Academic journal article Economic Inquiry

Tracking Customer Search to Price Discriminate

Academic journal article Economic Inquiry

Tracking Customer Search to Price Discriminate

Article excerpt

Suppose that every time you walked around the mall, somebody put a bar code on your shoulder and, as you walked into the shops in the mall, someone came up and scanned your shoulder and got the number ... went to a database, saying, "Ah, yes, that's Lesley who visited the shop next door 15 minutes ago." That's the level of surveillance that's going on on the Internet.

--Internet entrepreneur Jason Catlett in a 60 Minutes interview with Lesley Stahl ("Not As Private As You Think," aired November 28, 1999)


Transaction mechanisms in an electronic marketplace can be much different from those used in traditional markets. Predominantly, "brick and mortar" markets are described as sellers posting single offer prices that buyers either accept or reject through their purchase decisions. Without necessarily holding everything else constant, two buyers, each making concurrent purchases in the same store, pay the same posted price for an identical item. (1) In particular, a buyer that has visited multiple rival stores receives the same price as a buyer that impulsively purchases without comparing any prices.

In contrast, a combination of Internet technologies makes it possible to identify the search history of potential customers. Online retailers generally make use of a standard programming device that produces electronic files to tag individual customers with a unique identification. Commonly referred to as cookies, these small computer files are stored on the hard drive of the customer's computer. Developers created cookies so that a Web site could maintain state even though the HTTP protocol is stateless. This allows a Web site to recognize an individual and accommodate multiple-item purchases from an Internet retailer. (2) An unintended consequence of this technology is that other Web sites, namely rival sellers, could use a second Internet device called a Web bug to retrieve the identity of the Internet domain that placed a cookie, even though the specific information content of the cookie is typically encrypted. Two different types of Web bugs can ascertain the search history of a customer. A Web bug is a graphic--typically, one pixel by one pixel in size--on a Web page, and it is designed to monitor who is reading the Web page. Web bugs are virtually invisible because they are uncolored and so small. One type, an executable Web bug, is a file that monitors a machine's traffic and hard drive and periodically sends the information back to the Web site that planted the bug on the machine. The second type is not physically located on the machine and uses a technology called scripts (e.g., JavaScript, ActiveX, and Perl) to scan a hard drive searching for files. (3,4)

Because the process of identifying a customer and posting an offer price is electronic and immediate, sellers on the Internet can use this technology to post dynamic, customer-specific prices. (5) However, the major contributions to the rich literature on consumer search consider only models in which firms post a single price regardless of how informed the customer may be. For example, Braverman (1980), Varian (1980), Stiglitz (1979), and Salop and Stiglitz (1977) assume that a consumer is either fully informed or completely unaware of other stores' prices and that each firm posts a single price to any customer. Other work considers sequential consumer search, but again each firm posts a single price regardless of how informed the customer may be--see, for example, Stahl (1989), Stiglitz (1987), Rob (1985), Burdett and Judd (1983), Carlson and McAfee (1983), and Wilde and Schwartz (1979). (6) This article explores the fundamental changes to traditional sequential customer search in posted offer markets when firms have the ability to track customers and price discriminate accordingly.

Using controlled experimental markets, we investigate how the ability to track customers and offer differentially informed customers different prices affects the prices that buyers receive in aggregate. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.