The ant-based clustering algorithm needs to set several parameters and cluster for a long time.The ant- based clustering algorithm based on entropy, which uses entropy to amend the ants picking and dropping rules, can reduce the number of parameters and shorten the time of clustering. This paper intends to apply it to the customer segmentation, and compares the segmentation results with the traditional ant-based clustering algorithm. The experiment shows that the ant-based clustering algorithm based on entropy can accelerate customer segmentation.