Axel Bruns, Samantha Vilkins, Katherine M. FitzGerald, Tariq Choucair, Daniel Angus, Caroline Gardam, Kunal Chand, Laura Vodden, Klaus Gröbner, Katharina Esau, Carly Lubicz-Zaorski, and Ehsan Dehghan
- 7 Sep. 2026 – DigiWorld ECREA 2026 pre-conference, Brno
Presentation Slides
Abstract
The study of election campaigning on social media is increasingly complicated. This is due to both the significant diversification of platforms following the enxittification of Twitter, and the degradation or discontinuation of many Application Programming Interfaces (APIs). Twitter’s Academic API was decommissioned, Facebook's CrowdTangle replaced with the restrictive Meta Content Library; YouTube and TikTok APIs are unreliable and not globally available; Meta’s Ad Library and Google’s Ads Transparency Center are inconsistent; and federated platforms Bluesky and Mastodon raise new practical and ethical issues for data gathering.
We present new approaches to these major challenges. First, we discuss data donation as a complement to conventional API-based data gathering approaches: in particular, the new Mobile Observation of Advertising Toolkit (MOAT). MOAT’s mobile screen-recording infrastructure enhances platform observability by capturing mobile advertising data on Android devices; it offers a framework for the ethically sound gathering of organic and promoted social media content from participating users’ devices, capturing their lived experience of election campaigning beyond the whole-of-platform view that APIs (ideally) provide.
Second, we introduce the practice mapping method, an approach that addresses the fragmentation of datasets across diverse platforms. Practice mapping abstracts from the specific data formats and communicative affordances of individual platforms: conceptualising and encoding repeated communicative actions as practices, it enables comparisons of users’ (and campaigners’) activities across platforms. This creates networks connected by shared communicative practices, rather than only by direct interactions, and is especially valuable for platforms where direct interactions play only a very secondary role or are not captured in available data at all (e.g. TikTok, Instagram).
Such methodological evolutions offer practical solutions to the considerable challenges facing election campaign analysis on social media, extending available data and introducing new analytical perspectives.
