In the current wave of automotive manufacturing, understanding consumer preferences and needs and implementing precise marketing strategies can help reduce corporate costs and bring greater economic benefits to automakers. This article focuses on accurately understanding consumer purchasing behavior in a complex market environment, providing strong data support for new energy vehicle marketing strategies. This paper constructs a consumer purchasing behavior pattern recognition algorithm. By improving the SOM neural network algorithm to address the pendulum effect in samples, and combining it with the K-means algorithm, a new P2SOM-K-means clustering algorithm is proposed, which features a shorter learning cycle and faster convergence speed. Using the proposed algorithm to analyze the massive consumer data collected, the paper delves into key behavioral patterns of new energy vehicle consumers, including their purchasing reasons, preferences, purchasing methods, and usage channels. The paper categorizes the collected consumers into five groups, with the fifth group being the primary purchasers. This group is primarily characterized by individuals aged 26–35, company employees, and commuting needs. Finally, after deeply analyzing consumer behavior data, the paper proposes targeted strategies for optimizing marketing.
