The search-transfer problem: the role of weak ties in sharing knowledge across organization subunits

Administrative Science Quarterly, March, 1999 by Morten T. Hansen

SEARCH AND TRANSFER

Network Search

In a multiunit organization, a product development team situated in an operating unit may want to obtain useful knowledge residing in other operating units, but the team may not know that such knowledge exists in the organization and where it resides. Team members are confronted with the task of looking for and identifying useful knowledge in an organization in which knowledge is dispersed among subunits. Assuming that project team members are boundedly rational, they cannot easily amass and process a large number of opportunities for interunit knowledge sharing, however, and exhaustive intraorganizational searches will be very time-consuming, if not impossible. Multiunit firms, especially large ones with thousands of members, are complex organizations that make the search process difficult and uncertain. Existing relations that span subunits therefore become important because they serve as channels through which both useful knowledge and information about opportunities for knowledge use flow. Through interunit relations, project teams may hear about opportunities for knowledge use, even without having inquired, and have access to other subunits that have valuable knowledge or can point to other sources that do (cf. Burt, 1992: 13).

Search benefits of weak ties. Not all interunit relations are equally valuable in the search process. Following the argument originally advanced by Granovetter (1973), project teams with weak interunit ties - i.e., infrequent and distant relationships - are likely to have a more advantageous search position in the network than teams with strong interunit ties because their contacts are less likely to provide redundant knowledge. Nonredundant knowledge can be of two kinds. The first is new information relayed to a project team about opportunities for interunit knowledge use. Other subunits can point to specific knowledge residing in subunits to which the focal subunit has no direct ties. Here, a project team and its subunit's direct contacts act as bridges. Search is more beneficial to the extent that each direct contact can point to different types of opportunities. Obtaining information about the same opportunity twice is costly if the project team spent time in getting information it already had. A second type of nonredundant knowledge is project-specific knowledge (e.g., a software module) the direct contact itself can provide. Direct interunit contacts are less redundant to the extent that they provide different types of knowledge that can be used by a focal project team. Having two direct contacts that can provide the same software module is less helpful than being directly connected to two subunits that each possesses useful but different knowledge.

This argument rests on the premise that it is costly to maintain direct relations to other subunits. People in a subunit need to spend time cultivating relationships with other subunits and processing the incoming information from direct contacts. Because of these costs, people in a subunit can rarely afford to maintain relations with many other subunits, let alone maintain strong relations, which require more energy than weak ones. There are therefore significant opportunity costs in maintaining interunit relations: instead of having ties to two subunits that provide redundant knowledge, people in a subunit can develop a new relation that provides new knowledge.