Purpose Servitization of manufacturing is characterized by very complex decision processes within strongly unstable and uncertain decision contexts. Decision-makers are face situations of lack of internal and external information. This paper intends to develop a decision aid approach to support the management of servitization decision-making processes. Design/methodology/approach The scientific orientation of this research consists in working at improving the efficiency of the servitization decision-making process, by identifying factors of non-reliability, in order to propose remediation actions for the whole process. Improving the final decisions taken by the managers is considered as a consequence of the improvement of the decision-making process reliability. The method, based on modeling and evaluation, requires the specification of a decision process model for servitization, used as a basis to assess decision process reliability and diagnose the enterprise’s servitization decision system. Improving the final decisions made by the managers is considered as a consequence of the improvement of the decision-making process reliability. Findings Key added values: (i) to formalize a servitization decision-making reference model, (ii) to specify a reliability assessment applied to the decision system and (iii) to define a decision pocess reliability diagnosis procedure for servitization, illustrated in a case study. Research limitations/implications A direct perspective is to complete the focus on procedural reliability, by taking into consideration the subjective rationality of decision-makers in the reliability assessment procedure. Additionally, this reliability assessment method and diagnosis could become the basis of a larger risk management approach for servitization. Practical implications The diagnosis procedure proposed in the paper is dedicated to generating practical results for enterprise decision-makers, consisting in recommendations for decision process improvements, in the context of servitization. The approach is illustrated through an industrial SME case study. The practical implications are highly contextualized. Originality/value The key originality of this research is to tackle servitization complexity with a decision system modeling and diagnosis orientation, including the formalization of the notion of ‘decision process reliability’, and the specification and implementation of a quantitative assessment procedure.