Agent-based modeling and simulation of the decision behaviors of e-retailers

Purpose: Social media facilitates consumer exchanges on product opinions and provides comprehensive knowledge of online products. The interaction between consumers and e-retailers evolves into a collective set of dynamics within a complex system. Agent-based modeling is well suited to stimulate such complex systems. The purpose of this paper is to integrate agent-based model and technique for order performance by similarity to ideal solution (TOPSIS) to simulate decision behaviors of e-retailers in competitive online markets. Design/methodology/approach: An agent-based network model using the TOPSIS driven by actual price data is developed. The authors ran an experimental model to simulate interactions between online consumers and e-retailers and to record simulation data. A nonparametric test is used to conduct data analysis and evaluate the sensibility of parameters. Findings: Simulation results showed that different profits could be obtained for various brands under different social network structures. E-retailers could achieve more profits through cross-selling than single-selling; however, the highest profits can be achieved when some adopt cross-selling, whereas others use single-selling. From a game perspective, the equilibrium for price-adjustment frequency can be determined from the simulation data. Thus, price adjustment differences significantly affect e-retailer profit. Originality/value: This study provides new insights into the evolutionary dynamics of online markets. This work also indicates how to build an integrated simulation model with an agent-based model and TOPSIS and how to use an integrated simulation model and interpret its results.