Clustering in Market Research
One of the chief competencies of an effective market researcher is to identify groups within a population that may be more likely to purchase the researcher’s firm’s products. If groups are defined well enough, targeted marketing materials can lead to huge ROI. In practice, this process is referred to as clustering.
The basic approach goes like this:
1. Survey- The survey should be designed to provide multiple measures of propensity to purchase and preferences for the product. The survey should be administered to the population of interest and the sample size should be large enough to make statistically sound conclusions.
2. Analysis- Before the cluster analysis is performed, it is best to do a factor analysis on the survey to minimize the factors we cluster with. If several questions are found to measure the same thing, we should combine these before the cluster analysis is done. This analysis can all be done in SPSS. After the data reduction, do a cluster analysis and decide how many clusters are appropriate and save the cluster assignments. Now you can view the means of all the factors across the clusters.
3. Action-Naming the clusters helps companies target the clusters correct. Look for differences in the means of factors and name the clusters based on these differences. These differences are used to design your marketing strategies to ensure that the right customers are approached with the right product in the right way.