A Nash bargaining approach to retention enhancing bid optimization in sponsored search auctions with discrete bids

Ramakrishnan Kannan, Dinesh Garg, Karthik Subbian, Y. Narahari
2008 2008 IEEE International Conference on Automation Science and Engineering  
Bid optimization is now becoming quite popular in sponsored search auctions on the web. Given a keyword and the maximum willingness to pay of each advertiser interested in the keyword, the bid optimizer generates a profile of bids for the advertisers with the objective of maximizing customer retention without compromising the revenue of the search engine. In this paper, we present a bid optimization algorithm that is based on a Nash bargaining model where the first player is the search engine
more » ... d the second player is a virtual agent representing all the bidders. We make the realistic assumption that each bidder specifies a maximum willingness to pay values and a discrete, finite set of bid values. We show that the Nash bargaining solution for this problem always lies on a certain edge of the convex hull such that one end point of the edge is the vector of maximum willingness to pay of all the bidders. We show that the other endpoint of this edge can be computed as a solution of a linear programming problem. We also show how the solution can be transformed to a bid profile of the advertisers. Index Terms-Bid optimizers, Nash bargaining, advertiser retention, sponsored search auctions, internet advertising, mechanism design 2 Number of clicks the bidder's website receives for $1 spent to display the advertisement. For example, if the website receives 80 clicks for $100 spent, bang-per-buck is 0.8
doi:10.1109/coase.2008.4626517 dblp:conf/case/KannanGSN08 fatcat:ynmub6rqvbdm5l7n7h72uswriy