Then how can you know if building the feature made any sense?
Engineers don't usually get to decide which features to build.
Someone needs to have data to validate a feature has value. Otherwise the company is just operating on the GIGO principle.
I’ll admit, companies like that do exist and somehow survive!
I should think the vast majority of companies operate without data on the value added by specific product features. The cases where you can assign a dollar value to a specific feature are the exceptions, not the rule.
It's not necessarily a case of GIGO, either. Take VSCode, for example. Tons of useful features. How many of them can be linked directly to increases in Microsoft's revenue?
One would hope that the VSCode devs are trying to build a good text editor according to their own taste and intuition (while also listening to outside feedback). It would be a complete waste of time to try to 'validate' every feature added.