Customer portrait is the customer information that is labelled and digitised. Collecting and studying consumer data is highly helpful to enterprises because it empowers them to use these data to locate target groups and meet various customer needs. With the continuous rise of the e-commerce market, major e-commerce websites list providing customer value and being data-driven as company values. Traditional customer portraits rely excessively on objective and past experience; so, a simple but powerful customer portrait system is required to make the customer analysis platform simple and intelligent, and to make the data accumulated at ordinary times bring practical value to e-commerce companies. Based mainly on addressing this requirement, this paper proposes an effective customer portrait system for e-commerce. First, the research background and significance is expounded, which lays a theoretical foundation for this research. Then, the customer portrait system is designed, including the overall design scheme and functional framework of the system. The improved FCM algorithm is used as the basis for building the customer portrait model, and the performance of this algorithm is analysed and compared with that of the improved algorithm used in this paper.