Failures are not reported enough in the literature as the pressure of publishing successful stories has moved us away from considering that even unsuccessful stories can be equally relevant to increase knowledge in research. Analyzing failures can be a great learning opportunity for the community of researchers. In this paper, we present the outcomes of a research that investigated the challenges experienced when designing and running studies with children. The study began with a literature review which led to an inquiry focused on exploring failures within the Child-Computer Interaction community by interviewing 14 researchers. Interviews were analyzed in the vein of the thematic analysis approach. We deliver three themes (unpredictability, technology targeted at adults, children willing to achieve the goal), three action points (the context, the technology, and the activity) and a paradigm shift from "failures need to be hidden"to "failure is a learning opportunity".