My biggest issue with metrics and general "data" focus, is that it's incredibly hard to generate data about goodwill. eg we can prove that call to cancel reduce

My biggest issue with metrics and general "data" focus, is that it's incredibly hard to generate data about goodwill. eg we can prove that call to cancel reduces cancelations by 20%, but we have no way to measure how many customers now hate your company and tell their friends never to use your company, how many people would have returned but now won't, how many people just cancel 3 months later. In Facebook's case, we know they're seeing huge hits in active, engaged users, but their whole metrics ecosystem is built to optimize micro-engagement. Taken as a whole, those micro-optimizations are killing goodwill and turning off users, but it's much harder to measure that. So many decisions are made because variant b "won", without any discussion of the layers upon layers of factors that aren't easy to measure.