This paper presents an empirical investigation of how people appropriated Twitter for socio-political talk in response to a television (TV) portrayal of people supported by state welfare and benefits. Our findings reveal how online discussion during, and in-between, TV broadcasts was characterised by distinctly different qualities, topics and user behaviours. These findings offer design opportunities for social media services to (i) support more balanced real-time commentaries of politically-charged media, (ii) actively promote discussion to continue after, and between, programming; and (iii) incorporate different motivations and attitudes towards socio-political concerns, as well as different practices of communicating those concerns. We contribute to the developing HCI literature on how social media intersects with political and civic engagement and specifically highlight the ways in which Twitter interacts with other forms of media as a site of everyday socio-political talk and debate.