Answers are grounded in this paper and cite their sources.
Understanding the brand competition pattern is extremely important for automakers and dealers to identify emerging competitors and respond agilely to market changes. However, the traditional methods are subject to the analysts' experiences and analytical dimensions, which lack a reliable methodology and ignore the value of user behaviors. Therefore, in this paper, we incorporate the complex network into our model to explore and analyze the brand competition patterns in China's auto market with users' sales leads in detail. Our investigation involves 188,085,987 sales leads of 213 brands from a vertical auto website, and it has three important contributions. First, we propose an innovative visualization approach to help automakers quickly identify competitors and market positioning. Second, we analyze both the static and dynamic characteristics of the auto brand competition network, and reveal characteristics including the small-world phenomenon, disassortativeness, and preferential attachment. Finally, we discover the core-periphery structure in the network, and introduce a novel method to identify the core/periphery characteristics of brands with centrality metrics and K-Means clustering. Besides, features such as the number of sales leads, the number of models of the brand, and the sales price range are found to accurately predict the core/periphery characteristics of the brand.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2019) studied this question.