This research investigates how to break through intra-organizational data silos in multinational engineering companies by analyzing data sharing networks. Data occurs in many ways like numbers, statistics, and documents in structured and unstructured forms. Companies consider data as one of their most valuable assets for decision-making and to increase performance. Still, often they do not see the direct results of their data-driven investments back into the organization. The volume of generated data increases so rapidly that it exceeds our capabilities if not managed well, leading to organizational data silos. Engineering companies have to deal with unique and complex projects, temporary teams, fragmented departments, and lots of unstructured data. Multicultural teams that work from dispersed locations increase the probability of miscommunication and misinterpretation of project data. Data sharing between employees of three selected case projects within one multinational portfolio have been analyzed. First, the quantitative social network analysis in Python programming exposed the structures of and the node types in the data-sharing networks. Secondly, supported by the extensive literature review, qualitative in-depth interviews identified enabling and limiting factors that determine intra-organizational data sharing in the project life cycle. It was found that balanced networks contribute the most to the exploration and exploitation of intra-organizational data. Also, the multinational context is not acknowledged to limit data sharing, but weaker links do exist between geographically dispersed teams. Measures to create and maintain balanced data sharing networks are suggested on organizational, portfolio, and project level. Future research can focus on network dynamics over time, the effect of data sharing on project success criteria, and the effect of cultural dimensions on data sharing behavior.