[jl2718]

I have a theory that management can be connected to portfolio theory as a variance-reducing process. Outcomes are random but bounded in expectation by the efficient frontier. When Google started, they hired the smartest people and enforced zero management and zero hierarchy, and thus were operating at the high point of both expectation and variance of outcome. Instantiation of OKRs brought their operating point down to something more predictable, but at a tautologically lower expected outcome. They notice this and keep trying to remove management on special groups like X, but then some new brilliant management idea comes along, and they adopt it with roughly the same results as every other new management idea. It's like the investment industry before Benjamin Graham.