Along with the rapid development of e-commerce, social media network application development is rapid. According to relevant data, due to micro-blog users and online shopping consumers have a high degree of coincidence, in recent years, many power companies to use micro-blog to do their own products to promote sales of the network. Electronic business enterprises to social media network as a tool to promote their own brand of culture, to provide users with product-related knowledge and after-sales service, collecting the latest market consumer information, and establish links with potential users; to enhance their brand competitiveness, Low, return as high as possible marketing purposes. Social media network platform has its unique advantages, relative to the traditional communication platform, more open environment, user interaction and information accuracy and so on. And especially Sina micro-blog as the media from the era of the most prominent representatives of electricity companies can use its spread quickly and more accurately spread their own brand concept, the latest information on products developed by enterprises to strengthen ties with consumers and so on, Making the first time users receive the latest information and follow-up services and so on, and to this end to stimulate the purchase of consumer demand, and enterprises to maintain closer ties, enhance the trust relationship. Sina micro-blog to bring network users interactive experience can promote their emotional exchange and trust increase; and microblog own interesting to attract users to participate in various activities, desire, and for the harmonious relationship between communication. Therefore, Sina micro-blog for the generation of social media network on the consumer’s network trust more and more, and ultimately lead the consumer’s online purchasing decision-making behavior. In this paper, we use network trust as an intermediate variable to analyze how user’s cognitive ability, relationship strength and interaction affect consumers’ online shopping decision, and use multi-element Level regression of the open Sina micro-blog data set experimental validation.