The new round of power reform, energy internet and new generation digital technology has greatly promoted the development of integrated energy services. In order to develop a specialization integrated energy services plan, it is necessary to make a portrait of business scenes. Firstly, a set of indicator system is established, and typical daily load curves are extracted using Piecewise Aggregate Approximation method and Symbolic Aggregation Approximation method to calculate indicators to obtain the data base. Secondly, the clustering results of eight basic scenes are obtained using K-means cluster analysis method, namely, single-peak type which can participate in demand response, single-peak type which cannot participate in demand response), double-peak type which can participate in demand response, double-peak type which cannot participate in demand response, post-single-peak type and continuous power consumption type. Finally, integrated energy services solutions are proposed for the six business scenes respectively. The results show that the power consumption characteristics of each business scenes are very different, and it is necessary to develop specialization integrated energy service solutions for different scenes.