Growing businesses need e-commerce systems; marketers must help analyze data and make decisions. Data mining and decision support systems optimize marketing campaigns by analyzing customer preferences and purchase habits. With this combination, personalization, trend spotting, client segmentation, and cross-selling improve customer satisfaction, conversion rates, and Return on Investment (ROI). This paper aims to conduct a thorough study on the optimization of e-commerce marketing strategy based on data mining and decision support systems to solve the issues of high-cost investment in e-commerce platforms, low efficiency of online marketing, and improve the sales performance of e-commerce platforms. First, a Browser / Server mode (B/S) architecture based on e-commerce marketing system is built. The system is broken down into interface, functional, and support layers. The support layer includes multiple system interfaces to achieve data exchange across various functions inside the network marketing channel. The functional layer includes network marketing, service monitoring and management, basic information management, and other modules. Subsequently, the e-commerce platform utilizes the decision tree algorithm to optimize its online marketing strategy. This involves filtering the data available on the platform and subsequently computing the data structure and dissimilarity matrix. The ultimate objective is to determine the closest data class distance. The decision tree data mining algorithm is utilized to optimize the online marketing strategy of the e-commerce platform. The method proposed in this paper has been validated through experimental verification and has demonstrated a robust data throughput capacity. It has also been shown to effectively enhance the efficiency of online marketing. Furthermore, the system designed in this paper has the potential to enhance the economic benefits of e-commerce platforms.