The percentage of a department’s fees paid to its top law firms is one common way to describe law firm concentration. For example, 75 percent of a department’s fees went to 10 percent of its firms (See my post of May 3, 2006: at UPS, 25 core law firms represented more than 80 percent of its costs; Jan. 21, 2008: two firms paid the most; March 13, 2007: for 201 companies in 2005, the median number of law firms that were paid three-quarters of all outside counsel payments was 11 firms; and April 18, 2009: confusing data on concentration of spending.).
Convergence increases this clustering; clustering, however, can occur even with the opposite of convergence (See my post of Feb. 16, 2008: convergence with 26 references; and March 24, 2005: concerning concentration over convergence.).
An untested index of concentration compares the number of matters for which a firm was retained against its rank by number of matters (See my post of May 28, 2009: the h-index and concentration of matters.).
More prosaically, a third perspective on concentration looks at what percentage of the matters in an area of law a given firm handled (See my post of Oct. 22, 2006: the Gini coefficient of concentration.). You could adjust the number of matters handled by firm size, and see which firm gets relatively more of your work when firm size is held constant (See my post of March 9, 2009: fee concentration with firm size.).
A fourth view counts how many firms were paid a substantial amount – such as $100,000 or more during a year – as a percentage of all the firms paid or divided by billions of revenue to correct for company size. Bigger companies should set the threshold higher and some day there may be a standard for this descriptive metric such as a threshold of four times the median payment (See my post July 16, 2005: $100,000+ firms.).
Concentration of law firms and spending on them show up in other posts (See my post of Aug. 21, 2005: law firm usage map; and Dec. 7, 2007: counterpart of law firms concentrating their focus on a relatively small number of clients.).