Jeff Weinstein: Why a burning problem beats clever ideas

Jeff learned the hard way that beautiful, well-crafted products still fail if they don’t solve a top-tier, urgent problem. He now looks for signals like customers being genuinely blocked or frantic during outages as the litmus test for product-market fit before investing heavily in craft.

He treats any unsolicited customer message as a ‘P0 gift,’ responds immediately (even just to acknowledge), and builds text-message relationships with 5–10 archetypal customers. That small, high-signal cohort provides more direction than thousands of survey responses or second-hand research.

Jeff discounts friend and free-user feedback to zero and forces a ‘would you pay $X right now?’ conversation with target customers. The gap between ‘willingness to pay’ and actually paying exposes what’s truly valuable and prevents teams from chasing polite but low-value requests.

For Atlas, they used ‘companies with zero support tickets’ as a north star and drove it from 15% to 85%, which closely mirrored market-share gains. He pairs quantitative charts with qualitative tweets, names metrics in plain language, and centralizes them in one shared dashboard that teams check constantly.

Stripe emits a log event whenever something happens that would make a user’s day worse (e.g., 404s, delayed payouts, too many declines) and visualizes them as a stacked bar chart. This reveals unexpected failure modes at scale and gives teams a shared backlog of concrete pain to burn down.