In the realm of DAOs, there is a tension between decentralization such that anyone can participate and the concept of an organisation where there is structured manner of engagement alongside compensation tied to the work carried out.
Contributor incentives must keep that tension in mind. It is not intended to pay everyone who shows up or to manufacture work with the intention to reward participation. Instead, the purpose is to encourage a pipeline of contributors who are willing to engage with open problems facing the DAO and to fairly evaluate whether their contributions warrant compensation.
To understand what a contributor incentive program should achieve, we must first examine the different types of contributors’ engagements within a DAO before diving into the type of open problems they may tackle alongside how to evaluate the contributions.
Type of Contributors Engagement
Generally, contributor engagement can be understood across three categories including employees, consultancy, and freelance contributors.
Employees are individuals who are employed by formal entities such as the Arbitrum Foundation or the OpCo. They receive a salary and scope of responsibility within their respective organisation. It is generally assumed that employees are not qualified for a contributor incentive program as they are already engaged to participate in the Arbitrum ecosystem.
Consultancy is defined as individuals with a structured engagement, scope of work, and compensation that is tied to hours spent or the successful completion of work. The engagement is time-limited, it does not auto-renew, and there is a counterparty that monitors and periodically reviews their progress. The classic example is hiring an external contributor to join a committee and support the execution of a DAO proposal.
Freelance can be defined as having an interest to take on tasks, but there is no anticipated time commitment or a strictly structured engagement that tracks progress. For example, an ad-hoc worker may be willing to organise calls to discuss important issues in the DAO, spontaneously produce a report on a topic of interest, or be willing to provide feedback and lend their support to initiatives the DAO is working on. Going further, they may perform work or exhibit behaviours that extend beyond the DAO that helps push the Arbitrum ecosystem forward.
Focus on Freelance Contributors
We suspect an incentive program should primarily target freelance contributors. They represent a base of contributors where the expected scope of work is not well-defined, it is not covered by any compensation structure, and yet have a unique outsider perspective that may prove to be useful.
It is debatable whether consultancy-style work should be covered by an incentive program. They are often associated with DAO-approved programs with a scope of work, a recruitment process, and a budget to cover its cost. Firestarter programs to kick-start new initiatives may be more appropriate than an incentive program. This pays for concrete and substantial tasks to be scoped alongside a feedback loop to track progress.
This does not imply that freelance contributors cannot do substantial work. There should be funding options available to support this beyond an incentive program. The main point is to design a program that focuses more on rewarding collective action and voluntary work that the community as a whole can justify rewarding in a retrospective manner.
Incentivising Contribution in a DAO Context
We take this opportunity to cover the type of coordination tasks an incentive program will likely need to tackle or at least consider as part of its design. The overarching goal of an incentive program should be:
- Program Goal. Align contributors with behaviours that fosters a collaborative environment and improves the quality of DAO proposals that pass through the governance process.
Put another way, it can be viewed as a rewards program for contributors based on collective actions, with the intention to nudge certain behaviours that are perceived as helping to foster a collaborative environment.
With that in mind, there are four contributions modes that we can categorise for DAOs:
- Operational Stewardship. Keeping the DAO running with recurring calls, recording minutes, producing digests of proposals, etc.
- Program Execution. Allocating specific tasks and open problems to contributors.
- Exploratory Innovation. Encouraging contributors to organically publish new ideas and new problems that the DAO may need to tackle.
- Crowd wisdom. Extracting knowledge and feedback from contributors, especially if they are domain experts, to improve the quality of proposals that pass through the DAO.
We will tackle the first two problems here before focusing more time on exploratory innovation and crowd wisdom as that is generally what we consider to be at the heart of a contributor incentive program.
Operational Stewardship and Program Execution
Care must be taken for deciding how to reward operational stewardship. If incentives are attached to activities such as organising meetings without oversight, then the DAO may end up with ~50 weekly recurring calls and accidentally encourage unnecessary coordination.
We believe there should be a designated owner who decides if a recurring task is needed and whether compensation should be offered. The owner should periodically review the activity and decide if it should continue. We suspect the OpCo is best placed to make decisions on operational related tasks and it may be wise not to include operational-related tasks within the scope of an incentive program.
Program execution is typically accompanied with clearly defined scopes, a recruitment process, and compensation that covers the cost of the contributor. It can be implemented with a consultancy-style agreement from a DAO-approved program or a firestarter grant.
We expect an incentive program not to focus on program execution for well-known problems, but to focus on something a bit more challenging – advocating for the re-prioritisation of neglected problems for which no obvious funding path exists.
The absence of funding is rarely accidental and often reflects how a problem has been prioritised. Mission-critical problems are usually addressed by Arbitrum-aligned entities with dedicated teams and resources. Problems that are recognised as important, but not urgent, may sit in the backlog until human resources become available to tackle it.
What remains are problems that have not yet been fully recognised as important to address. It may be the case that an Arbitrum-aligned entity is not yet convinced that the return of investment justifies the time and money spent tackling it. This is likely to be true for many open problems, but there will be cases when an advocate in the community can present new evidence that demonstrates that an open problem’s priority should be re-assessed.
Put another way, an incentive program is not necessarily well-suited for funding the execution of programs to solve well-known problems. We envision it may be better to focus on rewarding contributors who can convince the community that an open problem matters and it deserves to be prioritised.
Incentivising New Ideas and Voluntarily Impact
This brings us to exploratory innovation which is arguably more important for an incentive program to cover than the execution of open problems.
Historically, contributors who proposed new ideas for the DAO and successfully shepherded them through the proposal process were often issued a reward directly from the proposal itself, which served as a means to validate the worthiness of the idea alongside rewarding the substantial voluntary work required to bring it to approval.
While rewarding contributors directly from a proposal ensured that exploratory ideas were not entirely uncompensated, it created a tension that contributors were required to undertake a disproportionate, often herculean, effort to have their ideas recognised. Moreover, it only rewarded the final act of drafting a proposal rather than earlier exploratory work including research, experimentation, and even sense-making, that enabled the community to evaluate whether the idea is worth pursuing in the first place.
The generation of new ideas alongside evidence for why it should be prioritised is precisely the behaviour an incentive program should encourage.
Accordingly, any mechanism designed to reward exploratory innovation should be selective and quality-driven. We expect the initial work by the contributor to be performed on a voluntary basis and rewards should only be issued where the work:
- Meets a high standard of quality,
- It is clearly justifiable to the wider community,
- Produces tangible next steps.
As we discuss later, we should focus more on the act of recognition than the financial reward attached to it.
Extracting Wisdom From the Crowd
The pursuit of extracting knowledge from domain experts to improve the quality of proposals has historically been tricky to achieve in the DAO.
Early delegate incentive programs, across different ecosystems, focused on rewarding participation such as attending governance calls or providing feedback on a public forum. Some programs offered rewards regardless of the quality of participation. Other programs evolved with a subjective review and points-based system as a means to filter quality from spam. Going further, there is even a desire to seek peer review from the DAO as a whole to rank contributors based on perceived impact.
We argue that these approaches have had mixed results. There is evidence that rewarding participation has attracted contributors who have made a meaningful impact in the DAO including pushing successful proposals through the process, contributors willing to take on consultancy-like tasks and eventually take up full-time employment.
On the other hand, there are several downsides that have emerged:
- Constrained participation formats. Contributors are forced to participate in a fixed-mode of communication such as calls or commenting on the forum, where genuine feedback is often obtained through multiple and different communication channels.
- Metric distortion (Goodhart’s law). When participation itself is financially rewarded, it ceases to be a reliable signal of value. It encourages low-signal activity, including AI-generated comments, which increases noise and becomes counterproductive for others who are seeking to positively contribute, such as excessive noise during the review process.
- Gaming of point-based systems. A filtering system with objective evaluation can be perceived as fair and equitable, but it tends to incentivise optimisation around scoring rather than substance. So much so, there is an overemphasis on point accumulation and regular disputes to maximise points. This has led to negative sentiment for previous programs and increasing mindshare to cancel incentive programs entirely.
- Insufficient domain-specific input. Despite increased participation through incentive programs, there is still a consistent lack of engagement from contributors with relevant domain expertise when reviewing proposals or stepping up as candidates for DAO elections.
Any new incentive mechanism should address the shortcomings of previous programs while continuing to attract meaningful participation in the governance process.
Direction For A Future Incentive Program
We envision that any future program should focus on rewarding a narrow set of contributor behaviours that improve governance outcomes rather than maximizing participation or activity.
Incentives should support targeted outreach by encouraging engagement from contributors who are not yet active in the DAO and attracting individuals with relevant domain expertise. This may involve issuing rewards for work outside of the DAO or introductory incentives that only last a short period of time.
We should generally move away from requiring contributors to push an entire proposal through the pipeline for their contribution to be deemed worthy. Instead, incentives should focus on recognising contributors who have helped the community re-prioritise initiatives, especially where important problems or opportunities were previously unknown or the problems were known but not considered a priority. This should encapsulate a range of activities from research reports, proposal drafts and active lobbying to key stakeholders.
The program should further aim to reduce noise in the governance process with an objective to make the proposal process digestible by domain experts and delegates who are charged with reviewing it. Interactions around proposals should be meaningful and structured in a way that enables domain experts to engage constructively rather than being crowded out by volume-driven participation. Additionally, the design should acknowledge that valuable contributions can emerge through different channels and formats. The program should incorporate mechanisms that can discover and surface contributions regardless if it was published on the forum, received via private communication, or left as a comment on a working draft.
Decisions on whether a contribution merits recognition should be localised and rest with the parties directly impacted by it. It should not be determined through a global approval process or a points-based system. Rewards should be discretionary and non-exhaustive with the intention to offer incentives as recognition for high-impact work rather than guaranteed compensation. In fact, it may be wise to probabilistically withhold payments (i.e., flip a biased coin), to further encourage contributors with an intrinsic motivation to participate. This may help change the question from “how much does this pay?” to “is this worth doing?”.
Altogether, a future incentive program should foster an environment that attracts new talent to participate in the DAO, meets contributors where they already operate, and surface valuable work without forcing participation into certain formats. Most importantly, it should start small in scale, with the ability to pivot its mechanisms when it fails to reward an important contribution, or worse, it rewards participation that the DAO as a whole does not deem justifiable.
Should There Be An Incentive Program At All?
We have mostly focused on the type of contributions that might be covered and a potential direction for a future program. However, there is an overarching question that should be answered, which is whether an incentive program is needed at all. Let’s take this opportunity to explore the different arguments.
The primary argument for an incentive program is that contributing to a DAO takes time, effort, and specialised expertise. Fair compensation can ensure that participation is not limited to those with excessive free time, but also attract domain experts whose insights would otherwise be inaccessible. Incentives can also lower the barrier to entry and entice participation from people who are not actively involved in the DAO. It is hoped that this enticement may help new contributors stick around and contribute further or act as a recruitment funnel for consultancy-style work or even full-time employment. Finally, there is thankless work undertaken by contributors that often goes amiss with no reward or recognition in which a dedicated program can fix.
On the other hand, the argument against an incentive program is that contributors should participate for reasons beyond an immediate financial reward. Contributors may be large token holders with a long-term view, projects that can benefit from proposals passed, or simply participate for the satisfaction of witnessing the direct impact of their actions. Additionally, the size of rewards an incentive program can reasonably offer may not be meaningful enough, especially for delegates representing companies, to justify the overhead of participation.
A further challenge lies in the evaluation of contributions itself.
Assessment can be messy and lead to public disputes. It can miss recognizing certain contributions and lead to unintended resentment. Poorly designed metrics risk triggering Goodhart effects, attracting contributors who optimize for rewards rather than impact. Evaluation also requires time and a dedicated program manager which introduces perception of gate keeping, agenda-setting or bias, and may discourage dissent or risk-taking among contributors.
Overall, there are merits to both positions. This reinforces the importance of clearly defining what behaviours an incentive program is intended to reward. Many disagreements around prior programs arguably stemmed from a narrow focus on rewarding participation and the use of global one-size-fits-all evaluation frameworks. A different design approach may alleviate some of the concerns raised to date.
That said, it remains entirely possible that a period without an incentive program, a deliberate detox, could demonstrate that other motivating factors for contributors to participate are sufficient. It may be the case that an incentive program, alongside all the work involved, may not be necessary at all.
Summary
An incentive program should align contributors with behaviours that fosters a collaborative environment and improves the quality of DAO proposals that pass through the governance process.
Any such program should primarily focus on freelance contributors whose work is not covered by an existing compensation scheme. It should not fund substantial, consultancy-like engagements that are better served by firestarter grants or DAO-approved programs, nor should it cover ongoing operational tasks that require clear ownership or periodic review.
There are specific behaviours that may be suited to an incentive program including:
- Contributors who can help the DAO re-prioritise open problems that should be tackled,
- Offer meaningful and impactful feedback on proposals that pass through the DAO,
- Recognise thankless work undertaken by contributors that can easily be justified to the wider DAO that it should be rewarded.
These behaviors are typically unstructured by nature and often performed by contributors on a voluntary basis. We expect the rewards to be retrospective in nature and often performed because of an intrinsic motivation by the contributor rather than the expectation of a reward.
In terms of design, a future program should take a vastly different approach to historical programs. It should avoid prescribing specific participation formats, support multiple modes of engagement, and rely on localised evaluation by those directly impacted by a contributor rather than a global, points-based system. It may be wise to implement probabilistic or artificial withholding of rewards to encourage contributions from people who are motivated to participate over maximum reward seeking.
Finally, we recognise that an incentive program may not be necessary at all. With no contributor incentive program currently in place, the DAO has effectively entered a detox period. We should take this opportunity to observe participation and whether existing motivations to participate are sufficient. If participation remains the same, or improves, then that will act as evidence that we should not implement a new program.
—
Keep in mind, everything in the above represents my own thoughts on a contributor incentive program and not representative of others at the Arbitrum Foundation (and more generally, Arbitrum aligned entities).
It is important to at least discuss the merits of a new incentive program, how it should be designed, and of course, whether it is needed at all. With that in mind, I’d invite everyone to read this post, drop your thoughts, and see how this discussion moves forward.