A latent class segmentation analysis of e-shoppers

We apply a latent class modeling approach to segment web shoppers, based on their purchase behavior across several product categories. We then profile the segments along the twin dimensions of demographics and benefits sought. We show that benefits sought can provide more diagnostic information than mere descriptive demographic profiling. Our study has some interesting findings that shed light on consumer perceptions and behavior with respect to online commerce. First, consumers are more concerned about web attributes that are associated with perceived losses than with web attributes that consumers associate with gains. Second, compared to other online purchase-related attributes, getting the lowest price does not appear to be a very important attribute for web shoppers. This finding may also indicate that prices on web pages are somewhat similar, and consumers are moving on to other criteria to continue their evaluation process. Third, demographics do not discriminate between web buyers even though that has been the traditional focus with the Internet. Fourth, there is a large segment of web surfers who dislike buying on the Web; the predominant reason for this appears to be their perception about the security of sensitive information. This segment feels that not only is it the most important of all attributes for online commerce, but also that the Web does a very poor job on this attribute/benefit. (C) 2002 Elsevier Inc. All rights reserved.