This case study will advise the reader on how to use video data publicly available online for research purposes and discuss the challenges and advantages of working with materials that preserve the actions of the participants who appear in the footage. As these actions constitute what we call “social life”, using videos that were not initially recorded for the purpose of research (that is, “naturally-occurring data” or, as I term them, “unmotivated materials”) is a powerful tool to inquire into how society operates and how people use language (and other conduct) to produce social facts. I will use here as example a case study on food delivery service in China, in which online videos posted on a Chinese social media called “Bilibili” were used to explore how food delivery drivers identified contingencies that challenged the social conventions embedded in their daily working practices and how these contingencies were repaired so that the order of their job could be reestablished. I will explain the step-by-step of this project, that is, how relevant data were identified and which actions were taken to guarantee a smooth and fast ethics review process. Also, I will present an example of how I work with online publicly available videos and discuss which implications the use of this type of material might entail for researchers with an interest in working with naturally occurring data.