<![CDATA[Purpose—This study addresses the growing disconnect between high social media engagement and actionable decision-making outcomes among Indonesian SMEs. While firms generate large volumes of interaction data, they often lack the analytical frameworks to interpret customer behavior. Design/methodology/approach—To address this issue, the study proposes a cognitive-behavioral model integrating natural language processing (NLP)-based sentiment analysis with engagement metrics. Findings—Using a dataset of 195,513 social media interactions, regression results indicate that sentiment significantly influences engagement (β = 0.43, p < 0.01), while engagement strongly predicts decision outcomes (β = 0.51, p < 0.01). Mediation analysis confirms that engagement partially transmits the effect of sentiment (indirect effect = 0.22). Research implication/limitation—The findings demonstrate that sentiment-driven engagement serves as a critical mechanism linking digital interactions to behavioral outcomes, offering a data-driven foundation for adaptive marketing strategies in SMEs. Originality/value—The novelty lies in conceptualizing engagement as an observable decision-making process rooted in cognitive sentiment dynamics.]]>
Copyrights © 2026
