A Model of Sponsored Results in Intelligent Recommenders and Search Engines

35 Pages Posted: 7 Sep 2008

See all articles by Hemant K. Bhargava

Hemant K. Bhargava

University of California, Davis

Juan Feng

University of Florida - Warrington College of Business Administration

Date Written: August 12, 2008

Abstract

Search engines, referrals services, advisors, and other forms of information gatekeepers and recommenders, are omnipresent in today's information-heavy economy. These services widely employ sponsored recommendations, wherein merchants pay the recommender in return for favorable placement in recommender's list. We develop and analyze an economic model of sponsored search on recommenders, in a setting where consumers cannot a priori discern low and high quality merchants or recall all the relevant merchants. Our model extends traditional advertising literature by incorporating an intelligent recommender who aims to advice and influence consumers, while also serving as an advertising platform for merchants. We find that an intelligent recommender not only makes consumers better-informed, but may also cause merchants to modify their price-advertising strategies in a way that favors consumers. Depending on the cost and quality differences between the merchants, the sponsored result may either serve a quality signaling role, or enable the recommender to exploit its leverage in making consumers aware of the merchant. As the recommender technology becomes more accurate at discerning firms' quality, the signaling demand reduces. However the recommender then becomes more capable of coercing lower quality firms to pay for inclusion in the search results. The recommender's overall advertising revenue increases; product prices drop due to a shift in the the higher quality firm's signaling effort from price to advertising; and consumer surplus increases. Our analysis assumes an advertising-based business model. When the recommender lacks sufficient pricing power over sponsored results, a higher accuracy level may lower advertising revenue and consumer welfare, hence the recommender should consider other revenue sources.

Suggested Citation

Bhargava, Hemant K. and Feng, Juan, A Model of Sponsored Results in Intelligent Recommenders and Search Engines (August 12, 2008). UC Davis Graduate School of Management Research Paper No. 04-08, Available at SSRN: https://ssrn.com/abstract=1264018 or http://dx.doi.org/10.2139/ssrn.1264018

Hemant K. Bhargava (Contact Author)

University of California, Davis ( email )

One Shields Avenue
Apt 153
Davis, CA 95616
United States

Juan Feng

University of Florida - Warrington College of Business Administration ( email )

Gainesville, FL 32611
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
101
Abstract Views
1,047
Rank
476,054
PlumX Metrics