WOM, eWOM and WOMachine: The evolution of consumer recommendations through a systematic review of 194 studies

Recommendation-based communication plays a pivotal role in consumer choices. From human sources to electronic word of mouth or different types of recommender systems, recommendations help consumers adopt or reject leads, and can be highly beneficial for organizations. In view of its relevance and the distinct characteristics the evolution of the topic, this paper aims to identify, summarize, and analyze the developments and impact of these recommendations on consumer decision making. To achieve this, 194 evidence-based studies were systematically reviewed. The results from a thematic synthesis showed that eWOM and recommender systems have a synergistic effect fueled by non-verbal cues of eWOM and accuracy of the system. Conversational recommender systems act similarly to WOM by encouraging explicit feedback. However, data privacy concerns make interactions towards these systems a difficult matter. Themes that emerged from WOM emphasized interpersonal relationships that are homophilous and with strong ties. Themes from eWOM focused on volume as a cue for popularity which increased credibility and trustworthiness. Finally, themes for automated recommendation center on usefulness and anthropomorphizing the recommender to build trust. Implications and future directions are provided.