This paper explores the use of membership categorisation practices by a high-profile celebrity public social media account that has been understood to create interest, attention and controversy across the UK (and wider European) media ecology. It employs a data set of tweets collected from a high-profile public ‘celebrity antagonist’ in order to systematically distinguish types of antagonistic formulation that have created different levels of interest within the social media community and beyond. Using classic ethnomethodological studies of banner headlines and other means of generating public interest and ‘making sense’, it respecifies high-profile antagonistic tweets as category formulations that display particular and regular membership category features that are reflexively bound to possible antagonistic readings, interest and controversy. In conclusion, it considers how such formulations may be grasped to represent resources that constitute ignition points within antagonistic communication and information flows that can be metaphorically understood as ‘digital wildfires’.