Leveraging online reviews for customer regional preference demand analysis: evidence from Chinese new energy vehicle reviews

Regional preference demand analysis can help enterprises timely capture emerging demands and trends in different regions, thereby promoting product innovation and optimising marketing strategies. However, existing research neglects the impact of regional features on customer demand preferences. To address it, this paper proposes an analytical framework of customer regional demand preference by developing a lexicon-based sentiment analysis method of online reviews and two-factor analysis of variance (ANOVA). We first develop a lexicon-based sentiment analysis method by constructing a sentiment dictionary and an auxiliary dictionary to define sentiment values regarding these attributes of individual online review. Then, considering decision-making psychology of customer, the sentiment values are adjusted based on prospect theory, and customer satisfaction is subsequently defined using adjusted sentiment values. Furthermore, customer demand preference is analysed from feature demand preference and product demand preference using two-factor ANOVA, in which regional feature mainly contains city level and geographic region. Finally, we perform an experiment using customers’ historical data collected from Autohome to explore customer regional demand preferences.