Investigating the Spillover Effect of Keyword Market Entry in Sponsored Search Advertising
Forthcoming at Marketing Science
63 Pages Posted: 6 Jan 2017
Date Written: 2017
As Internet advertising infomediaries nowadays provide rich competition information, sponsored search advertisers are becoming more strategic when selecting keywords. This paper empirically examines the spillover effects in advertisers’ keyword market entry decisions, that is, how an advertiser’s likelihood of using a keyword is affected by competitors’ keyword entry decisions. We develop a structural model to characterize advertisers’ keyword market entry decisions. We apply the model to a panel dataset of 1,252 laptop-related keywords mainly used by 28 manufacturers, retailers, and comparison websites that advertise on Google. Our analysis leads to several interesting findings. First, an advertiser’s expected position affects the nature of the competition. In particular, the spillover effect from below-ranked competitors is always positive, while the spillover effect from above-ranked competitors is either positive or negative. Second, the spillover effect from above-ranked ads is directionally affected by firms’ product-line characteristics: the effect among firms offering homogenous products (e.g. comparison sites) is negative, whereas the effect among firms with more differentiated products (e.g. manufacturers and retailers) is positive. Third, the spillover effect from above-ranked ads is directionally affected by firms’ positions in a distribution channel: the effect from upstream (downstream) on downstream (upstream) firms tends to be negative (positive). Finally, a downstream firm is more likely to learn new keywords from an upstream firm but not vice versa. Our counterfactual simulations demonstrate that the keyword-specific competition information provided by infomediaries can improve the search engine’s revenue by about 5.7%.
Keywords: Search Advertising, Competition, Distribution Channel, Entry Game, Keyword Selection, Infomediary, Internet Marketing, Bayesian Estimation
Suggested Citation: Suggested Citation