[swyx]

in an ideal world, yes it'd be nice to quantify. in practice, it would take extra dev time to quantify. time that we dont have. we have to rely on our engineers' subjective feelings about our engineering practices.

i understand the desire to be data driven. and your point about making everything a blocker. if you're at a point where you dont trust the judgment of your colleagues, you should drop everything and address that right away.

but, most of the time, operating in good faith works.

[qbasic_forever]

Be careful, this is how institutions bake-in bias. Subjective feelings are littered with implicit and explicit bias at every level.

At the end of the day it's all just people management problems. You're right it's a time sink for devs to quibble over and quantify prioritization--that's why I think it's more of a thing for management to assess and review. Trust devs to assess priority on their own, but verify with regular followup and review that those assessments are accurate. Business needs will change too and that might alter priorities--what was dust last week is now an enormous mountain because customer X demands the obscure feature no one cared about before.