Two-Stage Modeling of Customer Choice Preferences in Engineering Design Using Bipartite Network Analysis

Abstract

Customers’ choice decisions often involve two stages during which customers first use noncompensatory rules to form a consideration set and then make the final choice through careful compensatory tradeoffs. In this work, we propose a two-stage network-based modeling approach to study customers’ consideration and choice behaviors in a separate but integrated manner. The first stage models customer preferences in forming a consideration set of multiple alternatives, and the second stage models customers’ choice preference given individuals’ consideration sets. Specifically, bipartite exponential random graph (ERG) models are used in both stages to capture customers’ interdependent choices. For comparison, we also model customers’ choice decisions when consideration set information is not available. Using data from the 2013 China auto market, our results suggest that exogenous attributes (i.e., car attributes, customer demographics, and perceived satisfaction ratings) and the endogenous network structural factor (i.e., vehicle popularity) significantly influence customers’ decisions. Moreover, our results highlight the differences between customer preferences in the consideration stage and the purchase stage. To the authors’ knowledge, this is the first attempt of developing a two-stage network-based approach to analytically model customers’ consideration and purchase decisions, respectively. Second, this work further demonstrates the benefits of the network approach versus traditional logistic regressions for modeling customer preferences. In particular, network approaches are effective for modeling the inherent interdependencies underlying customers’ decision-making processes. The insights drawn from this study have general implications for the choice modeling in engineering design.

Comparing Utility-based and Network-based Approaches in Modeling Customer Preferences for Engineering Design

“…We select four explanatory vehicle attributes based on our prior studies (Fu et al, 2017). These are Price, Fuel Consumption, Power, and Make Origin country.…”

“…Nevertheless, utility-based preference modeling, such as the DCM, is limited when handling dependency of customer decisions (e.g., their decisions may be influenced by each other because of social relations) and collinearity of design attributes (e.g., vehicles with low prices are more possible to have smaller engine capacity) (Wang et al, 2016). To overcome these limitations, recent studies explored the capability of statistical network models in estimating customer preferences (Fu et al, 2017;Sha et al, 2018). Among existing network-based modeling techniques, the Exponential Random Graph Model (ERGM) is increasingly recognized as one of the most powerful analytic techniques (Snijders et al, 2006).…”

“…ERGM provides a flexible statistical inference framework that can model the influence of both exogenous effects (e.g., nodal attributes) and endogenous effects (network structures/nodal relations) on the probability of connections among nodes. In our prior work, ERGM has been used to study customers' consideration behaviors (Sha et al, 2017), forecast the impact of technological changes on market competitions (Wang, et al, 2016), model customers' consideration-then-choice behaviors (Fu et al, 2017), and predict products' co-consideration relations (Sha et al, 2018;Wang et al, 2018). While different techniques have been employed, to the best of our knowledge, no studies have compared the network-based approaches with the existing utility-based approaches in terms of their consistency, interpretability and predictive power in modeling customer choices.…”

Modeling Participation Behaviors in Design Crowdsourcing Using a Bipartite Network-Based Approach

“…Sha and Wang studied products' co-consideration relations with a unidimensional ERGM [43], and compared its predictive performance with network-based logistic regression model [44]. Fu et al [45] adopted the bipartite network setting of ERGM and studied the choice behaviors in support of engineering design. These studies have inspired us to further extend the capabilities of ERGM in engineering design, and to evaluate the feasibility of network configurations and the interdependence assumption in analyzing participation behaviors in design crowdsourcing contests.…”

A Network‐Based Approach to Modeling and Predicting Product Coconsideration Relations

“…In our recent study, Fu et al [43] developed a twostage bipartite network modeling approach to study customer preferences in making choices by decoupling the choicemaking process in two stages, the consideration stage and the choice-making stage. Wang et al [44] utilized a dyadic network analysis approach to predict product coconsideration relations based on exogenous factors, such as product attributes and customer demographics.…”