Twelve years after the publication of the seminal article by Tim Berners-Lee, James Hendler and Ora Lassila, which expounded the vision of a Semantic Web characterised by dynamic and large scale agent interoperability, the Semantic Web still distinctly lacks a “wow factor”. Many SW applications exist, but too often they are characterised by few data sources put together at compile time to drive some relatively simple user functionality. In many cases it is difficult to identify the competitive advantage that ‘being semantic’ affords these applications, compared to systems built using conventional technologies. Of course, one could argue that this is not necessarily a problem: the success of an area is measured in terms of its academic vitality and its impact on commerce and society. However, I would argue that there is actually a problem here and in my talk I will analyse these issues by examining how the notion of semantics is used in our community, highlighting the productive and unproductive uses of the term, and in particular describing the different ways in which semantics can be effectively exploited to provide added value to applications. The key message is that while there are many ways to exploit semantics to develop better functionalities, as a community we need to develop a better understanding (both fundamentally and pragmatically) of the value proposition afforded by the use of semantics. Without such understanding there is a risk that we will fail to take full advantage of the technologies that we are developing and the opportunities they create for us.