CGP 2.0 Updates and Renewal

[harsha]

Thank you so much @cylon for sharing your thoughtful comments and elaborating the benefits to Compound because of CGP 2.0.

[harsha]

Thank you @signal_corps for your support and sharing your experience as a CGP 2.0 grantee.

[harsha]

Thank you @kallolborah for sharing your thoughtful comments and support.

1. Based on the number of proposals received and the allocated budget for each domain, we have proposed adjustments to the budget, specifically increasing the grants budget for the Dapps and New Protocol Ideas domain. As you rightly pointed out, the proposed time commitments may vary depending on the number of proposals received for a domain and the domain allocators may exceed or work for less than 15 hours per week based on the proposal volumes. However, drawing from the learnings from CGP 2.0 and feedback from the domain allocators, we expect a domain allocator to allocate an estimated average time of 15 hours per week. While CGP 2.0 has consistently maintained an average communication turnaround time (TAT) of less than 48 hours and an average funding TAT of less than 2 weeks after the milestone has been completed, we are committed to maintaining a consistent TAT for all proposals and maintaining a high NPS score.

2. Thank you for providing your valuable suggestions. We agree that in order to ensure the success of CGP 2.0, it is vital to incorporate feedback from proposers and grantees. At the end of CGP 2.0, we gathered feedback from CGP 2.0 grantees through a survey and connected with them through calls, Discord to better understand their experiences and incorporate their feedback. Going forward, we will share feedback surveys with all proposers and grantees more frequently and provide additional opportunities, organise interactive sessions such as the demo day for proposer’s, grantees’ to share their feedback.

[kallolborah]

Thank you @harsha ! And I am happy to see that you already have taken into account the TAT, NPS and the feedback loop. I am happy to support and contribute to CGP 3.0.

[kevin]

I just want to start off by saying that it’s awesome to see all the developer activity the grants program has generated for Compound. This surge of engagement undoubtedly benefits the protocol by encouraging further enhancements and innovations. I would like to thank Questbook for starting up CGP 2.0.

That being said, while I understand the eagerness to start CGP 3.0, I think it’ll be beneficial to give the community some more time to digest the results of CGP 2.0. Given the recent conclusion of CGP 2.0, the community hasn’t had the opportunity to fully assess the impact of the funded projects and determine how best to enhance the program for CGP 3.0. There are also ongoing discussions by other parties, such as Alastor & w3s and Alphagrowth, that can shape the direction of CGP 3.0 and we should allow those discussions to develop before rushing to fund a specific program.

I would also like to provide my thoughts of the grants program from an objective bystander’s point of view by pointing out some areas of improvement for future grants programs.

Questionable value from current inbound approach

The lessons learned from CGP 1.0, as articulated by Larry Sukernik, emphasized the superiority of Request for Proposals (RFPs) in yielding valuable contributions for the protocol, in contrast to an inbound approach. To quote Sukernik:

@sukernik

When we launched Compound Grants, we thought people would know what to work on. In retrospect, we should have known that it’s easy to boil the ocean when thinking about what to work on. What happened is applicants would apply for grants for all sorts of projects, putting us in the position of assigning priorities to applications after they were submitted. Instead, the right approach is for grants programs to understand the level of priorities for the protocol and put out RFP’s that community members can begin working on. We didn’t do that at first, but we learned quickly. Now, we have a list of RFP’s that are a priority for the protocol to complete and encourage community members to work on them.

I want to preface the following section and say that I trust all involved parties (grantees, domain allocators) act in good faith for the protocol’s betterment. Yet, I reference specific examples not to call out specific projects, but to highlight how the current inbound approach falls short in optimizing protocol value.

Reviewing the projects funded by CGP 2.0, I question the value that some projects provide to Compound. Funded projects are occasionally redundant, given existing solutions. An example is the $30K grant to launch Comet on Optimism, which was approved despite Compound Labs already communicating the work being done there months prior. This grant was revoked after Labs cleared up the miscommunication, but it was concerning to see the cross-chain domain allocator’s lack of familiarity with the cross-chain work being done for the protocol. Another example is the Compound API kit grant to build a Typescript SDK for interacting with Compound, which already exists.

For novel projects not replicating existing work, there are certain projects that get paid a good amount of money and get to leverage Compound’s brand to launch a product that mainly benefits themselves rather than the protocol. I’d like to avoid providing specific examples here, but there are many projects that collect fees from users, none of which are given to the DAO. There are other projects that seem more like a solution in search of a problem, as is the case with cross-chain infrastructure projects that benefit greatly from being adopted by a blue-chip protocol like Compound.

Considering the significant portion of the DAO’s funds allocated to these grants, it’s wise for the DAO to reevaluate its fund allocation strategy, aiming to efficiently direct resources toward projects that generate maximum value for the protocol.

Lack of a dedicated quality oversight stakeholder

There are three major stakeholders as part of this CGP arrangement:

1. Questbook/Program Manager
2. Domain Allocators
3. Grantees

Although each of these stakeholders likely operates with good intentions, none are strictly incentivized to look out for the best interest of the DAO. Questbook/the program manager is paid to run the program. Domain allocators are paid to allocate their funds. Grantees get paid by hitting milestones. However, none of them are solely responsible for maintaining project quality, which opens the door to potential harm to the protocol’s reputation if issues arise.

This came to our attention when we realized a grant project received the green light from OZ to launch on mainnet, despite lingering security concerns tied to their contract. This situation raises significant concerns, as the protocol’s brand could suffer irreparable harm if any issues were to arise. While there seemed to be an implicit assumption that OZ, as the official security partner of the DAO, would uphold the quality standard, it appears this expectation was somehow overlooked. Given that any project launched via the grants program is backed by the DAO’s funding and implicitly carries the DAO’s stamp of approval, maintaining a high project quality benchmark is essential. Clearly defining the responsible entity for upholding this standard going forward is crucial to safeguarding the protocol, especially as more grant projects are shipped using the Compound banner.

Suggestions for improvement

Based on the issues presented above, here’s a list of suggestions to improve the grants program moving forward:

• Focus on RFP model instead of inbound approach

  • Have the active members of the community curate a list of projects to be funded, prioritizing projects that hold the highest potential value for the DAO. This guarantees that projects with the most positive impact on the protocol receive priority and attention.

• Designate OZ as the quality oversight stakeholder

  • Leverage the DAO’s existing relationship with OZ by designating them as an impartial security advisor that champions the protocol’s best interests. As a current security partner, OZ is well-positioned to uphold a quality benchmark for projects launched under the Compound name. Approval from OZ could even serve as a prerequisite for milestone payouts, promoting the delivery of high-quality code over expedited development.

• Tie milestone structure to value creation

  • Align milestone payouts with tangible protocol value creation rather than task completion devoid of substantial impact. Shift focus towards funding projects that genuinely enhance the protocol. Implementing performance-based milestones, such as payouts triggered by a Dapp driving TVL to Compound, can help achieve this goal. Another approach is to consider retroactive grants, rewarding projects based on their actual contribution to protocol value.

[harsha]

Thank you so much @kevin for sharing your valuable feedback and inputs. Really appreciate the encouraging words and your initial support to kickstart CGP 2.0. We recognise that there is a scope of further improvement in CGP 2.0, and believe that the active involvement of Compound Labs, community members, and other stakeholders will be instrumental in making CGP 2.0 renewal a success. Such actionable and constructive feedback regarding an individual initiative can be viewed as great progress in itself. We have actively taken feedback from active community members over Discord and Compound community calls on a frequent basis and have proactively worked towards resolving them. We invite additional comments on this initiative. Going forward, we will also ensure to amplify our efforts to seek feedback and provide additional opportunities for comments throughout the duration of the grants program rather than near its conclusion or renewal initiatives.

Furthermore, we are thankful for the opportunity to engage in discussions aimed at further improving the grants program and its design. This transparent approach, distinct from other opaque programs that lack adequate data points to facilitate open discussions about crucial improvements, has allowed us to seek actionable feedback from Compound Labs team and the community members.

RFPs vs CGP 2.0’s approach

Prior to the launch of CGP 2.0, each domain allocator wrote comprehensive guides outlining the RFPs as well as guidelines for crafting well-structured proposals. The detailed documentation can be found on Questbook’s dashboard just below the title of the domain.

The guides for each domain are referenced below.

• Dapps and protocol ideas: CGP 2.0 Domain: Protocol Ideas and dApps - Google Docs
• Dev Tooling: Notion – The all-in-one workspace for your notes, tasks, wikis, and databases.
• Security Tooling: Compound Grants 2.0 - Security Domain - Google Docs
• Cross Chain & Multichain Strategy: Compound Grants 2.0 - Multi Chain - Google Docs

Outbound vs Inbound

While each domain had well-defined RFP and proposal guides, we acknowledge that we can further scale our outreach and sourcing efforts. To accomplish this, we have significantly scaled our team and efforts to grow the number of builders submitting high-quality proposals. We already have an organic traffic of 20k+ builders and have now appointed dedicated resources to source proposals aligned with the domain-specific RFPs to increase our sourcing and outreach efforts. In addition to this, we worked and will continue to work with the Compound Labs team to participate and conduct hackathons at various events to increase awareness about the renewed grants program and source projects aligned with the domain-specific RFPs. This will include various initiatives, including guided sessions, workshops, and sourcing activities, which were previously implemented to attract talent and attention to CGP 2.0. Furthermore, the RFPs for the renewed program will be finalized only after actively taking feedback from the members of the Compound Labs and community members. We are open to receiving further feedback on scaling our outreach and sourcing efforts.

Reviewing the projects funded by CGP 2.0, I question the value that some projects provide to Compound. Funded projects are occasionally redundant, given existing solutions. An example is the $30K grant to launch Comet on Optimism, which was approved despite Compound Labs already communicating the work being done there months prior. This grant was revoked after Labs cleared up the miscommunication, but it was concerning to see the cross-chain domain allocator’s lack of familiarity with the cross-chain work being done for the protocol. Another example is the Compound API kit grant to build a Typescript SDK for interacting with Compound, which already exists.

The evaluation criteria of “Relevance to our ecosystem" was designed as an obligatory field for domain allocators to complete proposals under consideration. Its significance lies in carefully evaluating proposals, ensuring their alignment with Compound. This requirement emphasizes the selection of proposals closely tied to the Compound’s objectives.

Really appreciate you pointing this out. We’ve thoroughly addressed the situation regarding the Comet launch project, clarifying the matter with the proposer and resolving any miscommunication. As part of our efforts to prevent such occurrences, we’ve enhanced the role of the Program Manager. The Program Manager will now ensure that the accepted proposals are known to the members of Compound Labs every week, and over community calls along with syncing with domain allocators twice a week.

I’d like to avoid providing specific examples here, but there are many projects that collect fees from users, none of which are given to the DAO. There are other projects that seem more like a solution in search of a problem, as is the case with cross-chain infrastructure projects that benefit greatly from being adopted by a blue-chip protocol like Compound. Considering the significant portion of the DAO’s funds allocated to these grants, it’s wise for the DAO to reevaluate its fund allocation strategy, aiming to efficiently direct resources toward projects that generate maximum value for the protocol.

Thank you for highlighting this. Really appreciate it! With the Program Manager’s expanded role and the valuable input from Compound Labs, we are committed to prioritizing projects that align with the finalised RFPs and demonstrate a commitment to contributing their fees to the DAO. Additionally, we will actively seek inputs before finalizing evaluation rubrics to ensure that projects aiming to both generate and contribute fees to the DAO and provide value receive funding. We expect that the alignment of funded projects will improve significantly as we actively gather input from both Compound Labs and community members on domain-specific RFPs and evaluation rubrics.

Lack of Community Oversight

While we have and will continue to encourage community members to actively oversee the evaluation process and the status of each submitted proposal on Questbook, going forward, we will collaborate closely with approved proposals to facilitate the sharing of their proposal details and milestones with the Compound community. Some of the initiatives taken towards this direction can be seen here. Further initiatives will involve more frequent demo days, similar to the one conducted near the end of CGP 2.0, where the community can gain deeper insights into the ongoing projects and their impact to Compound. It is in the best interest of all stakeholders to deliver what is best for Compound to avoid any significant reputation damage.

This came to our attention when we realized a grant project received the green light from OZ to launch on mainnet, despite lingering security concerns tied to their contract. This situation raises significant concerns, as the protocol’s brand could suffer irreparable harm if any issues were to arise. While there seemed to be an implicit assumption that OZ, as the official security partner of the DAO, would uphold the quality standard, it appears this expectation was somehow overlooked. Given that any project launched via the grants program is backed by the DAO’s funding and implicitly carries the DAO’s stamp of approval, maintaining a high project quality benchmark is essential. Clearly defining the responsible entity for upholding this standard going forward is crucial to safeguarding the protocol, especially as more grant projects are shipped using the Compound banner.

We appreciate you raising this concern. This issue has been thoroughly addressed through in-depth discussions with the domain allocators, with special attention from the OZ members. As a result, the responsibilities of domain allocators have been extended to requesting an audit for the considered/accepted proposals, particularly those that involve Solidity code being deployed into production and directly impacting Compound. In order to streamline the code auditing process and avoid potential time-consuming challenges, the domain allocators will also provide assistance to the considered/accepted proposals by offering feedback on code quality and design.

We are dedicated to implementing quality assurance measures to safeguard the protocol and uphold and enhance Compound’s brand reputation.

Further Improvements

We acknowledge the potential for further enhancements, as highlighted above, and greatly appreciate the valuable feedback you’ve provided:

1. RFPs vs Inbound approach - Although all proposers were encouraged to review domain-specific RFPs before submitting their proposals, we will finalize the RFPs for the updated program after actively seeking input from both Compound Labs members and the broader community to ensure alignment with Compound’s focus areas and roadmap.

2. Designate OZ as the quality oversight stakeholder - We are actively collaborating with OZ on this initiative and have incorporated their feedback in this proposal to expand the role of domain allocators for improved quality oversight as initial steps.

3. Tie milestone structure to value creation - During CGP 2.0, all proposals were funded based only after the milestone was accomplished. We have taken feedback from a few Compound Labs members on this matter and have linked metrics such as TVL to the grants program’s success.

We extend our sincere gratitude for your valuable suggestions towards further improving CGP 2.0 We request you and the community to proceed and participate in the CGP 2.0 renewal, taking the following into consideration:

• Considering CGP 2.0 stopped reviewing proposals on June, 30th, a large number of small teams and proposals who could potentially contribute significantly to Compound are awaiting review on their proposals for over a month

• While we acknowledge and appreciate the initiatives undertaken by other entities to contribute to Compound’s growth, we believe that their focus is primarily directed towards business development, expansion, and funding proposals within this domain, rather than prioritising and funding core protocol-related and technical proposals. We also believe that they can complement CGP 2.0’s efforts should the community decide to proceed with their proposal. We are open to collaborating with them to accomplish what is best for Compound

[kevin]

@harsha, I appreciate your response and your acknowledgement of the areas where the CGP can be improved. I’d like to reiterate that ongoing discussions are taking place with other parties such as Alastor & w3s and Alphagrowth about improving the next iteration of CGP. It seems disingenuous to rush an on-chain proposal to allocate another ~$1M to the program before a) these discussions have finished and b) the community has had a chance to thoroughly access the value generated by CGP 2.0.

Allocating ~$1M is a substantial decision, and the DAO should exercise caution and due diligence to understand the potential return on investment before going through with this renewal.

[blake]

Hi Kevin,

I’ve been closely following the discussions around the Compound Grants Program and appreciate the depth of your insights. Addressing our Compound API Kit:

• Differentiation from Compound.js: Our Kit stands apart from Compound.js. While the latter offers basic functionalities, our Kit is tailored to provide advanced features, such as leverage, collateral-swap, zaps by integrating flash loans and swap aggregators, bridging the gap between simple actions and more intricate, intent-driven operations on the Compound protocol.

• Naming and Clarity: The naming of our project might have led to some confusion. A more fitting name, such as “Compound Intent Kit,” would better convey our aim to assist developers in crafting intent-centric applications without the complexities of smart contracts.

• Community Collaboration: Echoing Harsha’s sentiments, we’re deeply rooted in the belief of community feedback and collaboration. Our primary goal is to empower and foster innovation within the Compound community, and we’re always receptive to suggestions to better align with its vision.

Thank you for fostering such an open dialogue. We’re committed to the growth and success of the Compound community and look forward to continued collaboration.

[harsha]

Hi @kevin , we will withdraw the on-chain proposal. That said, we’d like to clarify that we have no intention of rushing this proposal. We have been taking active feedback from the community on every community developer call, and the feedback on the proposals through the discord grants channel very frequently throughout the duration of CGP 2.0. We have also explicitly requested and taken feedback from all the grant proposals both from the accepted and rejected groups.

Before submitting the renewal proposal for an on-chain vote, we also conducted a temp check vote to gauge the community’s sentiment towards the renewal proposal where we witnessed a clear majority towards the renewal. A similar process was followed for CGP hackathon proposal. Over the last 12 months, we have optimized the entire process, tooling, flow and communication to ensure every part of the grants program is as transparent as possible. We have always invited community participation on every community developer call, and discord chat. We intend to optimize for what’s best for Compound. If you as a representative of Compound Labs think the renewal of the proposal is not optimal at this point of time, we are open to withdrawing the on-chain renewal proposal to allow for further revisions before proceeding further.

a) these discussions have finished and b) the community has had a chance to thoroughly access the value generated by CGP 2.0

We respect these considerations and request you to provide specific timelines for concluding these discussions along with actionables. We assumed that we have taken detailed feedback from the community through demo day, forms, community dev calls, polls and replies on the forum post before submitting an on-chain vote.

If there are other parties who want to help with improving the grants program, we are always open to taking feedback or collaborating with them as long as the process is transparent and decentralized. We appreciate the involvement of Alastor & w3s and Alphagrowth in improving the next iteration of CGP and are open to taking their feedback and collaborating with them. We invite them to engage in a discussion on this forum post for potential collaboration. Additionally, we are receptive to revising domains, and allocations, and re-structuring processes with inputs from Compound Labs and the community through discussions and polls. We encourage Alastor & w3s, Alphagrowth and other interested parties considering to contribute towards CGP within a new domain while taking on domain allocator roles, to actively participate in community deliberations. We will initiate a community vote and gather feedback during a community developer call before presenting the refined proposal for consideration. Your involvement is essential in shaping our collective direction.

[allthecolors]

Thank you @kevin for laying out observations and suggestions for improving the next iteration of CGP. They overlap closely with many of the considerations that domain allocators like myself have been navigating within each domain and with each other in facilitating the program as a whole. I thought I could add some color here for the community about how the program has operated in practice this cycle, as I think it is relevant to how we might go about implementing some of the suggestions.

• On inbound vs RFP approach: In practice, CGP 2.0 operated on a hybrid model where sample RFPs were provided as a part a general call for proposals in each domain. Larry’s observation of relative chaos in the absence of RFPs with CGP 1.0 is valid, especially given the context of the heady days post-DeFi summer during which that program was in full swing. At the same time, during somewhat quieter times like now, there is value in permitting applications beyond a strict set of RFPs: it widens the scope of potential builders by allowing folks to bring their own ideas (which is often more intrinsically motivating than working on “someone else’s idea”); reduces the impression of scarcity (“someone else/faster/better will claim that bounty first, so I won’t bother”); and taps the broader community who may have better ideas within each domain than what we articulate as priorities for the protocol. In domains like mine, the vast majority of received proposals fell outside of the provided RFPs, and some could be argued to be duplicative – analytics stands out as an example – though I would counter that having multiple accurate analytics providers with different focuses and goals benefits the protocol from the standpoint of accessibility for sophisticated users. Token for token, I would have liked to see more of my domain’s support go to initiatives meaningfully and creatively introducing uninitiated communities to Compound, but the quality of proposals serving more sophisticated users was (perhaps not surprisingly) far higher on average. Community refinement of a key set of RFPs in each domain would be valuable, and I welcome ideas for stimulating that discussion. We struggled with limited community engagment on these in the development of CGP 2.0 despite robust discussion about what the structure of the program should look like.

• On tying milestone structures to value creation: The quality of milestone articulation in CGP 2.0 varied strongly from project to project, but overall I agree that there is room for improvement. Milestones in CGP 2.0 were proposed directly by the applicant and either approved or rejected along with the entire application. I often encountered situations where a project was supportable in principle, but the proposer’s milestones were strictly completion-oriented and therefore poorly aligned with the interests of the protocol, including many that did not include a mechanism for feedback from the community (which was articulated as a requirement in my domain’s program description). In many of these cases, I worked with builders to revise milestones as part of a complete proposal resubmission, which was clunky (and a fair amount of effort on both sides) but at least resulted in better-aligned milestones in most cases. It will not be hard to pull up examples where I let some poorly articulated milestones slide, often due to the aforementioned challenges of needing to revise and resubmit the entire application (requiring also a new eth payout address in Questbook) for any 3rd, 4th, etc round of milestone tweaks.
  Tying of milestones to value creation in the form of quantifiable protocol metrics is definitely preferable in principle to completion-based milestones. It also presents additional challenges on the ground that limited our use of them in practice:

1. for many projects, it is difficult to attach a timeline to KPIs within the authorization timeline of CGP. If a project articulates a plan to drive 1M USD equivalent stablecoin borrowing activity to the protocol, but it takes 9 months to do so, in the current model, CGP would not have a way to fund this milestone (and the builder would be stomaching an inordinate amount of risk to agree to it). There are some web3-native solutions to this class of problem (e.g. UMA’s KPI options), but these would still require holding CGP-allocated COMP in a smart contract potentially beyond the authorization period of the program. Furthermore, favoring projects with short-term KPI delivery could bias the program away from projects that build longer-term value that can’t be readily measured within the program timeline.

2. for other projects, metrics for value creation can be difficult to articulate: metrics for security, dev tooling, analytics, governance research, and education/outreach initiatives are all either more challenging to articulate or more challenging to verify. I think we’ve done an okay job with this for a majority of accepted proposals in my domain, despite its breadth, but I’m certainly open to suggestions for how some of them could be improved. The best place to do this is in the discussion thread on each proposal in Questbook, and ideally before the proposal is formally reviewed.

3. monitoring of the TVL or related KPIs may in some cases require additional development beyond the scope of the proposal; this can be added, but needs to be taken into account and represents a kind of embodied cost (albeit likely worthwhile) to sharpening milestones this way. Given that many projects in my domain were asked to trim features to stay below our per-grant limit, it becomes more difficult in some cases to justify adding on these costs.

• On lack of community oversight: I’m supportive of the idea of elevating OZ’s role in oversight, although it will require some close discussion with OZ about the parameters of that commitment within OZ’s ongoing B2DAO arrangement with the protocol. I disagree (civilly) with the suggestion that the CGP director and domain allocators are not directly incentivized to look out for the best interest of the DAO on account of their being paid by the DAO for their efforts. On the contrary, by analogy to a democratically elected politician, the allocators and program manager are directly incentivized to operate in the best interest of the DAO or else they will be voted out of office: if our performance had been truly misaligned with protocol interests, the conversation would more likely be about transferring responsibility entirely over to Alastor or Alphagrowth rather than exploring how they might work with the CGP 2.0 team to improve future rounds of Compound Grants.

Finally, I just wanted to note that while I respect @Harsha’s decision to afford the community additional time to review @kevin’s concerns and reflect any changes in a new on-chain proposal, the postponement is not entirely net-neutral because it exacerbates uncertainty of timeline and commitment among builders in the community, several of whom came to Compound specifically in response to the opportunities created by CGP 2.0. Some have already submitted proposal concepts in anticipation of a potential renewal. Others developed proposals in CGP 2.0 whose scope/cost exceeded what could be funded with the resources available in that program and chose to break their proposal into two or more components, some of which are planned (or even in progress) for a next iteration of CGP.

And one last note re: retroactive funding: while I agree that this could be a component of a next iteration of CGP, I have objections to the concept of CGP being a retroactive-first program. Retroactive funding has its place, but based on my experience interacting with our builders in CGP 2.0, most of the builders in our community are developing as a/the primary source of income. Retroactive programs that don’t commit support until after the fact are poorly aligned with their needs and would likely reduce both the quantity and quality of applications. Proactive funding is more inclusive and will help bring forward the best ideas for the protocol.

[kevin]

@blake thanks for clarifying. Again, I want to re-emphasize that I only used your project as an example of redundancy in funded projects, without assessing the project’s quality. I appreciate the nuances and the added features it offers on top of the existing Compound.js. A better example of redundancy would have been the funding of 3 analytics dashboards that @allthecolors had mentioned.

[kevin]

@harsha, while neither I nor Labs requested for the proposal to be cancelled, I appreciate the gesture of cancelling to allow for more constructive discussions to take place before a renewal is made.

@harsha

Before submitting the renewal proposal for an on-chain vote, we also conducted a temp check vote to gauge the community’s sentiment towards the renewal proposal where we witnessed a clear majority towards the renewal.

To be fair, it does look like the majority of the participants in the temp check vote are either directly involved with CGP 2.0 or are grantees who got funding from the program.

@harsha

We respect these considerations and request you to provide specific timelines for concluding these discussions along with actionables.

Respectfully, neither I nor Labs hold the authority to establish timelines here. My intention is solely to express the concerns shared by myself and some of my colleagues. Ideally, the service provider can consider these suggestions and address them before proceeding to a renewal.

If the community is happy with the grant program in its current form, there is nothing Labs can do to stop the renewal.

@harsha

We assumed that we have taken detailed feedback from the community through demo day, forms, community dev calls, polls and replies on the forum post before submitting an on-chain vote.

I admit we could have surfaced these concerns sooner. I wanted to wait for the conclusion of the grants program to thoroughly analyze it and provide feedback, as well as give the program a chance to run through completion.

As an outside observer, keeping tabs on the grants program was rather difficult since everything lived in the Questbook dashboard. Tracking the progress of the grants program required clicking into each project individually, a time-consuming task that only domain allocators are likely to do on a frequent basis.

One recommendation is to give weekly/monthly updates directly on the forums to convey which projects were funded and paid out. The Aave Grants program does a great job of communicating updates and CGP should strive to do the same.

@harsha

We will initiate a community vote and gather feedback during a community developer call before presenting the refined proposal for consideration.

Thank you. We look forward to helping to refine the program.

[kevin]

@allthecolors, thank you for your response and I appreciate all the work you have been doing for the Compound protocol.

@allthecolors

At the same time, during somewhat quieter times like now, there is value in permitting applications beyond a strict set of RFPs

Correct me if I’m wrong, but the grants program (CGP 1.0 included) has historically allowed applications that fall outside the scope of RFPs. I was not suggesting we only stick to RFPs, since the grants program should remain flexible to fund a diverse set of ideas. However, I think we may disagree on the extent of approving these non-RFP applications.

More often than not, applications that deviate from RFPs aren’t well-aligned with the program’s goals. While exceptions exist, these may be rare if the RFPs are well-defined. Allocators should lean towards rejecting such projects rather than approving them. Justifying funding of out-of-scope projects due to the lack of in-scope applications leads to a loss of program focus.

In essence, rejections of applications should be the norm. It is fine, and often expected, for a grants program to be selective. On a side note, it would be valuable to know the acceptance/rejection rate for CGP 2.0.

@allthecolors

In many of these cases, I worked with builders to revise milestones as part of a complete proposal resubmission, which was clunky (and a fair amount of effort on both sides) but at least resulted in better-aligned milestones in most cases.

That is good to hear, and I have seen examples of you reducing scope or altering milestones via the Questbook program. Unfortunately, as you stated, this process is cumbersome and requires resubmitting a new proposal, cluttering the dashboard even further. Hopefully, that experience can be improved going forward.

@allthecolors

Tying of milestones to value creation in the form of quantifiable protocol metrics is definitely preferable in principle to completion-based milestones. It also presents additional challenges on the ground that limited our use of them in practice:

You’ve outlined three difficulties with defining performance-based milestones and I agree with all of what you said. Forward-looking milestones (e.g. TVL) are certainly difficult to do. But perhaps there is a middle-ground, and this is heading towards my suggestion of an objective quality assurance stakeholder.

The problem with a lot of the milestones is they are completion-based. I’ve seen $4k be paid out for a system diagram + wireframe of an extension that is not reviewed by anyone and will likely never be used by the protocol. There are plenty of milestones that simply require a grantee to finish writing code, without any quality assurance on the code being written. In these cases, the grants program is paying for work to be done, without the expectation of ever using that work because there is no QA involved.

@allthecolors

I disagree (civilly) with the suggestion that the CGP director and domain allocators are not directly incentivized to look out for the best interest of the DAO on account of their being paid by the DAO for their efforts.

Apologies if I seemed accusatory; that was not my intention. This conclusion came from my previous point of milestones being weighted towards completion rather than quality. Consequently, I was seeing projects being paid out and launched with limited testing/auditing, which I wouldn’t expect a protocol like Compound to ever endorse. Given that these projects, especially public goods initiatives, carry the Compound DAO’s implicit approval, they should meet a stringent quality standard.

My concern around incentive misalignment for domain allocators is that allocators may feel the need to deploy most of their capital. You might disagree with me on this point, but let me ask you this: If all 100 applications did not meet your bar, but were not outright scams either, would you be comfortable rejecting all of them and telling the program manager you allocated $0 out of the $450k? I assume you’d feel pressured to allocate some funds at that point. However, a person solely focused on the protocol’s best interest may not allocate any funds.

@allthecolors

the postponement is not entirely net-neutral because it exacerbates uncertainty of timeline and commitment among builders in the community, several of whom came to Compound specifically in response to the opportunities created by CGP 2.0.

Since the recently cancelled proposal was for a ~$1M renewal of the program, I think it’s fair to evaluate how this new funding should be best spent. Builders who took part in the initial phase of CGP 2.0 should not be blocked by these discussions as the original agreed upon funds have already been allocated to the grants program.

Again, thanks your for detailed response and hope we can find a way to improve the next iteration of CGP.

[cmrn]

Hi, $8.5k mini grantee here (torque.fi).

First, thank you to all for the opportunity. Many DAOs don’t make such an effort.

To the Compound community, your feedback is valuable & appreciated. Please feel free to leave your review of our work here, on Questbook, or if you’d like to do so privately, please email hello at torque.fi.

It’s been a wonderful experience thus far & the plan is to make Compound proud. We’re almost ready for audits & Arbitrum mainnet launch. It’d be great to connect with OZ & we’re working on others. I’m witness to @allthecolors responsiveness & professionalism. Managing the most popular category is no easy task.

Our stablecoin update & related scope increase, although delaying us slightly, is a fresh attempt at an idea not so different from CompUSD years back. Actually, I don’t know what happened with it. Maybe someone can educate me. Since the announcement, there has only been validation, albeit not live.

Considering Compound has aligned with USDC, I’ll note we’re not in competition. USG is decentralized, returns earnings (if there is any) to holders via dynamic strategies, & is generally higher risk (comparable to FRAX or DAI). It’d be epic to discuss learnings from Compound’s research & development (possibly over Twitter Spaces)! Post-launch, Phantom (the strategy admin of Torque) is considering a proposal to shore up the Arbitrum USDC.e incentives if they remain as current without new USDC market live.

Of course, the market should aim to be profitable, but we need some spread like Polygon. If you don’t agree, do you propose we wait until the new USDC market is live? Thank you for the insights guys.

Let’s grow.

Sincerely,
Cameron

[harsha]

while neither I nor Labs requested for the proposal to be cancelled, I appreciate the gesture of cancelling to allow for more constructive discussions to take place before a renewal is made.

Thank you @kevin for sharing your comments. While we have been taking active feedback on CGP 2.0 throughout the last six months and on the renewal proposal very frequently, we want to ensure that we incorporate any additional feedback from all key stakeholders including the members of Compound Labs before proceeding for an on-chain vote.

To be fair, it does look like the majority of the participants in the temp check vote are either directly involved with CGP 2.0 or are grantees who got funding from the program.

We want to re-emphasize that we have invited active feedback and comments from different stakeholders very frequently throughout the duration of the grants program on Discord, developer community calls. Furthermore, we have invited feedback on the renewal proposal. While knowing the experience, feedback of stakeholders directly involved with CGP 2.0 is critical for further improvements, we have also explicitly requested and taken feedback from all the grant proposals both from the accepted and rejected groups.

Respectfully, neither I nor Labs hold the authority to establish timelines here. My intention is solely to express the concerns shared by myself and some of my colleagues. Ideally, the service provider can consider these suggestions and address them before proceeding to a renewal.

If the community is happy with the grant program in its current form, there is nothing Labs can do to stop the renewal.

We believe that any additional inputs and feedback from Compound Labs is critical to the success of CGP 2.0 renewal. This will ensure that the funded proposals align with the protocol’s priorities and value creation, prior to advancing to an on-chain vote for CGP 2.0’s renewal.

One recommendation is to give weekly/monthly updates directly on the forums to convey which projects were funded and paid out. The Aave Grants program does a great job of communicating updates and CGP should strive to do the same.

We appreciate your feedback and the work of Aave grants program to update the community. Going forward, we will be sharing updates on the status of each proposal across all domains once a week with the Compound community here. The Program Manager will ensure that the status of proposals (accepted, and funded) are known to the members of Compound Labs and the community once a week over forum posts and community calls. Questbook empowers anyone in the community to view the status of each proposal (even those that are rejected) along with the evaluation rubric, rubric scores and the feedback from each domain allocator. The Program Manager will include links to these proposals in the weekly forum updates, providing community members with the opportunity for a more comprehensive overview.

We extend our sincere gratitude for your valuable inputs and acknowledge that they will play a critical role in the impact and success of CGP 2.0 renewal. We are committed to incorporating additional inputs from key stakeholders including the members of Compound Labs before moving forward to the on-chain vote.

[kallolborah]

@kevin

I am one of the grantees of CGP 2.0 but I am writing here in general about some of your comments. I bring along more than two decades of experience of writing software, running new businesses, and rolling out software products for them.

1. RFPs vs Open calls
Are you trying to outsource work or inviting proposals that can bring in fresh perspectives ? RFPs may be good to attract ‘service providers’ but Open calls will perhaps attract entrepreneurs and new ideas, some of which may add a lot of value to Compound. Open calls with RFPs as guidelines may serve as a middle path, but Open calls may provide better signaling to attract bright new ideas, some of which you may not have even thought about. For example, returns by offchain borrowers of tokenized RWAs are higher than what onchain returns on staked collateral is on a lot of Defi platforms, and you may see a gradual withdrawal of liquidity from protocols such as Compound go to RWAs - I guess it will be hard to frame a RFP that will adequately capture this problem ?

2. TVL / Metrics
Defi is in its infancy. $42 billion of TVL is not great considering that there are hundreds of protocols and Dapps and that $42 billion is not going up. Who knows, TVL may not be the correct metric after all as capital efficiency goes up on Defi platforms. Locking in capital in a protocol may not be a good sign unless that capital is creating returns for liquidity providers. Perhaps, transaction volumes are a better indicator ? Again, this is something that submissions on open calls may shed some light on if they are given the flexibility of specifying their own set of metrics.

Finally, cancelling CGP 2.0 may have been done in a hurry and somewhat unilateral way, unless we are not aware of the need to take that decision in a way. There are some 20-30 grantees in the CGP and many of them (including us) have discussed doing work across CGP 3.0. Although, our interests are not greater than the interest of the Compound protocol, giving some consideration to grantees show the maturity of thinking of the team behind the CGP/Compound. As a core contributor, we expect you to lead the way in these matters in a thoughtful and mature way.

[robinnagpal]

We could possibly have a retro session either during one of the dev calls or as a separate call since many things were done well, but there are also areas for improvement.

Conducting a fully transparent process requires a lot of effort, and I believe Questbook and the domain allocators did an excellent job.

I personally feel we should establish two tracks:

1. RFP, which addresses the immediate needs related to rollouts, integrations, content, and growth, etc.
2. Another open track (or domains) for all kinds of ideas.

In the review or grant allocation process, we could introduce an additional step where domain allocators assist in shortlisting and refining the applications. A final review could then be undertaken by the domain allocator + someone from the labs or an individual who has been engaged with the compound community for some time, possessing both historical knowledge and a long-term perspective.

I believe this step will benefit grantees, as it allows them to specify the kind of support they’ll need after the grant project concludes, such as code reviews, audits, marketing, rollout, maintenance, and more.

[harsha]

Finally, cancelling CGP 2.0 may have been done in a hurry and somewhat unilateral way, unless we are not aware of the need to take that decision in a way. There are some 20-30 grantees in the CGP and many of them (including us) have discussed doing work across CGP 3.0. Although, our interests are not greater than the interest of the Compound protocol, giving some consideration to grantees show the maturity of thinking of the team behind the CGP/Compound. As a core contributor, we expect you to lead the way in these matters in a thoughtful and mature way.

Hi @kallolborah, thanks for sharing your comments. Really appreciate it! We canceled the CGP 2.0 renewal proposal’s on-chain vote to ensure that we incorporate any additional feedback from all key stakeholders including the members of Compound Labs before proceeding for an on-chain vote. Additional inputs and feedback from all key stakeholders is critical to the success of CGP 2.0 renewal. This will ensure that the funded proposals align with the protocol’s priorities and value creation, and contribute to the success of the next iteration. The budget for all already accepted proposals has been accounted for in the allocated funds section and the domain allocators will continue to review and make payouts for the accepted proposals/milestones upon the completion of respective milestones.

[harsha]

Thank you @robinnagpal for your kind words and the comments on further improving CGP 2.0. We’re actively working towards domain-specific RFPs, which we will soon be sharing with the community for their inputs. The renewal proposal will maintain its dual-track approach, including both approaches i.e. inbound (open) proposals to invite creativity and outbound (sourcing based on RFPs) proposals to ensure alignment and value creation.

Updates on the accepted, funded proposals across all domains will be frequently shared with the community on the forum. The Program Manager will share the status of proposals (whether accepted and funded) and ensure that the updates are effectively communicated to the broader community and members of Compound Labs through forum posts and community calls. This will increase visibility proactively and enhance community participation, giving the community an opportunity to voice concerns and seek clarity before milestone payouts are initiated.

[kevin]

@kallolborah

1. RFPs vs Open calls
Are you trying to outsource work or inviting proposals that can bring in fresh perspectives ? RFPs may be good to attract ‘service providers’ but Open calls will perhaps attract entrepreneurs and new ideas, some of which may add a lot of value to Compound.

I’m not asking for the removal of open calls. My previous posts pointed out the issues with the current inbound-centric approach. I suggested a more RFP-focused model that still allows for inbounds to find a good middle-ground between the two approaches.

@kallolborah

Finally, cancelling CGP 2.0 may have been done in a hurry and somewhat unilateral way, unless we are not aware of the need to take that decision in a way.

CGP 2.0 is not cancelled. As @harsha pointed out, the proposal to renew was cancelled by Questbook so they can take in the feedback and improve the program.

@harsha and the Questbook team are doing a great job of taking feedback and working on an improved proposal. These types of discussions should not be rushed. A large amount of funds is being asked for the renewal and we should ensure these will be effectively allocated.

@kallolborah

There are some 20-30 grantees in the CGP and many of them (including us) have discussed doing work across CGP 3.0.

Questbook already has the funds to pay out any grantee that was approved for the original scope of CGP 2.0. With GCP 3.0 not even being live yet, I don’t know why promises were made to grantees for funds that Questbook did not yet receive from the DAO.

Regardless, I understand your concern and want to reiterate the goal here is to improve the next iteration of the grants program rather than stop the program altogether.