> "would any user benefit from this"? [yes, no] This doesn't work... Unnecessary user accounts, telemetry, inefficient electron app etc etc: It's possible to g

> "would any user benefit from this"? [yes, no] This doesn't work... Unnecessary user accounts, telemetry, inefficient electron app etc etc: It's possible to genuinely argue they all ultimately benefit the user (e.g market research, developer efficiency). Some people would use this to push their agenda. It's also not a good idea to reject ideas that do not directly benefit the user, because many of those are good. What's missing is quantitative and qualitative assessment. Sticking to the decision tree format, that node needs to be a continuum. e.g "does-it-benefit-the-user" changes to "cost-to-user/benefit-to-user > x". You can also distinguish between quality of benefits and costs by weighting the inputs... and while we are at it "cost-to-bottom-line/benefit-to-user > x" would remove a lot of busy work. But realistically this is reasoned about in non mathematical terms, effects are nuanced and the more you dig the more you find. The best approach is human and imperfect, but can at least be considerate of the user experience as a whole. This I believe is the heart of the problem: Decisions being made in the abstract when not appropriate, e.g a department only considering the direct effects to that department and ignoring effects to the ultimate purpose of the department->company->service/product. Perhaps it's just the result of over siloed workers without any empathy or concern beyond their short term personal benefit.