A prediction study on E-commerce orders based on site search data

The search data of E-commerce site reflects millions of online consumers´ concerns and interests, as well as trends of their behavior, so it can also provide essential data basis for the study of daily number of E-commerce orders. This paper firstly establishes a conceptual framework, revealing that there is a certain correlation and lead-lag relationship between Ecommerce site search data and orders. Then, by empirical analysis of orders and search data of an E-commerce site, the paper processes a search data index and builds a prediction model based on it, which confirming that this relationship is statistically significant. Prediction results of the following seven days show that it works well, which can be helpful for Ecommerce enterprises to improve their inventory management and its steady development.