Amongst the plethora of methods that have been developed over the years to involve users, suppliers, buyers or other stakeholders in the design of new objects, co-design has been advertised as a way to generate innovation in a more efficient and more inclusive manner. Yet, empirical evidence that demonstrate its innovativeness is still hard to come by. Moreover, the fact that co-design workshops are gatherings of participants with little design credentials and often no prior relationships raises serious doubts on its potential to generate novelty. In this paper, we study the contextual elements of 21 workshops in order to better understand what co-design really yields in terms of design outputs and relational outcomes. Our data suggest that co-design emerges in crisis situations and that it is best used as a two-time intervention. We suggest using collaborative design activities as a way to bring about change through innovation.