Eliciting brand-related social media engagement: A conditional inference tree framework

Abstract

This study presents a machine learning approach using conditional inference tree (Ctree) to determine cognitive patterns that elicit consumer engagement into social media. Using the Ctree algorithm, a predictive model was computed using self-reported data on consumers' perceptions of brand equity and engagement into brand-related social media behavior from a sample of 1356 individuals. The predictive modeling analysis revealed 5 different cognitive patterns (rules) that stimulate brand-related social media behavior. Each rule comprises behavioral engagement discriminating low, medium, and high levels of consumption, contribution, and creation of brand-related social media content. Furthermore, based on the different patterns, the analysis portrait a typology of 5 subtypes of consumers according to their behavior, which by complementing the predictive analysis information may be used to stimulate different levels of consumption, contribution, and creation of brand-related social media content.