A class of its own: latent class segmentation and its implications for qualitative segmentation research

From an early age we categorise the myriad stimuli we are confronted with. This adaptive process continues throughout our lives, and finds a natural expression in marketers' desire to segment their consumers into different types. Argues that traditional attitudinal segmentations can prove disappointing when looking for differences between groups on behavioural or brand preference measures. Clusterwise regression, a form of latent class segmentation, offers an alternative approach to establishing meaningful market segments, and we present a case study using this technique within the health and beauty sector. Argues that the method's emphasis on the importance of different attributes to different groups of people could be usefully incorporated into qualitative group methodology to produce meaningful and robust market segmentation.