Should I Stay or should I Go? Bandwagons in the lab

•We experimentally test the model of Farrell & Saloner (1985).•The key question in the model is whether to stick with or abandon a status quo, where abandoning the status quo is irreversible.•The key ingredients in the model are type-uncertainty; endogenous timing of actions; and complementarities in actions.•The main deviation from the model observed in the data is a reluctance to lead. The reluctance to lead is particularly pronounced when leading is a conditional best response, rather than a dominant action.•This is explained well by an Agent Quantal Response Equilibrium.