Customer Segmentation Using Real Transactional Data in E-Commerce Platform: A Case of Online Fashion Bags Shop

Customer segmentation has been widely used in different businesses and plays important rules in customer service. How to get a suitable segmentation based on the real transactional data to fully mining the hidden customer information in the massive data is still a challenge in current e-commerce platforms. This paper develops a customer segmentation model for online shops and uses the real data from a fashion bag store as a case. This paper firstly conducts a data preprocessing to select the main customer features, then it constructs a segmentation model based on the Fuzzy C-Means algorithm, and finally accomplishes a customer prediction model using a probabilistic neural network to estimate new customer’s customer type. The results show that the customer samples are classified into three types, and the prediction accuracy is more than 90%. After that, this paper demonstrates the typical features of each type of customer and compares the new group features with the prior VIP groups. The ANOVA analysis test results show that the new groups have more significant differences than prior VIP groups, which means more effective segmentation results