My experience in a large ‘industry’ (not IT) is that we: a) are very bad at knowing how to measure performance b) mostly only measure performance as an individual contributor, and not as a manager or leader. (Which is to say: of course managers and leaders have their performance assessed, but it is often via the same tools/questions used for individual contributors, rather than specifically measuring skills in management or leadership domains) Then you overlay this with human factors such as nepotism, conflict aversion, politics, friendships… and of course we screw it all up overall. —- So, to your point, absolutely we’re bad at selecting people who have the aptitude to be a good manager or leader for such roles, and terrible at weeding out people who aren’t suited.