In order to solve the problem of optimizing the cost of data network layout in urban agglomerations, this paper uses the data of urban Internet industry size and population structure to build an evaluation model of urban network access value, and puts forward an optimization method of network layout in urban agglomerations based on network access value. According to the Internet usage scenarios, the network demand indexes of urban industry and consumer data are obtained, the network access value indexes of different cities are evaluated, and the minimum spanning tree algorithm is improved by combining the network topology, urban geographical location, and K-means clustering. The population and industry data of 21 cities in Guangdong Province are taken as examples to verify the method proposed in this paper. The layout decision results show that the cost is reduced by 66.59%. It is suitable for data network layout decision among urban agglomerations of different regions and different development levels, which ensures the urban development and achieves the lowest overall layout cost, and has high reliability and application value.