Decision making about launching (or not) an innovation project remains an uncertain task for managers, since each project have different characteristics representing a new challenge for the company. Indeed, the development and introduction of a new product involve changes on the way the firm designs, produces, distributes or sells its products. So decision-making tools are needed to take into account both, the project uniqueness and the product impact on the current and future technological systems. This article proposes a methodology to minimize the risks of failure of innovation projects, taking into account the degree of novelty and complexity. The aim is to determine the minimum threshold profile to develop optimally the project. The methodology is applied to two innovation projects with different degrees of novelty and complexity.