To understand the evolution of interaction between collaborative e-commerce vendor and consumers，we propose an agent-based computation model upon scale-free network based on vendor’profit model，the profit model is based on empirical result from commerce community，and program the system with NetLogo to simulate interaction between vendor and consumers． The experimental results show that better fineness are existing between experimental data and classic diffusion curves，it indicates that the experimental model can be used to simulate evolution of users’transfer behavior． If innovators or early adopters are dominant in network，online shopping can reach higher peak，and manager can get higher expectation profit while making lower management efforts，but late adopters are dominant in network，even significant efforts are made，lower online shopping lever can reach． Vendor can gain more profit by encouraging consumers to adopt several social tools，which can increase nodes of the network and change the network topology．