Cluster Analysis – Data Segmentation

Cluster Analysis is used to segment items or people into homogeneous groups. In marketing, segmentation often results from the need to differentiate based on positioning, and requires data relating to behavior and attitude. But it is also often used when analyzing performance, for instance of affiliate companies, points of sale, distributors, etc.; for clustering satisfaction of internal and external customers; and also for treating profiles of products, companies, geographic regions, and so on.

The many available clustering methods can be divided into two main groups:

• Hierarchical methods, which group data row by row and do not require specifying in advance the desired number of clusters.
• Partitioning methods, which assign items to a user defined number of clusters.