Data-driven Consumer Analytics from Social Media User-generated Content

This study examines how user-generated content (UGC) shapes consumer perceptions of peer advertising on social media. Drawing on cognitive learning theory, social exchange theory and social identity theory, the study employs a two-phase mixed-methods design combining an experiment and a survey to identify and validate key perceptual drivers. The findings indicate that peer-generated promotional messages, particularly those reflecting authentic, informal and experience-based communication, generate stronger trust and persuasion than traditional or influencer-based advertising. Five key drivers are identified: message informality, information quality, experimental message (EM), hedonic motivation (HM) and source derogation. The results further show that these factors are reflected in observable characteristics of UGC, including linguistic style, depth of information, emotional tone and relational cues among users. Such characteristics can be systematically examined using techniques such as text analysis and sentiment evaluation to better understand consumer responses. The study also conceptualizes these constructs as measurable analytics features derived from UGC, which can be operationalized using artificial intelligence-based techniques such as text mining and sentiment analysis. By linking perceptual constructs with analytics capabilities, the study advances an informed grounded theory approach to social media consumer analytics and provides practical insights for designing more effective marketing strategies.