Seminar (Dept. of Marketing)

Theme：Measuring the position effect of search rankings: Evidence from an online travel website

Speaker：Qingliang Wang（Assistant Professor, Northwestern Polytechnical University）

Time：2021-10-13 10:00

Address：Room 713, Mingde Business Building

Language：Chinese

Abstract：

Online platforms increasingly rely on ranked lists to display product information to consumers. These lists are typically constructed using machine learning algorithms that predict and optimize consumer responses to the ranked options. As a result, measuring the effect of a product’s position is challenging since the ranked list is endogenous. In this paper, we analyze the causal effect of a product’s position on consumer search outcomes using data from a travel search engine. The data is unique in that it provides information on the revenues that the platform made from the products listed, which we find correlate with the position of the product in the list. This feature of the data allows us to use revenues as an instrument for position and thereby measure the causal effect of position. In addition, we are able to quantify the estimation bias that arises from not correctly addressing the endogeneity of position.