With the continuous development of modern commerce, various emerging ecommerce platforms, including online shopping, have been widely used. In order to improve the efficiency of different types of user needs analysis and achieve accurate customer service of ecommerce platform, a customer service management method based on data mining and Hadoop platform technology is proposed. This method first uses weight estimation to classify customers deployed on the Hadoop platform, then performs the tagged classification of the products in the platform and implements parameter estimation of the userinterested product. Then using the Map Reduce parallel framework to optimize the traditional data association rules Apriori algorithm to reduce redundant subsets of the data. The simulation experiment results verified the feasibility of the customerprecise service management method proposed in the ecommerce platform, and it shows that the proposed method has good accuracy and operating efficiency.