Proposal: Revert the Delegate Incentive Program (DIP) to Version 1.5

[Instinct]

Proposal: Roll back the changes introduced in the Delegate Incentive Program (DIP) v1.6 and v 1.7

A Call for a Fairer, More Effective, and Accountable Arbitrum DAO
.

Note: This proposal has been revised based on community feedback. The goal is no longer to sunset the DIP, but to restore the fair and effective framework of v1.5.

TL;DR

This proposal calls for rolling back the detrimental changes introduced in the Delegate Incentive Program (DIP) v1.6 and v1.7. We advocate for a return to the v1.5 framework, which was previously approved by an on-chain vote and established clear, objective KPIs for delegate performance.

Last update: 04/09/2025

After in-depth analysis and considering community feedback, it's clear there is strong sentiment to continue the DIP. However, the program requires significant changes. 
The drop in quality and participation can be traced directly to the unsanctioned and poorly conceived changes introduced by @SEEDGov since version 1.6.

The proposal was edited to make the DIP more inclusive rather than sunsetting it.

The Core Issue

The Delegate Incentive Program (DIP) was created to encourage active and thoughtful governance participation. However, recent versions have distorted its purpose. Instead of fostering a healthy ecosystem, the changes introduced in v1.6 and v1.7 have made the program unfair, unstable, and subjective.

The program’s managers (PMs) at @SEEDGov have consistently ignored community feedback, applied rules retroactively, and created a system that penalises smaller delegates while failing to hold large, inactive ones accountable. Their proposed “fix” (v1.7) only exacerbates these problems by raising the entry barrier tenfold, cutting delegate compensation by 40%, and requesting a 25% raise for themselves.

The current version of the program is broken, its management has lost the community’s trust, and it is doing more harm than good. It’s time to restore the program that worked.

This proposal calls for a complete rollback of the DIP to the v1.5 framework.

The Problem: A System of Patronage, Not Performance

The DIP is no longer a fair system of rewards. It has become a system of patronage, where payments depend on the personal, opaque judgment of the Program Managers, not on clear, objective work.

1. The Rules Are Subjective and Unfair

The program has shifted from a predictable system to a “black box” where the PMs decide what has “quality” and “impact.”

• Guessing Games: Previously, delegates got points for explaining their vote (CR). Now, that clear rule is gone. Your explanation only gets points if the PM decides it was “impactful.” When delegate Paulo Fonseca pointed out that this was a “conceptually dangerous” metric that could reward bad behaviour, the PMs offered no clear definition, admitting it could only be judged “in hindsight.” This forces delegates to guess what the PMs want, which is not a healthy way to run a DAO.
• A “Caste System” for Delegates: The program introduced a “VP Multiplier” that punishes smaller delegates . If you have less than 4 million ARB, your score is automatically cut by up to 20%. @paulofonseca correctly called this a “sort of caste system,” and other delegates labelled it a “penalty” that makes it nearly impossible for new, hardworking delegates to get fairly compensated.

2. A Pattern of Mismanagement and Unprofessionalism

The program’s management has been marred by constant failures, making it impossible for anyone to rely on.

• Constant, Retroactive Rule Changes: The rules change almost every month. Worse, the PMs have applied major changes retroactively, in the middle of the month. As one delegate put it, it’s like “changing the rules of the game, literally in the middle of the game.”
• Late Results and Ignored Feedback: The monthly results are almost always late, and payments have taken as long as two weeks to be signed. When Paulo Fonseca repeatedly asked for proactive communication about these delays, he was met with dismissiveness. This lack of professionalism shows a fundamental disrespect for the delegates’ time and financial stability.
• Overstepping Authority: The PMs have suspended delegates for things they said on Twitter. This is a serious overreach of their power. As multiple delegates argued, their job is to manage the incentive program based on forum activity, not to police social media.

The Evidence: The Damage Is Real and Measurable.

These aren’t just opinions. The negative impact of the DIP is clear from the data and the words of the delegates themselves.

The Numbers Don’t Lie

The moment the unfair v1.6 rules were introduced, the number of delegates getting paid fell off a cliff.

Month	DIP Version	# of Paid Delegates	Total Payout (USD)
December 2024	1.5	49	$262,228
January 2025	1.51a	32	$205,534
February 2025	1.6 (New Rules)	21	$90,244
March 2025	1.6	25	~$100,000

Source: Figures discussed in delegate chat logs and forum posts.

The number of compensated delegates was cut by more than half as soon as the new rules took effect. One delegate calculated that 10 out of the 11 delegates who were disqualified in February were “small delegates” hurt by the VP multiplier.

The Program Fails Its Supporters’ Goals

Even delegates who supported the program’s goals, like L2BEAT, have seen it fail. @krst Urbański of L2BEAT stated the DIP should help top delegates afford staff and reward passionate, smaller delegates. Version 1.6 did the exact opposite: it punished smaller delegates and failed to hold large, inactive delegates accountable for their lack of participation. The program has failed to achieve the very outcomes that its key supporters envisioned.

The Program Fails Its Own Approved KPIs.

The original v1.5 proposal, approved on-chain, included specific KPIs set by @SEEDGov. The program is now failing to meet them.

@paulofonseca

As a reminder, these were the approved KPIs of the DIP, as per the onchain approval of the original DIP 1.5 version.

@SEEDGov

KPIs

In this new iteration of the DIP, we aim to establish the following KPIs:

• Achieve that 50 delegates receive incentives.
• Engage 100 delegates in the program.
• Achieve an average Total Participation (TP) of 80% among participants in the program within six months.
• Introduce improvements to the program after six months.

In the replies to this proposal, Paulo Fonseca has done a deep dive into the KPIs @SEEDGov hosted during the v1.5 vote

Delegates Are Losing Confidence

The most experienced delegates are walking away because the program is unstable.

“We’re now reducing involvement as we can’t know what (if any) compensation we will receive. Working up front without knowing if or how much you’ll be paid isn’t sustainable .”
— cr1st0f, representing ACI, a major protocol delegate

“Most professional delegates would prefer some baseline/bottom line assurance. No bottom-end confidence until two weeks after month-end isn’t consistent enough for many delegates.” — DonOfDAOs, delegate

The Final Straw: A “Fix” That Makes Everything Worse

After months of criticism, the PMs proposed an update (v1.7) that confirmed their disconnect from the community. Instead of fixing the problems, they doubled down on them:

1. More Elitism: They proposed raising the minimum VP for new delegates from 50k to 500k, effectively closing the door on new community members.
2. Pay Cuts for Delegates, Raises for Managers: They proposed a ~40% pay cut for all delegates while simultaneously asking for a 25% raise for themselves.
3. No Accountability: They proposed removing the ability for delegates to dispute subjective scoring, cementing their own power.

This proposal is not a good-faith effort to improve the program. It is an attempt to silence critics, entrench power, and reward the program’s managers at the expense of the delegates.

The Solution: Roll Back to v1.5

The community feedback has been clear: the DIP should continue, but not in its current broken state. The predatory changes from v1.6 and v1.7 must be removed. Rolling back to v1.5 will restore a stable, predictable, and inclusive program that allows delegates of all sizes to contribute meaningfully to the DAO.

@SEEDGov

At SEEDGov, we remain committed to continuously gathering community feedback and making the necessary decisions to ensure that Arbitrum DAO has the best Delegate Program in the space.

We ask that @SEEDGov honours this commitment and respects the community’s feedback by supporting this rollback.

By voting to revert to v1.5, we will:

1. Restore a More Effective Program: The v1.5 framework was more cost-effective, with a higher percentage of the budget going directly to a larger number of active delegates, rather than being concentrated among a few or covering bloated administrative costs.
2. Encourage Authentic Participation: Version 1.5’s clearer, more objective criteria rewarded consistent, good-faith participation. It provided a reliable foundation that encouraged delegates to invest their time and resources in governance.
3. Restore Fairness and Transparency: Reverting to v1.5 eliminates the unfair “caste system” and the subjective “black box” scoring. It re-establishes a level playing field where all delegates are judged by the same transparent standards.

Voting

Voting Strategy: Ranked choice voting
Voting Privacy: Shielded Voting
Voting options:

1. Sunset the DIP
2. Revert the DIP to v1.5
3. Keep the current version
4. Abstain

If the proposal passes and changes the status quo, the changes will apply to the last 3 months of the DIP: August, September, and October

[paulofonseca]

What a timing for a proposal like this!

Actually, I’ve been looking into the financials of the DIP 1.5/6/7 program and collected this data that can be checked here: DIP Accounting - Google Sheets

[SEEDGov]

Hi everyone,

We’d like to share a statement to clarify several claims made in this post:

@Instinct

The Delegate Incentive Program (DIP) was created to encourage active participation in governance. However, the program has failed. Instead of fostering a healthy ecosystem, it has become unfair, unstable, and subjective.

We don’t see any arguments supporting the claim that the program has “failed.” Labeling it as unfair is, ironically, a subjective assessment in itself.

As for instability, it’s worth noting that the program underwent minor changes in January, more significant updates in February, and then remained consistent through July. In August, the DAO approved (with over 150M votes in favor) the modifications that came into effect for the month.

On the matter of subjectivity, it’s important to remember that version 1.5 was approved with the explicit intention of introducing subjective elements to the assessment. So we find it inconsistent to point to subjectivity as a flaw, given that it was a deliberate and DAO-approved design choice.

@Instinct

A “Caste System” for Delegates: The program introduced a “VP Multiplier” that punishes smaller delegates . If you have less than 4 million ARB, your score is automatically cut by up to 20%. @paulofonseca correctly called this a “sort of caste system,” and other delegates labeled it a “penalty” that makes it nearly impossible for new, hardworking delegates to get fairly compensated.

First, the Voting Power Multiplier (VPM) doesn’t work as described. It follows a linear reduction model — below 4M ARB, the voting score begins to decrease proportionally. This change was introduced in v1.6 and later ratified in the v1.7 vote.

Second, the claim that the program creates a “caste system” or makes it “nearly impossible” for smaller/newer delegates to qualify is not supported by data. Looking at July’s results:

• 9 out of the 22 qualifying delegates had less than 500,000 ARB delegated.

• 4 of the top 5 scoring delegates — Tempetechie, Tekr0x.eth, StableLab, and Lampros DAO — are all under the 4M ARB threshold (and also under 500K).

It’s evident that some delegates have found ways to consistently add value to the DAO through various activities — and were recognized accordingly. Even delegates who don’t usually qualify, such as Hawheik and TodayInDeFi, earned spots in July due to improved contributions.

@Instinct

Late Results and Ignored Feedback: The monthly results are almost always late, and payments have taken as long as two weeks to be signed. When Paulo Fonseca repeatedly asked for proactive communication about these delays, he was met with dismissiveness. This lack of professionalism shows a fundamental disrespect for the delegates’ time and financial stability.

We’re unsure when Paulo was met with “dismissiveness.” Here is the message we sent in response to his concerns:

Regarding the July results, we expect to publish them between today and tomorrow. This month we were in a position to publish on the 15th, but an indexing error in the dashboard caused a two-day delay, and then the weekend followed.

With respect to the delay in payments, we would like to clarify that we are only the proposers; it is not our responsibility to sign or execute the transactions, so we are not the appropriate party to answer that question.

It is also worth noting that the 15th has always been an estimated date. It’s difficult to accurately predict what will happen over the course of the following year at the time of drafting the proposal. On more than one occasion, we have preferred to take additional time in order to improve the quality of the analysis rather than rush it just for the sake of being “on time.” It should also be taken into account that in February the individual reports were introduced, which require significant additional effort.

That being said, here are the delivery dates:

• November 2024: December 12th

• December 2024: January 14th

• January 2025: February 19th (staff was on holiday, as it is summer in Argentina)

• February 2025: March 18th (introduction of individual reports and v1.6 changes)

• March 2025: April 18th

• April 2025: May 14th

• May 2025: June 16th

• June 2025: July 23rd (ETHcc clearly impacted the timeline, and we notified the community of this in advance)

  

Except for June, we have never been more than four days past the expected release date — and in some cases, we delivered ahead of schedule.

It’s also worth noting that we do not control the Safe multisig that holds the DIP budget — this is publicly verifiable. We’ve been transparent about the causes of any delays, and we took Paulo’s feedback seriously. Also, we publicly committed to announcing any foreseeable delays in advance moving forward.

@Instinct

The Program Fails Even Its Supporters’ Goals

Even delegates who supported the program’s goals, like L2BEAT, have seen it fail. @krst Urbański of L2BEAT stated the DIP should help top delegates afford staff and reward passionate, smaller delegates. Version 1.6 did the exact opposite: it punished smaller delegates and failed to hold large, inactive delegates accountable for their lack of participation. The program has failed to achieve the very outcomes that its key supporters envisioned.

This seems to speak on behalf of krst from L2BEAT. We don’t recall him stating that the program has failed. Perhaps it would be more appropriate for each delegate to speak for themselves.

@Instinct

Delegates Are Losing Confidence

The most experienced delegates are walking away because the program is unstable.

This is another broad and unsupported claim. ACI is returning to the program through APE — a delegation that will now be able to qualify using the AAVE community wallet, which holds one of the highest VP amounts.

The alleged “loss of confidence” is hard to justify, given that just two weeks ago, the DAO overwhelmingly approved version 1.7 of the program.

@Instinct

The Final Straw: A “Fix” That Makes Everything Worse

After months of criticism, the PMs proposed an update (v1.7) that confirmed their disconnect from the community. Instead of fixing the problems, they doubled down on them:

1. More Elitism: They proposed raising the minimum VP for new delegates from 50k to 500k , effectively closing the door on new community members.
2. Pay Cuts for Delegates, Raises for Managers: They proposed a ~40% pay cut for all delegates while simultaneously asking for a 25% raise for themselves.
3. No Accountability: They proposed removing the ability for delegates to dispute subjective scoring, cementing their own power.

This proposal is not a good-faith effort to improve the program. It is an attempt to silence critics, entrench power, and reward the program’s managers at the expense of the delegates.

Once again, the claim that we’re “disconnected from the community” is contradicted by the fact that the 1.7 proposal was passed through a DAO-wide vote. We see this as an attempt to delegitimize the decision made by the community.

It’s also important to highlight that, despite the budget reduction introduced in v1.7, some delegates who previously received no compensation will now qualify under the new Tier X. For instance, Ultra, who has an impeccable voting record, was previously ineligible due to scoring below the 65-point threshold. Delegates like Ultra — and others we won’t name here — will now receive recognition and compensation for their consistent commitment to voting. We encourage a more careful evaluation of the actual outcomes of these changes as we move into the August results, considering that not even a month has passed since their approval, and we have yet to see how it works in practice.

Finally, we want to reaffirm that we’re actively working on Version 2.0 of the program, which will address certain design concerns and significantly reduce administrative overhead by simplifying the structure.

Throughout this entire process, our DMs have always been open. We’ve spoken with a wide range of delegates, stakeholders, AAEs, and other community members to gather feedback. We invite the author of this post to reach out directly as well — we’re happy to engage in conversation as we move toward the next iteration of the program.

[Hawheik]

@SEEDGov

Even delegates who don’t usually qualify, such as Hawheik and TodayInDeFi, earned spots in July due to improved contributions.

As I have been referenced here I will point out that the DIP v1.7 proposal was a significant contributing reason, though not the only reason, as to why I will no longer participate as a delegate going forward.

[DonOfDAOs]

I’ve discontinued participation in DIP explicitly because of program management. Frankly, and I say this only partly facetiously, I’d rather see a community-configured AI assess contributions or a literal lottery than whatever the current black box is. Or both, meet x criterion, you’re entered in the lottery. Gain entries from an AI-assessed quality bonus, more for being a large delegate, etc… Simple, easy, everyone knows the rules, each month you get what you get, and you don’t get upset. And, a fraction of the admin costs.

[0xTALVO.ETH_MTY]

Tbh, I’ve also felt discouraged from continuing with the DIP given the recent changes, that’s why I’m no longer active / sold my ARB

[dragonawr]

While I don’t agree with every argument posed by the OP, I do back some of his criticisms of the DIP. Specifically things like the retroactive enforcement of 1.6 version of the DIP, without any sort of reasoning beyond “just because”, which I did take issue with and seemed to impact overall Delegate participation across the forum.

I understand the program has to be fine tuned for efficiency and payouts should indeed be tied to meaningful contributions that at least elicit some form of debate within the DAO; however, low-voting powers delegates have a huge hill to climb that’s only continued to grow since the 1.6 version of the program.

Voting Power Multiplier is indeed a handicap, and while low-voting power delegates can qualify for compensation within somewhat reasonable standards, those with higher voting power can get away with the bare minimum or no contributions at all. I don’t see why we should pretend this is not the case.

The DIP 1.7v, while introducing much-justified changes in compensation amounts -for example-, only amplifies most of these issues.

@SEEDGov

The alleged “loss of confidence” is hard to justify, given that just two weeks ago, the DAO overwhelmingly approved version 1.7 of the program.

Lastly, I do find it disingenuous to point out the approval of DIP 1.7v as a sign that these changes are not unpopular, as we all know how voting power and the DAO work, and you can literally see a bubble map of how voting plays out. I would not be surprised to see more and more low-voting power delegates abandon the program and sell their ARB as a consequence.

I cannot say if this is the correct or wrong way to move forward, as that pretty much depends on what you or the program managers think should be the DIP’s objectives.

Personally, I won’t be moving away from the program as I still see value in it as a way to stay up to date with all things concerning Arbitrum, but I can see how the financials, time investment and subjective parameters would simply force out most non-Tier X delegates.

[cp0x]

It is interesting that I also summed up the results of the program today, but you already know that I was a constant critic of it and the last times I voted against the proposed changes
You can read my thoughts https://x.com/cp0xdotcom/status/1958978890258362874

And regarding the incentive system - I also have many thoughts on how to improve it all and I wrote many suggestions to SeedGov, hoping to use them
Apparently, it will be necessary to change the system radically after the end of this program in October

[Ignas]

What is the mission of Arbitrum DAO? To attract as many opinions and find creative solutions to hard problems?

Or to centralize and make DAO decisions as efficient as possible?

If the first option is important, then the DAO should reconsider a more inclusive campgin.

If efficiency matters, then dropping all incentives makes sense.

Personally, I haven’t received rewards since Feb, but still continued to work on the DAO.

I do feel that the incentives rules are too subjective scoring.

Without the VP multiplier penalty, my scores (from Feb to Jun) would consistently place me in tier 3. By July, it became clear that smaller delegates have no chance of being treated fairly.

And the data proves this isn’t just personal: the number of incentivized delegates has collapsed from 48 to only 22 since the controversial February changes. Participation rates also dropped sharply, from 72 in January 2025 to just 57 in July 2025.

The program has created a uncertainty where delegates cannot plan or commit resources. Instead of encouraging participation, it punishes the very people who are trying to contribute.

Maybe that’s the goal of it?

[karpatkey]

We recognise the significant time and effort required to contribute meaningfully as a delegate in Arbitrum governance. Given the scale of activity in the DAO, we believe there is clear value in providing compensation to delegates who consistently deliver quality contributions.

At the same time, we appreciate that designing an effective incentive program is inherently challenging. Striking the right balance between rewarding delegates, encouraging broad participation, and improving the DAO ecosystem is not an easy task.

For this reason, rather than fully sunsetting the program at this stage, we would like to see an alternative approach to delegate rewards proposed alongside this one. Having both options available would allow the DAO to weigh the trade-offs and collectively decide which path is best for sustaining long-term, high-quality governance.

[Saurabh]

The following reflects the views of GMX’s Governance Committee, and is based on the combined research, evaluation, consensus, and ideation of various committee members.

We align with Karpatkey’s perspective that both the DAO and its governance have evolved significantly. Seed Latam has invested substantial time, effort, and resources into improving this program and they have always been transparent about it. Also, one thing to note the program will end soon. In light of that, we believe sunsetting the program may not be the best course of action. Instead, we recommend exploring alternative paths forward and collaboratively working toward a feasible solution.

[Federico]

Participating in Arbitrum governance and voting with diligence is time intensive, and the DIP has been a meaningful incentive to encourage active participation. There’s real value in compensating engaged delegates, especially if the aim is to activate more voters and reduce quorum issues. That said, after the recent 40% cut, I worry the incentives may no longer be enough (especially for larger delegates) to justify the hours spent.

I also share the concern others have raised around the subjective criteria. While well-intentioned, it seems to nudge people toward formulaic forum posts that check the boxes, rather than adding genuine signal. This makes it harder to cut through the noise and ultimately hurts discussion quality.

I really appreciate SeedGov’s transparency and willingness to iterate here. I’m not in favor of sunsetting the DIP at this stage, but I’d like to see future versions lean more toward clarity and simplicity (e.g., by focusing primarily on voting activity) if the main goal is to activate larger delegates.

[404DAO]

Thank you for starting this discussion. Before diving in, we want to acknowledge the significance of the DIP and the work @SEEDGov has put into designing and iterating the program through governance’s evolution.

As a disclosure, we are current DIP participants. Over multiple iterations, we’ve adapted our contributions even as the program has become more demanding and compensation has been reduced.

If the DAO wants to attract and retain quality delegates and contributors, it must be intentional about how it incentivizes them—both financially and non-financially. Any future delegate incentive program should consider not just return on voting power (ensuring large VP delegates remain engaged and quorum is reached) but also return on people. This latter category includes both delegates and contributors, and today’s program doesn’t clearly differentiate between the two.

As @JoJo notes in [Constitutional] AIP: Security Council Election Process Improvements

@JoJo

As of today, the DAO is “passive” in every aspect.

• we don’t source protocols to participate to grant and incentive programs (as far as I know, I might be wrong with the recent example of the DRIP but is just too new to evaluate and there are not too many info around it)
• we don’t source people/member
• we wait for someone to show up, maybe after single initiatives of members of AF/OCL or personal efforts.

This is one of the big things that we need to change, in general. Not only in the security council. Otherwise we risk to slip into irrelevance.

On a similar note, we would like to highlight @krst’s comment in Proposal - Updates to the DIP, The Complete 1.7 Version:

@krst

As we mentioned during the DIP call, this statement so far is rather questionable. Almost all initiatives that are ongoing right now in the DAO originated from DAO contributors and delegates’ initiatives…

• There would be no DRIP if not for STIP and later variations, DETOX discussions, ARDC incentives analysis, and various proposals (like the IOSG wake-up call proposal) coming from the DAO to improve on the incentives.

We echo this sentiment and encourage SeedGov to work closely with the DAO to design mechanisms that continue to properly incentivize delegates and contributors—whose initiatives have been foundational to the DAO’s success—while still fulfilling their role as a program manager in ensuring expenses remain prudent and sustainable.

[paulofonseca]

I’ve been doing a little bit of digging into the DIP data, and I wanted to share what I found in more detail.

As a reminder, these were the approved KPIs of the DIP, as per the onchain approval of the original DIP 1.5 version.

@SEEDGov

KPIs

In this new iteration of the DIP, we aim to establish the following KPIs:

• Achieve that 50 delegates receive incentives.
• Engage 100 delegates in the program.
• Achieve an average Total Participation (TP) of 80% among participants in the program within six months.
• Introduce improvements to the program after six months.

Then, with DIP 1.7 that was approved with an offchain vote, the KPIs changed for the remaining 3 months of the program, as can be seen here with the introduction of new KPIs for DIP 1.7.

Either way, we should look into the performance of this delegate incentive program, and therefore of @SEEDGov as it’s program manager, as per the KPIs they themselves defined in October 2024, that are quoted above from the original proposal.

So during the first 9 months of the DIP, taking into account the results from November 2024 to July 2025, here is the performance data of the DIP:

KPI #1 – Achieve that 50 delegates receive incentives – FAIL

I’m assuming that this means 50 delegates receiving incentives, per month, and not 50 unique delegates in total for the duration of the program. Since the program was originally designed and funded to give incentives to up to 50 delegates per month, let’s assume it’s 50 delegates per month.
As I shared with the preliminary data in the post above, this KPI was not achieved.

The average number of delegates getting incentives, for the past 9 months, is 31.89, not 50. And the trend is not going up, as can be seen in the chart.

KPI #2 – Engage 100 delegates in the program – FAIL

I’m assuming that this means 100 unique delegates opting-in and qualifying to receive incentives, per month, for the duration of the program.
As can be seen, this KPI was not achieved.

The average of the number of unique delegates opted-in and qualifying, for the past 9 months, is 63.44, not 100. And the trend is not going up, as can be seen in the chart.
If, on the other hand, we consider that this KPI pertains to the total number of delegates that ever applied to the program in the official forum thread for that effect, up until the end of June (to be able to qualify for the July results) that cumulative number is 84 delegates, which is still not 100 unique delegates engaging with the program.

KPI #3 – Achieve an average Total Participation (TP) of 80% among participants in the program within six months – FAIL

This is the KPI that is the most important, in my opinion, for justifying Arbitrum DAO investing on Delegate Incentive Program, since this is a good proxy for quality participation of delegates in the DAO and overall engagement in the program.
As can be seen, this KPI was not achieved.

Total Participation was never at an average of 80% during the program. At the 6-month mark, the average TP was 62.16% and at the 9-month mark, it was at 56.55%, not at 80%. And the trend is not going up, as can be seen in the chart.

KPI #4 – Introduce improvements to the program after six months – FAIL

This is obviously a subjective KPI, and changes were indeed introduced even before the 6-month mark, and even retroactively. Unfortunately, none of those changes improved the program, as can be seen by the performance of the previous KPIs. Therefore, I consider that this KPI was also not met.

Conclusion

Running a Delegate Incentives Program is indeed challenging, and @SEEDGov has been iterating and changing the program to suit the changing landscape of the DAO. However, the changes introduced, retroactively, with DIP 1.6 in February 2025, clearly led to all the self-imposed performance metrics starting to trend down. This shows, that the design and management of this program by SEEDGov, since the introduction of the 1.6 version, was clearly not effective, according to the KPIs that were defined by themselves, that were never met during these 9 months of the DIP 1.5/6, and started abruptly trending down in February (the first month when the 1.6 changes were put into effect). Specifically, I believe the introduction of the officially named Voting Power Multiplier (which is not a multiplier at all, but is in practice a penalization for small delegates since it multiplies their voting power by a number smaller than 1 if they have less than 4M ARB delegated to them) changed the very nature of the program, and should be reverted.

Disclaimer: I’ve been a heavy participant in the program up until I was explicitly penalized in the May results. For 7 months, I’ve disputed results, I’ve offered feedback and suggestions for improvements, and I’ve now cross-checked all DIP transactions (they are all correct) and overall data, to better understand what happened with this program, and offer this analysis to the community. I’ve done this because I believe that Arbitrum DAO needs an independent analysis of the DIP that his not done by the Program Managers, which are obviously biased. The fact that, up until now SEEDGov have not reported on the progress of the KPIs that they themselves defined at the beginning of the program, is shocking to me. So that’s why I did this analysis. Please cross-check my data and give me feedback on this analysis. Thank you!

[baer.eth]

Gm @Instinct,

Thank you for putting this proposal forward. You’ve done an excellent job of capturing the deep frustration that many of us feel with the DIP in its current state.

As we discussed in DMs, I wanted to bring some of my feedback here to the forum to share with the wider community.

@Instinct

Proposal: Sunsetting the Delegate Incentive Program (DIP)

While your post correctly identifies the problems, my sense from speaking with other delegates is that the primary desire is not to eliminate the program, but to reform it into something that is fair, predictable, and genuinely serves the DAO.

Based on these conversations, here is what a successful and supported program would look like:

• 1. Predictability and Stability: Delegates need a system with clear, objective rules that are not constantly changing. We need to know that our effort will be rewarded based on a transparent framework, not on the shifting, subjective opinions of a Program Manager.

• 2. Rewarding “Skin in the Game”: The program should place a higher value on tangible contributions and commitment. This could mean higher minimum VP or ARB holdings, but more importantly, it means rewarding delegates who actively lead community initiatives, like the work done by @Tekr0x.eth , @TempeTechie and many others (ergo, non FDN/AAE lead initiatives but directly by the community.)

• 3. Reduced Subjectivity and Transparent Enforcement: The “black box” of subjective scoring must end. Rules should be applied consistently and transparently to everyone, without private enforcement or decisions that feel arbitrary.

My concern is that a vote to simply sunset the program is a vote without a clear plan for what comes next. It leaves a vacuum and doesn’t address the underlying need to incentivize high-quality participation. this has also been echoed my multiple delegates on the forum.

Perhaps a better path forward would be to work together with delegates like @paulofonseca or Us (@daoplomats) on a counter-proposal that outlines the principles of a DIP v2.0 one built on the foundations of predictability and fairness that the community is asking for.

[Instinct]

Gm everyone,

First, I want to extend a sincere thank you to everyone who has contributed to this discussion. The level of thoughtful feedback has been incredible and is a testament to the strength of this community.

---

A special thank you to @paulofonseca for his detailed, data-driven analysis. His work provides the objective proof that this program has failed to meet its own self-defined KPIs.
This data is the core evidence that prompted my proposal and shows that the issues we feel are not just subjective; they are measurable failures.

I also want to acknowledge the powerful testimonies from delegates like

@Hawheik , @DonOfDAOs , @dragonawr , and @Ignas . Their statements that they are leaving or have been pushed out by the program’s mismanagement and unfairness highlight the real human cost of the current system. When a program drives away talented and dedicated contributors, it has failed in its most fundamental mission.

The most important theme in this discussion is the question of

“Reform vs. Sunset,” thoughtfully raised by @baer.eth , @karpatkey , and @404DAO . Your concern that my proposal lacks a clear plan for what comes next is entirely valid, and I would like to clarify my intent.

My call to “sunset” the program was not meant to suggest an end to delegate incentives forever. It was a call for a necessary clean break. The history of this program, especially the flawed v1.7 “fix” (which cut delegate pay while increasing the manager’s, raised the entry barrier 10x, and removed accountability), has shown that incremental reform under the current structure is not working. We cannot build a strong foundation on a flawed one.

Therefore, I fully support the suggestion from @baer.eth. The productive path forward is to build a new program from the ground up, based on the principles we are all aligned on.

@paulofonseca have announced in the community call the other day that he is already working on such an initiative, let’s get this second phase of the proposal rolling.

---

Whether this vote to sunset the current program passes or not, the message from the community is clear: the status quo is unacceptable. I am happy to work with everyone here to build the replacement that Arbitrum DAO deserves.

[Instinct]

Gm @SEEDGov,

Thank you for your statement. While you have offered a defence of the program, it is contradicted by both the program’s own performance data and the direct testimonies of numerous delegates in this thread. To ensure a fair discussion, I want to clarify the record by presenting the evidence shared by the community.

@SEEDGov

We don’t see any arguments supporting the claim that the program has “failed.” Labeling it as unfair is, ironically, a subjective assessment in itself.

Let’s unpack your first statement
The Evidence: The program has failed to meet all four of its own original, DAO-approved KPIs . Delegate @paulofonseca’s detailed analysis shows a clear, data-driven failure:

@paulofonseca

KPI #1 – Achieve that 50 delegates receive incentives – FAIL

@paulofonseca

KPI #2 – Engage 100 delegates in the program – FAIL

@paulofonseca

KPI #3 – Achieve an average Total Participation (TP) of 80% among participants in the program within six months – FAIL

@paulofonseca

KPI #4 – Introduce improvements to the program after six months – FAIL

KPIs are something used to measure how good we are in reaching our overarching goal, in this case, ‘encourage active participation in @Arbitrum DAO governance’

A program that fails to meet any of its own success metrics can only be described as a failure. - Baer.eth

Why you should not use the newly proposed KPIs

1. Focus on Elites vs. Broad Participation

Original Goal: To be inclusive by engaging 100 total delegates and paying 50 delegates per month.

• New Goal: To focus exclusively on a small group of high-VP delegates (those with over 500k ARB)

Financial ROI vs. Community Health:

• Original Goal: To measure the health of the delegate community by achieving an 80% average Total Participation score.

• New Goal: To measure the program’s financial return on investment by maximising the “ARB per USD spent”
  Quantitative vs. Subjective Goals:

• Original Goal: The first three KPIs were simple, quantitative metrics (number of delegates, participation score).

• New Goal: While some KPIs are quantitative, one is now a subjective task for the Program Manager: to “identify” valuable contributors.

@SEEDGov

Second, the claim that the program creates a “caste system” or makes it “nearly impossible” for smaller/newer delegates to qualify is not supported by data. Looking at July’s results:

@SEEDGov

It’s evident that some delegates have found ways to consistently add value to the DAO through various activities — and were recognized accordingly. Even delegates who don’t usually qualify, such as Hawheik and TodayInDeFi, earned spots in July due to improved contributions.

I think i don’t have to say more, as the Delegates you have mentioned have come out and defended my claim.

Multiple delegates, including @Ignas and @dragonawr have added to the discussions saying how the program should be more inclusive. And there is solid proof of how your PM led to decline in contribution.

I think the most honourable thing to do is to take accountability and work on improving the program.

Finally

On Community Support and the v1.7 Vote

SEEDGov Claim: “The alleged ‘loss of confidence’ is hard to justify, given that just two weeks ago, the DAO overwhelmingly approved version 1.7 of the program.”

The Evidence: Equating a single vote with broad community support is disingenuous, as other delegates have pointed out.

• As @dragonawr noted: “I do find it disingenuous to point out the approval of DIP 1.7v as a sign that these changes are not unpopular, as we all know how voting power and the DAO work…”
• More importantly, the claim of community confidence is directly contradicted by the delegates who are now leaving. @Hawheik , whom you used as a success story, stated that v1.7 was a “significant contributing reason… why I will no longer participate.”
• @DonOfDAOs was even more direct: "I’ve discontinued participation in DIP explicitly because of program management

[stonecoldpat]

Hi all,

I’ve been following the DIP discussions over the past few months including the comments here.

As everyone is aware, there are problems with the DIP, and I’d summarise them as follows:

• Commingling of Intents. The program is a mixture of a delegate and a contributor incentive program that should be separate initiatives. A delegate incentive program should maximize dollars spent per vote whereas a contributor program should reward volunteers for positively supporting efforts in the DAO regardless of their ARB holdings. The commingling leads to exclusionary policies as people have been required to purchase 50k ARB to simply be eligible which acts as a barrier to contributors who can have a positive impact in the DAO. At the same time, we witnessed in January 2025 that ½ of all payments went to delegates with <100k voting power which demonstrates that it was incentivising contributors, but not necessarily voting power.

• Misuse & Farming. We have witnessed people trying to create duplicate accounts, on several occasions, with the intention of getting paid more than once per month from the program. Thankfully, we managed to catch them prior to rewards being paid out. Additionally, in a previous version of the DIP, it was common for people to craft replies using ChatGPT with the intention of maximising their payments and of course led to unnecessary spam which made the life of a delegate harder. This was the motivation to move towards more subjective interpretation of comments.

• Overpayment. We’ve witnessed payments of >$200k per month which can be annualised at $2.4m per year. This was a higher burn rate than many Series A cryptocurrency startups and it was the largest incentive program for any DAO by a significant amount. The mistake here was that the DIP should have started on a much smaller scale and then gradually build up the size of payments. It is much easier to grow a program than it is to shrink it. The goal should be to work out the optimal payment amount in return for the metrics that the program is optimising for.

• Rewarding Replicated Work. The initial program rewarded participation and not necessarily the strength of contribution. The only repetitive work that should be rewarded is a consistent voting record, and if desirable, rational for the vote. People should be rewarded for outstanding contributions that can be recognised by their peers which the current DIP program has attempted to pivot towards. For example, good deeds might include hosting governance calls, onboarding a big delegate or helping the DAO make a big impact at an event.

• Perception of Payments. People are viewing the DIP as a full-time salary as opposed to being rewarded for participating in the DAO. This perception should change. In my opinion, payments should be viewed as a “tip” or “thank you” from the DAO for doing a good job. There should be another overarching reason for why people are participating in the DAO beyond direct compensation. For example, they may have an interest in governance, represent the interest of a company or protocol, looking for work in a future proposal, or a gateway towards a full-time job in the space.

• Too many disputes & arguments. This thread, alongside the various disputes that arise every month when payments are published, is not healthy for the program. There is a joke in our DAO that people are spending more time on disputes than they are on contributing to the DAO. We need to fix this. The answer is likely to move away from mass-evaluation of comments/contributions to a recommendation system where peers in the DAO can make proposals to the program manager about good deeds in the DAO and the program manager picks the top ~5 recommendations to reward.

It is easy to point the finger at badly designed KPIs and then say the program has failed because they were not met. That is an issue with a waterfall model of governance and not necessarily with the program. We should be adjusting KPIs as the program is running and new information becomes available, especially for something as experimental as the DIP. This is something SeedGov has attempted to do with the different KPIs that emerged in the temperature check votes.

At the same time, the DIP has had some interesting and good outcomes:

• Attracted good contributors. We’ve seen some people pop up in the DAO who saw an opportunity to earn some pocket money and then stuck around because they found the project interesting.

• Nudged behaviour. People have carried out certain behaviours to remain eligible for payments. For example, delegate decided to cast early votes in the security council elections as this was an eligibility criteria. We should brain storm on other type of behaviours that might be positive for Arbitrum for a whole. For example, DIP could reward people who help amplify big announcements.

• Activated voting power. There are some large delegates where the monthly payment does make a difference on whether they prioritise reviewing a proposal and voting for it.

With the above in mind, there are times when I thought it was a good idea for the DIP to be shut down, especially with the endless disputes and arguments that pop up every month. At the same time, it does lead to the loss of some perks, that could be beneficial to the DAO as a whole.

The solution is just to revamp the entire program, including its purpose / structure / expectations by the community. It needs to start on a much smaller scale and build up over time.

I’ve spoken with SeedGov on this a few times now and they appear to have taken my comments on board to date. Additionally, the AF has been working with SeedGov on V2 to help overcome some of the issues identified and hopefully lead to a more effective DIP.

[paulofonseca]

Hey Patrick!

Thank you for sharing your perspective on the state of the DIP!

@stonecoldpat

It is easy to point the finger at badly designed KPIs and then say the program has failed because they were not met. That is an issue with a waterfall model of governance and not necessarily with the program. We should be adjusting KPIs as the program is running and new information becomes available, especially for something as experimental as the DIP. This is something SeedGov has attempted to do with the different KPIs that emerged in the temperature check votes.

It’s honestly a bit jarring to see the main spokesperson and the most recognizable and long-standing employee of the @Arbitrum Foundation publicly excusing a DAO program (that defined its own KPIs which were approved onchain) from not meeting those KPIs, after 9 months.

As I shared in my analysis above:

@paulofonseca

the changes introduced, retroactively, with DIP 1.6 in February 2025, clearly led to all the self-imposed performance metrics starting to trend down

It’s almost as if, somewhere in January/February, @SEEDGov decided to change the program in a dramatic way that would lead them to not reach their self-defined KPIs. Because honestly, before February, the data shows that they were actually on track to reach those KPIs. Again, nobody in the community asked them to introduce a Voting Power Multiplier (that is actually a penalty) with version 1.6. And there’s really no stated reason as to why those changes were made. And again, these dramatic changes in February were made retroactively, and without a temperature check.

So no, it wasn’t really a matter of “badly designed KPIs” or “an issue with a waterfall model of governance” or anything of the sorts. It was a deliberate and drastic change by the program managers which led them to not reach their KPIs.

We can speculate why that happened, but I don’t think that’s useful right now.

I think we should focus our energy on designing a next DIP program that is created from real community input, not designed by a self-serving service provider, behind close doors, and alongside the Foundation, as you said here:

@stonecoldpat

the AF has been working with SeedGov on V2 to help overcome some of the issues identified and hopefully lead to a more effective DIP

I believe the next DIP can and should be designed by the community of delegates, especially those that have offered immense feedback about the program, feedback that was not taken onboard at all, as this very proposal, and the comments of several delegates in here, is proof of.

So, I took the initiative to make the data analysis above, and also an anonymous feedback survey for delegates to share their true feelings about the DIP and what the design goals of the next DIP should be.

I’ll be running some calls next week to go over the survey results and workshop with the community the design goals for the next Arbitrum DIP.

[JoJo]

@paulofonseca

It’s honestly a bit jarring to see the main spokesperson and the most recognizable and long-standing employee of the @Arbitrum Foundation publicly excusing a DAO program (that defined its own KPIs which were approved onchain) from not meeting those KPIs, after 9 months.

Want to throw my two cents in here.

Both you and me, we do have a very different point of view on the DIP program, this is well known.

I am appreciating how you are trying to move from a critique such as “I don’t like it” toward something that is more quantitative and objective to discuss on top, even tho we are talking about a program that can be designed in a million different ways and that can satisfy a million different delegates in very different ways.

I won’t specifically challenge your analysis of KPIs. But I do want do invite you on a reflection: while it seems true, looking at your numbers, that some KPIs were not met, in a 1 year long program is also extremely easy to find a situation in which you establish certain criteria that then, applied to real situations or just a very fast moving context, don’t work as intended.

Why then we shouldn’t try to adjust the program, in a way that is perceived better, regardless if it can go against the KPIs?
We did had a vote a few weeks ago about it, and the majority voted in favour of these further changes. We didn’t had a vote for 1.6, but I am pretty sure the outcome would have been similar because I am fairly sure the PM discussed with several delegates this changes before pushing them.

I am not mentioning this to specifically defend the program; I think SeedGOV is doing a great job, and 1 year ago I told them several times that they would find themself being underpaid to carry on a task that would 1) be more cumbersome than what was looking like on paper 2) likely create discontent between some delegates, regardless of the decision taken, because of the economic and personal nature of it.
This is only my personal opinion of course, and I am not here to try and convince you about it.

But maybe, just maybe, is worth trying go “beyond” KPIs and understand what would have meant to stick to the initial recipe that, after a while, did show limits that ripples even outside our DAO.
There is of course a balance to keep in this decision, strong changes in programs voted have to have a consensus and a strong reason, and in this sense (hindsight 20/20) the 1.6 could have had a vote regardless of the fact that was in the power of the PM push for these changes without a temp check as per the original vote.

But even with that, it can make sense to adjust a program, that has a goal set 6 months, 9 months, 1 year ago, to a reality that 1) maybe doesn’t match what initially planned 2) can really change fast and at the speed of light cause we are in crypto.

And so, maybe, focusing on original KPIs being achieved or note means only focusing on the tree and not the forest.

My two cents, of course.