I was recently part of a meeting where the topic of discussion was a possible reorganization plan. But this was in a university, where—unlike in most companies—reorganizations don't happen very often (thank goodness).
As usual, someone suggested we do a “best practice” investigation by identifying how the other major schools were organized. The problem is, though, that the major schools are organized in several various ways, and there is no particular “best” practice.
Frankly, I've always thought that publicizing the “best practice” idea was simply a way to sell conference tickets and books. The problem, as I see it, is that it's not clear what it means for some practice to be “best”? Best at what? And by what standard? What's best for me might not be best for you, and so on.
But rather than relying on my personal opinion and conjectures about the problem at hand, we had the fortune of having a well-known research colleague who knew the data on organizational structures extremely well. If there really were a best practice, then it would have been borne out over thousands of careful empirical studies and observations. So we asked him what had been learned that we could apply right now.
Our colleague reflected on the question and said careful research shows that the best way to organize a department or business depends on… drum roll, please… what problem you are trying to solve.
Are you trying to solve conflicts between departments, or solve a problem with communications, and so on? The point is that it depends on your specific organization and your specific situation. In short, the concept of best practice was nonsense.