Anatomy of news media on Twitter

Abstract

Twitter, recently rebranded as X, has emerged as a major news source and key arena for political campaigns and marketing activities, becoming one of the primary sources of information for young people. In this context, leading digital media outlets have intensified their presence on the platform, shaped by its specific affordances. However, what characterises the behaviour of these digital media outlets on X, and what problematic aspects emerge? In addition, can the algorithmic response to these media be predicted? To address these questions, this study has adopted a mixed-method approach combining computational and interpretative techniques. Using Twitter’s API v2, a corpus of 34,464 tweets published by the four most influential digital newspapers in Spain was extracted and analysed. The discourse was examined in relation to platform affordances, including the algorithmic logic of visibility, which was modelled using negative binomial regression. The findings reveal a limited adaptation of digital newspapers to the platform’s distinctive language and communicative norms. Moreover, the analysis identifies significant problems related to editorial and advertising content, including difficulties in clearly distinguishing sponsored material from journalistic content. Practices of astroturfing are also detected—an ethically problematic strategy that is incompatible with the social role of the media. In addition, the study proposes a provisional model capable of predicting algorithmic performance with a high explanatory capacity (pseudo-R² = 0.76). Overall, the study contributes to a deeper understanding of the relationship between digital media and platform dynamics, while highlighting the urgent need for greater transparency from both platforms and media organisations to prevent the distortion of public opinion and the marginalisation of minority voices.