关键词: 搜索量, 消费品, 百度指数, 预测, VAR模型

Abstract: The traditional prediction of sales for consumer goods is often based on the sales data of consumer goods in the past time Since the data is often collected and published by the government, it is difficult to collect and has a time lag In order to study the relationship between the amount of network search and the demand for consumer goods, the network search traffic is used to predict the demand, and the sales volume and Baidu index of eight types of consumer goods with different characteristics are collected By using the co-integration test of econometrics and Granger causality test, the VAR model is established to verify that the predictive effect of the search traffic on consumer goods is related to the characteristics of consumer goods And the predictive effect of durable consumer goods is better than that of non-durable consumer goods, and the predictive effect of electronic consumer goods is also better than that of non-electronic consumer goods, and corresponding suggestions are made for related enterprises and government

Key words: search traffic, consumer goods, Baidu index, forecast, VAR model