Dependability is a super ordinate concept regrouping different system attributes such as reliability, safety, security, or availability and a key selling point of modern embedded systems. Dependable systems rely on mature quality management and development methods such as requirements / systems engineering and system analyses. In the automotive domain analysis methods for safety and security attributes at early development phases are well known and partially mandatory by domain standards. Nevertheless, approaches for analysis of serviceability attributes (the combination of reliability and maintainability) at early development phases are not yet available. Aim of the paper is to present a novel analysis method to quantify the impact of individual system parts on the overall system serviceability at early development phases. This approach bases on the concepts of state-of-the-art methods for safety and security analysis and extends their scope of application to serviceability feature quantification, thus enables consistent identification of system dependability target attributes. This, in turn, is a pre-requisite for ensuring a certain level of system dependability from start of development. In the second part of the document the application of the novel approach is demonstrated on an automotive training example of a battery management system.