The purpose of this paper is to introduce a solution to the problem of changing priorities of customer needs (CNs) in quality function deployment (QFD). Customer preferences and priorities are not very stable and they may change before products are ready for the market. Therefore, finding CNs accurately is a key to reach a higher level of customer satisfaction through improving products.
In the proposed model, a Markov chain is employed to model the changing priorities of CNs. The Markov chain finds a pattern of future CNs, the main inputs of QFD. The QFD method is applied to translate CNs into product requirements (PRs). The analytic network process (ANP) is attached to QFD to ensure that all the relations among the elements, inner and outer, are taken into consideration during the translation process. Thus, CNs are received and adjusted by a Markov chain.
The application of Markov chains for an ANP-QFD model develops an adequate method of finding a pattern of changing priorities of CNs. This pattern enables the ANP-QFD method to work independent of the initial CNs, and originates a Markovain ANP-QFD.
This study originates a stochastic ANP-QFD model. There have been several papers employing various tools and techniques such as the ANP or analytic hierarchy process for QFD to find accurate relations between PRs and CNs. While there are a few papers applying Markov chains to predict the future of the relations of QFD, there is no study which traces the changes in priorities of the CNs during the improvement process. This is addressed by applying a Markovian ANP-QFD. The model is validated through a case study.
