Service System Analytics: Cost Prediction

Services become every day more important for society. Cost prediction is one important aspect which has not received the required attention. Nonetheless, its impact on the economic success of service-based industries is considerable and, so, it cannot be overlooked. Therefore, in this paper we tackle the problem of cost prediction of service systems. Since previous work generally followed an ad hoc approach, we present a three layered approach based on customer factors to measure customer involvement since it is a good predictor for service cost. Service co-creation between customer and provider is the central concept of our costing approach. For service modelling we use Linked USDL, and to reason on costing data we use methods from service analytics. On the one hand, we reduce the uncertainty introduced by customers and can predict more accurate service costs. On the other hand, the systematic approach enables a better understanding of the costing problem and the service systems under study, and increases reusability of service analytics algorithms.