<p>The growing demand for new energy vehicles (NEVs) highlights a significant gap in existing recommendation platforms, which often fail to adequately match consumer expectations with available products. While numerous reviews influence consumer behavior, current platforms lack automated systems to analyze this information for personalized suggestions. This study addresses the new energy vehicle recommendation platform (NEV-RP) as a two-sided matching decision process, focusing on improving the satisfaction degree of supply and demand. To fill this gap, we propose a two-stage two-sided matching framework that considers consumer herd behavior. This framework utilizes Probabilistic hesitant fuzzy sets (PHFSs) and prospect theory to reflect real-world conditions. Initially, we define a consumer dynamic expectation function informed by consumer herd behavior, applying PHFSs to represent expectation and evaluation data. Subsequently, we construct a profit and loss matrix to calculate the comprehensive prospect value across both stages. Weights for each matching stage are determined based on market characteristics. Ultimately, we formulate and solve an optimization model aimed at maximizing matching satisfaction and platform profit. The framework’s effectiveness is validated through a case study of a NEV-RP, demonstrating its practical implications for enhancing consumer purchasing experiences and boosting trading volumes.</p>