The paper reports on ongoing research into the area of enterprise modeling. We have concluded so far that most enterprise models are static representations of enterprises that are expressed using diagrammatic schemes or computing simulation techniques. Their life cycle is usually very brief as they fail to adapt to changes that take place in an enterprise's dynamic environment. On the other hand enterprise models are directed to specific business functions, omitting the fact that relationships between various enterprise agents affect their operations. The paper reports on our attempt to provide a framework for assessing the impact of change, that is caused by the formulation or deletion of relationships between agents, on a wider environment. The framework is being tested on supply chains. For the sake of consistency the authors are referring to the automotive sector throughout the paper.