Why Hardware Startups Fail After the Prototype Works: Greg Kress (Pearl, YC W17)

Learn why working hardware prototypes still fail as businesses — and how to test demand, fund production, and recognize product-market fit. Greg Kress co-founded Pearl (YC W17), an office coffee startup, and is a Founder Advocate at Supabase. He joins Jason Gray to discuss hardware manufacturing, market timing, AI coding agents, hiring, and the milestones that help founders know when to change course. Chapters 0:00 The cost of solving the wrong problem 0:21 The YC coffee demo that could only brew once 5:07 Hardware inventions and unexpected design lessons 7:57 Why a working prototype is not a finished product 11:04 The motorcycle helmet story: testing assumptions 15:15 The value is in the problem, not the idea 15:39 Why Pearl shut down despite a working product 20:47 Market timing, flexibility, and Supabase 23:41 How AI agents change the choice of developer tools 28:23 Measuring progress without becoming stubborn 31:58 Hiring, culture, and trial projects 36:00 Focus and the real cost of founder time 39:00 Why venture hardware needs a tenfold improvement The Pioneer Fund Venture Partner Podcast explores the experiences and lessons of founders, operators, and experts in the Pioneer community. Subscribe for more conversations on startup building, leadership, and the founder journey. Learn more: Pioneer Fund: https://www.pioneerfund.vc/ Supabase: https://supabase.com/