With the development of mobile Internet, all aspects of our daily life have changed, including shopping, dining, entertainment, etc. In order to research the behavior of ordering food delivery, in this paper, we analyze users of Meituan and Ele, the two mainstream food delivery platforms, and obtain several groups with different behavior patterns by community detection. As we know, when people 1 order food delivery, they often contact the delivery guys by calling. Integrating the two behaviors can make up for incomplete information, reduce noise, then obtain more reliable community structures. In this paper, we construct a two-layer network, which are built based on the DTW similarity of behaviors, with data of user's behavior about accessing delivery platform and calling. The polarization trend of degree distribution confirms that there are consumer and working group. Feature integration algorithm is used to complete community division. By analyzing the different types of group, it is found that users with significantly different behavior characteristics can be distinguished.