Predictors of Customer Channel Choice in the High-Value Electronic Device Purchase Decision Process

Taralik, Krisztina and Törcsvári, Zsolt István (2024) Predictors of Customer Channel Choice in the High-Value Electronic Device Purchase Decision Process. MULTIDISZCIPLINÁRIS KIHÍVÁSOK SOKSZÍNŰ VÁLASZOK (1). pp. 143-170. ISSN 2630-886X

Preview

Text Taralik-Torcsvari.pdf - Published Version Available under License Creative Commons Attribution. Download (2MB) | Preview

Abstract

Difficulties and uncertainties in the customer journey reduce customer experience, while lower levels of uncertainty improve perceived control and trust, thereby increasing customer experience. Level of uncertainty experienced during the purchase decision may be related to the channel, product, or individual factors. The purpose of this study is to assess the relationships between the channel usage patterns of 415 online survey respondents and four hypothesized groups of predictors: demographic characteristics, the channel used by consumers in the earlier stages of the journey, shopping attitude and technological readiness (TR) factors during high-value electronic devices. Examining the channel usage preferences of the respondents, 6 types of patterns were identified, of which the 4 most frequently occurring patterns were retained in the further studies. A factor analysis was performed to reveal the close correlation between the explanatory variables compiled to measure the latent variables (within different dimensions of shopping attitude and TR). To identify the most significant explanatory variables multinominal logistic regression were used. In addition to the fact that all four groups of predictors contain factors that showed a measurable effect on the respondents' channel choice, it is important to highlight the effect of the need for physical touch and the channel used in the previous stage, which proved to be the strongest predictors. When evaluating our results, it should be noted that the four identified shopping pattern categories were represented in strongly different proportions in the sample, thereby significantly impairing the learning efficiency of the algorithm. Thus, our model can be considered applicable primarily to the "Blended" pattern category, which was overrepresented in the sample.

Actions (login required)

Edit Item