Abstract: To enhance the refined service level of power enterprises, a segmentation model based on customer electricity consumption behavior was constructed. By leveraging electricity consumption data to classify and model customer groups, differentiated marketing service strategies were designed in conjunction with typical consumption characteristics. Clustering algorithms were employed to categorize and identify customer electricity consumption behaviors, establishing a feature indicator system to match corresponding service solutions, which were subsequently validated through regional pilot projects. The research results demonstrate that the differentiated service model effectively improved customer response rates and power grid operational efficiency. This approach provides a feasible pathway for power enterprises to advance digital transformation and precision services.