Purpose - The paper aims to develop an evaluation model of the customer satisfaction index (CSI) in an R&D organization. A conceptual framework on customer satisfaction with a probabilistic approach has been attempted based on customer requirements and expectations in compliance with the clauses of ISO 9001:2008. Design/methodology/approach - A survey through a well-designed customer feedback data sheet has been used as an effective tool for the measurement of CSI. The questionnaire was framed on the basis of the requirements of a quality management system with advice to the customer for allotting grade points on a given scale to the quality parameters. The research model has been analyzed based on a fault-tree approach and the probability of failure of each quality parameter has been assigned on the basis of grade point average. Data analysis for the estimation of the probability of failure at a customer satisfaction level (CSL) was carried out based on the probability of failure of each quality element graded by the customers. The data were also tested through statistical inference of whether customer-to-customer satisfaction level differs or not. Findings - As a result of case study analysis, 88 percent of customers are fully satisfied. This gives significant information to the management process as well as providing a guiding tool for future improvements. The analysis was carried out based on a framed questionnaire graded by the customer and the result reveals that there is no significant difference between customer satisfaction levels. Research limitations/implications - This model can be used by any organization, irrespective of the number of customers participating, as well as the number of quality parameters being assigned in the customer feedback analysis. Originality/value - A literature review found that there are various approaches for evaluating a CSI. The paper describes how a newly-applied conceptual model based on the failure of CSL in the form of a fault-tree approach was designed and how the probability of failure of each element/parameters was assigned on the basis of a grade point average to evaluate the CSI, as well as the variation in satisfaction levels between customers being analyzed.