Which UGC features drive web purchase intent? A spike-and-slab Bayesian Variable Selection Approach.

Purpose – The purpose of this paper is to identify user-generated content (UGC) features that determine web purchase decision making. Design/methodology/approach – The authors embed a spike-and-slab Bayesian variable selection mechanism into a logistic regression model to identify the UGC features that are critical to web purchase intent. This enables us to make a highly reliable analysis of survey data. Findings – The results indicate that the web purchase decision is driven by the relevance, up-to-dateness and credibility of the UGC information content. Research limitations/implications – The results show that the characteristics of UGC are seen as positive and the medium enables consumers to sort information and concentrate on aspects of the message that are similar to traditional word-of-mouth (WOM). One important implication is the relative importance of credibility which has been previously hypothesized to be lower in the electronic word-of-mouth (e-WOM) context. The results show that consumers consider credibility important as the improved technology provides more possibilities to find out about that factor. A limitation is that the data are not fully representative of the general population but our Bayesian method gives us high analytical quality. Practical implications – The study shows that UGC impacts consumer online purchase intentions. Marketers should understand the wide range of media that provide UGC and they should concentrate on the relevance, up-to-dateness and credibility of product information that they provide. Originality/value – The analytical quality of the spike- and- slab Bayesian method suggests a new way of understanding the impact of aspects of UGC on consumers. [ABSTRACT FROM AUTHOR]