The regional or national innovation performance has been repeatedly measured in the literature; but it has so far not been discussed what this means, especially in relation to a region. What is the contribution of a region to innovation output? The usual approaches implicitly assume that higher innovation outputs per inhabitant, employee, or R&D employee can be assigned to a region. We argue that more insights are gained if we distinguish between various mechanisms that influence the innovation activities in a region. Different analyses need to be conducted, using different variables and including different local factors. Furthermore, we see no justification for using a linear dependence of innovation activity on the number of inhabitants or employees as a benchmark for performance. We use a method that takes into account these arguments and apply it to the Electrics & Electronics industry in Germany.