Correspondence: Paat Rusmevichientong, School of Operations Research and Industrial Engineering, Cornell University, USA. Tel: +1 607 255 9698; Fax: +1 607 255 9129; E-mail: paatrus@cornell.edu

¹Paat Rusmevichientong is an Assistant Professor in the School of Operations Research and Industrial Engineering at Cornell University. His research interests include data mining, information technology, and nonparametric algorithms for stochastic optimization problems with applications to supply chain and revenue management.

²Joyce Salisbury is Manager, Interactive Research Tool Development for General Motors, overseeing the company's efforts to use the Internet to supplement or replace traditional research methods. She has the responsibility for identifying and testing new methods and technologies to enhance GM's understanding of the automotive consumer. Salisbury holds a bachelor's degree in Manufacturing Systems Engineering from GMI Engineering and Management Institute (now Kettering University) and a master's in Engineering Management and a M.B.A. from Northwestern University and Kellogg School of Management. Salisbury serves as a liaison to many university projects and is a frequent speaker at industry conferences.

³Lynn T. Truss is a Staff Research Scientist in the Manufacturing Enterprise Modeling Group of the Manufacturing Systems Research Lab at General Motors R&D Center in Warren, Michigan. She has worked for GM R&D for over 21 years, and helped develop a GM-proprietary and patented expert system for the design of experiments, which is still used throughout GM today. She is currently leading the Enterprise Demand Sensing Research Program, which is a global team focused on improving demand sensing technologies and enabling the propagation of demand information throughout the manufacturing enterprise. Lynn received a B.S. in Mathematics from the State University of New York at Albany in 1982, and an M.A. in Statistics from Yale University in 1983.

⁴Benjamin Van Roy is an Associate Professor of Management Science and Engineering, Electrical Engineering, and by Courtesy, Computer Science at Stanford University. His recent research interests include dynamic optimization, machine learning, economics, finance, and information technology.

⁵Peter W. Glynn is the Thomas Ford Professor of Engineering in the Department of Management Science and Engineering at Stanford University, and has a courtesy appointment in the Department of Electrical Engineering. He is a Fellow of the Institute of Mathematical Statistics. His research interests include computational probability, queuing theory, statistical inference for stochastic processes, and stochastic modeling.

Received 25 May 2005.

Abstract

Developed by General Motors (GM), the Auto Choice Advisor web site (http://www.autochoiceadvisor.com) recommends vehicles to consumers based on their requirements and budget constraints. Through the web site, GM has access to large quantities of data that reflect consumer preferences. Motivated by the availability of such data, we formulate a non-parametric approach to pricing GM vehicles, highlight opportunities and challenges in using online data, and contrast our approach with existing methodologies and traditional data sources. Our analysis provides insights into the current pricing practice and suggests enhancements that may lead to a more effective pricing strategy.