Scientific-knowledge commercialisation has become a primary objective for universities worldwide. Science-industry exchange is a prerequisite for innovation. Collaborations are viewed as key to this objective. Despite government financing, artificially supporting the development of such networks has proved difficult. This study extends relationship marketing’s B-2-B model by using a unique multiple-stakeholder research design to explore why, how and by whom scientific-knowledge-commercialisation networks are established and managed. The study uses qualitative evidence from 82 stakeholders in 17 collaborative projects in Irish and German universities. This approach facilitates an analysis of the true value of the stakeholders’ roles. This provides for a holistic view, as opposed to prior research which reported findings based on analysis of one or two stakeholders. The study demonstrates how contextual differences impact on scientific-knowledge commercialisation in Ireland and Germany. The study finds that network capabilities are the main reason that collaborative networks are established. It finds that enduring networks are conduits for innovation and scientific-knowledge commercialisation due to continuously improving capabilities. Stakeholder retention is a catalyst for improving collaborative networks. Stakeholder retention results from stakeholder loyalty. Stakeholders become loyal because they are content with the overall relationship and quality of the commercialisation service. It is fundamental that stakeholders understand each other’s roles and motives as incongruities hamper network development. The findings highlight the central role that PIs play in building and managing relationships. The PI, like the entrepreneur, has to be ‘a jack of all trades’, taking on the roles of negotiator and project/ relationship manager. These roles are in addition to the traditional role of teaching, researching, acquiring funding, Ph.D. supervision and mentoring and administration. The findings suggest that PIs are better placed than TTO managers to act as boundary spanners between science and industry.