The advent of electronic commerce has led to its emergence as an indispensable ecosystem in recent years, particularly due to the numerous advantages it offers to consumers. Technological advancements have facilitated the ability to compare a wide range of products and alternative suppliers simultaneously, thereby enhancing the purchasing power of consumers. Consequently, suppliers offering the same product on various e-commerce platforms are compelled to devise diverse marketing strategies to attract users. Consumers, in turn, consider a range of parameters when making a purchase, including price, company evaluation, delivery time, packaging and payment terms. To increase their visibility, companies must also bear costs such as advertising. In the context of a competitive environment, it is imperative for firms to formulate strategies that will enable them to optimize their share of demand. In this study, a hybrid approach combining fuzzy logic and reinforcement learning methods is adopted for the purpose of strategic decision-making. This approach empowers firms to evaluate a range of variables, including price, advertisement, delivery time, promotion type, packaging quality and payment method. A total of 20-fuzzy-rules model the interactions among these inputs in line with expert knowledge, providing recommendations such as reducing price and advertising when market perception is low, and costs are high. The Q-Learning algorithm represents each firm’s state as a vector of fuzzy labels obtained from visibility, cost efficiency, and other parameters. Actions selected by the epsilon-greedy approach are integrated with the FIS recommendations at a 50% influence ratio. Simulation results demonstrate that this hybrid approach enhances the operational and financial performance of e-commerce firms. Overall, the study illustrates that the integration of FIS and Q-Learning provides an innovative and adaptive decision support system for e-commerce, offering a robust foundation for future applications.