The change from pre-descriptive 'waterfall' software process into iterative and incremental models has created a need for redefinition of software requirements engineering. Agile methodologies have emerged to support the paradigm shift by treating the symptoms: emphasizing change management and customer collaboration to embrace volatility of requirements and priorities whilst in development. However, it has been recognized that fast-paced agile development does not provide sufficient support for initial or long-term planning of the software product. Research and practitioner literature have started to address the need with the concept of a high-level definition of the software project's outcome: the software Product Vision.
In this thesis, uncertainty in new product development is studied from the perspective of Innovation Management. As a vehicle for reducing uncertainty in software projects, the concept of an software Product Vision (reason for the project's existence) is chosen to be examined from the viewpoints of New Product Development and Software Engineering literature. The work describes sources of uncertainty in software projects and explains the effects of a mutually understood software Product Vision on software project performance and end-product acceptance. Key parameters for an interdisciplinary and unified software Product Vision are identified by studying four existing and one emergent Product Vision models. Finally, a new Product Vision framework (InnCa) is created based on semantic analysis. The framework's applicability on software projects in evaluated in three participatory action research -case studies.
As a result, it is concluded that common parameters of a interdisciplinary 'Product Vision' can be identified. The framework created can be used to ideate, rapidly capture, iterate and analyze vague software ideas. It is applicable for sharing knowledge about the project's high-level goals amongst the project's stakeholders.
However, it is not argued in this thesis that the framework could be used in all kinds of projects and circumstances. While uncertainty in software projects is a chaotic and complex phenomenon, no 'silver bullet' can address all situations.The topic of software Product Vision may prove grounds for further research, possibly leading to practical tools for assessing and quantifying uncertainty about goals during a software project's trajectory.
