Cluster Analysis/Discriminant Analysis

If the question is “what segments are important to my business, and how do they think or behave?” then cluster analysis or discriminant analysis offers the answer. Both methods divide your marketplace into mutually exclusive and exhaustive segments based on similarity on a set of key criterion variables (demographics, attitudes, media usage habits, etc.).

Once segments are identified, any number of analytic methods can be employed within each segment so that critical differences in interest drivers, design preferences, or any other matter of interest can be identified and successful targeted marketing strategies developed.  When the segments are to be derived from similarities and differences across the criterion variables themselves, cluster analysis is the technique to use. If similarities in the way that the criterion variables relate to some other factor (customer satisfaction, for example) is of interest, then discriminant analysis is the way to go. Either way, you have the tools you need to get the different segments within your customer base the information they need concerning your product.