Cluster Analysis

Also known as: Clustering, K-means, Statistical Segmentation, Consumer Clustering

Statistical technique grouping cases (consumers) with similar characteristics into mutually exclusive clusters or segments.

Cluster Analysis is an unsupervised machine learning technique that groups cases (typically survey consumers or respondents) into groups or clusters such that cases within a cluster are more similar to each other than to cases in different clusters.

Most used algorithms in research: K-means (requires predefining number of clusters, very fast and scalable), Hierarchical Clustering (doesn't require predefining k, generates a dendrogram, better for exploration), and Gaussian Mixture Models (more flexible, allows clusters of different shapes and sizes).

It is the most used statistical technique for building market segmentations from survey data. At Atlantia, we apply clustering on U&A data, attitudes, and behavior to build actionable segmentations that go beyond traditional demographic segmentation.