The chapter reflects on the limitations of the discourse on networks in innovative clusters in light of the author’s own empirical findings and associated research. The conceptual part of the chapter stresses the importance of distinguishing between personal and formal networks, and between stages of network mechanisms. It also stresses the need for conceptual sensitivity towards individualized networks that may go beyond coherent communities. The authors reflect on weaknesses in previous research, and then elaborate on reasons why networks in clusters can be limited due to lack of perceived need or lack of opportunity to benefit from local networks. Subsequently, they show that spatial proximity tends to be important for the formation of networks, while it tends to be less important for actual knowledge exchanges. In fact, within the context of Norway, the authors illustrate that it is international networks that are related to innovativeness. Finally, they clarify the role of various types of proximities for innovation by finding empirical evidence for the so-called Goldilocks principle: a medium level of proximity delivers the best innovative returns to collaboration, while collaboration with partners that are either too close or too far may not be beneficial for innovation.