Online platforms such as Facebook, YouTube and Twitter offer a wide range of data for scientific research. Since many of the social media providers have set up application programming interfaces (APIs), extensive volumes of data can be collected automatically (Jünger, 2018; Keyling & Jünger, 2016). Social media data are attractive, inter alia, because they not only include already available communication, such as that from public media, but they also make organisational and interpersonal communication visible (Ledford, 2020). In addition, these data are process-generated (Baur, 2011, p. 1234), meaning that they are generated independently of scientific research and thus promise an authentic insight into human behaviour. 1 A wide range of studies in the social sciences exploit APIs for data collection and analysis. Thus, the establishment and development of APIs has significant implications for science.