Latent Class Segmentation

Latent Class (LC) segmentation operates under the assumption that there are groups underlying the data that give rise to segments. These groups are "latent" or not directly observable.

LC techniques use formal statistical modeling to get at these segments, unlike most other segmentation methods. Measurement of segment fit is more robust because of this. An advantage with LC segmentation is that we can designate a dependent variable and simultaneously conduct a segmentation and key driver analysis. This allows segments to be formed around the key drivers, a particularly useful device in certain circumstances.