The power of reciprocal knowledge sharing relationships for startup success

1. Introduction

In their study of the Boston biotech networks Owen-Smith and Powell (2004) contend that geographic proximity will support innovation, as it leads to knowledge spillover effects benefitting companies co-located in the same cluster. Repeating their study in the same Boston biotech cluster ten years later, we find the opposite: it is not location, that matters, but quality of the network interaction. While Owen-Smith and Powell constructed their network through formal alliances extracted from the Bioscan database, we built our network by manually polling R & D management about their interactions with their peers at other biotech startups in the Boston biotech cluster. We find that while location matters for the quantity of communication, the embeddedness of a startup’s R & D members in the core of their peer communication network matters for the innovative capability of their company. Alternating changes in leadership among communication partners is also important for facilitating innovation.

Research up to now has been inconclusive about what matters more for start-up business success: location, or communication. As has been shown repeatedly, geographical proximity cannot be analyzed in isolation; rather proximity should always be examined in combination with other dimensions (Boschma, 2005). Many arguments have been made qualitatively about the effectiveness of high-tech businesses clustering in geographic proximity, with the most prominent example being Silicon Valley in California (Saxenian, 2006). While evidence has been found for increased communication between start-ups co-located geographically (Gloor et al., 2008; Allen et al., 2009), the triadic link between increased communication and business success (Raz and Gloor, 2007) on the one hand, and co-location and business success on the other hand (Porter et al., 2005) has been little studied. This research tries to close that gap, by comparing the location of biotech start-ups with the amount and quality of interaction between them, and contrasting both with their innovative capabilities (measured by number of patent applications) (Figure 1).

To research this question we drew on a data set of interpersonal communication of senior R & D members in over 70 firms in the Boston biotech cluster, with information about firm innovation added. The Boston biotech cluster offers a unique testbed to study communication: first, the firms in the cluster all belong to the same...