
Experimentation Culture and Challenges

Marketplace lessons from Uber, Airbnb, Bumble, and more | Ramesh Johari (Stanford professor)
Topic 4 of 6
Published Nov 9, 2023 · Clip 40:22–53:54 (13:32)
Ramesh advocates for hypothesis-driven experiments focused on learning, not just wins, to avoid incrementalism and long runs that stifle innovation. He warns against judging solely on statistical significance, emphasizing accumulation of evidence and business context. Cultural shifts, like Bayesian priors incorporating past learnings, can reward broader impacts beyond immediate metrics.