Developing key performance indicators for variant management of complex product families

In this paper, we present a method for the development of key performance indicators for variant management purposes. It provides decision makers, product portfolio managers and - architects with a generally applicable way to gather information on how well a product portfolio and architecture perform in terms of balancing external and internal variety. The method employs tools from the field of product development and combines them with a data analytics perspective for general applicability. Each phase of the product life cycle is examined for its variant management goals, e.g. reduction of supply chain complexity in logistics or reduction of change request cycle time in engineering change management. The goals are analyzed for mutual influence in a cross-impact matrix and thereby classified into influencing factors, indicators, or critical factors. Indicators and critical factors are promising candidates for key performance indicators, and they are made measurable by linking them to available data in the company. Therefore, the existing database is analyzed for data objects, e.g. product components, and their relevant attributes, e.g. manufacturing costs. Each indicator is linked to suitable data. Based on the evaluated attainability and aggregability of data, the most suitable indicators and the correct level of aggregation are selected, e.g. business unit-, product-, component- or part level. In total, the indicators reflect all relevant aspects of variant management in the company and their implementation is prescribed by our method. We apply the method in a case study at MAN Truck & Bus AG, a German commercial vehicle manufacturer. Highly diverse market demands and accordingly complex product portfolios characterize the commercial vehicle industry. Using systematically developed performance indicators in variant management is crucial for manufacturers. The case study results are critically discussed with product architects as well as engineering management at MAN to show applicability and quality of the presented method.