Many members felt it slowed creativity, while others argued it prevented costly misfires and forced clearer learning loops. Early results showed fewer flashy launches but higher retention and faster iteration cycles across teams. Over time, experiment logs replaced opinions, revealing patterns in what users actually valued and ignored. The shift became clearly right when failed ideas were caught earlier and resources were reallocated more efficiently. Today it's seen as a core operating principle rather than a temporary process tweak.