Geodemographic segmentation: PSYTE Canada and PSYTE Quebec

Tom Exter, MapInfo Corporation
Ian Mosley, MapInfo Corporation

This paper describes the methodological framework and results of new geodemographic cluster systems for Canada. The national system includes 65 neighbourhood clusters, ten of which are specific to the Province of Quebec. A derivative cluster system specific for Quebec involves 36 clusters. These new Canadian cluster systems required extensive processing of 2001 Census data along with other source databases, including shopping centre locations, automotive purchase data, and “lifestyle” survey data. However, the key contribution this paper makes to the literature involves the use of neural net routines in combination with hierarchical clustering techniques in a two-stage methodological framework. The authors have shown that the clustering methods used in Canada (and by extension, Quebec) have wide applicability for similar clustering projects in other countries. The paper includes a discussion of geodemographic fundamentals, recent advances in clustering methodologies, and examples of how the results benefit business decision-making in a consumer marketing context.