Data-Campaigning on Facebook: Do Metrics of User Engagement Drive French Political Parties’ Publications?

Abstract

Research on data-driven campaigning has mostly focused on the strategies of central campaign teams. However, there is a lack of evidence explaining how parties and supporters use data-driven campaigning techniques to organise their social media campaigning. Do user engagement metrics influence the choice of campaign themes by encouraging political parties to concentrate their communication on issues that are most liked, commented on, and shared? Our study focuses on the use of Facebook by French political parties and their supporters during the 2022 presidential election campaign. We conducted a supervised content analysis based on machine learning to examine their Facebook posts (n = 17,060). Our results show that the issues prioritized by parties on Facebook may be different from those that are most prominent in their broader communications. In most cases, however, these themes are not chosen based on user engagement, even for parties that claim to have developed their base through digital channels. Instead, the choice of themes seems influenced by more traditional campaign strategies, such as the desire to capture the electorate of their closest rival. In our conclusion, we discuss the implications of these findings for the adoption of data-driven campaigning in digital election communication across Europe.

Keywords

• data-driven campaigning
• issue salience
• political communication
• political programs
• social media
• supervised learning
• user engagement