AA communication: on Flock team ANT packages

AA communication: on Flock team ANT packages

We’ve been looking to determine guidelines for the ANT packages usually requested by Flock teams as part of their Flock grant application. Team incentivization, token distribution, cooperation dynamics and Network ANT treasury management all are pretty strategic matters.

Determining a policy on this matter is tricky for different reasons:

  • ANT is currently highly undervalued
  • ANT’s value could fluctuate a lot in the future
  • ANT has a direct governance right over the network, which implies that token distribution has to be thought carefully
  • The Flock program is moving quickly, and it is hard to have an accurate guess on the number of teams that are going to contribute and for how long
  • ANT does not exacly look like the shares one would find in a traditional start-up
  • Flock teams have different internal structures and potentially use different means to incentivize their members

In this post, the association will detail the different options it has been exploring with Flock teams until now and determine its current ‘best option’. All community members are invited to discuss and add to these thoughts.

Determining ANT packages: planned policies vs. laissez-faire approach

The two main categories of approaches that can be adopted are either ‘planned’ or ‘laissez-faire’ policies.

Before diving into each category, here are important things to consider when one is thinking about ANT packages:

  • Incentivization - Is the package size large enough to provide a sentiment of ownership to the Flock team and its members?

  • Governance - Is the package giving a voice to the Flock team proportional to its involvement and trustworthiness? How much of the overall ANT supply and Network treasury does it take? What is the long term impact on ANT distribution?

  • Rewarding risk - Is the team rewarded for the risks taken? Are earlier teams better off than teams joining later?

  • Hamonization among teams - How does the package map among past Flock team packages and does it make sense? If a policy is adopted, how will the package of the 10th Flock team look like?

  • Team quality - Is the package the right size to reflect the quality/performance of the team and the value it brings to the network?

Planned policies examples

A strict policy has certain advantages: It is easier for Flock teams to justify their package. It is also easier for ANT holders to judge on packages by just looking at what the policy says.

The drawbacks of policies are obvious when addressing such an unpredictable and complex topic: you’re trying to model something that is ultimately very hard to predict. Therefore, it is highly probable that the policy has to be amended in the medium term, which questions the relevancy of such a system.

The employee package policy

In this perspective one considers packages mainly as ways to incentivise Flock teams.

Policy: provide teams with a fixed amount of ANT per team member. Optionally, team member package size can decrease over time from one Flock team to the next.

This policy plans for the following: team incentivization, Risk-reward, Harmonization among teams

The risk-reward policy

If we want to adopt a policy rewarding early Flock teams, we can say that from the Aragon One package of 1.675.000, package size would decrease every year, or every new Flock team by some multiplier, for example by 50%. With this approach we have a better control of ANT distribution in the Network.

This policy plans for: Rewarding risk taken by earlier teams, Gorvernance, Harmonization among teams

The ANT flow policy (mixed planned & laissez-faire)

A more sophisticated policy that could be adopted is the ANT Flow policy. This policy views package as instruments mainly at the team and Flock program level.

  • Each period (annually/bi-annually) an amount of ANT is issued to Flock teams from the Network’s treasury (without vesting)
  • The total amount issued can be
    • either determined by the Network through AGPs (similar to a tap mechanism)
    • or fixed and decreasing over time, similar to Bitcoin mining rewards
  • The ditribution of the FLow can be
    • the same for all Flock teams
    • split by ANT holders through a vote
  • A team is eligible to receive ANT Flows if:
    • It has been working as a Flock team for >=6 months
    • It has been approved at least twice by the Network

Besides ANT flows, teams can still negotiate an additional package as part OR separately from their Flock proposal AGPs if it thinks it should have more. This is the touch of laissez-faire.

The laissez-faire policy

The laissez-faire policy on ANT packages has the opposite characteristics vs. a planned policy.

It is a more costly process, both for the Flock team to elaborate and justify the package and for ANT holders to judge on the fairness of the package. It is also pushing the assessor to compare teams with each other, introducing potential competition between teams.

The advantage however is that this policy is more resilient to changing Network conditions.

How to think about ANT packages under a laissez-faire policy

Under this policy, teams will have to justify their package according to the different points detailed above. ANT holders can challenge the packages according to the same rationale. Metrics can be used to find a commong field of understanding.

NOTE: The metrics presented here are examples, not official metrics. Other metrics and reasoning can be substituted if considered more relevant.

  • Incentivization metrics - Package size, ANT/person
  • Governance - (ANT package)/(overall ANT supply)
  • Rewarding risk - $ value of the package on a projected future Network value
  • Hamonization among teams - assess ditribution curve of ANTs among teams using a (A1 package/(X+1)) like function. with X being either the year we’re in starting from 2018, or the number of packages already distributed to Flock teams
  • Team quality - This last criteria is more qualitative and can have an important impact on the ANT package’s size

Analysis of a current Flock team packages under a laissez-faire policy

Below are a couple of examples of ANT package analysis under a laissez-faire approach. Using the different perspectives and their metrics will help assess if a given package is fair and well constructed.

Aragon One

Aragon One has been approved to receive a 1,675,000 ANT package, with 1-year cliff and 4-year vesting.

Incentivization - This was ANT67K/person with the 25 employee target provided in the proposal. To date it is ANT98K/person for 17 team members. However, the team has a pool of unallocated ANTs for future employees and unexpected events.

Governance - This package represents 4.3% of the overall ANT supply. It is a sizable governance power, especially given the current ANV participation rate. It can be justified by the fact that A1 is the founding team and should keep an important weight in Network decisions.

Rewarding Risk - This is the most sizable package in the ecosystem and a substantial exposition to Aragon’s success being 4.3% of the Network’s value. If the Network is worth $200M that’s $8.6M. (Note that this $200m is a projection and not the current value.)

Harmonization among teams - A1 is the founding team and has no precedents to compare to. However we can expect that it sets a standard and that no team will get a larger ANT package unless exceptional circumstances arise.

Team quality - A1 is to date at the foundations of most pieces in the ecosystem, its design and the code in Aragon repos. It has the greatest technical knowledge and historical experience working on Aragon.

Autark

Autark has been approved to receive a 350,000 ANT package, with 1-year cliff and 4-year vesting. It is presenting a new Flock application with a new ANT package, taking into account its new role in the ecosystem.

Incentivization - Package is ANT50K/person for 7 team members. Although its hard to quantify the financial upside, it seems to be a reasonable amount to expose the team to the Network’s success.

Governance - 350,000 ANT represents 0.9% of overall ANT supply. This seems a bit weak considering that Autark is the second team joining the network after Aragon One and could logically have a greater say on Network decisions, especially as it is proving over time that its interests are aligned with the Network.

Rewarding Risk - Autark started to work as a Flock team 5 months ago under 0.6 Bella. There were not many active users in the Network and as the first Flock team onboarded after Aragon One, the team took significant risks entering a 6 month agreement. This should be reflected in the ANT package. For now, on a $200M worth network its $1.8M in value.

Harmonization among teams - If we aim at Flock teams having between 15% and 20% of the network’s governance power: A1 has 4.3%, Autark and Aragon Black should probably have the same amount adding both teams together (a 50% decrease on A1’s amount).

As a tool for this though process, one can use Amount = (A1 package/(X+1)), with X being either the year we’re in starting from 2018, or the number of packages already distributed to Flock teams.

Ideally following teams would undergo a smaller decrease to smooth out the curve and mimic a long tail distribution. We should also leave flexibility to readjust to unexpected events as information uncovers (ex: if the number of teams in the network is significantly higher/lower than expected).

Team Quality - Autark has accumulated significant knowledge of Aragon during its Nest membership, it has been ramping up the skills of the team and is now reaching a higher an higher level of expertise. The team is also hiring key roles to scale and deliver its future proposal.

Aragon Black

Aragon Black has taken a slightly different approach, asking a per year, per person amount and was approved to receive a ANT125K package on a 4 year vesting schedule for 6 months of work.

Incentivization - This was ANT12.5K/person over 4 years for 10 team members. If the team’s Flock membership was renewed under the same terms, they would receive 1M ANT vested over 4 years, hence a 100K/person amount. This is an important amount in comaprison to past practices. Please note that the package was decided during a discussion between theteam and AA at a time where practices on packages were not perfectly clear.

Governance - ANT125K is 0.3% of overall ANT supply, it is what the team will be vested by completing its first Flock membership. This is a pretty low amount if the team plans to keep contributing to Aragon.

Rewarding Risk - For now, the team will receive ANT125K for 6 months worth of work. Taking again the projection of $200M network value, this is a $600K package value for 6 months of work which seems reasonable. If the team renews its Flock membership and keeps going deeper into Aragon, it might want to look for a more sizable reward to the risks taken.

Harmonization among teams - A good way to think about ANT distribution among Flock teams is that it should have the following shape: (A1 package/(X+1)). According to this, AB entered the Flock program in Q2 2019, then the pacakage would be around 50% less than the initial A1 package amount.

Team quality - What is now the Aragon Black team initally working under Nest has been delivering what was promised on time when contributing to Aragon. While slightly under staffed in terms of Dev workforce, the team is currently working on strenghtening this aspect.

DISCLAIMER: The analysis above is one among multiple possible point of views. Some parameters like Team Quality and Risk could justify higer or lower packages. Moreover, because it is not the Aragon Association’s role to assess Flock teams, the AA will never formulate any strong opinions on the matter. This being said, applicant teams as well as ANT holders are invited to share their own thinking and debate as much as necessary.

Conclusion - TL;DR

Two options are available in order to allocate ANT packages to Flock teams: planned policies and laissez-faire. Each approach has different pros and cons.

For now, the Aragon Association is tilting towards a laissez-faire approach. In this approach, teams and ANT holders would be able to assess a package according to multiple dimensions at a given point in time.

Final word: as a general rule for all policies, Flock teams should ask for ANT packages only on their second mandate AND after 6 months spent contributing to Aragon. This is a buffer to foster trust and a better assessment of the team.

Feedback is welcome

4 Likes

A few comments/ questions:

Consider that Autark’s was a six month grant, while Aragon One’s was a one year grant. If we compare apples to apples then Autark actually received more ANT per team member than Aragon One:

Autark: 50k ANT per team member @ 6 month grant
Aragon One: 49.2k ANT per team member (if counting only the 17 current team members) @ 6 month grant

These numbers don’t seem right.

  • 125k divided by 10 team members is 12.5k each, not 25k each
  • Not sure where you get 1M ANT from? “Renewed under the same terms” implies they would get another 125k for 6 months of work, which would give the team 250k ANT total, not 1M ANT.

Are you including Autark and Aragon Black’s time in Nest here? Given that they asked for their ANT package with their first Flock proposal, not the second as suggested here. Or are you applying this policy to new Flock teams going forward, but making an exception for existing Flock teams?


Regarding the question of how to do ANT grants going forward: there is no right answer, but some answer will be more “right” than others. Although ANT packages have been negotiated in a relatively arbitrary/ ad-hoc or “laissez-faire” way so far, I think it makes sense to make the grants more formulaic going forward. These kinds of decisions have been called “more art than science” but I think there are good reasons to make it formulaic, such as fairness and lower overhead.

One approach is to divide the grants into tranches based on what year it is or what the headcount is. We could say something like “Flock team members in 2019 get X % of the network each, team members in 2020 get Y % each, and team members in 2021 get Z % each”, or something like “Flock team members 1-40 get X % of the network each, team members 40 - 100 get Y % of the network each, team members 100+ get Z % of the network each”. One challenge is that some teams (like Aragon One) may end up having fewer team members than projected, giving each team member more of the network than suggested; in this case I think the fairest thing to do is to let the team keep the ANT as long as they are meeting their agreed on goals. There’s a possibility they may fill the positions in the future anyways, and if they don’t then it can just be thought of as a performance bonus for doing more with less.

As to what X, Y, and Z should be in the above examples: Aragon One team members each received 0.247% of the total supply each, Autark team members received 0.126% each, and Aragon Black team members received 0.0315% each. If we double Autark and Aragon Black to account for one year proposals vs the six months they were granted, that brings them to 0.252% and 0.0631% each, respectively, for an average of 0.187% for each team member. Keeping in mind again that this does not account for what each individual team member is actually receiving, and noting that Aragon Black is thus far receiving about 1/4 what other teams have been granted.

We could start at the 0.187% of the total supply per team member and divide it in half for each tranche, however we decide to measure the tranches (in terms of # of team members or what year it is, or some other method). So tranche #1 gets 0.187% per team member, tranche #2 gets 0.0935% and so on. This would have to be factored in with the amount of ANT the AA has to grant as well, and also how this formula or a modified version of it may be applied to Nest grants and other forms of incentives.

I don’t have a strong opinion about whether to divide tranches based on headcount or what year it is. I do lean a bit towards headcount so that there isn’t a rush to get proposals in at the end of the year. There are some edge cases to think through like what if a proposal pushes the headcount into the next tranche (e.g. tranche #1 is team members 1-40, current headcount is 36, proposal is to add 6 more team members), or what if there are multiple proposals in a vote and one would push headcount into the next tranche but not the other (e.g. tranche #1 is team members 1-40, current headcount is 36, one proposal is to add three more team members and the other adds seven). But I think these edge cases can be solved for if we agree with this approach.

Efforts should probably be made to clarify the ‘intentions’ behind requested packages. I guess that for Autark as well as for A1:

  • Packages were thought as one-off packages for all their future Flock mandates.
  • Autark or A1 would not ask for new packages unless there was major changes in team size, perceived value of the team, etc.

Under this light the calculations above made sense. Your POV is still perfectly receivable as a way of thinking about packages.

Very true! messed up calculations, I’m going to correct that.

The logic used by Aragon Black effectively was ‘125K for 6 months of work’ (i.e. Aragon Black uderstood that it would be asking for new packages again and again for each mandate) contrary to A1 and Autark “one-off package with potential later adjustments” approach described above.

Things are pretty confusing as they are now. That’s why we translated Aragon Black’s package into ‘one-off’ terms so people could compare packages with each other.

I guess one important issue you are raising here is: do we want a one-off approach (A1, Autark for now), or new packages requested for each new Flock mandate (Aragon Black for now). From discussions AA had with Flock teams, the peference was tilting towards a one-off approach.

Exactly this.

The headcount approach

I think formulaic approaches make a lot of sense for these reasons. They also have the advantage that downsides are clear.

In the case of the headcount approach:

  • It may push teams to be more sizable in order to access more governance power
  • It also leaves less flexibility to attract great teams as time goes on and to account for Team Quality/performance.

Of course in the end there always are trade-offs. A mixed formulaic/ad-hoc approach could be worth considering too.

One quick note here is that Aragon One actually has a lot of unallocated ANT, for some reasons such as:

  • Unpredictability related to growing the team: We do have a strong commitment to keep the team small, but as I see Aragon One as a startup, things change fast and we don’t want to have our hands tied in case that we need to disburse more ANT to more employees or even contractors
  • Keeping some ANT in Aragon One, the company itself, as opposed to allocating it all, provides an inner incentive for the company to be aligned with the Network. Allocating ANT to employees incentivizes them to work on the Network, and having unallocated ANT for the company incentivizes the company itself. This is to cover edge cases, for example if all employees leave just after having their tokens vested, the company would still be incentivized to work on the Network
  • Governance power, so Aragon One itself can participate in the Network’s governance

So basically we are keeping unallocated ANT just in case and to be prepared for scenarios we cannot think of right now. I think when running a startup it’s a good idea to be paranoid and keep some buffer, rather than having to ask the Network again for ANT, which is something that I feel like Aragon One should avoid at all costs.

1 Like

I think we should leave it flexible - if a team ends up growing beyond what was expected when they received their first ANT grant, and the Network can afford to support that growth, they should be able to request more ANT to incentivize the additional team members. And in cases like Aragon Black, where they request much less than they actually intend to receive, then they have opportunities to request more later and there is an expectation that it is acceptable to do so. I think as long as there is a reasonable justification for requesting more ANT then teams should be free to do so.

Well teams would have to weigh that benefit vs the potential to get rejected for costing too much. And assuming that ANT is distributed from the team’s company to individual team members, there’s no guarantee they all vote the same way, so I don’t see an advantage or problem with thinking this way.

I’m not sure I understand this objection to the headcount method, can you explain more?

1 Like

In this policy, fixed and decreasing amounts of ANTs are issued to new members which leaves no flexiblity to offer higher/lower amounts to newcomers.

Ah right, that is either a positive or a negative depending on your perspective. An objective formula will ensure “fairness” for new teams, with less potential for Flock managers being accused of favoritism or corruption offering “sweetheart deals” to certain teams. But it would also eliminate the possibility for a team to have a specialized ANT package to match some specialized value they bring to the project. These perspectives have to be weighed against each other and compared to the values of the Aragon Network / Flock managers to see which policy would be more aligned.

1 Like