Author(s)

Haoyang Cheng

Corresponding Author

Haoyang Cheng

Abstract

As Internet technology develops and big data becomes prevalent, new retail business models gradually transform consumer habits and business operation modes. New retailing emphasizes the integration of online and offline, relying on data analysis to achieve precision marketing and consumer preference analysis. This paper explores the application of nonparametric statistical methods in new retailing, analyzing consumer behavior and optimizing marketing strategies. Nonparametric statistics overcomes the limitations of traditional parameter statistics on distribution assumptions. Furthermore, it applies to situations where data characteristics are unknown or do not conform to a specific distribution. This study uses nonparametric methods, including kernel density estimation, rank sum test, and Kruskal-Wallis H test, to analyze the transaction data from an e-commerce platform. The results indicate significant regional and temporal characteristics in consumer preferences. Additionally, this study employs the Bootstrap method to assess the impact of marketing activities and proposes corresponding precision marketing strategies. The research findings demonstrate that the consumer preference analysis realized by nonparametric statistical methods can effectively guide retailers to formulate precision marketing strategies to improve user experience and sales performance.

Keywords

New retailing; Consumer preference analysis; Nonparametric statistics; Kernel density estimation