This paper is about the possible negative impact of excessive collaboration on the performance of top employees. With the rise of participatory culture and developments in communications technology, management practices require greater conceptual awareness about possible outcomes of increased organizational interconnectivity. While there exists a sound theoretical basis for possible burdens brought by collaborative overload, the literature never really manage to measure and empirically test this phenomenon. We address this gap by developing a methodological framework for the identification of organizational actors at risk of operational capacity overload. Drawing on social network analysis as the widely applied approach for the estimation of employees' involvement in the information exchange networks, this paper describes potential personal and organizational causes leading to the emergence of collaborative overload. Relying on primary data gathered through a survey conducted among employees in a large insurance company, we present a testable model for overload detection. A second merit of the paper consists in finding a novel identification strategy for empirical works on cross-sectional network data, which often face the issue of endogeneity. This research suggests that active collaborative activity does not cause a decrease throughout every aspect of performance. We found that expertise sharing depends on a few key players who take core knowledge assets upon themselves and thus run higher risks of exposure to overload.