Research Summary: A Comparative Analysis of the Platforms for Decentralized Autonomous Organizations in the Ethereum Blockchain

Governance and Coordination
,
#1

TLDR:

  • This paper analyzes data from 72,320 users and 2,353 DAO communities on three main platform ecosystems across four dimensions: growth, activity, use of the voting system, and the funds owned by DAOs.
  • By active DAOs, Aragon is three times more active than DAOhaus, and eleven times more active than DAOstack; by active users, Aragon is 27 times more active than DAOhaus, and five times more active than DAOstack. (Faqir-Rhazoui Y., et al., 2021)
  • Within the scope of this research, Aragon is the biggest platform in many aspects, thus we may conclude that the ability of Aragon for customizing DAOs may reveal some essential features for developing DAOs. (Faqir-Rhazoui Y., et al., 2021)
  • The cost, decision-making system, and customizing functions of each platform may significantly cause the adoption and activity of DAOs on them.

Core Research Question

  • How is the performance of DAOs on the main platforms that provide DAO deployment as-a-service so far?

Citation

Link

Background

The authors break down the definition of DAOs further into interrelated parts:

  • Decentralized: A DAO exists on a public blockchain that is serverless, censorship-resistant, and cannot be controlled by a central party. A DAO and its activities must depend on the consensus of its participating members, which is achieved through a voting structure.

  • Autonomous: A DAO is autonomous because its existence relies on its members. The members determine the set of rules that the self-executing codes the DAO must follow.

  • Organization: A community that interacts towards shared and agreed-upon outcomes.

  • Aragon: It is the largest and most mature of the three platforms, with a variety of decision making apps (sets of smart contracts) to choose from as well as customizable organizational templates. The voting apps range from basic functionality and parameters in their native Vote app to more experimental ones such as Conviction Voting where individual preferences are captured using tokens and continually expressed as votes on relevant proposals. (Faqir-Rhazoui Y., et al., 2021)

  • DAOStack: It introduces the Holographic Consensus system in an attempt to address one of the scalability issues associated with DAO growth, where the requirement of an absolute majority becomes less practical as proposals increase. The idea is to use tokenomics as a filter on proposals where members stake coins to make bets on if a proposal will pass or not. Proposals that reach a baseline will pass via relative, not absolute majority. Thus in theory members are incentivized to be aligned with the DAOs best interest as their bets depend on choosing the right outcomes. (Faqir-Rhazoui Y., et al., 2021)

  • DAOhaus: It is built on top of the Moloch DAO framework. The Moloch v1 voting system does not require a quorum and a relative majority can pass a proposal. Proposals consist of a combination of requesting shares from the members or offering tributes which function as a donation or a deposit. In Moloch v2, one of the main updates is that non-DAO members can submit proposals for voting only through sponsorship (deposit) of a DAO member, where the sponsoring member will receive part of the deposit back after voting has been completed. The DAOhaus platform has a unique feature that grants members the the ability to “rage quit” which is the option to exit a DAO and keep all of their vested funds. If more than a ⅓ of members rage quit on a proposal, it will not pass. (Faqir-Rhazoui Y., et al., 2021)

  • xDai network: It is a sidechain of Ethereum that provides fast and inexpensive transactions compared to the Ethereum mainnet but sacrifices decentralization to some degree. As the gas price soars, the DAO platforms search for alternatives for Ethereum to avoid its expensive cost. The xDai is one of the successful cases. The table below shows the comparison of the prices of two DAOhaus operations and the average speed in both the Ethereum main net and xDai networks in October 2020. (Faqir-Rhazoui Y., et al., 2021)

Summary

  • The researchers start with an overview of the DAO platforms. Including the introduction of DAO and the three main platforms that provide users with customizing DAOs templates tools service.
  • The general approach of this paper is to retrieve and compare relevant statistics from platform inception to Nov 30, 2020. The data are split into four categories: growth, activity, voting system use and total funds.
  • Growth and activity are helpful indicators of adoption levels and the paper gathers available data on the aggregate and active number of DAOs and their users. Voting system use is measured by voter participation and proposal statistics. Total funds provide further insight into overall adoption.
  • The authors also include DAOs on both the Ethereum mainnet and xDai network in this analysis.

Method

  • The authors compare the performance of the three main platforms: Aragon, DAOstack, and DAOhaus by quantitative analysis. They aggregate data then give an explanation.
  • The authors accessed the blockchain data of these DAO platforms using the indexing protocol called The Graph. These platforms each have “subgraphs” that present the data as an API and the GraphQL language is used to fetch and run database queries.
  • The datasets produced and used in this paper are available in a Github repository and using an online visualization tool. Unfortunately the server for this tool is down as of 09/2021 but the source code for the tool can be found here. The DeepDAO web tool was used to gather cryptocurrency funding data.

Results

  • The table below shows the data of the three DAO ecosystems in terms of their number of DAOs, users and proposals. It’s noticeable that Aragon started using xDai in July 2020, DAOhaus started using xDai in July 2020, and DAOstack started using xDai earlier, in February 2020. The authors argue that this may explain why the adoption of xDai in terms of users and DAOs in DAOstack is significantly higher than in the other two platforms.

741×131 20.5 KB

(Faqir-Rhazoui Y., et al., 2021)

  • Growth
    • Two metrics are used for the comparison of the platforms concerning growth over time: the number of DAOs and the number of users.
    • Lacking data:

  1. The timestamp of the DAO creation for DAOstack.

  2. The timestamp of the user registration for Aragon DAOs.

  • These data do not reflect when a xDai DAO is new or the result of a migration process.
  • The authors point out that the number of DAOhaus’s users could be higher because 311 people used the ’rage quit’ option during the period analyzed and such an option is not available on other platforms, where users just abandon their accounts and do not use them.

857×625 140 KB

(Faqir-Rhazoui Y., et al., 2021)

863×632 127 KB

(Faqir-Rhazoui Y., et al., 2021)

  • Activity
    • The authors define the meaning of ‘active’ for both a DAO and a user as the same definition in a piece of their previous paper An overview of decentralized autonomous organizations on the blockchain that considers that “a DAO or a user were active in a given month if at least they performed an action in that month. The available actions to be performed depend on the platform”.
    • The authors point out that the approach followed may cause a highly conservative estimation of Aragon. So its number should be higher than the figures shows
    • The authors believe the peak in October 2020 of Aragon in figure 4 could be due to a migration to xDai.

859×641 167 KB

(Faqir-Rhazoui Y., et al., 2021)

856×638 173 KB

(Faqir-Rhazoui Y., et al., 2021)

  • Voting system
    • The authors take four metrics to compare the governance and voting system of each platform:
  1. The percentage of users who vote — for the observation of the engagement of the DAO community.
  2. The number of cast votes per voter — for the observation of how active voters are in terms of participation.
  3. The percentage of proposals that are approved— for the observation of how the voting system may influence the results.
  4. The percentage of positive votes among those cast.
  • The authors explain that the inactivity of a DAO with around 4,000 users may cause the percentage of voting users statistics for DAOstack, while there is a high number of inactive DAOs in Aragon.
  • The authors provide a potential explanation for the percentage of users who vote in xDai is smaller than in mainnet is that xDai is a younger alternative, even if it is cheaper.
  • Pertaining to the ratio of votes per voter, DAOhaus has an xDai ratio almost double that of the mainnet. The authors argue this could mean that xDai boosted participation.
  • In order to reduce costs, proposals may be discussed off-chain within the community, then be put to on-chain vote if it is likely to be approved. This explains the high percentages of approved proposals on every platform.
  • The authors present hypotheses for differences in of approved proposal percentages between every platform, while also pointing out that these conclusions should be validated through further studies:
    • DAOstack voting system: Because it requires either an absolute majority (51%), or enough staking for a proposal to be ”boosted” and thus able to be approved by a relative majority, it shows a lower number of approved proposals.
    • DAOhaus voting system: It requires no quorum, a relative majority can approve a proposal, so it shows a higher number of approved proposals. Furthermore, in DAOhaus v2, a proposal requires sponsorship of a community member, this also bats proposals that may be rejected.
    • Aragon voting system: Its standard voting app requires a minimum number of members to approve a proposal, so it makes sense that its approval rate is lower than DAOhaus. On the other hand, its voting system leads to lower rejection rates than the DAOstack system.

917×184 31.8 KB

(Faqir-Rhazoui Y., et al., 2021)

  • Funds
    • The table below shows the Top 10 cryptocurrencies in terms of DAO adoption, which is calculated by the number of DAOs that use them. In this summary, we also update the tables in the paper (Table 4 & Table 5) which were retrieved on the 1st of December 2020, and the latest accessible statistics on the same websites which were retrieved on August 2021.
    • The authors point out that it is important to bear in mind that the funds of a DAO are dynamic as it has inflows and outflows.
    • Many of those crypto-currencies are stablecoins (DAI, SAI, USDC, or USDT).

857×356 56.4 KB

(Faqir-Rhazoui Y., et al., 2021)

|506x269.0700726387376

(Retrieved August 2021)

852×321 58.6 KB

(Faqir-Rhazoui Y., et al., 2021)


(Retrieved August 2021)

Discussion and Key Takeaways

  • As Table 2 shows, Aragon is 10 times larger than DAOhaus and 79 times larger than DAOstack. While it is noteworthy that from Aragon’s 1,700+ DAOs (2,000+ including xDai DAOs) and 41,000+ users (68,000 including xDai), there are only 100 DAOs and 330 users that are active each month. So the authors conclude that by active DAOs, Aragon is three times more active than DAOhaus, and 11 times more active than DAOstack; by active users, Aragon is 27 times more active than DAOhaus, and five times more active than DAOstack. (Faqir-Rhazoui Y., et al., 2021)

Implications and Follow-ups

  • As the paper was written, eight of the top ten wealthiest DAOs rely on the Aragon platform, so Aragon’s ability for customizing DAOs may reveal essential features for developing DAOs. However, the number of active DAOs within Aragon has declined since May 2020. (Faqir-Rhazoui Y., et al., 2021)
  • DAOhaus had the highest percentage of proposals (92%) passed which may be due to its voting system. (Faqir-Rhazoui Y., et al., 2021)
  • The problems in Genesis DAO, a DAO aimed to promote the use of DAOs through DAOstack, may cause the stagnation of activity and adoption of DAOstack. (Faqir-Rhazoui Y., et al., 2021)
  • The arrival of Ethereum 2.0 which is expected to reduce gas prices, could have a significant influence on the relevant consequent quantitative research. (Faqir-Rhazoui Y., et al., 2021)

Applicability

  • This quantitative analysis of the three main DAO ecosystems shows how the different mechanisms and functions of those platforms may affect the adoption and activity of the DAOs on them. This information could provide industry with clues about the direction of DAO development.
  • The authors also try to explain every statistic in their paper. This could serve as a reference for those platforms to evolve.
10 Likes
#2

@Astrid_CH thank you so much for posting this fascinating summary, which is particularly interesting to me as I explore creating a DAO of my own. Could you please expound upon what some of the potential limitations of this study are? Are there other factors, such as major investments or marketing campaigns that might explain the difference between Aragon and the other two platforms? Are other reasons for Aragon’s ‘victory’ besides an appealing basket of features?

2 Likes
#3

Interesting research. Let me know more about DAOs.

1 Like
#4

@EasonC13 I’m so glad to know you like it. @jmcgirk Thanks for your questions and comments. From my own observation, some other factors that might explain:

  1. Aragon is the earliest one between them, which means that it could catch eyes with some blasting form. It also performs development progress at a good pace and updating of codes.
  2. Aragon has greater social media community attention, such as Twitter and Reddit users. For example, the followers of Aragon, DAOhaus, and DAOstack’s Twitter today (2021/11/6) are 90.4K Followers, 10.2K Followers, and 8,753 Followers, respectively.
  3. Aragon is more focused on providing general and easy-to-use modules that are both applicable to traditional centralized entities and decentralized communities, while others are more dedicated to providing infrastructures for decentralized forms.
  4. Aragon provides better UX for non-technical background users with clear and readable (from a non-technical perspective) UI.
  5. The promotion of their tokens, such as the number of exchanges their tokens listed, tokens’ capital, the advisors’ backgrounds, famous use cases projects, all further affect the investors’ confidence and users’ choice, and it’s just like a trust spiral. As the authors said, “Given DAOs are an early field relying on a novel technology, growth by early adopters is critical for the future mainstream adoption”. It’s probable that before something big happens, their development may follow rules like “The winner takes it all”.

There may be some potential limitations of this study:

  1. The authors mentioned that many DAO cases on these platforms are built for the purpose of experiments. I believe this may affect the result, especially for platforms that are easy to try.
  2. It seems that the paper omits Aragon Chain, which may affect a little bit in the context of the high gas fee on Ethereum.
  3. As the structures behind DAOs may not be unveiled thoroughly to the public, it couldn’t discuss the DAOs for decentralized communities and DAOs for centralized entities separately.
  4. The authors also mentioned some limitations of the collection of data, including (Faqir-Rhazoui Y., et al., 2021)
    xDai-
    “there is no available data about the proposals in xDai in Aragon”
    Growth-
    “the timestamp of the DAO creation currently is not available for DAOstack DAOs, while the timestamp of the user registration is not available for Aragon DAOs”, “our data does not reflect when a xDai DAO is new or the result of a migration process”
    Activity-
    “for Aragon xDai actions, we can only consider data from the Transaction app because the API does not provide data from the Voting app in xDai…may result in a highly conservative estimation in Aragon, where DAOs can be customized with different apps, and, hence, exhibit other types of activities”
    Voting system-
    “In the case of Aragon, DAOs may have multiple voting systems. However, for the sake of simplicity, we just retrieved data from the standard voting app”
    Funds-
    “DeepDAO does not provide information from all the DAOs in the considered ecosystems”
  5. The definition of ‘active’ may also play a significant role.
2 Likes
SCRF Contributor Recognition and Rewards Initiative
#5

The metrics prompted me to ask what a hypothetical product manager should prioritize to gain a larger market share. I’m still thinking and remain open.

However, that question would be only a piece of the whole picture. The next steps on improving a metric could be challenging.

Judging by Astrid’s previous comment, Aragon’s advantage does seem to go beyond the metrics included in the summary, such as a better developing experience, and more existing users.

Yet it’s not clear to me now what the key differentiating feature of these DAO platforms would be between each other. It feels like a million-dollar question.

3 Likes
#6

Thank you so much for a really detailed, interesting answer. It’s really interesting to think of this study in the context of the marketing, UX, and early mover advantage that Aragon has. What is the Aragon Chain? Is that their token? And how does it compare to $HAUS and other DAO tokens?

We’ve run a couple of different summaries about DAOs, including @kelsienabben and @quinndupont’s work. Nabben established a framework for defining and analyzing DAOs, what might her work have to say about the relative success of the three platforms? DuPont’s work looked at the aftermath of a notorious hack; how do the three platforms take security into account?

2 Likes
#7

Aragon Chain is a blockchain built for optimized Aragon usage that is scalable, high-throughput PoS, and fully compatible and interoperable with Ethereum. It’s interesting to think about how do these peripheral infrastructure projects affect the popularity of platforms. It needs deeper investigation.

Considering platforms present various functions and are ruled by different tokens, comparing tokens’ values we need to look into their functions and underlying assets. There are three tokens in Aragon ecosystem, including ANT, ANJ, and ARA. ANT is Aragon network governance token, the similar token in DAOhaus ecosystem is HAUS. ANJ is created for incenting and managing Aragon court’s jurors. ARA is the native token of Aragon Chain that takes ANT as its underlying asset.

1 Like
#8

It seems that the adoption time of xDai influences the user percentage which uses xDai. However, the Aragon has more xDai user because it is bigger and easy to attract users to use it on xDai. Therefore, the total user amount is critical to the influence of the blockchain technology adoption.

2 Likes
#9

Hi @EasonC13 , I agree with you, the total user amount plays a significant role in many facets. As @Twan said, the reasons for platforms adoption could be a million-dollar question.

In DAOstack, there are more DAOs on xDai than on mainnet, while the number of users on xDai is less than the number on mainnet, and the number of users who vote on xDai is only one-third of the number of mainnet. I’m thinking if there are many experimental DAOs on xDai in DAOstack that cause this result. But I can’t tell the motivation from the statistics, so this is still a hypothesis. I’m looking forward to more researches into the causes of their popularity, which may need some interview methods with communities participants.

1 Like
#10

Hi, @Astrid_CH, I read your summary carefully and could get valuable insights and trends of recent Ethereum DAOs. Thank you.
I agree that Aragon is one of the highest-profile and most capitalized DAO projects. Aragon works a bit differently than other DAOs, and it’s mainly due to its unique structure of decentralized digital jurisdictions.

One of the uniqueness is Aragon Court. The legality of the Aragon network is given and sustained by the Aragon Court. It is a judicial mechanism composed of smart contracts. Community members can participate in the Aragon Court’s judicial process as jurors, plaintiffs, or defendants, but they must put down a minimum stake of 10 000 ANJ tokens. ANJ is an ERC-20 token that only has use within the Aragon Court. Participants have to create a smart contract and deposit ANT tokens to mint ANJ tokens before applying for an Aragon Court role. The decision-makers in the Aragon Court are the jurors who are randomly selected from a pool of users previously enlisted in the process. Their decisions are made independently and anonymously, so the risk of tilting the scale of justice one way or another is minimal. Jurors receive incentives in ANJ tokens for their participation. Therefore, users are motivated financially to take part in the democratic process.

Aragon is working on issuing another ERC-20 token, called ARA, which will power the Aragon Chain. The Aragon Chain will be a second layer of the Aragon network that will use Proof-of-Stake (PoS) to enable quick, cheap, and low-risk transactions between the many communities Aragon platform. This protocol is still “under construction,” and its use of ARA tokens will be possible by bonding them to ANT tokens through smart contracts.

2 Likes
#11

Hi @stayhungry07212 , thank you very much for your valuable comment and detailed introduction about Aragon. As a potential user of DAO platforms and a legal researcher, I’m curious about how Aragon Court reduces the risk of judgment with prejudice or preference. In democratic countries, judiciary departments are often independent and ensure justice by giving judges economic and positional assurance, if a judge conducts malpractice out of personal considerations usually commits a crime as well. But I can only see the ways Aragon court takes to prevent injustice from their jurors are by staking tokens and a random choosing process. I have participated in a team that is preparing to launch a DAO, what our team members are discussing is that what if someone gets close to those jurors (if he knows who is a potential juror) and gives them incentives that are bigger than ANJ gives? When the decision-making mechanism is not “decentralized”, the trust issue on DAOs remains and the advantage of building it on blockchain would be blurred if the issue isn’t disposed of well.

2 Likes
#12

@Astrid_CH, The Aragon Court has the role to oversee the voting of community proposals, settle disputes between participants, and steer the network’s decisions according to its roadmap and align with the Aragon Manifesto’s core values. Aragon has the wild ambition of creating a purely democratic society where countless communities operate under their own governance rules, but still under the Aragon Manifesto. It sounds promising, encouraging, but idealistic. We know from real-world experience that a purely democratic system is nothing short of a utopia. Numerous societies have tried it, but they had to make small concessions to authoritarian rules because human nature makes it impossible for such a system to function for very long.

2 Likes
#14

Eugene’s overview of scrf’s roadmap on the 3.10.22 community call put DAO governance research as a priority. Using this article as momentum could be valuable. The voting activities in this forum appear very similar to DAO activities. Perhaps, SCRF would carry more weight in the DAO community if it were a DAO. Eugene mentioned Harmony Protocol as an avenue for creating a DAO. Harmony has an excellent blogpost - The Way of Harmony DAOs - about forming a DAO. Aragon is listed as a resource in the post. Harmony also offers a HRC-20 token [like an ERC-20 token but minted on Harmony] as a tool for a DAO.

2 Likes