Recap and reflection on Hypercerts Based Impact Funding experiment - The initiative was taken by GreenPill Network in collaboration with RaidGuild team.
The Raid Guild (RG) team forked the GrantsStack and helped us build a GPN-specific Hypercerts integrated tool which enabled chapters to mint GPN hypercerts for work done by the chapters, which was then meant to showcase their hypercert on the forked explorer page for GPN’s community round.
Hypercerts were minted on minter.greenpill.network.
The experiment was to see if having Hypercerts would have a correlation to the level of funding a project received.
Here is some of what we learnt:
- Hypercerts were minted for the work done over a 6 month period,
- The short amount of time gave us no real data that will measure the impact to the community.
- There was not enough quality control as any chapter could mint on their own without having to meet certain criteria. This resulted in a mixture of impactful hypercerts and some highly frivolous ones.
An interesting next step is setting up an evaluation flow. For example, a chapter steward could approve minted hypercerts and these attestations would go on chain.
With this being a forked version, we ran into some glitches that had us in a start-stop space for a few weeks. This meant there was a lot of manual time spent both on GPNs side and RG’s side. We also ran into some issues with hypercert visibility in the backend when using the forked manager.greenpill page as a Round Operator. We had to reject half the applicants and have them reapply because not all the data was showing properly. This is normal when experimenting with something, so it was to be expected. As an improvement, we could take a more guided approach in the minting process to make sure the applications are correct. This could be either hands-on or more rigid checks in the application.
Not all chapters grasped why minting hypercerts was important and what impact this could have, which was a challenge and most likely is what affected the quality control aspect of it. Having hypercerts in your application was also a requirement for our internal GPN community round. This also meant people were minting as a step vs a thoughtful proof of impact - which also potentially affected quality control. The design is that hypercerts loop impact creators in a larger value flow, but this is something very abstract and we can improve on this by providing more education.
We believe that the average donor might not have understood the importance of hypercerts and therefore might not have taken that into consideration when donating. This is an assumption, but it’s based on how the quality/proof of impact of the hypercert didn’t seem to play a big factor in if a project got funding or not.
Overlap with GG19. The GPN round ran on a forked version while GG19 was happening and this caused confusion as you needed to apply to the GPN round on the forked Grants Stack in order to have your hypercert appear and some chapters applied from the normal builder.gitcoin program page. This meant that while all Chapters/grantees had a hypercert, not all of them showed up in the explorer page. From a Round Manager perspective there was no way for us to know where they were applying from and have them reapply from the proper greenpill.builder page. In comparison, RG ran an experiment connecting the Grants Stack with Lens which was widely announced during EthCC. Such a major announcement pull people into your app, while the GPN experiment was more on a friends-and-family scale for the network. Putting more attention on the fork, or finding support from GitCoin to run a ‘beta’ version, could generate more traffic into these kind of experiments.
Our hypothesis was the chapters with more hypercerts would get more funding. Our sample size wasn’t good enough for real data collection as all chapters had to have a hypercert to be accepted into the round. It also doesn’t seem that the quality of the hypercert was taken into account during the donation process. This could also be due to the fact that this was an internal GreenPill Chapter round, and therefore many donors might (this is an assumption) have just been donating to their own community and 1 other, as we saw 403 contributions from 172 unique contributors, which means the average contributor donated to 2.3 projects.
We believe that hypercerts can and will have a positive impact on how people make donations within the GreenPill ecosystem, we just need more structure around the hypercerts eligibility/ minting criteria, and have more education around why they are important - both for projects and for community members making donations. We also need to work out the technical glitches to make a smoother process for everyone.
Technically, it was relatively straightforward to adapt the grants stack to support hypercerts. Since rounds, project and application, all store their metadata on IPFS and only the CID pointer is stored on chain, RG was able to modify the metadata being stored to contain an array or related hypercert IDs. When reading the metadata, any hypercerts found were displayed in the application.
While our experiment didn’t turn out exactly as we had planned, we do believe that hypercerts and other impact tracking tools will be helpful in assisting people make decisions around what projects to donate to. GPN is also interested in ways to track impact so that we are better able to fund our own community projects.