The purpose of this paper is to discuss the influence of data availability and quality within a utility to prioritise inspection and rehabilitation needs. Data are required in order to predict the structural condition of assets. Lack of data and budget limitation do not encourage utilities to evolve from reactive to proactive management. Methods and tools exist; however, improving operational practices will require the demonstration of data collection benefits. In this article, asset management approaches are presented and discussed regarding influence of data; from closed-circuit television reports elaboration to prioritisation of rehabilitation needs. Bottlenecks related to uncertainty, imprecision and incompleteness of data are identified, and the authors propose approaches to study these questions. This article also highlights the use of numerical experiment to simulate asset management scenarios, to experiment methods or to demonstrate the interest of new practices. Numerical experiment allows construction and use of virtual degraded databases derived from a completely known asset stock.