Cluster analysis methods will always produce a grouping. The groupings produced by cluster analysis may or may not prove useful for classifying objects. If the groupings discriminate between variables not used to do the grouping and those discriminations are useful, then cluster analysis is useful. For example, if grouping zip code areas into fifteen categories based on age, gender, education, and income discriminates between wine drinking behaviors, it would be very useful information if one was interested in expanding a wine store into new areas.
Cluster analysis may be used in conjunction with discriminant function analysis. After multivariate data are collected, observations are grouped using cluster analysis. Discriminant function analysis is then used on the resulting groups to discover the linear structure of either the measures used in the cluster analysis and/or different measures.