When Bell Labs began, only about 15% of its employees did research.
Of the two thousand technical experts, the vast majority worked on product development. About three hundred, including Clinton Davidson and Mervin Kelly, worked under Harold Arnold in basic and applied research. Bell labs was primarily not researchers
Most of the work was to create Continuous technology improvements - better telephone poles, more efficient vacuum tubes, cheaper cable sheathing.
For curiosity, I kept track of every single named employee mentioned in The Idea Factory - there were 64 of them. Sixty-four people over an approximately 70-year period, in an organization that employed thousands of individuals from day one. Obviously this number is skewed smaller by the narrative constraints of a book, the tendency to over-concentrate attribution on a few individuals, and the limits of historiography itself. Obviously hundreds of other people - technicians, support staff, managers - were crucial to the projects. However, it does mean that even within minority of people at Bell Labs doing research most of the work was not particularly notable.
There’s a question of counterfactuals - could you have gotten the output of Shannon and Shocklee without an organization the size of Bell Labs? Some parts were presumably irrelevant (the guys trying to eke the last 1% out of vacuum tubes? The guys figuring out how to make telephone poles last 5% longer?), but it is incredibly hard to say from from the outside, and may have been hard from the inside. I can only assume they didn’t support projects they knew to be worthless. The issue is that even on top of having the tacit and manufacturing around to act on the ideas we today recognize as valuable, The work at Bell Labs that was most valuable to AT+T and the work that was most valuable to the world was fairly disjoint. The latter subsidized the former.
This is an uncomfortable reality because it slams you in the face with the fact that Sturgeon’s Law applied even in arguably the most optimal situation possible - a monopoly whose massive profits literally depended on better technology and plowing some of the profit back into flashy research.