Press Releases

The Topmost Positions Are Not Always Most Profitable for Advertisers in Search Engine Auctions

NEW YORK--In their new research paper entitled, “An Empirical Analysis of Search Engine Advertising: Sponsored Search in Electronic Markets,” NYU Stern Professors Anindya Ghose and Sha Yang challenge the prevailing viewpoint that the topmost positions are always the most profitable ones for advertisers in search engine auctions. This is the first published study to quantify the impact of keyword type and length, position of the advertisement, and the landing page quality on consumer search and purchase behavior as well as on advertiser’s cost-per-click and the search engine’s ranking decision for different ads. This paper provides insight to managers about sponsoring retailer-specific keywords (e.g., “Walmart bedsheets”), brand-specific keywords (e.g., “Nautica bedsheets”) or generic keywords (e.g., “bedsheets”). With millions of available keywords and uncertain click-through rates associated with each keyword, identifying the most profitable set of keywords is a very challenging task for firms. This paper builds a comprehensive quantitative model that helps advertisers address this problem.

Based on a six-month panel dataset of several hundred keywords collected from a large nationwide retailer that advertises on Google, Professors Ghose and Yang found:

  • The value-per-click to an advertiser decreases with each position down the search engine results page, meaning that clicks from more prominent positions are more valuable than clicks from lower positions, because conversion rates also decrease.
  • Even though the more prominent positions on the search engine results page experience higher click-through or conversion rates, they are not always the most profitable ones. In fact, profits are often lower in the top positions than those in the middle positions due to the aggressive nature of bidding that increases the total advertisement costs (given the high click-through rates).
  • Search engines are accounting for both the current period’s bid price as well as prior click-through rates of the keywords before deciding the final rank of an advertisement in the current period, but the current bid price has a larger role to play in determining the final rank than ‘Quality Score’-related factors like prior click-through rates.
  • An increase in the landing page quality score of the advertiser by one unit is associated with an increase in conversion rates by as much as 22.5 percent and a decrease in advertiser cost-per-click.

The paper is forthcoming in Management Science and the full working paper is available at:

Professor Ghose is an expert in Web 2.0, user-generated content, online advertising and e-commerce, and Professor Yang’s research focuses on understanding household purchase behavior and market competition.

To speak with Professor Ghose, contact him directly at 212-998-0807, (e-mail is best); or contact Jenny Owen in NYU Stern’s Office of Public Affairs at 212-998-0561,

May 26, 2009: