The domain requirements of software projects often seem so specialized to developers that their original design does not incorporate any commercial-off-the-shelf (COTS) components. However, if major implementation problems are encountered at a later stage in the project, the integration of a COTS component that promises to solve those problems may become a desirable alternative to struggling on with the original implementation. While a number of methods and criteria have already been proposed for requirements engineering, risk assessment and candidate selection of COTS components, they were developed for application in the initial phases of a project and thus do not take into account the much tighter time and design constraints imposedin a later project stage. To spark discussion on necessary adaptations of the established methods, this position paper uses the example of a concrete project to illustrate the characteristics of “switch or struggle” situations and proposes an initial set of risk factors to be considered at that time.