Hey HN! Adam here. I’m a developer and founder, and I’ve shipped a lot of landing pages that looked fine but didn’t convert. The frustrating part wasn’t low conversion itself, but not knowing what exactly was wrong or what to test next without traffic data or paid CRO audits.

I built Crovise, a system that analyzes a landing page and outputs concrete conversion hypotheses. It reviews structure, copy, hierarchy, and UX patterns using static analysis combined with heuristics and LLM-based reasoning.

The goal isn’t to replace analytics or A/B testing, but to give a fast, opinionated pass similar to how a CRO engineer would review a page early on. There’s no session data or tracking involved yet, so all feedback is based on page content and layout.

I unlocked access for Show HN users so you can try it without signing up. I’m especially interested in feedback on the approach, blind spots, and cases where this kind of analysis breaks down.