Past studies of commercial, industrial, and institutional (CII) water use efficiency have been limited to simple normalization benchmarking measures. Simple normalization ratios (e.g., water use per employee) account for only one driver of water use and are often misleading because of the predominantly skewed nature of water use intensities. In order to improve on benchmarking systems to assess the water use efficiency of CII customers, this paper explores two popular methods used in the energy efficiency field: ordinary least squares (OLS) and data envelopment analysis (DEA). Parcel-level data from Austin, Texas, is used to demonstrate the methods, where a data-driven approach is employed to include average water use, building area, building value, year built, parcel area, and number of employees as measures of input, and annual sales as the economic measure of output.