The cross-border e-commerce market is booming in the context of digital economy globalization, and the accurate understanding of consumer behavior becomes the key to improve marketing efficiency. This study focuses on cross-border e-commerce consumer behavior characteristics under the perspective of digital economy globalization, and analyzes the shopping behavior data of 25,000 users provided by Tianchi Labs by using RFM customer segmentation model, entropy value method, factor analysis, and improved K-means clustering algorithm based on optimal K-value selection. The study extracted three public factors, namely, activity, purchase value and purchase intention, with a cumulative variance contribution rate of 83.45% through factor analysis, and constructed a cross-border e-commerce consumer behavior indicator system. The results of the cluster analysis show that cross-border e-commerce consumers can be divided into three categories: the first category of users (3079) with short consumption interval, high consumption frequency, low activity conversion rate, belonging to the important value customers; the second category of users (4209) with long consumption interval, low consumption frequency, low activity conversion rate, and low loyalty and satisfaction; the third category of users (17,712) with short consumption interval, low consumption frequency, low activity conversion rate, and remain active despite low loyalty. This study reveals the behavioral patterns and value differences of cross-border e-commerce consumers, providing data support and decision-making reference for platforms to develop differentiated marketing strategies.
