[Discussion] Safe DAO Resource Allocation Model (OBRA)


I’d consider adding something in addition to the $SAFE success reward for completion of initiatives that don’t go to plan. If a team fails to deliver wrt the metrics and KPIs outlined in the initial prop, they could recover some of the reward by presenting how and why they didn’t meet the expected goal so the rest of the community can learn & adapt.

I’m not sure if this needs to be an explicit guideline or could emerge as a cultural norm, but I think it could minimize KPI/ metric misalignment.

If something like this is implemented where a team can recover some of the reward by presenting progress to the goal, these guidelines should be explicitly outlined and not left to implicit emergence.

If this isn’t explicitly outlined and left to ad-hoc decisions, the general trend will be towards heavily rewarding failed initiatives. Having all compensation/reward schemas being explicit also helps contributors predict their own outcomes, which increases their engagement and willingness to participate.

1 Like

Yup! I envision the stablecoins cash payments getting streamed and then a pre-defined success fee in tokens paid out if all work was successfully achieved (to be pre-defined).

Yup, stablecoins or ETH if requested etc. (or even fully paid in $SAFE) but open to have the team to decide their compensation here.

100% Agree. I’d like to run the first wave of initiatives and then implement a patch to the process after rather than introducing more upfront complexity here. Want to layer on rules/complexity as needed over time vs. over speculate on the design decisions.


Every 4-month work/review period, the end vision ideally should produce a report card overview of the DAO’s efforts attributed to KPIs. The goal is to start slow and build up the org in a way so that we have become increasingly more efficient and to improve the output of metrics consistently.


Eventually, we’d be able to produce quantifiable metrics for CAC and LTV for safes. Once we do, we can more aggressively deploy capital to drive forward the growth of Safe DAO.


Operating model seems reasonable to me, I appreciate the laser focus on hard quantifiable KPIs in the report card @pet3rpan-1kx

I’d like to add that resources allocation should be structured to facilitate more qualitative objectives. Some goals are harder to define but is of sufficient importance to DAO’s core values. Some of which are:

  • measures of decentralization: decentralized UI, front end diversity, user owning their own data and/or run self-hosted transaction service
  • brand growth and education: important as Safe begins to differentiate from the legacy Gnosis Multisig branding
  • expansion into new use cases: applications that enable new non-financial use cases
  • development ecosystem diversity: pushing plugins development to the edge to grow a healthy engineering community instead of reliance on the core team

Also note: Safe DAO hard metrics is already impressive by industry standard, I’d like to see the resource allocation model focused on growing the pie in the long run- not to mention 4 months review cycle could restrictive to a more moon-shot initiatives


Hello Safe friends,

First of all, I would like to say that I love that blog post !! While reading it I had lots of flashbacks regarding the discussion while we are developing Inverter. Inverter is a funding tool that aims to increase accountability and transparency while reducing administrative effort.

The mechanism of Inverter embraces a very similar approach that this proposal suggests for fund management.

Inverter embraces milestone-based funding. Such that, while creating a proposal, the working team has to set milestones and deliverables. Each milestone will have a budget and duration. The milestone budget amount will only start streaming to the working team once they provide the Proof of Completion document.

As you have stated, “Without established KPIs or accountability, both contributors and funded projects lack clear goals and metrics to work towards.”

One of the novel solutions that Inverter brought here is that the “approval” process of the proof of work document is based on an optimistic approach. Unless there is a disagreement with the document, the document will be redeemed as satisfactory. However, if the community is not satisfied with the provided proof of completion document, they can interrupt streaming and withdraw their funds anytime. This allows a built-in accountability mechanism that does not require many approvals on each milestone which involves lots of bureaucratic work but still protects funders from under delivery of work.

One may be concerned that in DAO daily work activities, it is hard to perfectly define milestones with objective outcomes or there may be under deliveries due to external reasons, etc. However, the idea behind providing proof of deliverable completion documents is to maintain the consent of your funder, not complete all the promises without any compromise. For example, the writing guild may promise to provide 3 articles this month but publish 2. However, if the community believes that those 2 articles were very good and satisfied with their performance. All they need to do is not withdraw their funds, which signals that they are satisfied with the provided work.

Also, inside Inverter, communities can provide feedback on proof of deliverable reports. This may be a suggestion for future reports or a request for a new report or any sort of feedback. That way working teams will have constant pulse checks.

I believe one of the most important value add that Inverter can provide Safe DAO is the organization dashboard that it provides. Inverter provides an organizational dashboard that funding programs can see and get notified about

  • Who is already funding this proposal
  • Who is contributing to this proposal
  • Public discussions about the proposal in community forums
  • The proposal’s target funding goal
  • Established Milestones and Deliverables
  • Proof of Deliverable Completion reports
  • Feedbacks on Proof of Deliverable Completion reports from funder

Being able to navigate this important information for many proposals would help SAFE DAO to have a more transparent and easy time to manage its funding activities.

I am looking forward to hearing your opinion. I believe there is a very interesting value alignmentthere. I would love to invite the Safe community to be one of the early tester of the product in our close beta launch. We would love to have a user interview with Safe DAO members to improve our product based on your needs !!

Here you can check some documentation about Inverter:

Product Doc (Work in Progress): Introduction to Inverter - Inverter Network-Docs (WIP)
Latest Blog Post: The birth of Inverter Network, Team and Feature Updates- Part 2 — Inverter Network


Excellent proposal and discussion - having participated in many DAOs and defi projects for some time, I think you have hit the nail on the head for many DAOs in terms of improving the incentives and focus.


I will be releasing a final form of OBRA to be pushed for community governance over the next month!


I have read this a few times, slowly letting things percolate. Really appreciate you putting this thought out into the world.

On the first read I mostly connected with the idea of pushing initiatives to compete under the umbrella of a strategy, providing specific metrics to evaluate and compare them with.

I’m realizing now, this could actually be a bit broader. I’m very interested in seeing measurable competition between strategies, providing an evaluation of different approaches towards the DAO’s goals, different mental models.

This begs the questions: how are strategies selected, and how are good strategies recognized and rewarded? What if any DAO member could permissionlessly propose new strategies and trial initiatives within them to prove efficacy. Trial strategies would have less rewards and therefore less competition between incoming initiatives. More mature, proven strategies would have higher rewards, fostering intense competition within them. Strategy authors would receive increasing rewards as more successful initiatives are completed under their strategies.

Most importantly, trial strategies would provide innovators a permissionless space to iterate and improve the entire approach to meeting the DAOs goals. If the DAO has clear, measurable goals - a strategy could be evaluated simply as $ spent per unit of goal achieved.

Permissionless strategy creation and competition would seriously minimize the amount of things the DAO needs to agree on. The DAO decides on the Goal, and the most important metrics of success. Strategies and initiatives evolve permissionlessly on their own, forced to compete and deliver measurable success.


@JoelCares, your intuition is correct. The best way to decentrally manage a DAO is to get broader input from the community. Broader input statistically and empirically makes for more robust plans and stronger buy-in. Stronger buy-in on better plans leads to better results.

Separating objectives from the initiatives implemented to achieve those objectives is a great way to open up the process for more input. Let the community first focus on what they want to achieve or what the problems are than need to be solved. Then allow anyone to propose initiatives that they argue will achieve the objectives. Then the community can compare them and pick the ones they think are best. This is the practice of “open strategy” and is a proven best practice.

I know @pet3rpan-1kx plans to post the final proposal soon. OBRA and the report card approach is one way but it collapses goals (objectives) and a single initiative (described as a strategy). If not designed right this can cut off members of the community from producing alternative ways to achieve the desired outcome that may be better than the standalone proposal has.

1 Like

Hi folks – does anyone have an update where the DAO got to with this draft proposal?

I can provide some context. While working on an updated version of OBRA based on the comments here and based on feedback from others, it became clear that there are a few dependencies to resolve before a resource allocation framework like OBRA in its current version can be functional and effective.

For instance, the legal state of SafeDAO and its relation to the Safe Ecosystem Foundation (see also here) currently limits how SafeDAO could operate through OBRA. Within the Safe team, I’m working with legal and finance on a description of the status quo to give everyone a better idea of what these dependencies are.

Also, discussing OBRA within the Safe team led to related discussions around a broader product vision that SafeDAO could choose to pursue, with lots of open questions on details and timeline. From my side, it seems that some level of detail and timeline of these discussions should be defined first, before we define an implementation framework like OBRA (which then can be optimised for a more tangible short-term roadmap on what exactly OBRA will allocate resources to).

Once these discussions reach a point when they are ready to be shared here for discussion, they will be shared here, of course. In this case, OBRA opened a bit of a rabbit hole which just takes a bit longer to get to the bottom of.

These are just my2cents. @pet3rpan-1kx and others, please correct me if I’m wrong or missed something.


Spot on.

I’ve requested that the core Safe DAO team first needs to define the current project direction of the DAO before being able to apply a resource allocation framework. We should understand the current objectives of the DAO first before going forth with a resource allocation system.


More formally, I propose that as a DAO, we first need to clarify the following aspects before we can implement and ratify a resource allocation framework:

  • Product vision – The product vision defines an aspirational end state of a product. It defines what the product will achieve and the value propositions for end users and serves as an anchor point for how we assess the funding of new features, functionality, and products. We should look to establish a product vision based on an understanding of end-user needs and the market landscape. While ideally, we want to set a single canonical product vision, there may also be several product visions we may want to consider, explore or attempt to validate. Nonetheless, they should be initially defined first as a starting point. A product vision establishes the necessary clarity around Safe DAO’s motivations (the ‘WHY’) around tactical goals and surrounding strategy.

  • Product goals – These are specific tactical objectives that help achieve Safe DAO’s product vision. Goals should acknowledge the constraints of our available resources and the existing position/state of Safe DAO. With each product vision, there will be its own respective product goals that are nested under each. Product goals allow us to understand what we need to accomplish immediately (the ‘WHAT’). While typically product goals generally are coupled with a strategy on how to execute them, we want to enable the community to participate in bottoms-up ideation on the ‘HOW’ via OBRA. We expect many of OBRA’s strategy proposals to revolve around our established defined product goals.

  • Product metrics – Product metrics are data points that allow us to establish a picture of the current, past, and future state of Safe DAO’s progress toward goals and its product vision. Metrics aim to guide and inform, not necessarily be optimized at all times. A single metric alone will not provide the full state of reality, nor will a full set of metrics achieve that. Metrics should always be understood via the lens of qualitative assessment. Our initiative proposals should be measured against product metrics as much as possible.

We want to ensure a resource allocation framework supports the execution of the community’s outcomes, goals, and direction as opposed to working against it.


Hey! I’m new here, but after reading through it a few times this looks very promising, I’d love to learn how this works in practice. Thank you for putting this together, along with the governance references.

Perhaps I didn’t look hard enough, but does this say anything about what membership model(s) should be used / could fit well with this? How would one become a member or delegate and gain reputation in this model (if at all)?

1 Like

All initiatives are funded by streamed payments via a tool such as superfluid or sablier unless the sum of funds is < $10,000

You could add substake.fi to the streaming payment options here.

DAO would park some ETH in a substake vault. It uses stETH to generate yield. The initiative can only claim that yield. DAO can pull back the full deposit any time. So it actually doesn’t cost anything to the DAO to fund the initiative.

Introduction: Introducing Substake. A native web3 subscription payment… | by Substake | Medium

I think we do have some thoughts forming on the mission and goals in the post [SEP #4] SafeDAO Constitution. Also, a technical direction along those lines was mentioned here Draft: The Motivation and Future of Safe Contracts.

To add on, I would like to support initiatives in multiple L1’s and as such the metrics should not be defined too shallow.

1 Like

For people like me looking to see progress on this SEP, Please know that a new thread has been opened for this proposal which serves as a revised version

Check it out


Thanks for cross-posting :raised_hands:


This thread is closed as the proposal was resubmitted in an amended form in this proposal.

1 Like