The inverted U-shaped relationship between information entropy of keyword combinations and sales of digital products: Evidence from China Tmall

Abstract

As a result of a consumer's online purchase search task, the combination of keywords plays a crucial role in decision-making and sales on e-commerce platforms. The current research aims to examine how the information entropy of keyword combinations influences digital product sales. The data (N = 4755) was collected from Tmall.com, a prominent e-commerce platform in China, and the methods of natural language processing and regression were employed to process data. Our findings demonstrate that there is an inverted U-curve between the information entropy of keyword combinations that contains product name information entropy (PNIE) and brand name information entropy (BNIE) and sales instead of a linear relationship. Furthermore, logistics evaluation (LE) and seller service quality (SSQ) play differentiated moderating effects on the inverted U-curve relationship leading to various movement of turning points and shape changes of steepness or flatness. These findings contribute to understanding sales from the viewpoint of information entropy and providing compelling guidance for e-commerce enterprises on keyword advertising bidding, logistics, and service.