چكيده به لاتين
Nowadays, e-commerce websites offer product bundles with special discounts to sell more products, but bundling these products does not follow any special standard. In these websites, product bundles are the same for customers with different values. Meanwhile, it is against the fact that each customer creates different value for company and each section of the market has unique features, which affect the customer’s buying patterns.
In this research, we have presented a new model for direct marketing in e-commerce based on market segmentation. First, customers are segmented based on their loyalty to company. After that, each segment of customers is divided into different groups based on the demographic and geographical variables. Eventually, a product bundle is identified for the customers by clustering the purchased products in each group.
To validate the model, we have used classification models for predicting the customer's favorite product bundle. In this process, we have used customers profile data, transactions information and product categories to predict the customer’s favorite product bundle. As a result, we can do direct marketing by sending emails to the customers and recommend a suitable product bundle to them.
Keywords: Direct marketing, Market segmentation, Customer loyalty, Product bundling, Personalization, Electronic commerce
