> It's extraordinarily difficult to justify bugfixes and incremental improvements using exclusively analytics data Why would analytics be any less capable of c

> It's extraordinarily difficult to justify bugfixes and incremental improvements using exclusively analytics data Why would analytics be any less capable of conveying how these attributes of a software product affect customer purchasing decisions? If the metrics relate to e.g. retention, new users etc., these are independent of which aspects of the software cause them. A correlation is a correlation. Unless there is some reason to think product managers would be especially unaware of quality as a variable to observe.