Due to the commodisation and globalization of the markets, manufacturing companies have been pursuing new business models based on product- service bundles, which in literature have been investigated under the broad term of Product-Service Systems (PSS). Efficient provision of PSS is not immediate since it requires as a first condition a deep integration of product and service components and the related delivery processes. As regards durable equipment, maintenance is one of the most relevant and common services offered to the market. The identification of the best maintenance policy, and the expected improvement of availability and productivity of the industrial asset depends upon the knowledge, cognition, and expertise that the provider has on the machine and the related operating conditions. With this regard, it is fundamental to wisely collect and analyze data related to the machine and its delivery process, extracting useful information guiding the policy selection. By investigating the case of a durable equipment manufacturer, the paper focuses on the problem related to the creation of a proper engineering process and data infrastructure for supporting a gradual and robust transition towards a PSS-based business model. A dual perspective vision for ensuring a continuous improvement of maintenance services is proposed and discussed, based on a cross-analysis of the on-field collected data from the industrial asset and the maintenance delivery processes.