Discrepancy between SAFE token polls - what does it reveal?

It has been interesting observing the results roll in for three distinct polls all asking the same question, with slightly different wording and tone: should Gnosis issue a SAFE token?

Let’s dive in:

The question was first posed here in Gnosis Forum by @Anna and @lukas_gnosis, and has garnered 43 votes over the past 10 days. An overwhelming majority of these voters think we should implement the token, although detractors have been vocal in the comments to cite issues including incentive alignment and proposal scope.

It’s useful to note here that the majority of comments come from my fellow Gnosis employees, which makes sense as we are the initial and primary stakeholders impacted from these early decisions of the DAO! I wonder how much one’s “skin in the game” translates to measured analysis of impact, and likelihood to express that analysis thoroughly and openly in the forum?

• On Nov. 24, the day of GnosisDAO launch, @mkoeppelmann brought the proposal to the wider audience of his personal Twitter account by asking simply, “Gnosis Safe Token?”

The twitter response was a resounding Nay from a significantly larger pool of voters, but with little context gained from comments. Perhaps my hypothesis is correct about the “participation correlation” when one is directly impacted – and votes are tied to the identity revealed, as they are in the forum? Perhaps when there’s a wider pool from which opinions are being drawn, of which less voters have a vested interest in the results, it skews to “no”? Maybe the question comes across differently when coming from Martin than when generated formally from within Gnosis?

martin's poll

• With these hypotheses in mind, we posted the question the next day directly from the Gnosis Safe Twitter account
gnosis safe poll

Although we still have one day before the poll closes, it’s been interesting to see that with approximately 1/3 of the votes on Martin’s post, and double the votes on the forum post, the opinion so far is basically tied – with no comments to provide context.

Why do you think these discrepancies arise? Which shows a more accurate version of how we should proceed? Is there a way to ensure that the forums polls show the most accurate picture of what would be best for our DAO, and incentivize meaningful participation accordingly?

I would love to hear your thoughts!


My assumption would be that respondents on the Twitter pool are primarily Safe users, where participants in the proposal pool are GNO holders & Gnosis staff.

Safe users might be against a SAFE Token because they fear this being associated with fees or giving over control over their own funds to a decentralised governance system.

GNO holders might generally lean “yes” because this proposal establishes a way to capture value created through the Gnosis Safe project using the GNO Token.


I think those Twitter polls and even more the forum poll didn’t really cover the true users of the Safe.

Adding some numbers here:

  • Twitter poll Martin: 226 participants
  • Twitter poll GnosisSafe: 87 participants
  • Forum poll: 25 participants
  • On chain stats: Per day between 60 and 120 Safes make at least 1 transaction (source)
  • Off chain stats for the web interface: Per week, there are ~300 unique users (source, only includes users that accepted to get tracked, there are probably more)

Hence without being able to really prove it, I think we haven’t been able to the true Safe users to bring up their opinions in these Twitter polls and on this forum.


Great additions, thank you! This is useful for thinking about how we can engage more Safe users to participate in the forum.