In this paper, we develop a model of complementarity of knowledge flows in software ecosystems through two knowledge-acquisition mechanisms: a formal, fine-grained, contractual governance mechanism through inter- firm alliances and a nonformal, course-grained, noncontractual mechanism of spillover capture. In contrast to studies that focus solely on knowledge exchange in alliances, we focus on two mechanisms and test their additive and super-additive effects in the software sector. We examine the effect of a software firm’s position in the alliance network (formal, contractual mechanism) and patent citation network (nonformal, non- contractual mechanism) using two important network characteristics: reach and redundancy. We test our model using data on the packaged software industry during the period 1995 to 1999. Our results show that software firms’ sales performance is predicted by their positions within these two networks. Furthermore, we find that these network positions are additive and complementary in their impact on performance. Our results are potentially generalizable to other settings that have interdependent information and knowledge flows across organizational boundaries.