********************************************************************
Seminar
Department
of Systems Engineering and Engineering Management
The Chinese University
of Hong Kong
------------------------------------------------------------------------------------------
|
|
|
|
|
Title
|
:
|
Optimizing the reserve price in sponsored search advertisement auction |
|
|
|
|
|
Speaker
|
:
|
Prof. Wei Yang |
|
|
|
School of Business
|
|
|
|
Long Island University, C.W. Post, New York |
|
|
|
|
|
Date
|
:
|
July 2nd, 2008 (Wednesday)
|
|
|
|
|
|
Time
|
:
|
4:30 p.m. - 5:30 p.m.
|
|
|
|
|
|
Venue
|
:
|
Room 513
|
|
|
|
William M.W. Mong Engineering Building
|
|
|
|
(Engineering Building Complex Phase 2)
|
|
|
|
CUHK
|
|
|
|
|
------------------------------------------------------------------------------------------
Abstract:
|
Among various types of internet advertisement sponsored search advertising is one of the fastest growing sources of revenue. The mechanism used by Google and Yahoo! of selling keywords is called the generalized second-price (GSP) auction. Here is how it works. When a web user submits a search term (``query\'\') to a search engine, the search engine displays a certain number of relevant sponsored links on
result pages in addition to search results. If the user clicks on an advertising link at position $i$, he is directed to the advertiser\'s web site. The advertiser would be charged by the search engine for sending the user to the web page for an amount that equals to the bid of the one at position $i+1$ plus a minimum increment.
Search engines influence advertisers\' bidding by setting fixed or variable reserve prices to increase revenue. However, it is not clear to them what should be the best reserve price. The results obtained in the literature so far are less adequate. In particular, most of these models do not reflect the dynamic feature of the auction and may not necessarily offer the best strategy for search engines. Our analysis is built on the ``locally envy-free equilibrium\'\' defined in Edelman et al. (AER 2007) where an advertiser in one position has no motivation to exchange position with the advertiser right above him. In a GSP auction the advertiser occupying the last ad position pays the reserve price, and those above him pay higher prices for their
positions according to the equilibrium condition. Hence, the level of reserve price set by the search engine directly determines its revenue. Given any reserve price we assume that the search engine faces a sequence of bidders whose arrivals follow a compound Poisson distribution. We derive the analytical form of the search engine\'s expected revenue and show that if the value per click of advertisers is a variable with
increasing generalized failure rate (IGFR), the expected average revenue function is quasi-concave, and there exists a unique reserve price that maximizes the search engine\'s revenue from the GSP auction.
(A joint work with Youyi Feng and Baichun Xiao.)
|
-------------------------------------------------------------------------------------------
Biography:
|
Dr. Yang is an assistant professor in the business school of Long Island University at C.W. Post in the US. He graduated from a joint Ph.D. program between the Tepper business school and school of computer science of Carnegie Mellon in 2004. His research areas include the modeling and analysis of revenue management and vehicle routing. He teaches MBA programs in New York, Shen Zhen and Paris. Dr.
Yang is interested in providing useful, industrial-strength decision support tools using operations research models and algorithms. The companies he consulted include FlightOptions in Cleveland, Regional
Container Liner in Singapore and Sichuan Airline. Dr. Yang had a Master degree in Electrical Engineering from Tsinghua University, and a Bachelor in Automation from HuaZhong University of Science and
Technology in China.
|
************************* ALL ARE WELCOME ************************
********************************************************************