In the digital era, the activities of millions of users have become visible to news organizations and scholars alike through web analytics. Now it is possible to assess which articles users tend to click on as well as interact with on social network services. On an aggregated scale, these metrics give indications of how
collective attention is distributed across content types and genres. Communication researchers have used this kind of data to make inferences about the types of articles, news values, frames and genres that achieve a high degree of salience on the internet (for recent overviews, see Boczkowski & Mitchelstein, 2013; García-Perdomo, Salaverría, Kilgo, & Harlow, 2017). However, most studies have worked with either click patterns or social network interactions. Few studies have sought to integrate these data sources to assess collective attention from both perspectives at once (albeit see Bright, 2016). Therefore, there is great potential in, and need for, combining analytics data to provide new perspectives on collective attention online.
This paper does so by comparing popularity (news stories that receives most clicks on news websites) and virality (stories that users share most intensively on social network sites) of the same news content. The paper applies the two-faced concept of collective attention in an empirical case study of online news coverage during the Danish parliamentary election 2015. Choosing a period of heightened political activity, like an election, provide a critical case of collective attention, where we would expect popularity and virality to converge on political issues (Boczkowski & Mitchelstein, 2013). To assess collective attention, the paper constructs a baseline of all articles published on Danish news websites in the election period. This is combined with data scraped directly from news websites (to get the list of most popular articles) as well as a social media monitoring service (to get the most shared articles). Relying on operationalizations of frames and actors in news coverage, the paper analyses whether users gravitate toward, on one hand, the game-strategic or issue-oriented frames and, on the other hand, personality-driven versus party-based stories. The analysis finds great differences between what is read and shared, in particular when it comes to strategically-framed and issue-oriented coverage. The difference between popularity and virality is then explained with reference to the underlying mechanism guiding click behaviour and social sharing respectively. In the end, the paper discusses possible implications for users and news organizations.