Generative AI for Decision Making in B2C E-Commerce: A Conceptual Framework Based on Consumer Buying Behaviour

Abstract

Decision making in business-to-consumer (B2C) e-commerce connects activities performed by company employees and consumers before, during, and after a purchase. Existing generative artificial intelligence (GAI) applications usually address isolated tasks such as recommendation, forecasting, content generation, or chatbot communication. This study proposes a conceptual framework in which the five-stage Consumer Buying Behaviour (CBB) process provides a common foundation for GAI-supported decisions. The central component represents need clarification, information gathering, comparison of feasible offers, transaction completion, and post-purchase assessment as a recursive cycle with decision gates. The company component assigns GAI tools and agents to market analysis, product-information management, comparison, checkout support, and service improvement. The consumer component supports need specification, verified information retrieval, alternative comparison, transaction review, and post-purchase assistance while preserving consumer control. An exploratory pilot-validation protocol was applied using paired before-and-after questionnaires and participant-condition records from five employee roles and three consumers. Descriptive medians moved in favourable directions for employee accuracy, consumer decision quality, workload, and task time. The records demonstrate the calculation and reporting procedure but do not establish framework effectiveness. Interpretation is limited by the small sample, fixed condition order, and absence of task-level, technical, and longitudinal evidence.