In academia, social network theory and social network analysis are commonly employed to measure social capital. Social network analysis examines a defined set of actors, such as co-authors of research articles, and interprets the connections, or “edges,” between them. This approach; however, is limited as it often overlooks crucial qualitative and behavioral dimensions. In this context, social capital functions also as a catalyst or lubricant for cooperative work, requiring a more nuanced understanding. To address this gap, we conducted a study exploring researchers’ perceptions of social capital. Our hypothesis posited a potential misalignment between academics’ preferred practices for building social capital and their habitual behaviors. To test this, we piloted a survey targeting a community of 1,092 academics across six faculties—Health and Medical Sciences, Science, Social Sciences, Humanities, Law, and Theology—at the University of Copenhagen. The survey focused on their goal-seeking behavior in publishing new research. First we asked the researchers to report their habits, and then we surveyed their preferences. Collaboration preferences were analyzed through the lens of social capital theory, assessing three interrelated dimensions: cognitive, relational, and structural. This presentation will share key findings from the pilot study and situate them within broader research on collaborative work in academia.