Nowadays, an ever-increasing number of information and communication technology solutions (hardware or software based) are finding their way to the automotive sector. Vehicles are being transformed into electronic hubs of information, communication, entertainment and other applications. Prior to commercial deployment, every single of these solutions must undergo a scrutiny of technical tests, often in the field (i.e. on-road as opposed to simulation), in order to ensure safe operation and robust performance. ‘Robustness’ is here perceived as operating as close to the target specifications as possible and with minimum variance, under varying conditions (factors). Meeting this requirement given a limited amount of resources (human, financial, equipment etc.) available for on-road technical tests is often a serious challenge for both researchers and product developers. This study proposes an experimental design process, based on suitable statistical means, for minimising the number of technical tests required to optimise the performance robustness of an automotive service or product under development. The process is substantiated and exemplified for the case study of an electric vehicle consumption estimation product, but could also be used in a variety of other applications (such as navigation, infotainment, safety solutions and others).