Abstract We conduct a simulation‐based experiment to analyze how past experiences in a service facility system affect customers’ and service providers’ behavior. We study the problem faced by service providers in deciding by how much to adjust the capacity of their facility when captive repeat customers choose which facility to join based on their expected sojourn times. The customers’ decision‐making process differs from that of service providers in that the former is based on the customers’ experience, whereas the latter is based on the most recent information service providers have regarding demand. Customers use their previous experience and that of their neighbors to update their perceptions of the future average sojourn time, whereas service providers form perceptions of the future arrival rate. We use cellular automata to model the interaction between customers and service providers. We perform simulations to assess the way the customers’ and service providers’ decisions evolve and affect system behavior. Our findings demonstrate that the system we study possesses a certain degree of path dependence. The primary conclusion is that the more conservative service providers are regarding new information, the larger the market share they achieve, and the lower the probability that their facility will shut down.