What is the optimal company headcount?

I got a topic from Kade Ross (@kadeross) last night related to an article about the effect of increasing headcount on employee productivity.  You can read the article here.  The gist of the article is this:  for every 10% you increase company headcount, productivity drops by 6.3% for each employee.  Kade calls this the borg, a state of existence where nothing gets done and people look for ways to get nothing done.

In the article, the author puts the main cause of the productivity decline as issues with communication.  Dissemination of vision, company objectives, etc. get harder as the company gets larger.  I think that is some of it.  However, having worked at Symantec ( a company of 18,000 ), I think the cause of this is something far more harmful.  As companies grow in size, most introduce new layers of management ( bureaucracy ).  The more layers you have, the more you breed a culture of meetings as opposed to doing.  Meetings are required because you have to get everyone’s opinion, and make them feel a part of the process.  Meetings also allow the new management layers to carve out niches of independence and protection, where “protecting my turf” becomes the new rallying cry instead of “kicking ass.”

Kade puts the optimal headcount over/under at 35.  I put the over / under at somewhere below that, somewhere around 22.  It obviously depends on where the people are, but if the assumption for this exercise is that all employees are local, I think a small number allows everyone to be close enough to each other to hear what is going on and jump in on each other’s problems.  Assuming the standard 2-3 founders, that leaves 18 or so spots to fill with people that you know will kill it.  Most of us know that many people or know where to find them.

What do you think?  Can a big company somehow overcome the seemingly immutable laws of meetings, inefficiency and turf wars?  If so, how big a company can survive?  If not, what do you do when your company starts to hit these walls?