Academics who study the companies in an industry sometimes use a statistical technique called cluster analysis. “Cluster analysis procedures generally use simple mathematical measures of distance or likeness to group individual firms on the basis of how similar or dissimilar they are to one another. The average numerical values on the attributes for the different groupings are then used to infer what the different clusters represent.” The quote comes from Laura Empson, ed., Managing the Modern Law Firm: New Challenges New Perspectives (Oxford Univ. Press 2007) at 163 (by Peter Sherer). What you can learn from cluster analysis is influenced by differences in the magnitude and variance of the attributes.
The chapter cited describes how a cluster analysis was done of major US law firms. A general counsel might prepare a cluster analysis of the law firms retained by his or her law department. If done properly, the analysis might disclose unappreciated characteristics of their size, cost, location, breadth of services, period of representation, leverage and more.
Second, there may come a time when a sufficient set of descriptive benchmarks for legal departments is available. It would then be possible to do a cluster analysis to determine which law departments are similar to which other law departments and in what ways.