We Have Learned Nothing

25 years of startup methodology — Lean Startup, Customer Development, Business Model Canvas, Design Thinking — have failed to improve startup survival rates. Government BLS data and Crunchbase VC data both show zero systematic improvement in startup success over this period. The author argues these frameworks are self-defeating: once widely adopted, they cause founders to converge on identical strategies, eliminating any competitive advantage. Drawing on the Red Queen hypothesis from evolutionary biology, the author proposes that a true science of entrepreneurship must operate at a higher level of abstraction — not prescribing specific winning tactics, but generating frameworks for continuous differentiation. Citing Feyerabend's 'Against Method,' the author argues the only durable principle is that there are no fixed rules, and that any method, once institutionalized, becomes the next thing to reject.