THE EFFECT OF VISUALIZATION OF CHOICE OPTIONS ON ACTUAL CHOICE

Consumers have to choose among a great variety of products with different attributes such as brand, design and price. Companies are eager to find out why consumers choose their product or why they do not. The product choice can depend on the product’s characteristics, but it can also depend on the way the attribute levels are visualized. A clear understanding of what drives the consumer’s product choices can lead to several improvements, resulting in higher sales and market shares. The goal of this thesis is to investigate how different visualization approaches affect consumer choice behavior. The choice behavior is investigated using a choice-based conjoint (CBC) study for the choice of a credit card. Each respondent has to decide which of the product concepts shown in several successive choice tasks he prefers. Because people can react differently to a company’s specific credit card offering, individual-level “part-worth” utilities are allowed for each respondent. A utility is a value representing the attractiveness of each feature in a conjoint study. Individual-level part-worth utilities allow for easy segmentation as they provide a way to detect different groups of respondents. This study uses the hierarchical Bayes (HB) estimation because of its ability to provide reasonable estimates for the utilities, based on only a few choices by each respondent. The software that is used for the estimation procedure is Sawtooth’s implementation of CBC HB. To verify the results produced by Sawtooth, a HB estimation procedure has been implemented in the open source language R as well. In this thesis, it is proven that visualization techniques have an effect on consumer choice behavior. Therefore, they are included in the design of market research exercises in order to increase realism, external validity, and to provide recommendations to optimally communicate a product line-up. The results were presented at the 23rd AMAs Annual Advanced Research Techniques (ART) Forum on June 25, 2012 in Seattle (WA, USA). The content of this thesis is based on an internship at the SKIM, an international market research company specialized in conjoint analysis. The internship was jointly supervised by the Chief Methodology Officer G. Loosschilder, Research Director K. van der Wagt, Project Manager C. Borghi and Prof. Dr. P. Goos.