There is a growing interest in socio-technical networks that encompass people, process and technology. They capture connections among technical artifacts and human resources. Existing studies have shown that social networks among resources that collaborate to work on a business process instance have an impact on the performance of the instance. Hence, the social network among the resources should be aligned with the technical work structure. Thus, we propose to identify empirically the specific features of social networks (e.g. size, closeness, density, etc.) that have the most significant impact on the performance of a process. Then, we can leverage these significant social network features to optimize dynamic assignment of tasks to resources for better performance. In general, organizations can use this two-part approach to evaluate their resource social networks and shape their resource networks to achieve socio-technical network alignment. We use real data from business processes in the IT incident management domain for this study to demonstrate our proposed methodology.