The ultimate goal of innovation policy is to improve (or to at least maintain) current levels of innovation performance. But the question is, "what is innovation performance?" What does it refer to and how can it be best assessed? To begin with, the term "innovation" has itself come to refer to a wide range of different phenomena. For example, distinctions are commonly drawn between:

• New to the world innovations and adoption of existing (new to the firm) innovations
• Radical and incremental innovations
• High-tech and low-tech innovations
• Product and process innovations
• Technological and organisational/marketing innovations

Rarely is this full range of phenomena included in the framings of innovation used by policy makers and analysts. In particular, the second types of innovation in most of the above bullet points tend to be neglected, despite their significance (as recognised in countless studies of innovation processes). An important starting point for innovation policy therefore is to adopt a fuller framing of the different types of innovation and to appreciate the "patterns of innovation" that characterise the system under consideration.

Another related aspect of innovation performance concerns its purpose. It is not particularly useful to view innovation as an end in itself. Rather, innovation should be considered more a means to achieving higher level goals, such as increasing firm profitability, improving healthcare outcomes, contributing to the adoption of "greener" lifestyles, etc. (the ultimate goals of innovation are, of course, wealth creation and improvements in quality of life). It is important for policy makers and analysts to appreciate the diversity of ends to which existing innovation efforts are directed – meaning the variety of goals pursued, the balance between them and their disparity – as policy interventions designed to shift towards alternative, more desirable ends could be easily frustrated by path dependent "lock-in" effects.