Promises, promises: how consumers respond to warranties in Internet retailing.
In recent years, Internet retailing has generated an increased level of attention. A large portion of the existing marketing literature focuses on the potential impact of the Internet and its implications on retailing (e.g., Mathwick, Malholtra, and Rigdon 2001; Stewart and Zhao 2000). Some research involving online shopping suggests that consumers' perceptions of risk play a major role in determining patronage decisions: Nonstore shopping is perceived to be more risky than in-store shopping (Van den Poel and Leunis 1999). This is because shoppers lack the opportunity to physically examine or test the products and they fear not getting what they want (Mitchell 1999). The general deduction derived from these studies is that consumers are likely to perceive higher risks in online than in in-store shopping.
From extant marketing research, we know that extrinsic cues influence consumers' perceptions of product quality (Purohit and Srivastava 2001). For example, warranties as an extrinsic cue in the form of statements about product quality or performance are an important part of a firm's marketing strategy (Barsky 1995), based on the assumption that consumers will perceive a product to be of higher quality when such external cue statements are present versus when they are not. However, we also recognize that some researchers have found warranties to be less effective as a market signal for certain market/product situations (Gerner and Bryant 1981). Legally, such representations made by sellers with respect to the character or quality of an article are known as express warranties (Dowling 1985). These legally enforceable warranties facilitate marketing by reducing the risk of purchasing a product that turns out not to meet reasonable or promised expectations (Hamilton and Petty 2001). Most of the studies on extrinsic cues have been almost exclusively devoted to their individual effects on perceived product quality and value in the context of in-store retailing. Yet, some researchers have pointed out that various extrinsic cues interact with one another (Erevelles, Roy, and Vargo 1999; Purohit and Srivastava 2001).
Understanding these relationships and their implications is crucial to both Internet marketers and consumers. Internet retailers can use such knowledge to decide which extrinsic cue to use to signal quality or to reduce perceived risk on the part of consumers. From a consumer affairs standpoint, a better understanding of how Internet retailers are signaling quality can enhance the efforts to disseminate information to Internet shoppers. Consumers can then use the information to be more informed when evaluating purchases in an environment where they typically will be incurring greater risk by not having an opportunity to examine the merchandise before purchasing.
The primary objective of our research is to determine how warranties influence perceptions of risk associated with online purchasing. Specifically, our interest is in risk reduction, i.e., whether the presence of a Web site warranty influences consumer perception to the extent that the risk associated with online shopping is reduced. However, in addition to risk reduction, we also are interested in ancillary variables that are affected by warranties and examine how perceived product quality and purchase intentions can be increased by the use of warranties as an extrinsic cue.
While warranties represent the focal extrinsic cue, we further address two other extrinsic cues that are central to risk reduction for consumers, specifically, brand name and retailer reputation. (Although we differentiate between implied and express warranties in our general discussion of warranties, we shall use the term warranty to refer to express warranty information throughout this paper.) In our analysis of the extrinsic cues' influence on consumer responses, we particularly focus on interaction effects, i.e., how warranties interact with retailer reputation and with brand name in the context of Internet marketing.
Perceived Risk of Internet Retailing
In recent years, electronic commerce has experienced rapid growth (Forrester Research 2001; Market Wire 2005). Internet retailing, a large component of electronic commerce, owes its popularity to certain consumer benefits not available through other non-store shopping modes. These benefits arise primarily from the structural characteristics of the medium, such as its interactivity, and its 24-hours-a-day accessibility (Alba et al. 1997; Levy and Weitz 2001).
The Internet environment is also unpredictable. At times, online retailing is seen as the wave of future marketing, while at other times Internet retailers have experienced disappointing results. Despite its rapid growth and the apparent overall attractiveness of the industry, actual industry performance has caused some analysts to question its growth potential (Sauer and Burton 1999). For example, critics doubted that electronic commerce would ever contribute even 10%-15% of total retailing revenue (Hof 2001). Many online companies (e.g., eToys) were successful initially but eventually failed to survive, thus validating the views of some of the industry naysayers (Riera 2002). However, despite those early "caution" signals of industry potential, more positive signals have emerged in this dynamic environment, e.g., Amazon today is now profitable and E-bay is a phenomenon. According to an eMarketer report, E-Commerce in the U.S.: Retail Trends, reported on in Market Wire (2005), U.S. e-commerce revenues jumped more than 25% last year, and annual double-digit range gains are expected to continue through 2008. According to the report, it is estimated that U.S. consumers will spend $84.5 billion online in 2005 alone for retail goods and services and that number will grow to $139 billion by 2008, numbers which are expected to significantly outpace total retail spending over the next four years. The report also notes that based on a 2003 survey, 79% of all online retailers were profitable, up from 70% in 2002.
Many factors contribute to the ebb and flow of Internet retailing and its growth potential. One factor that has dampened online consumer enthusiasm is the privacy issue, i.e., consumers are less likely to buy over the Internet if they are concerned about the privacy of their personal information (Miyazaki and Fernandez 2001). However, there are other important concerns that extend beyond just privacy. For example, other primary reasons for not shopping online that go beyond privacy issues include the inability to touch merchandise, the uncertain credibility of the Internet retailer, and one reason of particular interest to the present study, the aversion 2to risk and perceived risk (Balabanis and Vassileiou 1999).
In the sections that follow, we provide a brief conceptual discussion of three primary dependent variables of interest, namely, perceived risk, product quality, and purchase intentions. Then we discuss conceptually the extrinsic cues impacting these variables that act as risk relievers in the Internet retailing context, namely, warranties as the primary risk reliever and retailer reputation and brand name as associated risk relievers.
Perceived Risk, Product Quality, and Purchase Intentions
Perceived risk is commonly defined as a two-dimensional construct comprising the uncertainty involved in a purchase decision and the consequences of taking an unfavorable action (Bettman 1973; Cox and Rich 1967; Dowling and Staelin 1994; Mitchell 1999). Another stream of extant marketing literature also has shown that perceived risk is made up of six components, namely financial, product performance, psychological, physical, social, and time risks (e.g., Peter and Tarpey 1975). Our study takes the perspective of the former stream in the context of nonstore shopping (e.g., shopping through catalogs, direct mail, telephone, and door-to-door salespeople), which is more risky than in-store shopping (Akaah and Korgaonkar 1988; Van den Poel and Leunis 1999). Consumers who shop in-store have the opportunity to reduce uncertainty by personally inspecting or testing the merchandise; by comparing different sizes, colors, or styles; and by referring to a salesperson (Dowling and Staelin 1994).
Purohit and Srivastava (2001) studied the effect of multiple cues on judgements of product quality in physical retailing. Based on their cue diagnosticity framework, they found an interactive effect of manufacturer reputation, retailer reputation, and warranty information. In particular, evaluations of product quality did not change when warranties were offered for unknown brands sold via retailers with weak reputations. However, when strong (or weak) brand names were sold via retailers with strong reputations, the presence of warranty information improved perceptions of product quality. In our study, we take the research further by exploring the potential interactive effects of brand name, Web site reputation, and Web site warranty information in an online shopping context with respect to perceived risk, product quality, and purchase intention.
While actual purchase is the best outcome measure of the use of extrinsic cues in online shopping, purchase intention is a predictor of subsequent purchase and is what we focused on in this study as the online shopping extrinsic cue (Grewal et al. 1998). When consumers make buying decisions, they adopt different risk reduction strategies, which impact their actual buying behavior (Dowling and Staelin 1994). Thus, consumers are less likely to shop on the Internet if they are unable to adopt an effective means of reducing perceived risk.
So far, we have reviewed perceived risk, product quality, and purchase intention as the variables of interest that are impacted by extrinsic cues to relieve risk. In the next section, we discuss the extrinsic cues themselves and their role as risk relievers.
Extrinsic Cues as Risk Relievers
Dowling and Staelin (1994) predicted that consumers engage in risk-relieving activities in order to reduce their level of perceived risk and hence their feelings of discomfort. These risk relievers (also known as extrinsic cues) include seeking information from formal and informal sources (e.g., obtaining information from a friend who has used the product), limiting the set of alternatives to well-known brands, spending with retailers with good reputations, trying the product prior to purchase, and reducing the amount that could be lost (e.g., by setting a price limit on purchases/by insisting on warranties) (Akaah and Korgaonkar 1988; Dowling and Staelin 1994; Van den Poel and Leunis 1999).
Nonstore retailers such as direct marketers also have promoted risk relievers such as money-back guarantees, warranties, free samples, and pre-purchase trials (Akaah and Korgaonkar 1988; Levy and Weitz 2001). An analysis of existing research yields a ranking of the various types of risk relievers as shown in Table 1. As indicated, of the top six risk relievers, "warranty/money-back guarantee" is ranked first, with "brand/manufacturer reputation" and "retailer reputation" ranked second and third, respectively. Based on that research, we propose that warranty information provided by the Web site should be the focal extrinsic cue, in addition to examining brand name and retailer reputation as other extrinsic cues that consumers use to reduce their perceived risk in online shopping. Below, we provide a more detailed discussion of warranties, brand name, and retailer reputation.
Warranties: Definitions, Regulation, and Role as Extrinsic Cue
The Uniform Commercial Code recognizes two kinds of warranties: (1) implied and (2) express warranties. An implied warranty consists of (1) a warranty of merchantability and (2) a warranty of fitness/purpose. A warranty of merchantability requires that a product be fit for the purposes for which it is ordinarily used. A warranty of fitness/purpose, on the other hand, places the onus on the seller to recommend a product that can fulfill the needs a customer has laid out. The Uniform Computer Information Transactions Act further expands implied warranties to include (1) a warranty of noninfringement and (2) a warranty of quiet enjoyment. The former requires the licensor to ensure that the product licensed is not in violation of any intellectual property laws. The latter requires that the licensor not interfere with the licensee's use of the product.
Express warranties, on which we focus in this research, are created "by any affirmation of fact, description of goods, sample or model from the seller that is made part of the basis of the bargain" (Hamilton and Petty 2001). This means that if a buyer made a purchase decision after reading a description of what the product purports to be able to do, there is an express warranty that the product must conform to this description. An exception to this is the puffery that is made in advertising; the seller's own opinion of their product does not constitute an express warranty. Express warranties can take many forms, from purely a description of what the product purports to be able to do to a full warranty, also known as a guarantee. Full warranties are the voluntary programs of repair, refund, or replacement that are attached to a product upon its sale.
Most countries have laws to guide how businesses can use warranties. For instance, in the United States, warranties are enforced by the Uniform Commercial Code Warranty Provisions ([subsection] 2-312 to 2-318), and the Federal Magnusson-Moss Warranty Act (15 U.S.C. [subsection] 2301-2312). The former covers all transactions involving goods, while the latter is limited to just consumer goods (Hamilton and Petty 2001). Warranties also extend to a limited number of people beyond the initial purchaser of a product. The Uniform Commercial Code has recently been further extended by the Uniform Computer Information Transactions Act, which deals with transactions involving information and computer software. Uniform Computer Information Transactions Act makes use of a licensing model rather than a goods transaction model that is better suited for the rapidly changing information technology environment.
Researchers have examined if extrinsic cues used in other shopping contexts apply in online shopping situations. Although these studies have investigated the relative importance of warranties and other risk relievers, they have not addressed the combined effect of these various cues on perceived risk in a different shopping context, which is online. For instance, while in the context of traditional retailing, Dowling found that a warranty has no effect on perceived risk (Dowling 1985), Van den Poel and Leunis (1999) found that warranties are the most important risk relievers for online shoppers.
Warranty information performs an important function for marketers by serving as a persuasive sales variable and by protecting sellers from unreasonable claims (Purohit and Srivastava 2001). While manufacturers have used warranty information as a competitive weapon to differentiate their products from their competition (Innis and Unnava 1991), extant literature on the use of warranty information by Web sites as an extrinsic cue has been limited. Thus, our research seeks to advance the understanding of the effect of a Web site warranty on a consumer's perception of risk and quality in online shopping.
Warranty information influences product perceptions and behavioral intentions either directly or indirectly through the reduction of perceived risk (Dowling and Staelin 1994; Erevelles, Roy, and Vargo 1999; Shimp and Bearden 1982). Warranties also have an impact on perceptions of product quality (Erevelles, Roy, and Vargo 1999). The general proposition is that under conditions of uncertainty, consumers' perceptions of quality are affected by warranty through the signaling of product reliability (Kelley 1988). Purchase intention greatly increases when warranties are offered because the perceived risk of purchase is reduced and the quality of the product is assured (Van den Poel and Leunis 1999).
Retailer Reputation and Brand Name as Extrinsic Cues
Like a brand name, a retailer's reputation can be interpreted as the perception of quality associated with it. As an example, the name "Nordstrom" evokes an image of a luxurious store environment, high levels of customer service, and high-quality merchandise (Grewal et al. 1998). Retailer reputation, as a risk reliever, has been featured in many studies of well-established physical retailers (e.g., Akaah and Korgaonkar 1988; Derbaix 1983). Consumers' perceived risk of purchase was lower for "brick and mortar" retailers who possess a reputation for providing good service and quality products than for unknown retailers (Purohit and Srivastava 2001). In addition, several empirical studies have shown that when the perception of a store or retailer's name is more favorable, buyers' perceptions of product quality also will be higher (Dawar and Parker 1994; Dodds, Monroe, and Grewal 1991). Chu and Chu (1994) demonstrated that manufacturers are able to rent the reputation of a retailer in order to signal product quality. A retailer's reputation affects consumer buying decisions as consumers are more likely to purchase from established and reputable retailers than from unknown retailers (Akaah and Korgaonkar 1988; Tan 1999).
Hypotheses for Extrinsic Cues as Risk Relievers
In this research we generate two sets of hypotheses for extrinsic cues as risk relievers. The first set of hypotheses deals with our focal extrinsic cue, warranty information, and its effects on perceived risk, product quality, and purchase intentions when interacting with Web site reputation. The second set of hypotheses deals with warranty information and its effects on perceived risk, product quality, and purchase intentions when interacting with brand name. We feel the development of hypotheses focusing on interaction effects, as opposed to more simplistic main effects, would provide greater insight to our understanding of warranties in an Internet context, similar to other recent studies focusing on interaction effects in an Internet contexts, e.g., see Chu, Choi, and Song (2005) for a study of interactions among infomediary reputation, manufacturer brand, and retailer brand in an Internet context.
Interaction between Warranty Information and Retailer Reputation
Shimp and Bearden (1982) hypothesized interaction effects between physical retailer reputation and warranty. They argued that, for a reputable company, consumers' attitudes toward a product are positively related to the quality of the warranty offered. However, for an unknown company, better warranties would actually result in negative attitudes. This idea was based on the assumption that when an attractive warranty is offered by an unknown company, consumers will respond negatively because they think the warranty is "too good to be true (TGTBT)." On the other hand, an attractive warranty claim should be less likely to initiate defensive cognitive responses if made by a reputable and trusted retailer. Although Shimp and Bearden (1982) failed to obtain empirical support for their TGTBT hypothesis, this effect was found in a more recent study that involved an in-store retailer (Purohit and Srivastava 2001).
Boulding and Kirmani (1993) also investigated this interaction effect within the framework of economic signaling theory. Although their research was not conducted with respect to retailers (electronic or otherwise), their results have relevance to our current interests. An application of their results to our context suggests that when a retailer with a strong reputation uses a warranty and a retailer with a weak reputation does not use a warranty, consumers use the presence of a warranty for the retailer with a strong reputation and the absence of a warranty for the retailer with a weak reputation to infer product quality. This reasoning is based on the assumption that the consumer will view the weak retailer in contrast to the strong retailer, i.e., a strong retailer is able to offer a warranty, whereas a weak retailer is assumed to not be able to provide such a warranty. This is a slight modification of the Shimp and Bearden (1982) TGTBT hypothesis. We in turn propose the following hypotheses:
H1: When dealing with online retailers who have strong reputations, consumers will perceive a lower level of risk when a warranty is present than when a warranty is absent. However, when dealing with retailers with poor reputations, consumers' perceptions of purchase risk will not vary with the presence or absence of a warranty.
H2: When dealing with online retailers with strong reputations, consumers will perceive higher product quality when a warranty is present than absent. However, when dealing with retailers of poor repute, consumers' perceptions of product quality will not vary with the presence or absence of a warranty.
H3: When dealing with online retailers with strong reputations, consumers' purchase intentions will be higher when a warranty is present than when it is absent. However, when dealing with retailers of poor repute, purchase intentions will not vary with the presence or absence of a warranty.
Interaction between Warranty Information and Brand Name
A brand is a name, term, symbol, sign, or design to identify goods or services and a means to differentiate them from those of competitors (Keller 2002). In making product choices, consumers often use brand names as risk relievers (Grewal et al. 1998; Purohit and Srivastava 2001; Rao and Monroe 1989). In online shopping, where consumers typically cannot see or feel the product and have no intrinsic cues to rely on, the brand name serves as a heuristic for the evaluation of the product. By providing an established brand name, Internet marketers can increase the likelihood that consumers will purchase the product from them because of the lower levels of perceived risk (Tan 1999). Brand name also influences consumer preferences and purchase intentions in in-store retailing contexts, and we propose that the effects will carry over to the online shopping context (Cobb-Walgren, Ruble, and Donthu 1995). A strong brand name, like a reputation, is a signal of quality, so it reduces risk for customers (Keller 2002). We suggest that
H4: When dealing with online retailers who market well-known brands, consumers' perceptions of purchase risk will be lower when a warranty is present than absent. However, when dealing with retailers marketing unknown brands, perceived purchase risk will not vary with the presence or absence of a warranty.
H5: When dealing with online retailers who market well-known brands, consumers' perceptions of product quality will be higher when a warranty is present than absent. However, when dealing with retailers who market unknown brands, consumers' perceptions of product quality will not vary with the presence or absence of a warranty.
H6: When dealing with online retailers who market a well-known brand, consumers' purchase intentions will be higher when a warranty is present than absent. However, when dealing with retailers who market unknown brands, purchase intentions will not vary with the presence or absence of a warranty.
To be certain that privacy/disclosure risks were not dominating other types of risks (which were the types we manipulated in the experiment which follows), we first conducted a pretest of the various types of risk (see Table 2). This exploratory pretest on a convenience sample of 74 Internet shoppers (i.e., people who have made online purchases) required them to identify the types of risks commonly associated with online purchases. The results indicate that although privacy/disclosure risks were elicited, they ranked 8th (credit card security) and 10th (loss/theft) in a list of most salient risks.
We designed an experiment involving 240 Internet users to test the effects of the three risk relievers on consumers' perceived purchase risk, perceived product quality, and purchase intentions in an online shopping context. Subjects were recruited via e-mail invitations to visit our experimental Web site, which was featured via a link. For their participation, they were given an option to enter in a lottery, which offered shopping vouchers as prizes and were randomly allocated to one of the treatment conditions. The sample size for each treatment cell was 30. The sample profile was similar to that of the Internet user population in terms of gender, age, and education.
We employed the following 2 x 2 x 2 between-subjects factorial design: Brand Name (well-known, unknown) x Retailer Reputation (strong, weak) x Warranty Information (present, absent). The stimuli development consisted of various stages. We sought feedback from 12 Internet users to select the product category for use in the actual experiment based on verisimilitude of perceived risk and also to ensure the realism of our Web site offering. Mobile phones were selected as the product category for the experiment. We recognize that using only mobile phones in this research design limits the findings for the present study to this particular product category. A future research opportunity would be to cross-validate our results by including other, diverse product categories.
To be consistent with other studies of the effects of risk relievers on product evaluations, we used actual brand names and retailer names in our experiment (Dodds, Monroe, and Grewal 1991; Shimp and Bearden 1982). We did this so that the subjects would conjure perceptions of reputation and knowledge in their evaluation of the retailers and products. We developed eight different Web sites, which enabled the manipulation of the three independent variables. Subjects who responded to our e-mail invitation were randomly directed to only one of the eight Web sites.
Appendix 1 shows the screen shot of a condition with Web site warranty for an unknown brand of mobile phone marketed by a retailer with weak reputation (one of eight versions). The content and layout for all the Web sites were identical, except for the differences in brand names, retailer names, and warranty information that were manipulated. Each Web site also contained a picture of a mobile phone not available in the market to prevent a confounding effect of product ownership. A limitation of keeping the content and layout of each Web site identical is that we could not assess the influence of Web site qualities such as attractiveness on consumers' attitudes and purchase intentions. Given that previous studies have shown that such qualities can have an impact (Balabanis and Vassileiou 1999), another opportunity for future research would be to incorporate varying levels of Web site attractiveness in the research design.
Warranty information was manipulated at two levels--"express warranty" and "no express warranty." For the former, information on the express warranty by the online retailer was included in proximity of the product profile. In the other condition, no express warranty information was provided on the Web site. An alternative strategy would have been to include a phrase to indicate that there was no Web site warranty for the mobile phone. However, the presence of such a phrase may sensitize subjects to warranty information (Innis and Unnava 1991). Thus, subjects either processed warranty information or were not primed to its nonexistence.
Similarly, retailer reputation was manipulated to distinguish between a strong reputation and a weak reputation. A pretest was undertaken to select two mobile phone Web sites that were distinguishable on the basis of reputation. Twenty-five subjects were shown a list of online retailers and were required to rate their reputation on a 7-point differential scale, ranging from "strong" to "weak" reputation. Strong reputation implied that the retailer was a known entity, while weak reputation referred to a relatively unknown retailer (hence without prior reputation). The results presented in Table 3 indicate that among the set of retailers, Retailers V and Y were rated as having the strongest ([bar.x] = 5.96) and weakest ([bar.x] = 2.48) reputations, respectively, and that this difference was significant (t = 9.89, p < .001). Based on these results, Retailers V and Y were chosen as the two mobile phone retailers for the actual experiment.
Brand name was manipulated at two levels: well known and unknown. Through a pretest, two brands of mobile phones were identified that were distinguishable on the basis of brand name and brand quality. In this pretest, 25 subjects were shown a list of mobile phone brand names and were required to rate their perceptions of each brand name based on two 7-point differential scales. The scales were anchored by "well-known brand name-unknown brand name." Table 4 shows that Brand D ([bar.x] = 6.48) and Brand C ([bar.x] = 2.88) were rated as the strongest and weakest brand names, respectively, and that this difference was significant (t = -13.94, p < .001). Hence, Brands D and C were chosen to represent the manipulation for brand name.
In this research design, we focus on three extrinsic cues. However, we also recognize that intrinsic product attributes (e.g., nutrition content in cereal, style of cars) also can have the capability to signal quality and reduce purchase risk. Literature on cue utilization has proposed that intrinsic cues have a greater influence on consumers' product evaluations than extrinsic cues (Zeithaml 1988). As such, it would be interesting in future research to investigate whether intrinsic or extrinsic cues are more effective in relieving risk in online shopping. Additional research also could provide insight into the effects of other extrinsic cues (e.g., price, promotions, and level of advertising) in online shopping. Also, other extrinsic cues could be incorporated, e.g., studies suggest cues relating to privacy (e.g., privacy notices and seals) influence risk perception in online shopping (Caudill and Murphy 2000). As noted above in our description of the three main risk relievers (brand name, retailer reputation, and warranty information), we explored each risk reliever in this study as a two-level factor, e.g., the presence or absence of warranty information. As Boulding and Kirmani (1993) have suggested that warranty length and warranty scope might have different impacts on consumers' evaluation of products, additional research could replicate this study and manipulate the warranty length (e.g., 30 days versus one year) and warranty scope (e.g., limited versus full warranty) to provide insight into their appropriate use.
Scales for the Dependent Variables
Perceived purchase risk was measured as a function of two components-the certainty of an undesired event (such as a purchased product not performing as expected) and its consequences (Bettman 1973; Cox and Rich 1967; Dowling and Staelin 1994; Mitchell 1999). Following these extant research, the certainty variable was measured on a 7-point scale, "1" denoting "very certain" and "7" denoting "almost never certain." The consequence variable was measured on a 7-point scale with "1" denoting "no danger" and "7" denoting "a great deal of danger." The certainty and consequences components were multiplied together to form a measure of perceived purchase risk--see Mitchell (1999) for a detailed explication on the tradeoffs of the multiplicative versus additive arguments for the two components in providing a measure of overall perceived risk.
The other two dependent measures, perceived product quality and purchase intentions, were captured as follows: (1) perceived product quality was measured using 7-point Likert statements such as "this product appears to be durable" (Dodds, Monroe, and Grewal 1991) and (2) intentions to purchase via the Internet were measured using a single item, given that many consumer behavior studies have used a single-item purchase intention measure, e.g., see Spears and Singh (2004) for a comprehensive listing of commonly used purchase intention measures.
Reliability Analysis and Manipulation Checks
Cronbach's Alpha was used to assess the reliability of the variables. Only one dependent measure, perceived product quality, had multiple items. The coefficient alpha was 0.95, so the variables were deemed to be reliable and fit for use.
Manipulation checks were conducted in the following manner. At the end of the experiment, subjects evaluated the brand name (very strong name to very weak name and very high quality to very low quality), retailer reputation (very strong reputation to very weak reputation), and warranty information (very attractive to very unattractive) on 7-point differential scales. The measures used for these manipulation checks were based on similar studies by Dodds, Monroe, and Grewal (1991) and Innis and Unnava (1991).
Results of a 2 x 2 x 2 analysis of variance on the manipulation check of brand name indicate significant main effects of brand name (F = 307.93, p < .0001) but no significant main effects of the other manipulations on the brand measures. Similarly, the results of the manipulation check of retailer reputation reveal a significant main effect on retailer reputation (F = 315.43, p < .0001). There were no significant main effects of the other manipulations on the retailer measures. Finally, the results of the manipulation check of warranty information indicate a significant main effect of warranty information (F = 214.98, p < .0001). Furthermore, there were no significant main effects of the other manipulations on the warranty measure. Thus, the brand name, retailer reputation, and warranty information manipulations were deemed as successful.
Multivariate Analysis of Variance
Preliminary analysis revealed no significant differences for the two covariates across all the eight cells (p > .4). Attitudes toward online shopping and search costs were held constant across all the treatment groups in this experiment. Data collected for this experiment were analyzed using Multivariate analysis of variance (MANOVA) and subsequently followed up with analysis of variance based on past studies on the effects of risk relievers (e.g., Dodds, Monroe, and Grewal 1991). Bartlett's test of sphericity indicated that the variables were correlated among each other (p < .001), hence justifying their grouping and the use of MANOVA on these three dependent measures simultaneously. The MANOVA and univariate results are reported in Table 5, and the t-test results of the cell means are provided in Table 6.
The MANOVA results indicate a strong support for the interaction of retailer and warranty (F = 7.52, p < .001), and support for H1-H3. The interaction effects are graphed in Figures 1-3. In addition, the main effect of brand name (F = 46.94, p < .001) was found, but no interaction between brand name and warranty information was seen.
[FIGURES 1-3 OMITTED]
Interaction Effect of Retailer Reputation and Warranty Information (H1-H3)
There was a significant interaction between retailer reputation and warranty information for perceived purchase risk (F = 4.31, p < .05). Specifically, when retailer reputation was strong, consumers perceived lower purchase risk ([bar.x] = 14.00) compared to when warranty information was absent ([bar.x] = 19.97, p < .01). However, when retailer reputation was weak, perceived purchase risk did not differ when warranty information was present ([bar.x] = 21.22) and when warranty information was absent ([bar.x] = 21.47, p > .8). Hence, the results support HI. This interaction is represented in Figure 1 and Table 6.
For perceived product quality, the interaction between retailer reputation and warranty information also was found to be significant (F = 20.35, p < .001). When retailer reputation was strong, the presence of warranty information produced higher perceived product quality ([bar.x] = 5.00) than when no warranty information was provided ([bar.x] = 4.37, p < .001). When retailer reputation was weak, the absence of warranty information ([bar.x] = 4.28) led to higher perceived product quality than its presence ([bar.x] = 3.91, p < .05). Thus, the results partially supported H2. This interaction effect is represented in Figure 2 and Table 6.
Finally, there was a significant interaction between retailer reputation and warranty information for purchase intentions (F = 15.12, p < .001). Specifically, when retailer reputation was strong, the presence of warranty information ([bar.x] = 3.98) led to higher purchase intentions compared to when warranty information was absent ([bar.x] = 2.90, p < .001). When retailer reputation was weak, purchase intentions did not differ when warranty information was either offered ([bar.x] = 2.17) or not offered ([bar.x] = 2.45, p > .2). Thus, these results supported H3. This interaction is represented in Figure 3 and Table 6.
Effect of Brand Name
No interaction effect for perceptions of brand name and warranty information was found on perceptions of purchase risk, product quality, and purchase intentions (hence no support for H4 to H6). The univariate results in Table 5 revealed a significant main effect on perceived purchase risk (F = 22.09, p < .001), perceived product quality (F = 187.23, p < .001), and purchase intentions (F = 23.40, p < .001). This main effect suggests that a strong brand name conjures lower perceived purchase risk, higher perceived product quality, and higher purchase intentions than a weak brand name.
Our results provide evidence that when shopping online, consumers seem to demonstrate parallel and divergent responses to risk relievers (brand name, retailer reputation, and warranty information) to those that have been successfully utilized in other shopping environments. Contrary to expectations, the presence or absence of a brand name was found to influence perceived risk, product quality, and purchase intention as main effects whereas the interaction with warranty information was insignificant. However, for Web site reputation, its interaction with warranty information was highly significant as hypothesized. In other words, retailer reputation moderates the effect of expressed warranty, whereas the brand name has main effect only. This suggests differences in the nature of the two risk relievers in an online context. Brand name knowledge appears to be a strong influencer of risk perception regardless of warranty information (i.e., consumers perceive well-known brands to be relatively of good quality and hence less risky without seeking further risk-relieving cues). Retailer reputation, on the other hand, appears to be more tenuous as a risk reliever. For Web sites with positive reputations, the presence of warranty information lowers perceived purchase risk, enhances perceptions of product quality, and increases purchase intention. For Web sites with weak reputations, warranty information has no impact on consumers. Furthermore, retailers with weak reputations were unable to impact purchase intentions and were even at risk of lowering product quality judgments.
Our findings, as shown in Tables 5 and 6 and Figures 1-3, partially support the validity of our conceptual framework. Next, we discuss their implications.
Our research shows that for Internet shoppers, perceptions of risk are influenced by a number of risk relievers, specifically, brand name, retailer reputation, and Web site warranty information.
An important contribution of this study to theoretical development is its demonstration of the interaction between online retailer reputation and Web site warranty information on perceptions of purchase risk, product quality, and purchase intention. Our results are consistent with the TGTBT effect that was first suggested by Shimp and Bearden (1982) and later observed by Purohit and Srivastava (2001) in an in-store context. Furthermore, consumers interacting with an unknown Web site might even doubt the retailer's ability to back up warranties. Also, our results demonstrate that perceived risk and purchase intentions made no difference for Web sites with weak reputations that provided warranty information.
There was even a more pronounced TGTBT effect for product quality: not only was there no improvement in perceptions of quality among customers of retailers with weak reputations who offered warranties but there was an actual decay in consumer evaluations of product quality. However, as we hypothesized, the TGTBT effect did not apply to reputable Web sites. Perceptions of risk were lower and perceptions of product quality and purchase intentions were higher for retailers with strong reputations who provided warranty information as compared to retailers with strong reputations who did not offer such information. The effects of these three risk relievers hold distinct and important implications for the interaction of online consumers with established marketers as well as with marketing start-ups.
An established Web site, by virtue of its record of past performance and quality, provides consumers with a great degree of assurance, and as a result, consumers are more likely to purchase from them. In contrast, while brand name information elicits positive responses on its own, it does not interact with warranty information. In other words, online shoppers do not necessarily consider Web site warranties to be an important consideration when making decisions about well-known brands. It should be noted that we are not suggesting that our results indicate that consumers do not expect established brands to be backed up with warranties. It is just that brand information seems to be assessed by consumers as independent of the warranty. This result seems to be consistent with the earlier findings of Gerner and Bryant (1981). They found that appliance warranties across many different brands varied very little, thus making warranties less effective as a market signal. This could account for the lack of a significant brand and warranty interaction effect.
Our findings indicate that the use of warranties by leading Web sites as a means of differentiation is effective; hence, reputable Web sites should offer warranties to lower online consumers' perceptions of purchase risks, increase their perceptions of product quality, and heighten their purchase intentions.
In contrast to the effect warranties have on the business of reputable Web sites, the use of warranties by online marketing start-ups and companies with weak reputation is negligible and may even be more harmful than helpful. Such companies can consider marketing products with strong brand names to aid risk relief and "rent" the reputation of brand name products. Since our findings support the idea that a strong retailer reputation influences patronage decisions, instead of simply offering warranties, less reputable online marketers should initially invest in building their reputation.
Public Policy Implications for Consumers
For several decades, public policy makers have been concerned that consumers receive fair and equitable treatment in the marketplace by not being subjected to situations where a purchased product does not meet the specified conditions and claims of the marketer. Such concerns have stimulated a variety of various forms of warranty legislation, e.g., the Magnuson-Moss Warranty and Federal Trade Commission Improvement Act 1975 (Dowling 1985). It is not only in the United States but also in other parts of the world that the issue of warranties is a major concern of pubic policy makers. For example, in Europe the recently enacted Consumer Guarantees Directive establishes rules roughly comparable to those in the United States (Hamilton and Petty 2001).
Despite the overall attention that public policy makers have directed toward warranties, little attention has been given to warranties in an online environment. This is understandable given the rapid growth and popularity in electronic commerce only within the past few years. However, the structural characteristics of online shopping compared to in-store shopping can create an environment where legislative guidelines may have to be reevaluated. This is of particular importance in terms of consumer interest as more and more consumers engage in online shopping. It is critical that the attention given by public policy makers to warranties in a non-Internet environment also be afforded to the Internet environment in order to protect consumer interest.
From our study, there are a number of implications for public policy makers to consider that can benefit consumers. Since we found that explicit warranty information can positively interact with brand reputation in influencing consumers' purchase decision, it is important for regulators to be watchful of how major retailers are utilizing warranty provisions to their advantage and to ensure that these retailers honor these explicit warranties in such a manner that is not detrimental to consumer interest. As for new Internet retailers, they may seek to highlight additional offers and enhanced warranty information to attract consumers. Consumers thus should be forewarned that in the event that the product does indeed not perform, they may face relatively greater barriers in seeking remedies from unknown Web sites than reputable ones.
Warranties represent an attempt to ensure that consumers are fully informed about the product, and the public policy implications of more information is that there should be less chance of individuals misallocating their resources during purchase (Dowling 1985). One of the structural characteristics of the online shopping environment is that it is relatively easy to communicate such warranty information to online customers. However, as indicated by the results of our research, there can be a differential impact on consumers' perceptions of the Web site warranty information based on the reputation on the retailer. Also, as noted above, online shopping still is perceived as more risky compared to in-store shopping. While warranties are viewed as a type of risk reliever, and one that public policy makers have embraced for non-Internet environments, we would encourage public policy makers to move deliberately in enacting any new warranty legislation that would impact the somewhat uncharted waters of Internet retailing. The challenge will be to provide legislation that will put consumers at ease with online shopping and simultaneously continue to spur, and not hamper, the growth of electronic commerce.
Sample Web Site for Weak Web Site Reputation (Retailer Y), Unknown Brand (Brand C), with Warranty Condition
Akaah, I.P. and P.K. Korgaonkar. 1988. A Conjoint Investigation of the Relative Importance of Risk Relievers in Direct Marketing. Journal of Advertising Research, 28 (4): 38-44.
Alba, J., J. Lynch, B. Weitz, C. Janiszewski, R. Lutz, A. Sawyer, and S. Wood. 1997. Interactive Home Shopping: Consumer, Retailer and Manufacturer Incentives to Participate in Electronic Marketplaces. Journal of Marketing, 6 (July): 38-53.
Balabanis, G. and S. Vassileiou. 1999. Some Attitudinal Predictors of Home-Shopping through the Internet. Journal of Marketing Management, 15 (June): 361-385.
Barsky, J. 1995. Guarantee; Warranty; Loyalty. Marketing Tools, 2 (6): 72-74.
Bettman, J.R. 1973. Perceived Risk and Its Components: A Model and Empirical Test. Journal of Marketing Research, 10 (May): 184-190.
Boulding, W. and A. Kirmani. 1993. A Consumer-Side Experimental Examination of Signaling Theory: Do Consumers Perceive Warranties as Signals of Quality? Journal of Consumer Research, 20 (June): 111-123.
Caudill, E.M. and P.E. Murphy. 2000. Consumer Online Privacy: Legal and Ethical Issues. Journal of Public Policy & Marketing, 19 (Spring): 7-19.
Chu, Wujin, Beomjoon Choi, and Mee Ryoung Song. 2005. The Role of On-line Retailer Brand and Infomediary Reputation in Increasing Consumer Purchase Intention. International Journal of Electronic Commerce, 9 (Spring): 115-127.
Chu, W. and W. Chu. 1994. Signaling Quality by Selling Through a Reputable Retailer: An Example of Renting the Reputation of Another Agent. Marketing Science, 13 (Spring): 177-189.
Cobb-Walgren, C.J., C.A. Ruble, and N. Donthu. 1995. Brand Equity, Brand Preference, and Purchase Intent. Journal of Advertising, 24 (Fall): 25-40.
Cox, D.F. and S.U. Rich. 1967. Perceived Risk and Consumer Decision Making--The Case of Telephone Shopping. In Risk Taking and Information Handling in Consumer Behavior, edited by Donald F. Cox (487-506). Boston: Harvard University Press.
Dawar, N. and P. Parker. 1994. Marketing Universals: Consumers' Use of Brand Name, Price, Physical Appearance, and Retailer Reputation as Signals of Product Quality. Journal of Marketing, 58 (April): 81-95.
Derbaix, C. 1983. Perceived Risk and Risk Relievers: An Empirical Investigation. Journal of Economic Psychology, 3 (1): 19-38.
Dodds, W.B., K.B. Monroe, and D. Grewal. 1991. Effects of Price, Brand and Store Information on Buyers' Product Evaluations. Journal of Marketing Research, 28 (August): 307-319.
Dowling, G.R. 1985. The Effectiveness of Advertising Explicit Warranties. Journal of Public Policy & Marketing, 4, 142-152.
Dowling, G.R. and R. Staelin. 1994. A Model of Perceived Risk and Intended Risk-handling Activity. Journal of Consumer Research, 21 (1): 119-134.
Erevelles, S., A. Roy, and S.L. Vargo. 1999. The Use of Price and Warranty Cues in Product Evaluation: A Comparison of U.S. and Hong Kong Consumers. Journal of International Consumer Marketing, 11 (3): 67-91.
Forrester Research. 2001. Digital Marketing Will Reach $63 Billion in the US, While Online Advertising Alone Will Rise to $42 Billion Worldwide by 2005. Forrester Research Inc., Cambridge, MA.
Gerner, Jennifer L. and W. Keith Bryant. 1981. Appliance Warranties as a Market Signal. The Journal of Consumer Affairs, 15 (Summer): 75-86.
Grewal, D., R. Krishnan, J. Baker, and N. Borin. 1998. The Effect of Store Name, Brand Name and Price Discounts on Consumers' Purchase Intentions. Journal of Retailing, 74 (Fall): 331-352.
Hamilton, J. and R. Petty. 2001. The European Union's Consumer Guarantees Directive. Journal of Public Policy & Marketing, 20 (Fall): 289-296.
Hawes, J.M. and J.R. Lumpkin. 1986. Perceived Risk and the Selection of Retail Store Patronage Mode. Journal of the Academy of Marketing Science, 14 (Winter): 37-42.
Hof, R.D. 2001. Amazon's Go-Go Growth? Gone. Business Week, (February) 3719: 39.
Innis, D.E. and H.R. Unnava. 1991. The Usefulness of Product Warranties for Reputable and New Brands. In Advances in Consumer Research, 18, edited by Rebecca H. Holman and Michael R. Solomon (317-322). Provo, UT: Association for Consumer Research.
Keller, K.L. 2002. Building, Measuring, and Managing Brand Equity. 2nd edition. Englewood, NJ: Prentice-Hall.
Kelley, C.A. 1988. An Investigation of Consumer Product Warranties as Market Signals of Product Reliability. Journal of the Academy of Marketing Science, 16 (Spring): 72-78.
Levy, M. and B.A. Weitz. 2001. Retailing Management. Boston: McGraw-Hill/Irwin.
Lynch, P.D. and J.C. Beck. 2001. Profiles of Internet Buyers in 20 Countries: Evidence for Region-Specific Strategies. Journal of International Business Studies, 32 (4): 725-748.
Market Wire. 2005. U.S. E-Commerce Growth to Significantly Outpace Total Retail Spending. New York. http://web.lexis-nexis.com (accessed July 19, 2005), Market Wire Inc.
Mathwick, C., N. Maholtra, and E. Rigdon. 2001. Experiential Value: Conceptualization, Measurement, and Application in the Catalog and Internet Shopping Environment. Journal of Retailing, 77 (1): 39-56.
Mitchell, V. 1999. Consumer Perceived Risk: Conceptualisations and Models. European Journal of Marketing, 33 (1-2): 163-195.
Miyazaki, A.D. and A. Fernandez. 2001. Consumer Perceptions of Privacy and Security Risks for Online Shopping. Journal of Consumer Affairs, 35 (1): 27-44.
Peter J. Paul and Lawrence X. Tarpey. 1975. An Investigation of Perceived Risk in the Brand Level. Journal of Marketing Research, 13 (May): 184-188.
Purohit, D. and J. Srivastava. 2001. Effect of Manufacturer Reputation, Retailer Reputation, and Product Warranty on Consumer Judgments of Product Quality: A Cue Diagnosticity Framework. Journal of Consumer Psychology, 10 (3): 123-134.
Rao, A. and K. Monroe. 1989. The Effect of Price, Brand Name and Store Name on Buyers' Perceptions of Product Quality: An Integrative Review. Journal of Marketing Research, 26 (August): 351-357.
Riera, J. 2002. Bumper Christmas Heralds Coming of Age for e-Tailers. Retail Week, February 8.
Roselius, T. 1971. Consumer Rankings of Risk Reduction Methods. Journal of Marketing, 35 (January): 56-51.
Sauer, C. and S. Burton. 1999. Is There a Place for Department Stores on the Internet? Lessons from an Abandoned Pilot. Journal of Information Technology, 14 (December): 387-398.
Schiffman, Leon G. and Leslie Lazar Kanuk. 1987. Consumer Behavior. Englewood Cliffs, NJ: Prentice-Hall.
Shimp, T.A. and W.O. Bearden. 1982. Warranty and Other Extrinsic Cue Effects on Consumers' Risk Perceptions. Journal of Consumer Research, 9 (June): 38-46.
Spears, Nancy and Surendra N. Singh. 2004. Measuring Attitude Toward the Brand and Purchase Intentions. Journal of Current Issues & Research in Advertising, 26 (Fall): 53-66.
Stewart, D.W. and Q. Zhao. 2000. Internet Marketing, Business Models, and Public Policy. Journal of Public Policy & Marketing, 19 (Fall): 287-296.
Tan, S.J. 1999. Strategies for Reducing Consumers' Risk Aversion in Online Shopping. Journal of Consumer Marketing, 16 (2): 163-180.
Van den Poel, D. and J. Leunis. 1995. The Impact of Price, Branding and Money-Back Guarantee on Store Choice. Proceedings of the 8th International Conference on Research in the Distributive Trades. Milan, Italy: Universita' Bocconi, pp. B4.21-B4.29.
--. 1999. Consumer Acceptance of the Internet as a Channel of Distribution. Journal of Business Research, 45 (March): 249-256.
Zeithaml, V.A. 1988. Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence. Journal of Marketing, 52 (July): 2-22.
May O. Lwin is an assistant professor at the Division of Public and Promotional Communication, School of Communications and Information, Nanyang Technological University, Singapore (firstname.lastname@example.org). Jerome D. Williams is the F. J. Heyne centennial professor, Department of Advertising, College of Communication, with a joint appointment in the Center for African and African American Studies, The University of Texas at Austin, Austin, TX (email@example.com).
The authors are extremely indebted to Geraldine R. Henderson for her insightful comments and helpful recommendations on analysis and conceptualization in earlier versions of the paper.
TABLE 1 Risk Reliever Rankings Warranty/ Brand/ Date Money-Back Manufacturer Published Guarantee Reputation Purohit and Srivastava 2001 1 1 Van den Poel and Leunis 1999 1 2 Tan 1999 4 3 Van den Poel and Leunis 1995 1 2 Dowling and Staelin 1994 1 1 Akaah and Korgaonkar 1988 1 2 Schiffman and Kanuk 1987 1 -- Hawes and Lumpkin 1986 1 -- Roselius 1971 -- -- Average ranking 1.38 1.83 Information Source/ Retailer Reference Reputation Price Group Trial Purohit and Srivastava 1 -- -- -- Van den Poel and Leunis -- 3 -- -- Tan 2 -- 1 5 Van den Poel and Leunis -- 3 -- -- Dowling and Staelin -- -- -- -- Akaah and Korgaonkar 4 3 6 5 Schiffman and Kanuk -- -- -- 1 Hawes and Lumpkin -- -- -- -- Roselius 1 -- -- -- Average ranking 2.00 3.00 3.50 3.67 TABLE 2 Types of Risks (Pretest 1) Number of Respondents Type of Risk Elicited Rank Reporting Financial/overpayment 1 44 Improper/infrequent/difficult use 2 41 Poor quality 3 28 Defective 4 23 Obsolescence 5 20 Adverse selection 6 16 Incompatibility 7 11 Credit card security 8 7 Fulfillment risk 9 5 Loss theft 10 3 Warranty conditions 11 5 Customer service issues 12 4 TABLE 3 Mean Scores for Retailer Names (Pretest 3) Retailer Names Reputation Mean Scores Retailer V 5.96 Retailer W 3.44 Retailer X 5.72 Retailer Y 2.48 Retailer Z 3.12 Note: Higher means indicate better reputation. TABLE 4 Mean Scores for the Selection of Brand Names (Pretest 2) Brand Names Brand Name Mean Scores Product Quality Mean Scores Brand A 5.24 5.04 Brand B 3.84 4.32 Brand C 2.88 3.36 Brand D 6.48 6.32 Brand E 3.96 4.40 Note: Higher means indicate better-known brand name and better product quality. TABLE 5 MANOVA and Univarate Results Univariate analysis MANOVA Perceived Risk df F Significance F Significance Brand 2 46.49 <0.001 22.09 <0.001 Retailer 2 16.86 <0.001 10.01 <0.01 Warranty 2 1.76 >0.1 5.09 <0.05 Brand x Retailer 4 1.74 >0.1 0.44 >0.1 Brand x Warranty 4 0.59 >0.1 0.16 >0.1 Retailer x Warranty 4 7.52 <0.001 4.31 <0.05 Brand x Retailer x 8 1.15 >0.1 1.72 >0.1 Warranty Univariate Analysis Perceived Purchase Product Quality Intentions F Significance F Significance Brand 187.23 <0.001 23.40 <0.001 Retailer 28.31 <0.001 41.59 <0.001 Warranty 1.40 >0.1 5.18 <0.05 Brand x Retailer 1.47 >0.1 1.76 >0.1 Brand x Warranty 1.33 >0.1 0.58 >0.1 Retailer x Warranty 20.35 <0.001 15.12 <0.001 Brand x Retailer x 1.13 >0.1 0.009 >0.1 Warranty TABLE 6 Retailer Reputation and Warranty Interaction (Comparison of Cell Means) Strong Retailer Reputation Warranty Warranty Present Absent Mean Mean Delta Perceived purchase risk 14.00 19.97 5.97 ** Perceived product quality 5.00 4.37 0.63 ** Purchase intention 3.98 2.90 1.08 ** Weak Retailer Reputation Warranty Warranty Present Absent Mean Mean Delta Perceived purchase risk 21.22 21.47 0.25 Perceived product quality 3.91 4.28 0.37 * Purchase intention 2.17 2.45 0.28 * p < .05; ** p < .01. Note: Higher means indicate greater perceived risks, better perceived quality, and greater purchase intentions.
|Printer friendly Cite/link Email Feedback|
|Author:||Lwin, May O.; Williams, Jerome D.|
|Publication:||Journal of Consumer Affairs|
|Date:||Dec 22, 2006|
|Previous Article:||Are we making the grade? A national overview of financial education and program evaluation.|
|Next Article:||The influence of individual factors on the effectiveness of message content in antismoking advertisements aimed at adolescents.|