Retrofunding via Rounds and Spam

Lessons from the first few days of Purple's latest Retro Funding Round

Purple is currently running its current iteration of Purple's Retro Funding Round via Rounds in the /purple channel. It is a 10 ETH Round, thanks to a 5 ETH contribution from Purple and a 5 ETH contribution from Octant. You have 7 days left to submit your entry and anyone who has made any demonstrable impact on Farcaster can apply by posting their impact into the channel and tagging it #retropgf. Purple token holders and approved Octant voters then vote for their favorites via a Farcaster like. The funds are then distributed based on the number of Likes one receives.

This is our second Retro Funding Round via Rounds and our first one under new Grants Lead, Emre Ekinci. One thing I liked about our experiences last time is that Purple went from rewarding initially 5 people, to 15, and in the last round, over 50 people received at least some reward and recognition for their contribution thanks to using Rounds.wtf. I expect that number to be much bigger this time.

I have nothing but praise for Rounds.wtf and its team. I actually have a hard time writing about it because I think it's truly a transformational new crypto-social primitive that when I sit down to write about it I just see a big galaxy brain meme in front of me as I think of all its nuances and aligned incentives. Maybe just watch this 20m podcast with its founder, Seneca. It is excellent.

Our Retro Funding Round is designed to capture a wide range of impact from media, to channel hosting, to software development, and more. The key is that applicants must have had a demonstrable impact on Farcaster. Simply minting an NFT and sharing it on Warpcast once doesn't qualify. This brings us to our main challenge: spam and low-effort applications.

Current Challenges:

  1. Spam and low-quality submissions: Since we're providing a meaningful token reward, we've seen an influx of spam, AI-generated posts, and low-effort submissions.

  2. Moderation burden: Emre and I have been aggressively moderating the channel, hiding posts, banning bad actors, and checking to ensure they didn't receive any likes in Rounds. This is a time-consuming process, occurring probably 15 times a day by me and I have no idea how many times by Emre.

Potential Solutions:

  1. OpenRank moderation: Automod could filter based on OpenRank, potentially helping to filter out bots while letting real people through. However, determining the appropriate OpenRank threshold requires further investigation. Does anyone have a suggested OpenRank number to use a justification for it? I would turn this on today.

  2. Moderator-approved posts: Automod could be set up to show only posts liked by a moderator on the main feed. While labor-intensive, this would result in a clean feed of quality submissions to vote on. If we had a paid staff, it might be worth doing it this way to make sure all votes are going to high quality submissions.

Spam is becoming a problem across various grants programs. At CharmVerse, we're collaborating with Google on an automated spam check phase. For Optimism Retroactive Public Goods Funding (RetroPGF), we have a team of badgeholders who review each proposal to filter out spam from hundreds of submissions.

While this issue isn't unique to Rounds, it's a crucial consideration when designing a Rounds campaign. As we continue to refine our approach, we're open to suggestions and feedback from the community to make our Retro Funding Rounds more efficient and effective.

Loading...
highlight
Collect this post to permanently own it.
thoughts onchain logo
Subscribe to thoughts onchain and never miss a post.