RESEARCH

Abstract Academic interest in the popularity and Research on Consumer Behaviour in Online success of online has been increas- ing. Although much research has been Auctions: Insights from a Critical Literature carried out in an attempt to understand online auctions, little effort has been made Review to integrate the findings of previous research and evaluate the status of the XILING CUI, VINCENT S. LAI AND CONNIE K. W. LIU research in this area. The objective of this study is to explore the intellectual develop- ment of consumer behaviour in online research through a meta-analysis of the published auction research. The findings of this study are based on an analysis of 83 articles on this topic published mainly in information systems (IS) journals between 1998 and 2007. The results indicate that the consumer beha- viour research on online auctions can be categorized into three major areas – facil- itating factors, consumer behaviour and auction outcomes. Based on this literature INTRODUCTION (Hannon 2004), and more research- review, directions for future research on e- ers are devoting effort to the area. auction consumer behaviour are discussed, Online auctions have become an Among the various interesting online including potential new constructs, unex- increasingly popular and efficient e- auction topics, consumer behaviour is plored relationships and new definitions and commerce method of facilitating the fast becoming one of the most pop- measurements, and suggestions for metho- participation of Internet users in ular research areas, and many studies dological improvements are made. trading activities through flexible pri- in this area have already been carried Keywords: , e-auction, online cing processes, convenient access and out. For example, research efforts consumer behaviour the availability of a large variety of (Bichler et al. 2001, Leloup and products. Although online auctions Deveaux 2001, Townsend and are not appropriate for every type of Bennett 2003) have been dedicated Authorspurchase, and they operate in a to investigating the fundamentals and manner different to traditional sales general development of Internet auc- Xiling Cui channels (Bapna et al. 2001b, 2003b, tions. There is also more specific ([email protected]) is a PhD candidate in MIS at the Chinese University of 2004), researchers (such as Dai and online auction research, such as stu-

Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 Hong Kong. Her research interests focus Kauffman 2002 and Pinker Seidmann dies on technology or service adop- on electronic commerce, online and Vakrat 2003) believe that they tion (Bosnjak et al. 2006, Hu et al. auctions, and IT investment analysis. can be used as an effective tool to 2004, Stafford and Stern 2002, Vincent S. Lai reduce purchase prices, save time, Zhang and Li 2006), beha- ([email protected]) is a streamline the bidding process and viour (Ba et al. 2003, Bapna et al. Professor in MIS at the Chinese enable a worldwide selection of sup- 2001b, 2004, Borle et al. 2006, University of Hong Kong. His Research pliers and products. However, if Easley and Tenorio 2004, Namazi focuses on IS adoption and diffusion, online auctions are not used judi- and Schadschneider 2006, Roth and virtual collaboration, electronic commerce, and global IS strategy. His ciously and their risks are not carefully Ockenfels 2002), reputation or trust articles have been published in CACM, assessed, then these types of cyber- (Ba et al. 2003, Gregg and Scott

2008 Electronic Markets J MIS, DSS, I&Mt, EJIS, EJOR, JIT, space purchases can have adverse 2006, MacInnes et al. 2005, Melnik

ß among others. effects, such as damaging supplier and Alm 2002, Standifird 2001, Connie K. W. Liu relationships, choosing incapable sup- 2002), the winner’s curse (Bajari and ([email protected]) is a pliers, driving out qualified suppliers Hortacsu 2003, Oh 2002), and auc- PhD student at the Chinese University Volume 18 (4): 345-361. www.electronicmarkets.org DOI: 10.1080/10196780802420752 Copyright and underestimating the total costs tion mechanisms (Budish and Takeya- of Hong Kong. Her current research associated with the lower purchase ma 2001, Hann and Terwiesch 2003, focuses on institutionalization, ERP, price (Hartley et al. 2006, Jap 2003). Kauffman and Wang 2001, Mathews imitation and IS adoption. She has previously published in Journal of In recent years, online auctions 2004, Mathews and Katzman 2006, Database Marketing & Customer have received wider acceptance, parti- Spann et al. 2004, Spann and Tellis Strategy Management. cularly among larger organizations 2006, Standifird et al. 2004–5). 346 Xiling Cui, Vincent S. Lai and Connie K.W. Liu & Consumer Behaviour in Online Auctions: A Review

However, most of these previous studies of consumer identical items and has been used for perishable commod- behaviour in online auctions have been carried out ities such as flowers and fish. A sealed first-price auction largely on a piecemeal basis, thus inhibiting the accepts bids in a concealed fashion, and the bidder with emergence of cumulative findings. Fortunately, a few the highest bid wins the auction. A sealed second-price recent studies have brought together some of the earlier auction, which is also known as a , is similar investigations on online auctions. For example, Pinker et to a sealed first-price auction, except that the highest al. (2003) reviewed the current business and research bidder wins, but pays only the second-highest bid. This issues in online auctions and made a plea for more auction type is designed to alleviate bidders’ concern of research into B2B auctions. Chakravarti et al. (2002) overpaying for the products they bid for. In online performed a literature review of pre-Internet auctions settings, multi-unit auctions may extend the sealed that can provide insight into online auction research. second-price mechanism to third-price or even further Baker and Song (2007) reviewed the existing studies of down the line, and the bids are not necessarily sealed. Like single-item online auctions and proposed a model for in the multi-unit auctions in eBay and Taobao.com, future online auction research. Despite all of these bidders do not only bid on prices but also on the quantity efforts, it is still critical that researchers develop a they want to buy. All winning bidders pay the same price unifying framework and synthesize the current findings per item, the lowest successful bid price. This is also called on consumer behaviour in online auctions to provide a Dutch Internet Auction. structure for the state of the knowledge that exists in this In addition to these mechanisms, new auction mechan- field. The identification of the controversial points in isms have also emerged to facilitate online auctions. For these research findings and a distillation of the manage- example, name-your-own-price (NYOP) is a variant of the rial implications of this new area of knowledge sealed first-price auction mechanism, in which a bidder (McKinney et al. 2002) are also essential. As a first step submits a bid for an auctioned item, and the sellers towards the synthesis of the research findings on respond by either accepting or rejecting it. The group consumer behaviour in online auctions, and to deepen buying auction mechanism has also developed rapidly due and develop the theoretical foundations of this area of to the ease of Internet access. In group buying auctions, research, it is critical that we examine previous efforts to bidders submit an order bid rather than a price bid. The identify trends for new research opportunities. selling price drops as more buyers place their orders, which This paper aims to present a more comprehensive aggregates the power of buyers to gain volume discounts overview of the current research literature on consumer (Chen et al. 2006, Kauffman and Wang 2001). behaviour in online auctions in the hope that more valid Other than these mechanisms, online auctions have new research topics and directions can be derived. This features added during the derivation process from their literature review includes multiple-item auctions with an traditional counterparts. These new features fall into three emphasis on consumer behaviour. The research ques- categories, auction setting, bidding behaviour and service tions posed by this paper address the following. facilities, as shown in Table 1. In an auction setting, the ending rules of an online auction can either be ‘hard close’ 1. What types of consumer behaviour have been or ‘soft close’. With the former, the auction ends at a fixed investigated in online auctions? Can these findings time that has been made known to all bidders (Namazi

Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 be synthesized to provide a structure for the state of and Schadschneider 2006). In a soft-close English the knowledge in this domain? auction, in contrast, the auction can be extended for a 2. What bidding strategies, if any, have been system- set period of time with each new bid. Probably due to the atically and successfully adopted in online auctions? hard-close rules and the longer duration of online 3. Based on the literature review, can research gaps be auctions, a unique and interesting sniping phenomenon identified for possible future research endeavours? has also emerged in online auctions. In this strategy, bidders wait until nearly the end of the auction to submit their bids (Borle et al. 2006, Roth and Ockenfels 2002). FUNDAMENTALS OF ONLINE AUCTIONS In addition, online auctions also offer unique services that are only possible with Internet technologies. For Four basic online auction mechanisms have been adopted example, reputation systems provide sellers and buyers from traditional offline auctions – , Dutch with the opportunity to rate each other and leave auction, sealed first-price auction, and sealed second-price comments after each transaction; reference prices are auction. An English auction begins with the lowest provided by the online auction intermediary (or the acceptable price and then increases the bids until no sellers in some cases) to help buyers compare the bidder will increase it further. The winner is the one with different prices on offer; and buyout options are provided the highest bid, if price is the only criteria. A Dutch by sellers to give buyers an alternative method of auction, in contrast, begins with a high asking price that is acquiring an item for a fixed price without having to gradually reduced until someone is willing to accept it. wait for the entire process to be completed. In addition, Usually the is designed for multiple online auction sites also offer escrow services through a Electronic Markets Vol. 18 No 4 347

Table 1. Comparison between online auctions and traditional article pool. This process began by reading the abstract of auctions a paper to determine its relevance to the research topic. The papers that were selected individually were then Traditional Online compared and discussed, and a final list was compiled after auction auction much deliberation. The selection process took nearly two months and resulted in a final research pool of 83 papers, Auction Auction space Limited Almost as shown in Table 2. These papers were then reviewed setting unlimited separately by the two authors, although weekly meetings Auction types Many More were conducted to consolidate the findings. Auction duration Short Long Auction ending Soft close Mainly hard close RESEARCH RESULTS Bidding Entry to an auction Not easy Easy behaviour Perception of Easy Difficult The review of the selected articles suggests that the study potential bidders of consumer online auction behaviour diverges widely in Multiple-item bidding No Yes terms of both its scope and focus. It is therefore impossible Snipe bidding No Yes to summarize and synthesize all of the findings of prior Service Reputation systems No Yes studies. As anticipated, many of these studies produced facility Escrow services No Yes contradictory or inconclusive findings, thus increasing the Reference prices No Yes difficulties the authors faced in synthesizing the literature. Buyout option No Yes Consequently, our survey is not an exhaustive one. We intend merely to pinpoint the major findings of the prior studies to highlight some of the critical efforts that have neutral trusted-third-party (TTP) to guarantee the been undertaken in the pursuit of knowledge on online transaction security of buyers and sellers. With this auctions. Of the 83 online auction papers on consumer service, the buyer deposits money into the safekeeping of behaviour reviewed, the research foci and topics can be a TTP, pending on the satisfaction of the contractual summarized and categorized into three main categories: conditions. Once these conditions have been fulfilled, 1) facilitating factors; 2) consumer behaviour; and 3) the TTP sends the money to the seller. auction outcomes, with the first category affecting the second two. The following sections of the paper provide a LITERATURE REVIEW RESEARCH METHODOLOGY detailed discussion of these three categories.

Although the research on consumer behaviour in online Facilitating factors auctions has been carried out across several disciplines, namely, information systems (IS), marketing, economics Facilitating factors refer to the factors that facilitate and psychology, the aim of this paper is to provide a review

Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 online auctions and affect consumer behaviour and the of the literature on consumer behaviour research on online auctions within the IS discipline. A two-stage process was auction outcome. Of the studies reviewed, the facilitat- followed in searching the available pool of articles – article ing factors identified can be further classified into three identification and article analysis. In the article identifica- categories, based on the different perspectives of the tion stage, a thorough search was carried out on the intermediary, the seller and the buyer. ProQuest database for the 10-year period from 1998 to 2007, using ‘electronic auction’, ‘e-auction’, ‘online Intermediary factors. Auction intermediaries can auction’ and ‘Internet auction’ as the search keywords. manipulate many variables to enhance their competi- The abstracts of the top 10 MIS journals (Rainer and tiveness and achieve greater success in their auction Miller 2005) were also manually searched for the same operations. Of these variables, the technology property of period in case not all of the articles were found in their online platforms is clearly the most crucial variable. ProQuest. Only articles that were available in full were For example, the effects of the site architecture and the retrieved. Comments, notes, columns, book reviews, informational content and design of auction sites have conference papers, working papers and news reports were been found to be critical to buyers’ ratings and excluded, as were papers that address , perceptions of usability and to their subsequent overall technical algorithms, B2B auctions (because businesses site preferences (Kwon et al. 2002). In addition, exhibit different types of behaviour than do consumers) researchers have also applied the Theory of Reasoned and auction design. Action (TRA) and the Technology Adoption Model At the article analysis stage, two authors were indepen- (TAM) to validate this correlation between technologi- dently involved in selecting the relevant papers from the cal characteristics and auction site adoption (Bosnjak 348 Xiling Cui, Vincent S. Lai and Connie K.W. Liu & Consumer Behaviour in Online Auctions: A Review

Table 2. Papers on consumer behaviour in online auctions

Year Journal Reference

1999 American Economic Review Lucking-Reiley 1999 2000 Information Technology and Management Bapna et al. 2000 Journal of Industrial Economics Lucking-Reiley 2000 Marketing Letters Wilcox 2000 Quarterly Journal of Electronic Commerce Herschlag and Zwick 2000 2001 Communications of the ACM Bapna et al. 2001b Decision Support Systems Bapna et al. 2001a Economics Letters Budish and Takeyama 2001 Electronic Commerce Research Bichler et al. 2001; Leloup and Deveaux 2001 Journal of Management Standifird 2001 Journal of Management Information Systems Kauffman and Wang 2001 Marketing Letters Dholakia and Soltysinski 2001 2002 American Economic Review Roth and Ockenfels 2002 Artificial Intelligence Magazine Ockenfels and Roth 2002 Behaviour and Information Technology Kwon et al. 2002 Economic Inquiry McDonald and Slawson 2002 Electronic Markets Standifird 2002 European Journal of Information Systems Rafaeli and Noy 2002 International Journal of Electronic Commerce Oh 2002; Stafford and Stern 2002; Ward and Clark 2002 International Journal of Information Management Wang et al. 2002 Journal of Industrial Economics Melnik and Alm 2002 MIS Quarterly Ba and Pavlou 2002 2003 Brazilian Electronic Journal of Economics Mathews 2003 Communications of the ACM Bapna 2003; Townsend and Bennett 2003 Decision Support Systems Ba et al. 2003 Management Science Bapna et al. 2003a; Hann and Terwiesch 2003; Pinker et al. 2003 Information Systems Research Bapna et al. 2003b Journal of the Operational Research Society Brint 2003 Journal of Consumer Psychology Ariely and Simonson 2003 Journal of the Academy of Marketing Science Hartley et al. 2006; Jap 2003 Journal of Computer Information Systems Ottaway et al. 2003 Psychology and Marketing Gilkeson and Reynolds 2003 Rand Journal of Economics Bajari and Hortacsu 2003 2004 Applied Economics Letters Dodonova and Khoroshilov 2004 Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 Electronic Markets Mollenberg 2004 Information Systems Research Hu et al. 2004; Pavlou and Gefen 2004 Journal of Consumer Research Kamins et al. 2004 Journal of Economics Mathews 2004 Journal of Interactive Marketing Bruce et al. 2004; Heyman et al. 2004; Spann et al. 2004 Management Science Easley and Tenorio 2004 MIS Quarterly Bapna et al. 2004 2004–5 International Journal of Electronic Commerce Standifird et al. 2004–5 2005 Electronic Commerce Research and Applications Kauffman and Wood 2005 Electronic Markets Rafaeli and Noy 2005 International Journal of Consumer Studies Cameron and Galloway 2005 International Journal of Electronic Commerce MacInnes et al. 2005 Internet Research Yang 2005 Journal of Interactive Marketing Dholakia 2005 Journal of Marketing Research Park and Bradlow 2005 Management Science Carare and Rothkopf 2005; Ding et al. 2005; Terwiesch et al. 2005 Marketing Science Dholakia and Simonson 2005 Organizational Behaviour and Human Decision Ku et al. 2005 Processes The Rand Journal of Economics Ariely et al. 2005 Electronic Markets Vol. 18 No 4 349

Table 2. Continued

Year Journal Reference

2006 Economic Theory Mathews and Katzman 2006 Decision Support Systems Antony et al. 2006; Chen et al. 2006; Lin et al. 2006; Onur and Tomak 2006; Zhang and Li 2006; Zhang 2006; Information & Management Amyx and Luehlfing 2006 Information Technology and Management Kauffman and Wood 2006 International Journal of Electronic Commerce Gregg and Scott 2006 International Journal of Operations and Production Hartley et al. 2006 Management International Journal of Modern Physics Namazi and Schadschneider 2006 Journal of Consumer Behaviour Stern and Stafford 2006 Journal of Economics & Management Strategy Houser and Wooders 2006 Journal of Marketing Spann and Tellis 2006 Statistical Science Borle et al. 2006 2007 Journal of Electronic Commerce in Organizations Baker and Song 2007 2008 Decision Support Systems Bapna et al. 2008

et al. 2006, Stafford and Stern 2002) and have identified three effects that exist simultaneously in group confirmed it to be significant. buying: the ‘positive externality effect’, the ‘price drop In addition to technology properties, auction inter- effect’ and the ‘ending effect’. The externality effect mediaries can also manipulate their auction sites by postulates that existing orders can stimulate new orders, offering different auction mechanisms, for example, the which means that a greater number of existing orders traditional English auction, the Dutch auction or certain can potentially generate more new orders, whereas the innovative online auction mechanisms, such as name- price drop effect suggests that prices are more likely to your-own-price (NYOP) and group buying auctions. We fall when they are closer to the next discount level, and find that English auctions have been investigated most the ending effect means that bidders are more likely to frequently, encompassing research into bidding strate- place orders during the last three hours of an auction. gies (Bapna et al. 2004), revenue comparison (Bapna et Chen, Chen and Song (2006) have argued that when al. 2001a), bidding behaviour (Ariely and Simonson economies of scale are considered, group buying 2003) and fraudulent acts (Kauffman and Wood 2005). auctions outperform fixed price mechanisms if the seller Dutch auctions have received far less attention because is risk-seeking. they are far less common among the popular auction In addition to technology properties and auction sites. However, contrary to intuition, an interesting mechanisms, intermediaries can also manipulate their Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 study that applied a decision theory model and a game service facilities to improve their service and performance theory model to investigate bidding behaviour suggests levels. The types of facilities that have been researched that bidders in Dutch auctions prefer to purchase objects include TTPs, reference points, reputation systems and sooner and at a higher price to economize on the cost of buyout options. The digital certificates issued by TTPs return when it is positive (Carare and Rothkopf 2005). have been found to be effective in forcing market In the NYOP research, experienced bidders are found to participants to behave honestly (Ba et al. 2003). exhibit lower frictional costs, which are the implicit and Effective TTPs have also been found to engender trust, explicit costs associated with market transactions, than not only in a few reputable sellers, but in the entire are inexperienced bidders (Hann and Terwiesch 2003). community (Pavlou and Gefen 2004). In the reference Research also suggests that a large number of NYOP point research, Dholakia et al. (2005) found that consumers do not exhibit rational decision-making, as participants submit fewer, lower and later bids when predicted by an economic model (Spann and Tellis reference points are provided. Reference prices have also 2006). Consumers who place many bids with rather been found to have a negative association with over- long inter-bid times are more likely to bid irrationally. bidding (Amyx and Luehlfing 2006) and a direct impact However, if they are willing to haggle extensively, then on the final bid price (Kamins et al. 2004). they can, on average, obtain a product at a lower price The reputation systems research suggests that the (Terwiesch et al. 2005). negative feedback posted in the reputation systems of Research on group buying, another new mechanism online auction sites can predict future incidences of in online auctions, has also generated many interesting auction fraud and that experienced bidders are better at findings. Kauffman et al. (2001), for example, have using this information to avoid such fraud (Gregg and 350 Xiling Cui, Vincent S. Lai and Connie K.W. Liu & Consumer Behaviour in Online Auctions: A Review

Scott 2006). The research on the buyout option has also increase the perceived utility of auction products provided some interesting results. For example, it has (Kauffman and Wood 2006), although other research been found that a buyout price is appealing to risk-averse has found that product pictures are not critical (Ottaway bidders (Budish and Takeyama 2001). Nevertheless, if et al. 2003). The research on starting prices suggests the seller and the bidder are risk-neutral, then the former that a low starting value can draw in more bidders, but will choose a buyout price that is sufficiently high to that their bids are likely to be relatively low, thus ensure that the option will never be exercised (Mathews eventually leading to a low final price – a phenomenon and Katzman 2006). However, time impatience on called the ‘anchoring effect’ (Ariely and Simonson either side of the transaction can motivate the seller to 2003). Anchoring effect means that people’s valuation offer an option price that is low enough to be exercised usually starts from some starting value, named ‘anchor’, when bidders do not discount future transactions and is adjusted up or down based on their assessment or (Mathews 2004). Empirical observation has also con- beliefs (Tversky and Kahneman 1974). In online firmed that time-impatient bidders are likely to choose auctions, the starting price often works as the anchor the buyout options (Standifird et al. 2004–5). Table 3 and then has certain impact on the bidders’ bidding summarizes this discussion of facilitating factors. behaviour and the final price, which has also been empirically confirmed through a field experiment with Seller factors. There are two research sub-streams Bidz.com (Dodonova and Khoroshilov 2004). related to sellers: auction settings and the seller’s Researchers have also found that bid increments have a reputation in online auctions. Auction settings usually positive effect on bidder participation when the incre- refer to product characteristics (such as the product ment is below a certain value, but to have a negative picture and description), starting price, bid increments, effect when the increment is above that value (Bapna, auction duration and auction time. In general, research Goes and Gupta 2001b). As for auction duration, some on auction products has already confirmed the correla- researchers found it to be positively associated with the tion between product characteristics and auction site final price (Brint 2003, Lucking-Reiley 1999) while ratings (Kwon others did not find such an effect in their studies et al. 2002). The inclusion of product pictures and (Gilkeson and Reynolds 2003, McDonald and Slawson descriptions on auction sites has also been confirmed to 2002). Interestingly, some even obtained contrary

Table 3. Intermediary factors affecting bidding behaviour

Sub-stream Factors Research findings References

Technology Theory of Reasoned Action/Technology Generalized to the online Bosnjak et al. 2006; Stafford and Stern properties Adoption Model (TRA/TAM) auction setting 2002 Information design Related to the buyers’ overall Kwon et al. 2002

Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 preferences Auction English auction Most research is conducted in mechanism this context Dutch auction Bidders prefer to bid sooner Carare and Rothkopf 2005 Name-Your-Own-Price (NYOP) Experienced bidders are more Hann and Terwiesch 2003; Spann et al. rational 2004; Spann and Tellis 2006; Terwiesch et al. 2005 Group buying Network effect/price drop Chen et al. 2006; Kauffman and Wang effect/ending effect 2001 Service facility Trusted-Third-Party (TTP) Enhances bidders’ Ba et al. 2003; Pavlou and Gefen 2004 honest/active behaviour Reference point Reduces bidding fever; Amyx and Luehlfing 2006; Ariely and increases sniping; mediates Simonson 2003; Brint 2003; Dholakia and starting price on final price Simonson 2005; Kamins et al. 2004 Reputation system Helps to predict future auction Gregg and Scott 2006 fraud Buyout option Insure risk-aversion/impatient Budish and Takeyama 2001; Mathews bidders 2004; Mathews and Katzman 2006; Standifird et al. 2004–5 Electronic Markets Vol. 18 No 4 351

findings that auction duration is negatively correlated Bidder factors. The two major research sub-streams with the final price (Ariely and Simonson 2003). They related to bidders that we have identified are the argued that it is probably due to the trumping of the bidder’s own attributes and the social influence of ‘arrival process effect’ by the ‘waiting time effect’. The other bidders. Bidder attributes mainly concern the arrival process effect means that the longer an auction bidder’s reputation, experience and emotions. lasts, the greater the likelihood of the auction site being Researchers have found that a bidder’s reputation has hit by new bidders, thus resulting in the accumulation of an impact on the likelihood of a dispute, but that this more bidders and an increase in the competitive pressure effect is less significant than the seller’s reputation on the auction’s closing price. The waiting time effect, in (Houser and Wooders 2006, MacInnes et al. 2005). contrast, is the utility of delay during the auction period Although some online auctions may ask the buyer and and results in a low final auction price. Another possible the seller to rate each other’s reputation after each explanation for the conflicting results was found in a transaction, most sellers tend to be lenient in their recent study that the relationship between the auction feedback, which eventually results in unreliable ratings duration and the final price may be impacted by sellers’ for bidder reputations. Probably due to this phenom- reputation (Bapna et al. 2008). A consumer may prefer a enon, research has found that bidders’ reputation short auction to a long one, because it may result in ratings are rarely used as a reference for bidding fewer search, delay and monitoring costs during the behaviour in online auctions (Dholakia 2005). course of the auction (Kauffman and Wood 2006). However, experience is considered important in Research also suggests that the timing of an auction can determining bidders’ behaviour and the strategies they have a significant effect on the total number of bids and adopt during auctions. In other words, bidders gain hence on the final price. A ‘weekend effect’ has also been experience from previous auctions, which allows them identified. This effect suggests that bidders are willing to to learn the bidding strategies that are more likely to be pay a higher price for an auction item if the end of the successful. However, when strategic dominance is auction coincides with a weekend (Kauffman and Wood weak, bidders’ experience has a weaker impact on their 2006). bidding behaviour (Wilcox 2000). On the emotional Studies on the seller’s reputation have also generated front, research has reported that bidders experience quite a few interesting research findings. An analysis of strong emotions in terms of their excitement at the reputation data in eBay’s feedback forum suggested winning and frustration at losing, and that these that sellers’ reputations follow a lognormal distribution emotions change dynamically as a function of the (Lin et al. 2006). Empirical studies have identified that outcomes of their previous bids (Ding et al. 2005). they have a consistent, statistically significant and Another popular research stream from the bidder’s positive impact on the final prices achieved at auctions perspective is social influence, which investigates the (Ba and Pavlou 2002, Houser and Wooders 2006, social effect of submitted bids on the bidding inten- Melnik and Alm 2002, Ottaway et al. 2003) and that tions of others. Kauffman et al. (2006), for example, this area deserves to be investigated further found that bidders tend to pay more for items that (McDonald and Slawson 2002). Standifird (2001), other bidders are also interested in. This is called the

Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 along the same lines of research, found that a positive ‘herd effect’. Heyman et al. (2004) found that a greater reputation on the part of a seller has only a mild number of participants in the early stages of an auction influence on the final bid price, but that a negative will lead to more active bidding at a later stage, and reputation rating has a critical influence on it. In they called this the ‘quasi-endowment effect’. This addition to its effect on the final bid price, other effect contributes to the sense of ownership that research has indicated that a seller’s reputation can bidders develop during an auction, even though they also have a significant, negative impact on the do not yet have legal claim to the item in question. In likelihood of a dispute in online auction transactions addition to these herd and quasi-endowment effects, (MacInnes et al. 2005). In addition to reputation researchers have also explored the effect of social ratings, Standifird (2002) has also investigated reputa- facilitation, which is the tendency for a bidder to tion types by comparing the auctions conducted perform better in terms of bidding behaviour and through three online auction venues, CNET, eBay performance in the presence of other bidders. Rafaeli et and Amazon, which represent three different types of al. (2005), for example, identified the presence in reputation schemes. The results indicated that a much virtual environments of social facilitation, which higher final bid price was likely to be achieved when motivates a stronger purchasing decision among computer-related items were auctioned through customers. They also indicated that the social presence CNET or eBay rather than through Amazon, demon- of other bidders in online auctions affects not only the strating that different reputation types may engender price, the timing and the number of bids, but also the different auction outcomes. Table 4 summarizes our auction’s market outcomes. Table 5 summarizes our discussion of the seller factors. discussion of the buyer factors. 352 Xiling Cui, Vincent S. Lai and Connie K.W. Liu & Consumer Behaviour in Online Auctions: A Review

Table 4. Sellers’ factors affecting bidding behaviour

Sub-stream Factors Research findings References

Auction setting Product Enhances bidding intention Kauffman and Wood 2006; Kwon et al. description/Image 2002; Standifird et al. 2004–5; Zhang and Li 2006 No image effect found Ottaway et al. 2003 Starting price Positively associated with final price Ariely and Simonson 2003; Brint 2003; Gilkeson and Reynolds 2003; McDonald and Slawson 2002 Negatively associated with the # of bids; Ariely and Simonson 2003; McDonald and mediated by price comparison/anchoring Slawson 2002 effect Negatively associated with auction Gilkeson and Reynolds 2003 success Bid increment Positive effect on active participation Bapna et al. 2003b As it increases, it first increases the Bapna et al. 2003a; Bapna et al. 2001b; auctioneer’s revenue, then reduces it Bapna et al. 2003b Duration Negatively related to the # of bids/final Ariely and Simonson 2003 price Positively associated with final price Brint 2003; Lucking-Reiley 1999 No effect on auction success or final Gilkeson and Reynolds 2003; McDonald and price Slawson 2002 The upper relationship is impacted by Bapna et al. 2008 sellers’ reputation End timing Weekend effect, midnight effect Kauffman and Wood 2006; McDonald and Slawson 2002 No effect Brint 2003 Sellers’ attributes Sellers’ reputation Positively related to intention to Ba and Pavlou 2002; Houser and Wooders bid/final price 2006; Kauffman and Wood 2006; McDonald and Slawson 2002; Melnik and Alm 2002; Ottaway et al. 2003; Standifird 2001; Standifird et al. 2004–5 Moderated by the item price Bruce et al. 2004 No effect on the final price Ariely and Simonson 2003; Gilkeson and Reynolds 2003 Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 Positive and negative (or high and low) McDonald and Slawson 2002; Standifird rankings 2001

Consumer behaviour. Consumer behaviour is the aspect researchers have applied adoption theories, including that we want to highlight in this literature review. We have TRA and TAM, to evaluate the practical utility of online identified a large body of research related to consumer auctions, and their findings have validated the applic- behaviour, which suggests the importance and wide ability of these theories to explaining and predicting the acceptance of this research topic in online auction subsequent adoption of the technology by bidders research. Due to the wide range of research issues explored (Bosnjak et al. 2006). Other researchers, such as in these publications, we have classified them into four Stafford and Stern (2002), have integrated a set of groups – adoption, entry decision, bidding behaviour and different theories, including TAM, affinity theory and bidding strategy. involvement theory, to predict consumer inclinations to use auction sites. Their empirical results, too, have Online auction adoption. Researchers have already proved the theoretical validity and applicability of the identified the general reasons for the adoption of online integrated model in their clear and consistent explana- auctions. Lower transaction costs, lower purchasing tion of this adoption phenomenon. In addition to the prices and the inclusion of a wider pool of suppliers antecedents of online auction adoption, the barriers to are some of the primary factors that support the its adoption have also been investigated. These findings adoption of these auctions (Carter et al. 2004). Some suggest that inadequate auction knowledge and Electronic Markets Vol. 18 No 4 353

Table 5. Bidders’ factors affecting bidding behaviour

Sub-stream Factors Research findings Reference

Bidders’ attributes Reputation Less effect than the seller’s reputation MacInnes et al. 2005 Experience No effect on final bid price Ottaway et al. 2003 Causes more strategic bidding Borle et al. 2006; Wilcox 2000 Positively related to auction success Gilkeson and Reynolds 2003 Emotion Excitement of winning and frustration Ding et al. 2005 with loss Social influence Network effect Others’ bids stimulate bidding, even when Dholakia and Soltysinski 2001; Kauffman and shilling occurs Wood 2006; Kauffman and Wood 2005; Ku et al. 2005; Stern and Stafford 2006 Social facilitation Affects the price/market outcomes Rafaeli and Noy 2005

concerns over information security are two of the major technology adoption (Kwon et al. 2002) or bidding barriers to online auctions for potential buyers (Hartle et behaviour (Kauffman and Wood 2006). al. 2006). Researchers have not only used the general technol- Bidding behaviour. In this paper, bidding behaviour ogy adoption models to explore the adoption of online denotes general, aggregated behaviour across all phases auctions, but have also investigated specific auction of the auction, rather than behaviour that is confined to technology services, such as payment choice and escrow a specific phase of the auction (Ariely and Simonson services, and their subsequent adoption and use. For 2003). In this line of research, researchers (Ariely and example, Zhang and Li (2006) studied payment choice Simonson 2003, Kamins et al. 2004, Kauffman and in online auctions by investigating product attributes, Wood 2006, Pavlou and Gefen 2004) have used both trader characteristics and payment attributes. Their the number of bids and the final bid prices as proxies to results indicated that product attributes have a stronger measure general bidding behaviour, mostly because of influence on payment choice than do trader character- their simplicity and convenience. With few exceptions, istics, while a seller’s reputation has no significant the research on online auctions has presumed bidder influence on the actual payment choice. In contrast, rationality. However, the empirical findings on the Antony et al. (2006) investigated the determinants of online behaviour of bidders have suggested that many escrow service adoption and showed that, in an of them bid in excess of their pre-set limits, which is experimental setting, such market factors as the fraud called ‘bidding fever’. A ‘winner’s curse’ has also been rate and the product price are important determinants of identified in the online setting. This is the tendency for the adoption of escrow services. The seller’s reputation the winner to incur a loss because of his or her inaccurate also had an indirect effect through the buyer’s risk assessment of an object’s resale value (Ku et al. 2005). Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 perception of the adoption of escrow services. Although both of these phenomena cause bidders to bid in excess of their limits, auction fever is primarily emotional in nature, while the winner’s curse stems Entry decision. The first decision a consumer needs to largely from uncertainty over the value of an object. make before an auction is whether to participate. This Theoretical validation of the rational choice, escalation decision could be influenced by his or her interest in the of commitment and competitive arousal models have auction item or in the auction site and most probably also demonstrated that irrationality plays a critical role in involves a value assessment (Ariely and Simonson 2003). bidding behaviour (Ku et al. 2005). An entry decision may be made at different stages of the In addition to irrational bidding behaviour, research- auction – at the beginning, in the middle or near the end ers have also turned their attention to such fraudulent – depending on the arrival time of the bidders. However, bidding conduct as shill bidding – the seller bidding on the entry decision is intrinsic and difficult to observe, his or her own auctioned item using secondary registra- because bidders do not have to act immediately even tions or associates to artificially drive up the bid price after they have decided to enter an auction; therefore, (Namazi and Schadschneider 2006). Kauffman et al. some researchers have used the bidder’s first bid or (2005), for example, believe that fake bids submitted by subsequent bids during the auction as the convenient sellers or their associates facilitate the herd effect. They proxy for an entry decision (Bajari and Hortacsu 2003). explain this from the valuation signal perspective, that is, There is very little research that focuses exclusively on bidders may view a large number of bids as a signal that the entry decision. Most of the studies that address it are an item is worth more than what they were originally combined with research on either online auction willing to pay for it. Dholakia and Simonson (2005) 354 Xiling Cui, Vincent S. Lai and Connie K.W. Liu & Consumer Behaviour in Online Auctions: A Review

have suggested that explicit reference points can be used and it is an attractive alternative for participants who by online auction participants to help them avoid falling value their time greatly (Bapna et al. 2004). However, into the shill bidding trap caused by herd behaviour bias agent bidding is still not very popular, perhaps because by encouraging them to be more cautious and risk- bidders do not fully understand the process or because averse and to avoid a bidding frenzy. In the presence of they like to retain the flexibility to revise their reservation explicit reference points, bidders may make less risky price at all times (Easley and Tenorio 2004). choices and adopt more controlled bidding strategies, Opportunists frequently adopt a sniping or snipe bidding thus leading to a lower probability of overpaying and a strategy (Borle et al. 2006, Roth and Ockenfels 2002). greater incidence of sniping. This means that they watch a timed auction and then place a winning bid just before it ends to prevent other Bidding strategies. A bidding strategy represents a series bidders from outbidding them or driving the price of interrelated types of bidding behaviour that a bidder higher (Namazi and Schadschneider 2006). Sip-and- adopts with a certain purpose or tactic in mind across dippers, who practice a special variant of sniping, place different phases of an online auction (Ariely and Simonson one early bid to establish their time priority and then 2003). Park and Bradlow (2005) have developed an snipe during the auction’s closing stages to compete at integrated model that incorporates four key components the margin. This strategy is normally adopted in special of the bidding process – whether people win an auction, multi-unit auctions (auctions that include multiple items (and if so) who bids, when they bid and how much they of the same product) that require a time priority. bid. Bapna et al. (2001b, 2003b, 2000, 2004), in In addition to the five aforementioned bidding contrast, classify bidders into five types based on the strategies, Lucking-Reiley (2000) also reported a bid number of bids and the entry and exit times: evaluators shielding strategy in which one (or more) bidder(s) (early or middle), participators, agent bidders, opportu- aim(s) to deceive other bidders. In this strategy, the nists and sip-and-dippers. Evaluators refer to bidders who bidder offers a low bid on an item and then follows it up place just one bid at an early stage (or sometimes in the with another bid – using a false identity – that is high middle stages) of an auction; participators are bidders who enough to keep rivals from entering the auction. Then, bid throughout the auction, thus increasing the price by just before the end of the auction, the bidder retracts the the minimum bid increment; agent bidders use online high bid that was made using the false identity, and the bidding agents to bid at the minimum level required to low bid wins. outbid the current highest bidder until the bid exceeds the reserve price they set earlier; opportunists arelatebidders Auction outcome. Many online auction studies adopt who act only towards the end of an auction; and sip-and- auction outcome as a dependent variable in their dippers are a combination of evaluators and opportunists. empirical or theoretical enquiries, using either an These five bidder types tend to apply unique strategies economic or non-economic perspective. The economic in the bidding process. Evaluators, for example, employ perspective normally explores the final auction price, the a strategy (Easley and Tenorio 2004) to auctioneer’s revenue, the winner’s curse and consumer submit bids that are higher than required and usually surplus, while the non-economic perspective investigates only once. This strategy can minimize the time and cost auction success and the likelihood of winning. Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 of monitoring (Bapna et al. 2004) and is effective in auctions with fewer overall bids (Avery 1998, Bapna et Economic perspective. Economic auction outcomes can al. 2004, Easley and Tenorio 2004). Research has also be further classified by whether they take the seller’s or found that experienced bidders tend to use a jump the buyer’s perspective. Researchers have explored a bidding strategy in the early stages of an auction (Borle wide spectrum of determinants of auction outcomes et al. 2006). Participators, in contrast, normally adopt a from the seller’s perspective, ranging from the auction ratchet bidding strategy (sometimes called participatory setting and participant attributes to bidding behaviour. bidding, successive bidding or pedestrian bidding) to By and large, these research findings demonstrate that bid by the minimum bid increment (Brint 2003). starting price, the number of bids, auction duration and Although ratchet bidding has seldom been specifically the seller’s reputation may have effects on the final price focused on in studies of online auctions, Borle et al. (Ariely and Simonson 2003, Gilkeson and Reynolds (2006) consider it to be of strategic importance in online 2003). The auctioneer’s revenue, another economic auctions. Agent bidders prefer to use an agent bidding outcome that concerns sellers, has been found to be strategy that takes advantage of the automatic bidding positively related to the insurance premiums paid (Ward software provided by auction intermediaries to place and Clark 2002), the shipping price paid, the auction their bids. On most online auction sites, this software length, the number of bids (Onur and Tomak 2006) and allows bidders to set a reserve price and then follow the the bid increment (Bapna et al. 2001b, 2003a, 2003b, minimum bid increments, when necessary, in the course Onur and Tomak 2006). It has also been found to be of the bidding process (Bapna et al. 2004). This strategy marginally affected by the auction’s opening bid (Bapna can save the time and cost of tracing and monitoring, et al. 2001b, 2003a, 2003b). In research that considers Electronic Markets Vol. 18 No 4 355

the two ending rules, the hard-close and the soft-close, on consumer behaviour in online auctions. The varieties hard-close auctions have been found to yield lower of behaviour displayed, including technology adoption, revenue, especially when the item value is low (Onur and entry decision, bidding behaviour and bidding strategy, Tomak 2006). in turn, can also be used to explain some of the studies From the buyer’s perspective, the common phenom- that have focused on auction outcomes. In other words, enon of the winner’s curse in traditional auctions also our literature review suggests that the current research exists in online auctions (Amyx and Luehlfing 2006, on consumer behaviour in online auctions has already Bajari and Hortacsu 2003). Research suggests that the formed itself into a structure that can be used to guide occurrence of the winner’s curse is positively associated future investigations. Although a cumulative tradition with retail price dispersion and auction type, but not has not yet emerged, the existing body of research with product value (Oh 2002). In multi-unit auctions, provides a foundation to guide and enhance future researchers identified the existence of consumer surplus research endeavours on online auctions. (Bapna et al. 2001b, 2004), which is the excess price a consumer would be willing to pay for a product rather than go without it, and it is operationalized as the FUTURE RESEARCH DIRECTIONS difference between the highest bid (price which a bidder would be willing to pay) and the winning bid (the price Although our review of the online auction literature has they actually paid). In the studies of Bapna et al. (2001b, covered a vast area and has sometimes cut across 2004), different bidding strategies were also found to disciplines, the work in this area is still being developed, result in different economic outcomes. Evaluators fare and plenty of research opportunities remain to be worst, participators fare best, and opportunists lie explored. By examining and analysing previous studies, somewhere in between in terms of consumer surplus. we have been able to identify some gaps in the Agent bidders, who appear later than other types of knowledge, as depicted in Figure 2, including potential bidder, have been found to be the best at maximizing new constructs, unexplored relationships between con- surplus and are even better than participators. structs and new measurements and definitions of these constructs. Moreover, theoretical and conceptual devel- Non-economic perspective. The non-economic perspec- opments are crucial for this relatively new area of work, tive investigates auction outcomes using auction success and a resolution of the sometimes conflicting findings as an indicator. Success means that the auction results in will contribute substantially to the accumulation of a deal – that is, it draws at least one bid and that the final knowledge. This section provides a detailed discussion of price is higher than the reserve price, if any (Lucking- some potential future directions for research on con- Reiley 2000). Gilkeson et al. (2003) studied the sumer behaviour in online auctions, based on the three determinants of auction success, and their results show main categories. that both the starting price (due to the anchoring effect) and the reserve price of the seller have negative effects on Knowledge gaps related to facilitating factors. The auction success, while the number of unexpected bids services that are unique to the online environment and bidder experience have a positive impact. require further exploration. For example, escrow Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 The likelihood of winning is a non-economic outcome services, which are also called TTP services, are usually variable that has been investigated from the bidder provided by intermediaries to enhance the building of perspective. Bapna et al. (2004) indicated that oppor- trust and transactional equity between sellers and tunists and sip-and-dippers are generally more eager to buyers, and they are believed to have a significant win, and thus have a higher likelihood of winning, while effect on consumer behaviour in online auctions. others tend to be more cautious. However,onlyafewpapershaveaddressedtheimpact of these services on bidding behaviour (Pavlou and Summary of research findings. This extensive literature Gefen 2004), the research findings, of which, remain review of consumer behaviour in online auctions to be confirmed by separate investigations in order to suggests that the prior research endeavours in this field make them generalizable to different areas of applica- can be categorized into three sub-streams: facilitating tion. Because trust is an important issue in the online factors; consumer behaviour; and auction outcome. A environment, in which there is no face-to-face interac- detailed analysis of these three sub-streams reveals that tion and fraud is commonplace, it is worth devoting they have certain intended or unintended relationships more resources to considering ways to enhance the that form themselves into an integrated framework, as level of trust between participants and to investigating depicted in Figure 1. The facilitating factors, which are the strength and fragility of trust. explored from the intermediary, seller and bidder From the perspective of sellers, product type could well perspectives, explain the effects of technology properties, affect the bidding behaviour of consumers. People service facilities, auction mechanisms, auction setting, participate in auctions for different reasons: some believe sellers’ attributes, bidders’ attributes and social influence they can save money, while others look for rare items 356 Xiling Cui, Vincent S. Lai and Connie K.W. Liu & Consumer Behaviour in Online Auctions: A Review

Figure 1. Research framework for online bidding behaviour

among the vast range of goods offered online. Therefore, of consumer behaviour, thus giving researchers another it is not difficult to imagine that consumers may employ direction to explore. different strategies when they acquire different types of From the perspective of bidders, product value can also products, such as commodities (where saving money is the affect auction participant behaviour. Generally speaking, sole purpose of bidding) or collectables (where obtaining a when they bid on an expensive item online, people tend rare item is the only concern). An investigation of these to be more cautious and perhaps less prone to the herd different product types and the bidding behaviour effect, whereas they may be more impulsive when it associated with each of them would thus be an important comes to bidding on a less expensive item. However, not step in synthesizing consumer behaviour in different many auction studies have addressed this issue. contexts and scenarios. In addition, new products and Finally, because online auctions attract people from all second-hand products may also engender different types over the world, it is very likely that cultural factors or Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010

Figure 2. Possible research opportunities. Note: Italics: suggested research areas. Electronic Markets Vol. 18 No 4 357

differences play a part in shaping the online behaviour of Another research gap in the consumer behaviour consumers. It is possible that certain facilitating factors context is the possible antecedents of bidding strategy. may be regarded as quality factors in the East, while they While many researchers have explored this topic, most of may be regarded as non-essential in the West (for them have focused solely on the patterns of bidding example, the provision of a socializing platform within strategies or combined such analysis with the effects of the auction site). Research in this area would not only these strategies on the auction outcome. Unfortunately, enrich our knowledge of consumer behaviour in online the existing research on the factors that may affect the auctions, but would also extend the scope of cultural adoption of different bidding strategies is much weaker. research. It would also be interesting to examine how Some researchers have discussed possible factors, but people from collectivist cultures respond to online none so far has conducted empirical work to validate auctions in comparison to offline auctions, as the online their claims. Moreover, bidding strategies that aim to environment is more individualistic in nature. defraud, such as shill bidding and bid shielding, need to be explored in more depth to prevent or reduce their Mediating and moderating effects on consumer occurrence. behaviour. This line of research is still relatively new; therefore, the discovery of new constructs is essential to The measurement of auction outcome. The variety of its growth, both in depth and in breadth. Although constructs for auction outcome needs to be broadened. research has addressed the direct effects on consumer Previous research has used different measures and behaviour, no paper has yet explored the possibility that definitions for auction outcome. For example, some different types of relationships, such as mediating and studies have used auctioneer revenue to gauge the moderating effects, may exist. It is possible that a new success of the outcome, while others have tried to construct, such as motivation, may mediate the relation- measure this construct by looking at the benefit to the ship between certain facilitating factors and consumer consumer. In addition to using the dividing line of behaviour, because external causes do sometimes have economic and non-economic outcomes, online auction the ability to affect attitudes towards certain issues and, research can be investigated in an alternative way by in turn, alter the behaviour that is adopted. Similarly, using three levels of outcomes: intermediary-, transac- such personality traits as risk preference may have a tion- and participant-level outcomes. Intermediary-level moderating effect on the relationship between certain outcomes refer to outcomes that are directly related to facilitating factors and consumer behaviour. In fact, the intermediary or the auctioneer. Transaction-level certain bidder attributes that are currently thought to outcomes are related to the transaction process of one have a direct effect on consumer behaviour may more auction, including its final price and the total number of accurately be regarded as moderating factors. This is bids in it. Participant-level outcomes are the outcomes because certain attributes, such as experience, which from the bidder’s perspective, such as customer gain/ clearly has some kind of impact on consumer behaviour, loss, customer satisfaction and entertainment fulfilment. have resulted in conflicting findings in the literature. This three-level classification system for outcomes, Therefore, the impact of such attributes may or may not combined with classification based on the economic be direct, and thus they warrant further investigation. versus the non-economic perspective, forms a two- Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 dimensional 263 matrix framework (Table 6) that may Research opportunities related to consumer be used in future auction outcome research. The behaviour. New research areas within the consumer differentiation of these research alternatives may lead behaviour context include the establishment of a clear to the compilation of a more organized and compre- definition of the ‘entry decision’. This is essential before hensive list of variables and constructs to represent and future research can be built upon it. As previously measure auction outcome. In fact, the auction outcome discussed, the entry decision may not necessarily equate research topics that have been studied to date could be to a bidder’s first bid, although the alternative is difficult classified into this framework to help identify those to measure and can be problematic when it comes to topics that have already been researched and those that verifying the results. Nevertheless, the early decision- require additional study. As indicated in Table 6, the making process, which ultimately determines a bidder’s non-economic perspective has received less attention subsequent behaviour, requires more attention from than has the economic perspective. Some of the topics researchers. Attention to this issue may even open the that could be considered for future research include door to a whole new area of research, including intermediary/auctioneer satisfaction, bidder satisfaction conceptual studies, the development of models and the and the bidder’s entertainment fulfilment. measurement of creation and verification. This new research area would require extensive background Insights into research methodology. In addition to the knowledge and the ability to utilize theories and knowledge gaps that exist in each part of this integrated resources from a wide range of disciplines, including framework, there are also methodological imbalances psychology, sociology, IS, marketing and management. that could easily be improved. First, the research 358 Xiling Cui, Vincent S. Lai and Connie K.W. Liu & Consumer Behaviour in Online Auctions: A Review

methodologies used could be extended in scope. To Table 6. Auction outcome classifications date, most studies have been conducted by analysing real data from auction sites. Such objective data, although Economic Non-economic they are highly valued in empirical research, also have certain limitations. For example, such personal attributes Intermediary-level Revenue of the Satisfaction of the as intrinsic thought cannot be reflected in observed data. intermediary/ intermediary/auctioneer* Moreover, most of the observed data provided by online auctioneer auction houses have been collected as individual auction Transaction-level Final auction The number of bids Auction units of data, meaning that, although these data can price success clearly describe the patterns of each auction, they cannot Participant-level Winner’s curse Winning likelihood Customer track the overall behaviour of each individual. These Consumer satisfaction* Entertainment limitations could easily be solved through the use of surplus fulfilment* surveys, a tool that has been widely used to capture individual-level variables and constructs in other research Italics*: suggested research areas areas, to make the research on online bidding behaviour more comprehensive and convincing. on the verbal thought processes and objective responses of Another important area of research is the establishment participants. of theoretical models. It is clear from the literature review carried out here that data-driven research is still the main research format in online auction studies. Only a few CONCLUSION papers have developed a theoretical framework or model first and then followed it up with confirmatory tests. This paper attempts to make a contribution to the Clearly, data-driven studies are important for any research research on consumer behaviour in online auctions. It topic, particularly when it is in its infancy. However, reviews the extended literature on online bidding additional confirmatory studies are needed once a research behaviour and classifies the studies into three different topic has grown and matured. Widely-accepted, well- categories. Within each category, these studies are tested theories from more mature fields, such as market- summarized and compared. Explanations and clarifica- ing, management, psychology or economics, can be tions are given where possible. A research framework for borrowed, applied or generalized to establish a theoretical online bidding behaviour, based on previous studies, is backbone for research in this area, thus making it more presented in the hope of guiding future research. sophisticated and credible. The theories that have been Knowledge gaps and research opportunities have been employed in online auction research to date include image derived from this framework, and suggestions for future theory (Dholakia and Soltysinski 2001), the TAM research directions are made. (Bosnjak et al. 2006, Stafford and Stern 2002), the TRA As Zahedi (2002) concluded, synthesis papers are ‘a (Bosnjak et al. 2006), affinity theory (Stafford and Stern needed scholarly undertaking’; it is unfortunate that they 2002), involvement theory (Stafford and Stern 2002), have been largely overlooked by IS researchers. We hope information cascades theory (Dholakia and Soltysinski that the review presented in this paper provides a Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 2001), signalling theory (Gilkeson and Reynolds 2003), comprehensive summary of the research on consumer prospect theory (Ward and Clark 2002), the winner’s behaviour in online auctions, that our synthesis has been curse (Amyx and Luehlfing 2006, Bajari and Hortacsu thorough and thoughtful, and that our presentation of 2003, Oh 2002), and the anchoring effect (Ariely and the issues and research gaps has been carried out in an Simonson 2003), among others. In addition to the unbiased manner, bearing in mind that the proposed separate application of these theories in online auction definitions, variables, constructs and methodologies are research, there is a need for a parsimonious theory that can by no means exhaustive. The aim of this paper is to link all of the various theories (Baker and Song 2007). provide readers with an overall picture of what has been done in this research area to date, how the various Insights into auction design. Although auction design is studies can be related to one another, and what can be not the main focus of this review, we can still draw some done to advance this research in future. research directions that may result in the maximization of the economic and non-economic outcomes for all players, especially buyers. For example, to reduce the winner’s curse in online auctions, Vickrey auctions could be used, References particularly for certain common value auctions. NYOP Amyx, D. A. and Luehlfing, M. S. (2006) ‘Winner’s Curse and auctions may be useful in counteracting anchoring effects. Parallel Sales Channels – Online Auctions Linked Within e-Tail To investigate these topics, future (laboratory) studies, Websites’, Information & Management 43(8): 919–27. such as experiments, could be used to test the effects of Antony, S., Lin, Z. and Xu, B. (2006) ‘Determinants of these modifications, both individually and in combination, Escrow Service Adoption in Consumer-to-consumer Online Electronic Markets Vol. 18 No 4 359

Auction Market: An Experimental Study’, Decision Support An Empirical Investigation Using eBay Online Auction’, Systems 42(3): 1889–900. Statistical Science 21(2): 194–205. Ariely, D., Ockenfels, A. and Roth, A. E. (2005) ‘An Bosnjak, M., Obermeier, D. and Tuten, T. L. (2006) Experimental Analysis of Ending Rules in Internet ‘Predicting and Explaining the Propensity to Bid in Online Auctions’, The Rand Journal of Economics 36(4): 890–907. Auctions: A Comparison of Two Action-theoretical Ariely, D. and Simonson, I. (2003) ‘Buying, Bidding, Playing, Models’, Journal of Consumer Behaviour 5(2): 102–16. or Competing? Value Assessment and Decision Dynamics in Brint, A. T. (2003) ‘Investigating Buyer and Seller Strategies in Online Auction’, Journal of Consumer Psychology 13(1&2): Online Auctions’, Journal of the Operational Research 113–23. Society 54(11): 1177–88. Avery, C. R. (1998) ‘Strategic Jumping Bidding in English Bruce, N., Haruvy, E. and Rao, R. (2004) ‘Seller Rating, Price, Auctions’, Review of Economic Studies 65(2): 185–210. and Default in Online Auctions’, Journal of Interactive Ba, S. and Pavlou, P. A. (2002) ‘Evidence of the Effect of Trust Marketing 18(4): 37–50. Building Technology in Electronic Markets: Price Budish, E. B. and Takeyama, L. N. (2001) ‘Buy Prices in Premiums and Buyer Behavior’, MIS Quarterly 23(3): Online Auctions: Irrationality on the Internet?’, Economics 243–68. Letters 72(3): 325–33. Ba, S., Whinston, A. B. and Zhang, H. (2003) ‘Building Trust Cameron, D. D. and Galloway, A. (2005) ‘Consumer Into Online Auction Markets through an Economic Incentive Motivations and Concerns in Online Auctions: An Mechanism’, Decision Support Systems 35(3): 273–86. Exploratory Study’, International Journal of Consumer Bajari, P. and Hortacsu, A. (2003) ‘Winner’s Curse, Reserve Studies 29(3): 181–92. Prices and Endogenous Entry: Empirical insights from eBay Carare, O. and Rothkopf, M. (2005) ‘Slow Dutch Auctions’, Auctions’, Rand Journal of Economics 34: 329–55. Management Science 51(3): 365–73. Baker, J. and Song, J. (2007) ‘A Review of Single-item Carter, C., Kaufmann, L., Beall, S., Carter, P., Hendrick, T. Internet Auction Literature and a Model of Future and Petersen, K. (2004) ‘Reverse Auctions – Grounded Research’, Journal of Electronic Commerce in Organizations Theory from the Buyer and Supplier Perspective’, 5(1): 42–68. Transportation Research 40(3): 229–54. Bapna, R. (2003) ‘When Snipers become Predators: Can Chakravarti, D., Greenleaf, E., Sinha, A., Cheema, A., Cox, J. Mechanism Design Save Online Auctions?’ C., Priedman, D., Ho, T. H., Issac, R. M., Mitchell, A. A., Communications of the ACM 46(12): 152–8. Rapoport, A., Rothkopf, M. H., Srivastava, J. and Zwick, R. Bapna, R., Goes, P. and Gupta, A. (2000) ‘A Theoretical and (2002) ‘Auctions: Research Opportunities in Marketing’, Empirical Investigation of Multi-item On-line Auctions’, Marketing Letters 13(3): 281–96. Information Technology and Management 1(1–2): 1–23. Chen, J., Chen, X. and Song, X. (2006) ‘Comparison of the Bapna, R., Goes, P. and Gupta, A. (2001a) ‘Comparative Group-buying Auction and the Fixed Pricing Mechanism’, Analysis of Multi-item Online Auctions: Evidence from the Decision Support Systems 43(2): 445–59. Laboratory’, Decision Support Systems 32(2): 135–53. Dai, Q. and Kauffman, R. (2002) ‘Business Models for Bapna, R., Goes, P. and Gupta, A. (2001b) ‘Insights and Internet-based B2B Electronic Markets’, International Analyses of Online Auctions’, Communications of the ACM Journal of Electronic Commerce 6(4): 41–72.

Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 44(11): 42–50. Dholakia, U. M. (2005) ‘The Usefulness of Bidders’ Bapna, R., Goes, P. and Gupta, A. (2003a) ‘Analysis and Reputation Ratings to Sellers in Online Auctions’, Journal Design of Business-to-Consumer Online Auctions’, of Interactive Marketing 19(1): 31–40. Management Science 49(1): 85–101. Dholakia, U. M. and Soltysinski, K. (2001) ‘Coveted or Bapna, R., Goes, P. and Gupta, A. (2003b) ‘Replicating Overlooked? The Psychology of Bidding for Comparable Online Yankee Auctions to Analyze Auctioneers’ and Listings in Digital Auctions’, Marketing Letters 12(3): Bidders’ Strategies’, Information Systems Research 14(3): 225–37. 244–68. Dholakia, U. M. and Simonson, I. (2005) ‘The Effect of Bapna, R., Goes, P., Gupta, A. and Jin, Y. (2004) ‘User Explicit Reference Points on Consumer Choice and Online Heterogeneity and its Impact on Electronic Auction Market Bidding Behaviour’, Marketing Science 24(2): 206–17. Design: An Empirical Exploration’, MIS Quarterly 28(1): Ding, M., Eliashberg, J., Huber, J. and Saini, R. (2005) 21–43. ‘Emotional Bidders – An Analytical and Experimental Bapna, R., Jank, W. and Shmueli, G. (2008) ‘Price Formation Examination of Consumers’ Behaviour in a Priceline-like and its Dynamics in Online Auctions’, Decision Support ’, Management Science 51(3): 352–64. Systems 44(3): 641–56. Dodonova, A. and Khoroshilov, Y. (2004) ‘Anchoring and Bichler, M., Field, S. and Werthner, H. (2001) ‘Introduction: Transaction Utility: Evidence from On-line Auctions’, Theory and Application of Electronic ’, Applied Economics Letters 11(5): 307–10. Electronic Commerce Research 1(3): 215–20. Easley, R. F. and Tenorio, R. (2004) ‘Jumping Bidding Borle, S., Boatwright, P. and Kadane, J. B. (2006) ‘The Strategies in Internet Auctions’, Management Science Timing of Bid Placement and Extent of Multiple Bidding: 50(10): 1407–19. 360 Xiling Cui, Vincent S. Lai and Connie K.W. Liu & Consumer Behaviour in Online Auctions: A Review

Gilkeson, J. H. and Reynolds, K. (2003) ‘Determinants of Lin, Z., Li, D., Janamanchi, B. and Huang, W. (2006) Internet Auction Success and Closing Price: An Exploratory ‘Reputation Distribution and Consumer-to-consumer Study’, Psychology and Marketing 20(6): 537–66. Online Auction Market Structure: An Exploratory Study’, Gregg, D. G. and Scott, J. E. (2006) ‘The Role of Reputation Decision Support Systems 41(2): 435–48. Systems in Reducing Online Auction Fraud’, International Lucking-Reiley, D. (1999) ‘Using Field Experiments to Test Journal of Electronic Commerce 10(3): 95–120. Equivalence Between Auction Formats: Magic on the Hann, I.-H. and Terwiesch, C. (2003) ‘Measuring the Frictional Internet’, American Economic Review 89(5): 1063–80. Costs of Online Transactions: The Case of a Name-Your- Lucking-Reiley, D. (2000) ‘Auctions on the Internet: What’s Own-Price Channel’, Management Science 49(11): 1563–79. Being Auctioned, and How?’, Journal of Industrial Hannon, D. (2004) ‘Online Buying Gathers Steam, One Buyer Economics 48(3): 227–52. at a Time’, Purchasing 130(16): 40–3. MacInnes, I., Li, Y. and Yurcik, W. (2005) ‘Reputation and Hartley, J. L., Lane, M. D. and Duplaga, E. A. (2006) Dispute in eBay Transactions’, International Journal of ‘Exploring the Barriers to the Adoption of e-Auctions for Electronic Commerce 10(1): 27–54. Sourcing’, International Journal of Operations and Mathews, T. (2003) ‘A Risk Averse Seller in a Continuous Production Management 26(2): 202–21. Time Auction with a Buyout Option’, Brazilian Electronic Herschlag, M. and Zwick, R. (2000) ‘Internet Auctions: Journal of Economics 5(2). Popular and Professional Literature Review’, Quarterly Mathews, T. (2004) ‘The Impact of Discounting on an Journal of Electronic Commerce 1(2): 161–86. Auction with a Buyout Option: A Theoretical Analysis Heyman, J. E., Orhun, Y. and Ariely, D. (2004) ‘Auction Fever: Motivated by eBay’s Buy-it-Now Feature’, Journal of The Effect of Opponents and Quasi-endowment on Product Economics 81(1): 25–52. Valuations’, Journal of Interactive Marketing 18(4): 7–21. Mathews, T. and Katzman, B. (2006) ‘The Role of Varying Houser, D. and Wooders, J. (2006) ‘Reputation in Auctions: Risk Attitudes in an Auction with a Buyout Option’, Theory, and Evidence from eBay’, Journal of Economics & Economic Theory 27(3): 597–613. Management Strategy 15(2): 353–69. McDonald, C. G. and Slawson, V. C. (2002) ‘Reputation in an Hu, X., Lin, Z., Whinston, A. B. and Zhang, H. (2004) ‘Hope Internet Auction Market’, Economic Inquiry 40(3): 633–50. or Hype: On the Viability of Escrow Services as Trusted McKinney, V., Yoon, K. and Zahedi, F. ‘‘M.’’. (2002) ‘The Third Parties in Online Auction Environments’, Measurement of Web-customer Satisfaction: An Information Systems Research 15(3): 236–49. Expectation and Disconfirmation Approach’, Information Jap, S. (2003) ‘An Exploratory Study of the Introduction of Systems Research 13(3): 296–315. Online Reverse Auctions’, Journal of the Academy of Melnik, M. I. and Alm, J. (2002) ‘Does a Seller’s Ecommerce Marketing Science 67(3): 96–108. Reputation Matter? Evidence from eBay Auctions’, The Kamins, M. A., Dreze, X. and Folkes, V. S. (2004) ‘Effects of Journal of Industrial Economics 50(3): 337–49. Seller-supplied Prices on Buyers’ Product Evaluations: Mollenberg, A. (2004) ‘Internet Auctions in Marketing: The Reference Prices in an Internet Auction Context’, Journal of Consumer Perspective’, Electronic Markets 14(4): 360–71. Consumer Research 30(4): 622–8. Namazi, A. and Schadschneider, A. S. (2006) ‘Statistical Kauffman, R. J. and Wang, B. (2001) ‘New Buyers’ Arrival Properties of Online Auctions’, International Journal of under Dynamic Pricing Market Microstructure: The Case of Modern Physics 17(10): 1485–93. Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010 Group-buying Discounts on the Internet’, Journal of Ockenfels, A. and Roth, A. E. (2002) ‘The Timing of Bids in Management Information Systems 18(2): 157–88. Internet Auctions: Market Design, Bidder Behaviour, and Kauffman, R. J. and Wood, C. A. (2005) ‘The Effects of Artificial Agents’, Artificial Intelligence Magazine 23(3): 79–87. Shilling on Final Bid Prices in Online Auctions’, Electronic Oh, W. (2002) ‘C2C Versus B2C: A Comparison of the Commerce Research and Applications 4(1): 21–34. Winner’s Curse in Two Types of Electronic Auctions’, Kauffman, R. J. and Wood, C. A. (2006) ‘Doing their Bidding: International Journal of Electronic Commerce 6(4): 115–38. An Empirical Examination of Factors that Affect a Buyer’s Onur, I. and Tomak, K. (2006) ‘Impact of Ending Rules in Utility in Internet Auctions’, Information Technology and Online Auctions: The Case of Yahoo.com’, Decision Support Management 7(3): 171–90. Systems 42(3): 1835–42. Ku, G., Malhotra, D. and Murnighan, J. K. (2005) ‘Towards a Ottaway, T. A., Bruneau, C. L. and Evans, G. E. (2003) ‘The Competitive Arousal Model of Decision-making: A Study of Impact of Auction Item Image and Buyer/seller Feedback Auction Fever in Live and Internet Auctions’, Organizational Rating on Electronic Auctions’, The Journal of Computer Behaviour and Human Decision Processes 96(2): 89–103. Information Systems 43(3): 56–60. Kwon, O., Byung, K., Choong-Ryuhn and Lee, E. J. (2002) Park, Y.-H. and Bradlow, E. T. (2005) ‘An Integrated Model ‘Impact of Website Information Design Factors on for Bidding Behaviour in Internet Auctions: Whether, Who, Consumer Ratings of Web-based Auction Sites’, Behaviour When, and How Much’, Journal of Marketing Research and Information Technology 21(6): 387–402. 42(4): 470–82. Leloup, B. and Deveaux, L. (2001) ‘Dynamic Pricing on the Pavlou, P. A. and Gefen, D. (2004) ‘Building Effective Online Internet: Theory and Simulations’, Electronic Commerce Marketplaces with Institution-based Trust’, Information Research 1(3): 265–76. Systems Research 15(1): 37–59. Electronic Markets Vol. 18 No 4 361

Pinker, E. J., Seidmann, A. and Vakrat, Y. (2003) ‘Managing Behaviour’, International Journal of Electronic Commerce Online Auctions: Current Business and Research Issues’, 9(2): 167–76. Management Science 49(11): 1457–84. Stern, B. B. and Stafford, M. R. (2006) ‘Individual and Social Rafaeli, S. and Noy, A. (2002) ‘Online Auctions, Messaging, Determinants of Winning Bids in Online Auctions’, Journal Communication and Social Facilitation: A Simulation and of Consumer Behaviour 5(1): 43–55. Experimental Evidence’, European Journal of Information Terwiesch, C., Savin, S. and Hann, I.-H. (2005) ‘Online Systems 11(3): 196–207. Haggling at a Name-Your-Own-Price Retailer: Theory and Rafaeli,S.andNoy,A.(2005)‘SocialPresence:InfluenceonBidders Application’, Management Science 51(3): 339–51. in Internet Auctions’, Electronic Markets 15(2): 158–75. Townsend, A. M. and Bennett, J. T. (2003) ‘Living and Rainer, K. and Miller, M. (2005) ‘Examining Differences Across Bidding in an Auction Economy’, Communications of the Journal Rankings’, Communications of the ACM 48(2): 91–4. ACM 46(12): 351–3. Roth, A. E. and Ockenfels, A. (2002) ‘Last-minute Bidding Tversky, A. and Kahneman, D. (1974) ‘Judgment under Uncert- and the Rules for Ending Second-price Auctions: Evidence ainty: Heuristics and Biases’, Science 185(4157): 1124–31. from eBay and Amazon Auctions on the Internet’, Wang, K., Wang, E. T. G. and Tai, C.-F. (2002) ‘A Study of American Economic Review 92(4): 1093–103. Online Auction Sites in Taiwan: Product, Auction Rule, and Spann, M. and Tellis, G. J. (2006) ‘Does the Internet Promote Trading Type’, International Journal of Information Better Consumer Decisions? The Case of Name-Your- Management 22(2): 127–42. Own-Price Auctions’, Journal of Marketing 70(1): 65–78. Ward, S. G. and Clark, J. M. (2002) ‘Bidding Behaviour in On-line Auctions: An Examination of the eBay Pokemon Spann, M., Skiera, B. and Schafers, B. (2004) ‘Measuring Card Market’, International Journal of Electronic Commerce Individual Frictional Costs and Willingness-to-Pay via 6(4): 139–55. Name-Your-Own-Price Mechanisms’, Journal of Interactive Wilcox, R. T. (2000) ‘Experts and Amateurs: The Role of Marketing 18(4): 22–36. Experience in Internet Auctions’, Marketing Letters 11(4): Stafford, M. R. and Stern, B. (2002) ‘Consumer Bidding 363–74. Behaviour on Internet Auction Sites’, International Journal Yang, K. C. C. (2005) ‘Consumers’ Attitudes Toward of Electronic Commerce 7(1): 135–50. Regulation of Internet Auction Sites: A Third-person Effect Standifird, S. S. (2001) ‘Reputation and e-Commerce: eBay Perspective’, Internet Research 15(4): 359–77. Auction and the Asymmetrical Impact of Positive and Zhang, J. (2006) ‘The Roles of Players and Reputation: Negative Ratings’, Journal of Management 27(3): 279–95. Evidence from eBay Online Auctions’, Decision Support Standifird, S. S. (2002) ‘Online Auctions and the Importance Systems 42(3): 1800–18. of Reputation Type’, Electronic Markets 12(1): 58–62. Zhang, H. and Li, H. (2006) ‘Factors Affecting Payment Standifird, S. S., Roelofs, M. R. and Durham, Y. (2004–5) Choices in Online Auctions: A Study of eBay Traders’, ‘The Impact of eBay’s Buy–It–Now Function on Bidder Decision Support Systems 42(2): 1076–88. Downloaded By: [Schmelich, Volker] At: 14:54 24 March 2010