PurposeThis paper aims to investigate how the product architecture and service architecture methodology can be applied in strategic maintenance optimization to reduce the non-value-adding variance of maintenance, decrease the complexity and ensure alignment in maintenance practices in asset-intensive companies. The proposed maintenance architecture model will make it possible to make data-driven decisions regarding how the equipment should be grouped and maintained.Design/methodology/approachThe research approach is considered exploratory, and the main research strategy is a case study. The maintenance architecture model is developed based on the product architecture methodology and then tested in three different cases in the oil and gas sector.FindingsThrough the maintenance architecture model, it is possible to pair a quantitative data-driven approach with qualitative understanding of dependencies between equipment, maintenance actions and maintenance work management processes, enabling a more holistic and top-down data-driven approach to improving maintenance, than what currently exists in literature.Originality/valueThe proposed model provides a contribution to the understanding of maintenance and is positioned at a detailed level, different from other maintenance improvement models. This model is focused on the main drivers of maintenance that can be utilized at the strategic level compared to optimization of maintenance for individual pieces of equipment.