The future of nuclear fusion as a viable energy source has two major hurdles to overcome. Firstly, there are the daunting and complex technology and physics issues to be resolved before a power plant capable of breeding its own fuel and producing an excess of electricity can be built. Secondly, fusion must offer a useful and economically competitive product to energy markets worldwide, where it will compete with renewable sources and energy storage. However, that future energy market also places a premium on aspects of generation such as reliability of electricity delivery, and the avoidance of externalised costs which might arise from distributed generation such as wind farms or large-scale energy storage required to match seasonal variations in supply and demand. Fusion potentially has an advantage in these areas and an optimised environmentally-friendly energy system will need a mixture of technologies in order to be clean, reliable, cheap, and power-dense enough not to compete excessively for land. DEMO will act as a technology demonstrator for a fusion power plant, providing relevant, in-situ proof of operation of materials and components, and of viable strategies for fuelling and component replacement. However, as a first of a kind and with inevitable performance margins built into the design due to the uncertainties associated with first integrated operation of all plant systems, DEMO will not be optimised for commercial availability or minimum electricity cost, but rather to produce the data required to achieve those in a full fusion environment. This paper reviews the features needed for commercial operation of a fusion plant, and how they can be achieved based on DEMO operational experience and parallel technology development.