§ 01TL;DR
Donation rate rose from 2% to 8% after clarity improvements.
We ran a content audit and moderated usability study on Octant, a community-driven platform for funding public goods on Ethereum. Nine findings shaped the V2 roadmap. The research revealed that users admired the product but couldn't understand it, and that the assumed audience didn't match the real one.
§ 02The brief
Octant lets users lock Golem tokens, earn staking rewards, and redirect a portion to projects that strengthen the Ethereum ecosystem. The Golem Foundation's product and design leads had tracked behaviour in Matomo and run internal feedback sessions. Users were curious but quickly confused or disengaged. They needed structured external research to find out where and why.
I want to understand how Octant works and where my rewards go
so that I can confidently participate and feel good about my contribution
§ 03The work
We started with a content and UX audit of the platform, identifying specific issues: vague homepage messaging, missing tooltips, and jargon that made sense internally but not to users. Selected fixes were applied to a staging environment so we could test how small copy changes affected comprehension during sessions.
Research design
We chose moderated usability testing with semi-structured interviews. Surveys would have given broader reach, but with a small user base and complex mechanics we needed depth. Unmoderated testing wasn't viable because the web3 wallet setup required real-time troubleshooting. We built three hypotheses from the audit: terminology was the primary barrier, users would misunderstand "public goods", and the 90-day epoch cycle would blunt emotional engagement.
Recruitment and sessions
We recruited ten participants across Portugal, France, and Germany. All were active in web3 but had never heard of Octant. Each participant needed a working MetaMask wallet on the Sepolia test network before their session. We managed that onboarding end to end. Sessions ran 30-45 minutes over Google Meet. Zero no-shows.
Analysis
Nine themes emerged consistently. We cross-referenced these against the audit findings and our hypotheses. Some confirmed what we expected: terminology slowed comprehension across the board, and Match Funding was universally misunderstood. Others were new: Patron Mode was mistaken for a premium feature, the metrics page confused personal and community data, and the emotional payoff of donating was completely missing.
§ 04What we found
It's cool... once you get it.
Users admired the product's design and polish. They described it as "clean", "elegant", "better than most dapps". But comprehension got in the way of confident use. Every participant thought Match Funding boosted their personal rewards. It doesn't. It boosts the project's donation. When every label requires interpretation, users spend cognitive effort on language instead of decisions. The cumulative effect is exhaustion and disengagement.
§ 05Impact
All nine findings were accepted by the product and design team and directly shaped V2 priorities. In subsequent epochs, the proportion of users who donated their rewards rose from roughly 2% to 8%. Multiple factors likely contributed, but the timing aligned with the clarity improvements the research recommended. Correlation, not causation, but a strong signal.
The research revealed that the assumed user base didn't match the actual recruitable audience. That reframed how we thought about positioning and messaging entirely. Head of Product, Golem Foundation