Research and innovation policies have been increasingly ascribed to regions as a con- sequence of devolving power and resources from the national level to the local level. Local administrators have been empowered with new and challenging responsibilities but often lack the necessary instruments and knowledge to adequately evaluate the under- taken initiatives and to operate consistently with the European and national dimensions. In this respect impact assessment brings in a new metrics that is neither exclusively related to the scientific and/or economic value of the innovation results nor to a support function to other policies. Experts and/or ad hoc organisations are often appointed by local administrations with the task of describing the dimensions of such impact - qualifying (and quantifying) its descriptive parameters and identifying interested parties. So far this approach has not produced robust results as for the causality links generated, the additionality issue and the role of the local dimension, while the available tool-box is quite poor. This paper is intended to illustrate the results of a pilot experience carried out in the last few years in Trentino. The analysis takes into account the relations among policy makers, available tools, resources, and the role of the public administration and of professional evaluators. Conceptual and interpretative approaches and tools need further development, but the mayor weakness seems to come from the relationship between customer-and evaluators. A new market based on new rules, professional roles and shared behaviours is needed in order to address a correct evaluation pathway and to analyse policies and activities within a common frame. [ABSTRACT FROM AUTHOR]