Founders, executives, and managers all have to hire for roles they have never done themselves. Most people anchor a career in one craft. They were a great engineer, a great salesperson, or a great designer, and then they have to build a team around work they have never practiced. A founder hires across every function the company needs: individual-contributor engineers, salespeople, marketers, specialists like revenue operations, someone to run finance. Early on the company is too small for a real executive team, so they hire specialists whose jobs they barely know; later they build out functional leaders for sales, marketing, engineering, product, and finance. The chance that one person has done all of those jobs is zero.
Knowing what great looks like is the calibration you do before you evaluate anyone. You use the MOC to define the mission, the outcomes, and the competencies of the role, and then you figure out what the best people in the world at that job actually do: the behaviors they have, the values they hold, what their week looks like, who they hire. Without it, the interview defaults to whoever is most impressive in the room.
It is not doing the job yourself, and it is not having done it. You will never have been a great CFO before you hire one, but you do need to know a thing or two about finance. For any role, it is hard to know what a 4 or a 5 out of 5 looks like if you have not gotten yourself to a 1 or a 2 [2]. A CEO building software has to understand how software gets built; a CEO building shoes has to get into the materials and the manufacturing, or the odds of making a great shoe are low.
Ultimately, knowing what great looks like is a virtue with a vice on either side of it [1]. On one side is abdication: hiring someone and checking out, with no way to monitor the work or even tell whether it is any good. On the other is perfectionism: believing that even great people should have no flaws, and ending up with a team of excellent people you are constantly disappointed in.