Development of an influencer-brand congruence and complementarity measurement framework for recommender systems

Abstract

Currently, there is a growing interest in identifying the prerequisites of a successful influencer brand collaboration and for brands to determine the most appropriate matches among online creators. While previous studies have addressed this challenge using credibility or attitude homophily measures, this study presents a new perspective by combining existing findings with recommender system mechanisms. Through the application of exploratory and confirmatory factor analysis, we delineated eight factors and 54 items that describe the personalities of both brands and online creators. Subsequently, by employing principles from collaborative item-based filtering, we utilized interval and binary similarity measures to compute congruence scores between a brand and a content creator based on the prerequisite of the Kamins match-up hypothesis. Our findings indicate that Euclidean distance serves as the most efficient congruence measure for the proposed recommender system. It demonstrates superior capability in capturing a broader range of nuances compared to other tested measures, including the Pearson correlation coefficient, Jaccard similarity, and simple matching.