This paper builds on the established finding that the performance of scholars depends on their interpersonal networks. Until now, these networks have largely been measured by analysing the credits and acknowledgements on their publications, especially their co-authorships. First, it seeks to clarify inconsistencies in existing findings by providing a comprehensive analysis of the effects of co-authorship among the overall population of actively publishing researchers from Germany. Second, it acknowledges that co-publication is only one very formal and explicit form of academic networking and develops a new indicator based on an academic’s inferred co-presence at conferences. Comparing the impact of these two different aspects of networkedness, we find that hierarchy and influence play a stronger role in determining a scientist’s performance in the context of informal networks than they do when considering formal co-publication networks.