The significant uptake of residential and commercial solar photovoltaic (PV) systems worldwide has sparked a keen interest in monitoring the performance of these systems. In this paper, we develop a data-driven framework to systematically and holistically characterise the performance of a PV installation in the field. We demonstrate the efficacy of the proposed data-driven framework by applying it to PV generation data obtained from a large commercial building in northern Australia. In the wake of only limited site-specific information available from the building, as is often the case in practice, we show how that information can be combined with other publicly available data sources to obtain deeper insights into the performance of the building's PV system. We believe that our framework serves as a valuable starting point for PV owners to ascertain how their systems are performing once it is deployed in the field.