Special Session Summary Online Shopping Environments: Empowered Consumers Or Browsers Beware?

Gerald Haubl, University of Alberta
[ to cite ]:
Gerald Haubl (2001) ,"Special Session Summary Online Shopping Environments: Empowered Consumers Or Browsers Beware?", in NA - Advances in Consumer Research Volume 28, eds. Mary C. Gilly and Joan Meyers-Levy, Valdosta, GA : Association for Consumer Research, Pages: 451-453.

Advances in Consumer Research Volume 28, 2001     Pages 451-453



Gerald Haubl, University of Alberta

Against the background of the rapid growth of electronic commerce and, more specifically, consumer online shopping, it becomes increasingly important to develop an understanding of how consumers process product information and make purchase decisions in electronic environments. Two important characteristics of artificial marketplaces are that they allow for a high degree of personalization of information and that they greatly facilitate stimulus-based, as opposed to memory-based, decision making. While these and other properties of electronic shopping environments may be highly beneficial to shoppers in some respects, they also provide new possibilities for influencing consumers’ preferences and, ultimately, their purchasing decisions.

The focus of this special session was on the question of how susceptible consumers’ shopping behavior in digital marketplaces is to being influenced in a significant way by characteristics of the information environment, in particular characteristics that can be very easily manipulated by vendors. The paper by HSubl and Murray examines this question in the context of electronic recommendation agents, and shows that the mere inclusion of selected attributes in the calibration and sorting algorithm of such agents can alter consumers’ preferences and that this preference-construction effect may persist even beyond agent-assisted shopping encounters. Te paper by Mandel and Johnson investigates the idea that Web page backgrounds may change preferences by influencing attribute importance through associative priming, and provides empirical evidence that this type of priming effect may occur in digital environments. Finally, the paper by HSubl and Popkowski Leszczyc examines the effects of different characteristics of Internet auctions that are under the vendor’s control on important auction outcomes, and shows that bidders tend to discount some unambiguously relevant pieces of information while unduly relying on certain types of irrelevant information.

The papers that were presented in this session are complementary in that each of them approaches an important mechanism by which consumers’ shopping behavior in digital marketplaces may be influenced. Taken together, the three papers enhance our understanding of human decision making in electronic environments which, given the rapid growth of electronic commerce and online shopping, is becoming an increasingly important area of consumer research.



Gerald HSubl, University of Alberta

Kyle B. Murray, University of Alberta

The objective of this paper is to examine the role of electronic recommendation agents in connection with the construction of preferences by decision makers in online shopping environments. Following HSubl and Trifts (2000), an electronic recommendation agent is conceptualized as a software tool that (1) calibrates a model of the preference of a consumer based on his/her input and (2) uses this model to make personalized product recommendations (in the form of a sorted list) in a decision task based on its understanding of the consumer’s preference structure.

Interactive artificial marketplaces allow for the unbundling of product information from the actual products. In such electronic environments, the constraints of physical space no longer govern the organization of information (Johnson, Lohse, and Mandel 1999). In addition, online shopping represents a retail format that dramatically increases the potential for consumers to make choices within highly interactive environments (Alba, Lynch, Weitz, Janiszewski, Lutz, Sawyer, and Wood 1997). Therefore, it is important to obtain a better understanding of how consumers process product information and make purchase decisions in artificial marketplaces that are effectively unconstrained in their presentation of information.

The interactive nature of artificial marketplaces allows for highly customized and personalized information environments. Johnson et al. (1999) argue that "because the decision environment can actually influence how ... preferences are constructed, it can influence what is chosen" and that "if, in fact, such manipulation can be #customized’ at the individual level, the potential for influencing choice is very significant" (p. 20).

A recommendation agent is an example of a tool that can personalize and customize information at the individual level. Evidence from HSubl and Trifts (2000) suggests that consumers tend to rely on such an agent to help them reduce their search costs and to improve their decision quality. Given this reliance on a recommendation agent to assist or even lead in the customization of the shopping environment, and the impact the environment can have on preference construction, it is particularly interesting to examine the potential influence that a recommendation agent can have on preferential choice processes.

Along these lines, we propose that a recommendation agent that considers only a subset of the attributes that are relevant in a particular product category and choice situation may induce a preference-construction effect that is reflected in a change of the relative importance that consumers attach to different attributes. Specifically, our key hypothesis is that, everyting else being equal, the mere inclusion of an attribute in a recommendation agent will render this attribute more prominent in consumers’ purchase decisions. The results of Experiment 1 provide support for the existence of this mere-inclusion effect in an agent-assisted shopping task.

Furthermore, we suggest three possible explanations of the mere-inclusion effect; it might be: (1) a direct consequence of the format of information presentation, (2) the result of feature-based priming, or (3) a reflection of decision makers inferences about relative attribute importance. These alternative explanations are investigated in Experiment 2. The results suggest that this type of preference construction is due primarily to consumers’ inferences about attribute importance. We also find that the mere-inclusion effect persists beyond the agent-assisted shopping experience and into subsequent preferential choice tasks in which no recommendation agent is available.


Alba, Joseph, John Lynch, Barton Weitz, Chris Janiszewski, Richard Lutz, Alan Sawyer, and Stacey Wood (1997), "Interactive Home Shopping: Consumer, Retailer, and Manufacturer Incentives to Participate in Electronic Marketplaces," Journal of Marketing, 61 (July), 38-53.

HSubl, Gerald and Valerie Trifts (2000), "Consumer Decision Making in Online Shopping Environments: The Effects of Interactive Decision Aids," Marketing Science, 19, 1, in print.

Johnson, Eric J., Gerald L. Lohse, and Naomi Mandel (1999), "Designing Marketplaces of the Artificial: Four Approaches to Understanding Consumer Behavior in Electronic Environments," working paper, Columbia University.



Naomi Mandel, Arizona State University

Eric J. Johnson, Columbia University

This paper examines the idea that Web page design can change preferences by influencing attribute importance. Specifically, preference construction (Payne, Bettman and Johnson, 1993) may be influenced by Web page backgrounds through associative priming, which occurs when a person retrieves an item from long-term memory and activation spreads automatically to other related items in memory (Herr 1989). When a subject looks at a Web page, he or she may experience this priming from stimuli such as the page’s background pictures or color, which may affect the subject’s judgments about the product’s attributes and whether to buy the product.

Most importantly, this paper explores the nature of priming effects on information search, and examines the possibility that priming changes external search as well as internal retrieval. If we believe that priming works by increasing the accessibility of product related information, we might suggest that its effects are primarily limited to memory-based choice. However, it is possible that priming leads to differences in search, thereby impacting stimulus-based search as well. In addition, experts and novices may differ on the richness of their internal representation of products, which may produce differences in the mechanism of priming.

Experiment 1 showed that visual stimuli can manipulate salient features. For example, subjects who were shown car descriptions after reading an introduction with a re flaming background rated safety significantly more important in automobile purchasing than did those who saw a green background with pennies. Experiment 2 replicated these results with a large sample, and also studied differences in the decision making process. The priming affected the order in which subjects examined attribute information, which in turn influenced their choices. Experiment 3 found that priming affected the choices of both experts and novices, but through different mechanisms. Novices spent an increased amount of time looking at information about the attribute on which they were primed, affecting their subsequent choices. However, for experts, information search did not mediate the effect of priming on choice.


Herr, Paul M. (1989), "Priming Price: Prior Knowledge and Context Effects," Journal of Consumer Research, 16, 67-75.

Payne, John W., James R. Bettman, and Eric J. Johnson (1993), The Adaptive Decision Maker, Cambridge: Cambridge University Press.



Gerald HSubl, University of Alberta

Peter T. L. Popkowski Leszczyc, University of Alberta

Due to the rapidly increasing prevalence of Internet auctions, particularly person-to-person auctions, as a selling format, developing an enhanced understanding of individuals’ bidding behavior in such auctions is critically important. While there has been a considerable amount of mostly theoretical research on auctions in economics (e.g., Milgrom and Weber 1982), the psychological aspects of bidding in auctions have received very little attention in the literature to date. This paper investigates the effects of several characteristics of Internet auctions on two important auction outcomes, (1) the number of bidders and (2) selling prices. The focus is on auction characteristics that are under the control of the vendor. The primary objective of this work is to examine whether individuals’ bidding behavior in Internet auctions reflects some of the context-induced biases that have been observed either in the context of fixed-price marketplaces or in laboratory studies of consumers’ value judgments. The paper is based on two controlled field experiments involving sets of real-world public auctions conducted on a large Internet auction site using ascending-bid auctions with predetermined end times.

Experiment 1 focuses on the effects of specifying a fixed price component (e.g., a shipping-and-handling charge), as well as of the magnitude of such a component, on bidding behavior and auction prices. From a consumer-behavior standpoint, it is important to understand how consumers process information about such fixed price components in the context of Internet auctions. Classical economic theory predicts that consumers take fixed price components into account perfectly when expressing their willingness to pay for a product in the form of a maximum bid in an auction. Thus, market demand should be invariant to the magnitude of the fixed price component, and the total price paid by the winner of an auction (i.e., the sum of the fixed price component and the winning bid) should be unaffected by the magnitude of this fixed charge.

Recent work by Morwitz, Greenleaf, and Johnson (1998) suggests that the mere patitioning of a given total price may decrease consumers’ recalled total costs and increase demand relative to an all-inclusive price. These authors proposed cost-benefit trade-offs in connection with the processing of information and anchoring-and-adjustment processes of information integration as potential explanations of why consumers might react more favorably to a partitioned than to a combined price. Since merely altering the presentation of a full price through different types of partitioned pricing may affect market demand for a product, it is of interest to examine whether a similar phenomenon can be observed in the context of Internet auctions that involve fixed charges. Based on the work by Morwitz et al. (1998), we predict that, when constructing their willingness to pay for a given product, bidders in Internet auctions will fail to fully take into account a fixed component of the product’s price.

The magnitude of fixed price components was varied systematically in a controlled field experiment that involved auctions of sets of collectable postage stamps. The results suggest that, as predicted, the fixed price component has a significant effect on the total selling price such that the latter increases as the magnitude of the fixed price component increases. The results of Experiment 1 suggest that bidders in real-world Internet auctions fail to fully take into account a fixed price component. Everything else being equal, consumers buying in online auctions tend to pay more for a given product when a fixed price component is included, and this effect increases as the magnitude of the fixed component increases.

Experiment 2 examines the effects of providing different types of external reference-price information on bidding behavior and selling prices in auctions. While there is ample evidence suggesting that external reference prices affect consumer purchase decision making in fixed-price markets (e.g., Mayhew and Winer 1992), their role in the context of auctions has not been explored to date. One important price-related piece of information that may be part of the stimulus environment of an Internet auction is the seller-specified reserve price. Potential bidders may use the reserve price as a cue when making inferences about a product’s value to them. That is, a reserve price may serve as a signal for value. Therefore, we predict that seller-specified reserve prices have a positive effect on bidders’ willingness to pay for a product, and that higher reserve prices will lead to higher selling prices in auctions. At the same time, however, an increase in reserve price will also lead to a decrease in the number of bidders. The latter may, in turn, have a negative effect on selling prices. Therefore, reserve prices should have a dual effect on selling prices in Internet auctions. First, higher reserve prices are expected to have a direct positive effect on selling prices through their role as a signal for product value. Second, they should have an indirect negative effect on selling prices that is mediated by a reduction in the number of bidders.

The magnitude of the signaling effect of seller-specified reserve prices should be positively related to the amount of ambiguity that bidders perceive with respect to the value of a product to them. That is, consumers will rely more heavily on reserve price as a cue for value when their ability to assess the subjective utility of a product is poor. One way of reducing the ambiguity about a product’s value is to provide potential bidders with an external reference price that is not under the complete control of the seller. Catalog values or other published types of price information may be used for this purpose. We expect that the availability of such an objective piece of reference-price information will reduce the signaling effect of reserve prices.

As in Experiment 1, a controlled field experiment using real-world auctions of stamps was conducted to test our hypotheses regarding the effects of external reference prices on auction outcomes. The magnitude of seller-specified reserve prices and the availability of objective reference-price information were varied systematically. A predicted, the inclusion (and increasing magnitude) of a reserve price reduces the number of bidders. In addition, consistent with our hypothesis, selling price increases as reserve price increases. Finally, the positive effect of reserve price on selling price may be diminished by the availability of an objective reference price.

This paper represents a first effort aimed at developing an understanding of consumers’ bidding behavior in person-to-person auctions on the Internet. The results of two controlled field experiments suggest that individuals who place bids in real-world Internet auctions are subject to seemingly irrational biasesBthey tend to discount some unambiguously relevant pieces of information while unduly relying on certain types of unimportant, and possibly irrelevant, information. These findings demonstrate that bidders’ valuations of products may be affected by characteristics of Internet auctions that can be specified by the seller in a largely arbitrary fashion, such as fixed price components and reserve prices. This suggests that auction markets on the Internet should not be uncritically classified as highly efficient. From a consumer-welfare standpoint, the findings reported here lead us to conclude that the potential for systematically manipulating consumer bidding behavior in Internet auctions appears to be large.


Mayhew, Glenn E. and Russell S. Winer (1992), "An Empirical Analysis of Internal and External Reference Prices Using Scanner Data," Journal of Marketing Research, Vol. 19. (June), 62-70.

Milgrom, Paul and Robert J. Weber (1982), "A Theory of Auctions and Competitive Bidding," Econometrica, Vol. 50, No. 5, 1089-1122.

Morwitz, Vicki G, Eric A. Greenleaf, and Eric J. Johnson (1998), "Divide and Prosper: Consumers’ Reactions to Partitioned Prices," Journal of Marketing Research, Vol. XXXV (November), 453-63.