Personas have become an important tool for Human-Computer Interaction professionals. However, they are not immune to limitations and critique, including stereotyping. We suggest that while some of the criticisms to personas are important, the use of personas is open to them in part because of an unquestioned focus on explicating user needs and goals in traditional persona research and creation. This focus, while helping designers, obscures some other potentially relevant aspects. In particular, when the goal of the product or software being designed is associated with social and political goals rather than with bringing a product to the market, it may be relevant to focus personas on political aspirations, social values and the will or capacity of personas to take action. We argue that it is possible when producing personas (and associated scenarios) to partially move away from representing needs and embrace personas which more explicitly represent political or social beliefs and values. We also suggest that a phenomenographic approach to user data analysis is one way to achieve this. We provide empirical evidence for our position from two large-scale European projects, the first one in the area of Social Innovation and the second in the area of eParticipation.