This article reports on the methodology for setting the Mexican User Satisfaction Index for Social Programs (MUSI-SP) as tested in seven national social programs. The evaluation is based on Structural Equation Modeling (SEM). How satisfaction takes the central place of the SEM, which postulates its causes and effects, contributes to the increased validity and reliability of satisfaction indicators that allow benchmarking between social programs. The MUSI model is an adaptation of the American Customer Satisfaction Index (ACSI) model. The MUSI methodology includes qualitative and quantitative techniques. The estimation model is by the Partial Least Squares (PLS). In each of the seven social programs, no statistical evidence was found to reject the main relationships postulated by the ACSI’s model: that Perceived Quality impacts Satisfaction, and Satisfaction impacts Trust. The improvement opportunity areas were also identified for each program. These results give valid and reliable feedback to public policies.