Dependency analyses have become crucial in today’s enterprise architecture practices, which usually face complex IT landscapes with highly interdependent applications. In such environments, a deep understanding of the application’s context is essential to determine its qualities and project its further evolution. However, method support for making this context a tangible IT landscape management part and thus facilitating quantitative decision making still seems expandable. Based on the representation as a network of applications and their relations of data exchange, this paper therefore suggests ways to support the IT landscape’s examination through network analysis. We develop this approach based on a combination of theoretical explanations, past empirical findings, and experiences taken from the architecture practices of four sample organizations. We illustrate and evaluate our approach with a short case study. Our approach, developed and illustrated in close alignment with insights from actual practice, thus offers ideas and advice for researchers and practitioners alike.