A-priori and post-hoc segmentation in the design of healthy eating campaigns

Abstract

Nutrition-oriented public health campaigns – both communication and intervention initiatives – often target particular demographic groups, for example schoolchildren, adults at workplaces, older women, presuming that the members of these groups are homogenous with respect to healthy eating. Although such practice may be justifiable from the practical point of view, it is unclear how effective these implicit segmentations are. In this study the authors argue that it is important to transcend demographic boundaries and to further segment demographic groups. A study with 13–15-year-old adolescents is used to prove that segmentation generates valuable insights for planning promotion of healthy eating. Four types of predictive segmentation models are compared – a one-segment model, an a-priori segmentation based on demographic variables, an a-priori segmentation based on behavioural variables and a post-hoc segmentation. The results of the study show that it is useful and also ethical to differentiate people using segmentation methods, since it facilitates reaching more vulnerable segments of society that in general resist change. It also demonstrates that post-hoc segmentation is more helpful for designing healthy eating campaigns.

Keywords:

• segmentation
• healthy eating
• public health campaigns
• social marketing

Notes

^1. In Models 2–4 a two-step (or nested) segmentation is conducted, as this is further segmentation of one demographic group. However, for reasons of simplicity this will not be referred to as a hybrid segmentation in the remaining part of the paper.