Measure What Matters: A Dual Outcome Service Quality Model for Government Service Delivery

Abstract

Measuring customer service quality evaluations has been important since the rise of the service industry and many models in this area have been published. Most models focus on one outcome with a set of predictors. These outcomes are often ill defined and concepts are used interchangeably causing issues in creating good and consistent measures of quality. In this study we develop a new model combining multiple outcome variables and a series of predictors to show the interdependent nature of service outcomes. We test the model using machine learning based on survey responses from 3702 Dutch people. The results indicate that two types of outcome variables are important; quality of the outcome and satisfaction with the process. Each is predicted in different ways by four dimensions. This means governments could benefit from a better specification of the desired outcomes of service delivery and targeted measurement approaches.