From product instinct to stakeholder alignment in an afternoon
Loft's home and library screens were identical: both just showed a flat list of bookmarks. The original design assumed a recommendation engine and user-generated content that never made it to MVP, so engineering defaulted to a plain list. Two screens, same thing.
The obvious fix was a recommendation engine: analyze bookmarks, run personalization logic, surface what's relevant. But that meant real backend infrastructure, LLM costs per user, and a cold-start problem for anyone new to the app. Not the right investment at this stage.
The question became: what makes the home screen feel personal and useful using only data that already exists for every user?
Two hours in Lovable and V0. The layout surfaced four things built entirely from existing data:
- Top collections - top 3 by bookmark count, showing where their attention actually lives
- Recent bookmarks - the last few saves, so users come back to read what they added
- Top tags - a quick reflection of their own reading patterns
- Quick actions - new collection shortcut and Loft tips
No new backend endpoints. No LLM calls. Everything pulled from data the app already had.
A screenshot alone doesn't move a team. I recorded a short walkthrough video and paired it with a lean PRD: the reasoning behind each section, what it replaces, and edge cases for new users with no bookmarks yet. Both shared together so the team saw the idea and the thinking at once.
One round of feedback, one iteration. The designer applied the full design system and engineering shipped it. It's live today as Loft's home screen.
The prototype wasn't the deliverable. It was the tool that made the decision possible without long back-and-forth, multiple review meetings, or asking engineering to build anything speculative first.