Cloud computing services have evolved to a sourcing option that promises a wide range of benefits, such as increased scalability and flexibility at reduced costs. However, many enterprise applications are subject to strict requirements -- e.g. regarding privacy, security and availability -- and are embedded into complex enterprise IT architectures with a multitude of interdependencies. For these reasons, many decision makers have developed a sceptical stance towards cloud computing. A solution might be a hybrid (local/cloud infrastructure) approach where only suited components are migrated to a cloud infrastructure. This, however, has significant architectural consequences that need to be taken into account. This contribution suggests a cloud migration framework that will be implemented as an IT-based decision support system based on modelling the interdependencies between components. The approach is illustrated with the example of Business Intelligence (BI), i.e. integrated approaches to management support. The underlying decision model would particularly consider data transfer volumes, performance, sensitivity of cloud based data repositories, as well as exposure to public networks. The potential of such an approach is illustrated with a selected set of BI scenarios. Based on this, conclusions are derived and generalised for approaches taking into account deployments on both the local premises and cloud infrastructures.