Data-Driven Dynamic Network Modeling for Analyzing the Evolution of Product Competitions

Abstract

Abstract Understanding the impact of engineering design on product competitions is imperative for product designers to better address customer needs and develop more competitive products. In this paper, we propose a dynamic network based approach to modeling and analyzing the evolution of product competitions using multi-year product survey data. We adopt Separate Temporal Exponential Random Graph Model (STERGM) as the statistical inference framework because it considers the evolution of dynamic networks as two separate processes: formation and dissolution. This treatment allows designers to investigate why two products enter into competition and why a competitive relationship preserves or dissolves over time. In an open market, the available products to customers are continuously changing over the time, posing challenges for conventional modeling methods concerning fixed product input. Consequently, we propose to leverage “structural zeros” in STERGM to tackle the problem of modeling varying product competitors as nodes in dynamic networks. We use China’s automotive market as a case study to illustrate the implementation of the proposed approach and its benefits compared to the static network modeling approach based on Exponential Random Graph Model (ERGM). The results show that our approach identifies the driving factors associated with product attributes and current market competition structures for the change of competition in both formation and dissolution processes. The insights gained from this paper can help designers better interpret the temporal changes of product competition relations and make product design decisions with the aid of dynamic network-based models.