we identified were a result of a lack of human resource and time on behalf of the leadership of the Core Unit. These issues are actively being mitigated, and we have already seen improvements in the latter month of the quarter.

Future risks to the GovAlpha core unit and Governance in MakerDAO remain and will require mitigation in coming quarters. The most important of these include:

  • Governance participation.
  • Core unit operation.
  • Managing the growth of the MakerDAO community.

Thanks to MKR Holders and the wider Maker community for your continued confidence in GovAlpha. We’ll do our best to make Q3 stand out.

20 Likes

Thank you GovAlpha for your efforts last quarter and for maintaining such high level of contributions and transparency.
Speaking as a relatively small maker holder I really appreciate all your core unit does for the DAO.

3 Likes

Are you able to provide more colour as to why it has been hard to motivate whales to get out and vote? If not, no worries–I can sort of think of reasons why. Delegation can’t come soon enough!

I was under the impression the new Algo had cut out the fat. Or are you referring to getting it increased to 10,000 DAI per week? Sort of read your comments below but not sure if its budget related or what exactly will make the GovAlpha team satisfied. But please don’t waste time answering this because I know you’re busy.

Maybe start taking Friday to Sunday off and have @prose11 take Saturday to Monday off? I don’t know–we don’t want you to burn-out Long. Take it easy. No worries. Let things play out. Ask the community for help. we are here.

This is good and I love the honesty. We All learn as we go along. No worries. It can only get better.

This is awesome!! Totally loving all the goals for Q3. And don’t stress out–your Team can get there when it gets there. All is good.

:100: :100: :100:

You and Payton are awesome! If you can find a third that would totally spoil us. But if you can’t then we have won the lottery by finding y’all – “I Rather be lucky then good”

Why don’t you outsource this to someone who practically lives her like @PaperImperium the Venezuelan Kids (can’t remember their tags), or even myself (I can behave :crossed_fingers:t4: ) – and then in 2022 hire someone from overseas. I’ve spent some small time in just about every Telegram group and I have found two (2) that have impeccable community service. And on both occasions they have been a team based in India that has the most pleasant attention-span -and they never ever lose their cool, no matter how silly of a question people post.

All-in-all I want to personally thank the GovAlpha CU–you folks are exceptional. Please, please let us know how WE the Maker Community can help. Just ask, don’t be shy. We are willing and able.

5 Likes

Thanks for the big post.

Could you elaborate on your concern around voter participation?

  • Is your concern a general one that “participation is important for a community”?
  • Are we not hitting quorum? (In which case, maybe lower the quorum? what is the quorum?)
  • Something else?

Thanks.

Thanks, that means a lot.

I think ultimately the issue is diffusion of responsibility. There are lots of MKR Holders, why vote when someone else will do it? Also, in the case of VC’s the MKR isn’t owned by those doing the voting (at least usually.) So there is a level of separation between exposure to the consequences of decisions (or the lack thereof.) Finally, VC’s particularly have LOTS of checks and balances to ensure that individuals can’t ‘go rogue’ so their voting processes are heavyweight and involve multiple individuals.

Sure, there have been incremental improvements, but we had a lot of high hopes for using it in a more targeted way (incentivising signal and MIP engagement, for example). We’ve been working on that for coming up on six months lol.

I guess I was just hoping we could have done more with it, and more quickly.

Haha, so have weekends? Yes, we do try to have those, and usually we succeed. I think we’re in a better place now that we have both Prose and myself as facilitators, and more contributors. I am less worried about this than I was.

Hah, yep this has always been the plan (to delegate management to contributors). Originally I outsourced it to Payton. Only that backfired slightly when he got more busy. @s_ben does do a lot of the heavy lifting here, so I’m not too worried.

Thanks! We appreciate the kind words and support.

Sure. So, my concerns are both general and specific.

  1. Decentralization is the name of the game. Multiple projects in DeFi aim to be decentralized, but the majority of tokens are owned by large VC firms (this is also true for Maker). Furthermore, all token voting is plutocratic rather than democratic by necessity. Having a solid, active and invested community lets us actually be decentralized in fact, even if vote-weight on-chain is not as well distributed as we would like.
  2. We don’t have quorum’s exactly, but yes, sometimes votes take longer than they perhaps should do. Additionally the total voting MKR is not a very high percentage of the total existing MKR. It’s not a horrible percentage, but it could be much better.
  3. More participation gives decisions more legitimacy. This is really important in the longer term. If we can’t get consensus and buy in around important decisions, then the DAO is going to spend a lot of time on arguments, and less time on actually doing things.
  4. More voter participation protects the protocol against Governance Attack.
5 Likes

And comments here:

I think the biggest issue is that while there is the mcdgov.info governance info data portal there has been no single report (that I know of) doing an analysis of the state of MKR much less the state of MKR in governance. I understand why when reading the report above. GovAlpha basically doesn’t have the resources to put together such a report. I think without such a report the whole move on delegation is kind of this hail mary on improving participation. I am going to hazard a prediction here. More MKR may enter the governance contract but it will be dominated by whales. Top 100 wallets hold 82.65% of all MKR. Now this includes governance, the 84K MKR, AAVE etc. Removing these the wallets in the top 100 account for 79.99% of all remaining MKR. There are more contracts (I have only looked carefully at top 18 wallets so far) so these numbers should modify but they are illustrative of:

Also when I look at the total list of wallets participating in governance something >80% (done by eye not formally by date) of all wallets appear to have voted on something in the past 4 weeks. In the governance wallet the top 17 wallets account for 90% of the possible vote.

When I started looking at this I realized what is missing from GovAlpha is a kind of ‘state of MKR’ report that tries to look carefully at not just what 80% of MKR is doing (the top 100 wallets) but tries to engage with the MKR in about 400 wallets that composes 15.6% of all MKR out there, and wonder whether anything should be done about the remaining 5% attributed to all other wallets.

I want this to sink in a bit here when we think about delegation, governance participation, etc.

  • Top 100 etherscan.io wallets have 80% of all MKR that could vote.
  • Next 400 wallets have 15.6% of all MKR that could vote.
  • All other wallets have no more than 5% of all MKR.
2 Likes