Ek navigated the music streaming industry by maintaining unusually well-calibrated probability estimates in a domain filled with optimistic overconfidence. While competitors assumed they could negotiate favourable licensing terms with record labels (overestimating cooperation probability) or that consumers would pay premium prices (overestimating willingness-to-pay probability), Ek built Spotify's model around conservative probability assumptions — low per-stream margins, high churn risk, and uncertain label negotiations. By assuming things were less likely to go well than they felt, Ek designed a business model resilient to the downside scenarios that sank competitors like Rdio and Tidal. For founders, Ek demonstrates that slightly pessimistic probability estimates often produce better business designs than slightly optimistic ones.