Dynamical prediction model of consumers’ purchase intentions regarding anti-smog products during smog risk: Taking the information flow perspective

Abstract

A growing number of residents have been exposed to city smog in recent years, resulting in rapidly increased sales of anti-smog products and even the occurrence of panic buying. To better understand the public’s consuming behaviors in this context and predict the trends of purchase intention, a dynamic prediction model during smog risk is proposed based on the protective action decision model (PADM) and the health belief model (HBM). Through a questionnaire survey, we measured the public’s psychological perception, including perceived severity and susceptibility to smog risk, and the perceived benefits and cost of three popular anti-particulate matter 2.5 (PM2.5) air purifiers. All perceptual beliefs with different group characteristics can be predicted by support vector machine regression algorithm. From the information flow perspective, this study introduced four time-varying functions to update the dynamic changes in public’s perception and improve the predictive power. We collected online sales data on three kinds of air purifiers during the 2016–2017 period. The simulation results indicated that the proposed model can be used to dynamically predict the sales of air purifiers in a smog risk context. For governments, the findings contribute to regulating the public’s irrational consuming behavior by risk communication and risk early warning, the firms can formulate corporate strategies on the pricing and marketing of anti-smog products.