The optimal product line design problem consists of predicting the acceptance of new products in the market, prior of the actual production phase. It requires a unique blend of techniques and resources as it deals with multiple parameters and hard optimization problems. This paper proposes a service oriented architecture which outlines the basic features that a marketing decision support system should provide in order to address this problem. Each feature corresponds to a distinct service while the overall functionality is achieved though service orchestration. The proposed breakdown of the problem into distinct services fits intrinsically the way that marketing managers anticipate the optimal product line process, while it allows them the opportunity to get a solution without getting obstructed of the various, complex optimization parameters and needing to coordinate the whole process.