Comment by curiousgibbon

It's entirely nonsensical to use O as a lower bound though. You could have two companies no growth whatsoever in value and correctly state that one has O(n) growth rate and the other has O(n^2) because a constant is both O(n) and O(n^2) (and O(n!) and O(exp(n^n)) ...). The author is trying to argue that there's some separation between two hypothetical startups' growth rates and as such an upper bound on one, say O(n), and a lower bound on the other, say Omega(n^2), is warranted. It sounds like you're not entirely an expert despite your recently-passed final. Strange concept, eh?