Holistic Modeling, Measuring & Management of Product Variety-Induced Complexity : A Supply Chain Case Study in the Material Handling Equipment Industry

Sammanfattning

Over the past decades, globalization and mass customization have continued to increase supply chain complexity in manufacturing firms, particularly through product variety-induced complexity (PVIC). While broader portfolios can increase market reach and yield competitive advantages, they also create hidden operational and managerial effects that are difficult to measure with traditional ERP systems. This thesis develops a practical measurement model to quantify PVIC across supply chain domains and support product portfolio management decisions. Using a Design Science Research strategy, the model was developed through a combination of literature, company data, practitioner interviews and iterative evaluations in a case study of a global material handling equipment manufacturer. The model integrates upstream, internal, and downstream complexity drivers using established approaches such as Weighted Sum Model structure, variable weights derived from the Analytic Hierarchy Process, entropy-based measures and operational indicators to quantify and manage the multidimensional nature of PVIC. In doing so, the study adopts the theoretical lens of Organizational Information Processing Theory for understanding how fragmented PVIC-related information can be consolidated into decision-support outputs that strengthen managerial information processing capacity. As a proof of concept, the proposed measurement model was implemented at the case company using Python and Google Data Studio dashboards. It applies a leave-one-out calculation logic to estimate the PVIC contribution of individual options available on two base machines in the case company’s product portfolio. The model uses several data points to capture the fact that PVIC is not concentrated in one department, but spread throughout the SC and emerges through the combined effects of sourcing, R&D, production, planning, quality, and sales processes arising from the numerousness, diversity and interconnectedness of available product variants in the portfolio. Sensitivity analyses and practitioner interviews were conducted to evaluate the validity, internal consistancy and robustness of the model. The findings revealed that PVIC measurement in isolation is not enough to assess its strategic or dysfunctional nature. By combining complexity measures with both commercial value and strategic reach indicators, the model can more effectively support managerial decisions to either absorb strategic complexity by keeping product variants, or to reduce dysfunctional complexity by removing product variants. Furthermore, this combination presented in one user-friendly output supports more productive cross-functional PVIC assessment thereby reducing some of the negative managerial effects also associated with PVIC. An additional valuable finding shows that entropy-based metrics, while mathematically sound in capturing dispersion uncertainty, should be applied selectively and in combination with more tangible, effort-based metrics to more accurately capture the full extent of PVIC in a practical context. Overall, the study bridges the gap between academic understanding of PVIC and its practical management in a real-world context by proposing a holistic approach to modeling and measuring PVIC to support managerial decision-making.