We understand software innovation as concerned with introducing innovation into the development of software intensive systems, i.e. systems in which software development and/or integration are dominant considerations. Innovation is key in almost any strategy for competitiveness in existing markets, for creating new markets, or for curbing rising public expenses, and software intensive systems are core elements in most such strategies. Software innovation therefore is vital for about every sector of the economy. Changes in software technologies over the last decades have opened up for experimentation, learning, and flexibility in ongoing software projects, but how can this change be used to facilitate software innovation? How can a team systematically identify and pursue opportunities to create added value in ongoing projects? In this paper, we describe Deweyan pragmatism as the philosophical foundation for Essence – a software innovation methodology – where unknown options and needs emerge as part of the development process itself. The foundation is illustrated via a simple example.