Measuring Motivations for Online Opinion Seeking
Ronald E. Goldsmith, David Horowitz
Florida State University
Abstract
Online interpersonal influence or electronic word-of-mouth ("eWOM") is an important aspect of ecommerce. Consumers give and seek opinions online in much the same way as they do offline, thereby affecting the sales of many goods and services. To further the understanding of eWOM, the present study used data from a survey of 309 consumers to develop a 32-item self-report scale measuring consumer motivations for online opinion seeking. This study revealed eight distinct factors. Consumers seek the opinions of others online to reduce their risk, because others do it, to secure lower prices, to get information easily, by accident (unplanned), because it is cool, because they are stimulated by off-line inputs such as TV, and to get pre-purchase information. A second study using data from 109 consumers showed that: (1) the scales measuring these motivations are free from social desirability response bias and acquiescence, (2) other consumers' information is more important than advertising, and (3) consumers are likely to continue to seek WOM online, thereby confirming its importance in ecommerce.
Introduction
As the Internet becomes increasingly a locus
for consumption, we observe both differences and similarities
between online and offline consumer behavior. The unique asynchronous
and interactive nature of cyberspace gives consumers unparalleled
access to information, wide product and brand choice, the ability
to make price and quality comparisons as never before, and the
opportunity to interact with companies and with other consumers
in many different ways (Negroponte and Maes 1996). These interactions
are conducted via e-mail, instant messaging, homepages, Blogs,
listservs, forums, online communities, newsgroups, chat rooms,
hate sites, review sites, and social networking sites (Goldsmith
2006). Consumers clearly have a variety of means to communicate
with each other online to share information and feelings about
products and brands. Thus, the venerable topic of personal influence
now includes its extension into cyberspace as online interpersonal
influence or electronic word-of-mouth (eWOM).
In its early days, the net was seen as an opportunity for marketers
to communicate with consumers and to even engage them in two-way
communications (e.g., Feather 2000). Only gradually did it become
apparent that consumers were using the Internet to communicate
with each other. It is not surprising that consumers use the
new medium to exchange product related information in much the
same way as they do offline. What is different about cyberspace
are the: (1) variety of avenues or means by which consumers
can exchange information, (2) anonymity and confidentiality
online through which consumers don't have to reveal their identities
when seeking and giving advice, (3) physical cues used to assess
the identity of others which are lacking, (4) freedom from geographic
and time constraints that make cyberspace a global community
paralleling the local physical one, and (5) permanence of online
conversations (Gelb and Sundaram 2002; Kiecker and Cowles 2001).
Thus, eWOM deserves investigation as an extension of traditional
interpersonal communication into the new realm of cyberspace.
Bell and Song (2004) show that offline social contagion (social interaction and passive observation) influences the adoption of a new online retail service, just as social network theory predicts (Birnie and Horvath 2002; Guadagno and Cialdini 2004). Researchers, however, have only just begun to study online interpersonal communication, especially where it involves consumption. Questions arise about which means are used, which ones are most important, how consumers feel about online eWOM, what effects eWOM has on their behavior, and how marketers influence eWOM for their own purposes. The focus of the present study is on the reasons why consumers seek information and advice online. This is important because the better marketers understand why consumers engage in eWOM, the better they will be able to manage the manner in which eWOM influences purchase decisions. Theory oriented research will then have a tool for testing propositions derived from social influence theory in the Internet context. Thus, the purpose of the present study is to elucidate the motives for seeking eWOM and to develop a reliable and valid operationalization to measure these motives.
We seek to make a contribution to the theoretical aspects of social influence by providing evidence that eWOM is motivated in a significant way in much the same fashion as is offline interpersonal communications. This will permit the extension of those theories into the new communications milieu of cyberspace. Managerially, we hope that our findings can be of use to marketers who seek to manage eWOM by providing insights into consumer motives. Using this information, marketers should be able to improve their interactions with consumers online and direct eWOM toward promoting selected brands, websites, and companies. From a methodological perspective, we develop a measure of consumer motives for seeking opinions online that complements the scale developed by Hennig-Thurau et al. (2004) to measure motivations for supplying opinions online. We provide evidence for the psychometric soundness of our scale. These aspects include scale dimensionality, internal consistency, criterion-related validity, and freedom from response artifacts.
Literature Review
Lazarsfeld and Katz pioneered the study of interpersonal influence, which describes the flow of information and influence from person to person in social systems (Katz and Lazarsfeld 1955; Lazarsfeld, Berelson, and Gaudet 1944). Since their early work, a large body of research has appeared describing the many aspects of this important social phenomenon (e.g., Keller and Berry 2003; Weimann 1994). Marketing and consumer researchers have created a thriving research area of their own devoted to studying how WOM influences consumption (e.g., Arndt 1967; Dichter 1966; Haywood 1989; Whyte 1954). Several lengthy descriptions of "buzz management" have also appeared (e.g., Gladwell 2000; Hughes 2005; Rosen 2000), indicating managerial interest in both offline and online word-of-mouth.
As soon as the World Wide Web made it possible for ordinary people to use the Internet, they began to associate with one another online, creating virtual communities. Companies were quick to use these communities to gain competitive advantage, increase market share, create new businesses, and promote their wares. Rayport and Sviokla (1994) described the revolution the Internet was bringing to the marketplace and how businesses had to reconceptualize their notions of the value proposition, competitive advantage, and marketing strategy in the "marketspace." They did not, however, realize that among the changes brought by online interconnectivity was the potential for consumers to connect with each other in new, powerful, and flexible ways. Armstrong and Hagel (1995, 1996) proposed that commercial enterprises should go beyond simply advertising on the Internet and try to organize online communities to suit their own purposes:
We believe that commercial success in the on-line arena will belong to those businesses that organize electronic communities to meet multiple social and commercial needs. By creating strong on-line communities, businesses will be able to build customer loyalty to a degree that today's marketers can only dream of and, in turn, generate strong economic returns (Armstrong and Hagel 1996, p. 135).
Although Armstrong and Hagel did not explicitly discuss WOM or eWOM, their recommendations for managerial efforts to organize consumers online clearly implied an active strategy to monitor eWOM and attempt to guide it (see Lindgreen and Vanhamme 2005).
More recent managerial discussions have emphasized the role of the Internet in building brands (Breakenridge 2001), in creating and managing customer relationships in ways that take advantage of the personalization features of the Web (Osenton 2002), and adapting business models to "hybrid" consumers who are actively involved online (Wind, Mahajan, and Gunther 2002). These reformations, however, view the information flows as almost exclusively oneway, B2C. While they recognize the importance of virtual communities and of customer feedback, they do not explicitly discuss the importance of eWOM (C2C) to their strategies.
Marketers, however, have been avidly using the Internet to promote interpersonal communication. For example, stories ranging from the box office success of The Blair Witch Project (Streisand 1999), the meteoric diffusion of Botox (Ries and Ries 2002), and even the cult-like status enjoyed by KrispyKreme donuts (Serwer 2003) have each been attributed not to the products' advertising, but rather to their promoter's use of non-traditional means to create buzz. The music industry has been especially aggressive in encouraging eWOM through viral marketing, web postings, and product sampling (Humphries 2004).
Thus, the two thought streams, interpersonal influence and ecommerce, have merged into the nascent study of eWOM. Working independently, a few researchers saw the potential for creating a better understanding of online consumer behavior by noting the confluence of these ideas. Senecal and Nantel (2001) wrote one of the first scholarly papers to describe "online interpersonal influence" and to discuss eWOM as a parallel to offline WOM. They proposed a framework for the study of this phenomenon including descriptions of types of online interpersonal communications, the role these influences play on consumer decision making, and propositions for empirical research on the topic (see also Kiecker and Cowles 2001).
Smith, Menon, and Sivakumar (2003) reported some of the first empirical findings in an experimental study using scenarios. They studied the influence of peer recommendations on decision making, emphasizing the role of trust in an anonymous environment. They found that two important source variables, expertise and tie-strength of the recommender, were related to decision outcomes. These influences varied, however, depending on the shopping motives of the respondent. Where shopping motives were hedonic in nature, tie-strength was more important than expertise, but when utilitarian motives were dominant, both tie-strength and expertise were important. Moreover, trust mediated the impact of the peer recommender's characteristics on their perceived influence, propensity to search for product information, and willingness to recommend the opinions of the peer.
Godes and Mayzlin (2004) analyzed the content of posted messages from Usenet newsgroups. Their goal was to assess how much the amount and dispersion of these communications had on the success of new TV shows. They found that a measure of dispersion (how much eWOM takes place across different online communities) explained a good proportion of the shows success. They demonstrated that the availability of eWOM records creates new methods of studying WOM that promise to give new insights into how interpersonal communications take place.
Walsh, Mitchell, and Wiedmann (2004) surveyed
German consumers online to measure how many were eMavens or
online information disseminators. They found few demographic
differences between mavens and non-mavens, but there did appear
to be different types of eMavens with regard to their motivations
for spreading WOM online. De Bruyn and Lilien (2004) designed
a field study in which students were encouraged to spread WOM
online via email. They tracked this flow of influence through
the stages of referral: awareness, interest, evaluation, and
final decision. They concluded that certain characteristics
of the social tie, tie-strength, perceptual affinity, demographic
similarity, and source expertise had differential impacts depending
on the stage of the decision process. Tie-strength exclusively
facilitated awareness, perceptual affinity triggered recipients'
interest, and demographic similarity had a negative influence
on each stage.
Senecal and Nantel (2004) performed an experimental study of
consumers' use of online recommendation sources and their influence
on product choice. Their results showed that the consumers who
consulted product recommendations (from three sources, including
other consumers) selected the recommended products twice as
often as consumers who did not consult any recommendations.
Interestingly, the online recommendation source labeled "recommender
system," the automated personalization programs common
to many eBusinesses, was more influential than the "human
expert" and "other consumers" recommendation
sources. In addition, they found that the type of product being
recommended influenced the use of information source. Recommendations
for the experience product (wine) was were more influential
than the recommendations for the search product (calculators).
Bailey (2005) studied consumer use of product review websites where consumers post reviews of consumer products. His survey showed that many consumers consult these sites and place a great deal of importance on the information and opinions found there. His results also indicated that being an opinion leader oneself seemed to lead consumers to greater awareness and use of product review websites and that men were more likely than women to do so. Although Bailey (2005) gathered some information about the factors leading consumers to seek eWOM (as additional information, to give assurance, to get other consumers' views, as a primary source of information, came upon it by chance, and referred to it by another), these motives were elicited as open-ended questions, so they were not quantified or systematic. Moreover, they were limited to the motives for seeking opinions from product review sites.
Of especial importance to our study, Hennig-Thurau et al. (2004) sought to delineate the motives of consumers who share their opinions online. The Henning-Thurau et al. (2004) theoretical framework builds upon the work of Balasubramanian and Mahajan (2001) and identifies five categories of eWOM communication motives. The first of these categories is focus-related utility, which is the utility a consumer experiences when he or she makes a contribution that adds value to the community. Secondly, when a consumer uses the contributions that other community members have provided to benefit themselves personally, consumption utility is created. Approval utility relates to the satisfaction a consumer feels when he or she is commended by others. Moderator-related utility occurs after a third party helps make it easier for a community member to make a complaint. Finally, homeostasis utility is founded upon the desire people have for equilibrium or balance in their lives.
Our study complements their study by completing the opinion giving/opinion seeking model with a measure of the motives for seeking eWOM. This is important theoretically because motivations are the reasons for behavior (Hawkins, Best, and Coney 2004, p. 355), and any study of an aspect of consumer behavior, such as online word-of-mouth, would be incomplete without accounting for the reasons it occurs. Motivations might play a direct role in shaping specific aspects of eWOM, they might indirectly affect behavior operating through mediators, and they might act as moderators of other relationships. Thus, measuring them is an important element in understanding these social processes.
Method
A critical incident study (CIT) (Bitner, Booms, and Tetreault 1990) was performed to obtain qualitative data regarding salient instances in which eWOM had influenced a recent consumer purchase. In this qualitative study, the undergraduate students of an introductory-level marketing course from a major Southeastern U.S. university were instructed to first think of and to then provide a written description of a specific instance in which they had made a purchase based on eWOM, or "information that was obtained online from other consumers." In order to ensure subjects recalled instances in which they were influenced by eWOM, a brief lecture on eWOM was given to the class, and written instructions in the questionnaire describing eWOM were provided.
The qualitative phase of this study is considered important because it elicited open-ended answers regarding specific incidents in which consumers had been influenced by opinions from other consumers through an online medium. The responses to the qualitative study contained specific details regarding a purchase in which participants had been influenced by eWOM. Eighty-six usable responses were collected and were coded through the use of content analysis (Tax, Brown, and Chandrashekaran 1998). To code the data, we followed an inductive categorization method in which recurring factors found in a text passage are identified (Spiggle 1994). In our case, one coder processed the written statements in order to identify and categorize consumer motives for online opinion seeking. The coder was familiar with content analysis procedures but unfamiliar with the opinion seeking literature. Because of this, the coder was unfamiliar with any framework or conceptualization a priori and the procedure was objective and data-driven.
After the content analysis was complete, one of the authors reviewed the results with the coder in order to make sure the proper procedures were followed and to examine the categories that were identified. These categories are shown in Table 1.
Table 1. Content Analysis Results
The statements from the qualitative study also are shown in
Table 1. These are typical statements from each of the seven
categories identified in the content analysis. For each motive
category, the original language from the verbatims was rewritten
as complete sentences to be as grammatical as possible while
preserving as much as possible the sense of the original statements
so that they would express the ideas of the consumers in the
qualitative phase. These formed the 39 questionnaire items.
A questionnaire containing 39 items was used to measure the customer motivations to seek opinions online. In addition to the 39 items developed to measure motives for seeking opinions online, items measuring demographics and behaviors were also included in the questionnaire.
After pretesting with several student and adult respondents, the questionnaire was distributed to a new sample of undergraduate business students in a marketing class at a large Southeastern U.S. university. Each student was instructed to complete the questionnaire and also to obtain a completed questionnaire from two adults (non-students), one male and one female. The questionnaire ended by asking for a first name and phone number so that a follow-up contact could verify that the questionnaire was completed by someone other than the student. Students were considered appropriate participants in the study because they have universal access to the Internet on the campus, they are familiar and frequent users of the Internet (including buying), and because they are active seekers of interpersonal information (Weiss 2003). We deliberately gathered data from older, non-students in order to enhance the demographic variability in the sample. Thus, although we used a convenience sample, an effort was made to collect data from a realistic group of consumers with wide differences in demographics that would be large enough to provide a strong test of the factor structure of the scale.
Sample
The sample size for this study was 309. There were 167 (54%) women and 142 (46%) men. Their ages ranged from 18 to 58 years, with a median age of 22 years. There were 235 (76%) whites, 34 (11%) Hispanics, 28 (9%) African-Americans, and 5 (2%) others. Most of the sample, 54.8%, described their or their family's socio-economic status as either top corporate or lesser corporate. Twenty-eight percent described themselves as middle management or small firm owners, and the remaining 17% said they were skilled craftsmen, average workers, service workers, or semiskilled. Thus, the sample consisted largely of upper and middle class white college students and adults. Tests showed that the demographic characteristics were largely unrelated, with the sole exception that the men were slightly older (22.6 years) than the women (21.5 years) (t = 2.1, p = .034). Although not representative of the general population, the sample contains a large proportion of the typical users of the Internet, namely younger affluent consumers.
Exploratory and Confirmatory Factor Analyses
The 39 items assessing motives for online opinion seeking were first factor analyzed by submitting them to the SPSS principal axis factor analysis program to determine the dimensionality of the entire set of items. This is the recommended procedure for scale development by several authorities (DeVellis 1991; Netemeyer, Bearden, and Sharma 2003; Spector 1992). The Kaiser eigenvalue criterion of 1.0 was used as was a minimum cutoff for the factor loadings of .4 (Netemeyer, Bearden, and Sharma 2003; Pett, Lackey, and Sullivan 2003). The initial factor solution was rotated using the oblique method because the factors were presumed to be non-orthogonal, thus yielding a more veridical solution than an orthogonal rotation.
The results of this analysis showed that some of the items did not meet the criteria for inclusion in the scale, and thus these items were removed one by one as the analysis was repeated. Four items were removed in this fashion. This yielded an eight-factor solution that possessed the properties of simple structure (see Pett, Lackey, and Sullivan 2003, p. 132). Each factor was defined by at least two items. Items loaded highly (> .4, with one exception) on only one factor and nearly zero on the others (the largest cross loading in the matrix was -.215).
The 35 remaining scale items were factor analyzed by submitting them to the AMOS structural equation modeling program to evaluate the dimensionality, construct validity, and construct reliability of the scale. Specifically, we estimated a single measurement model containing all eight dimensions proposed in the previous section. We used a CFA approach in order to simultaneously evaluate the relations of each item with the factor representing the construct they were supposed to measure (their convergent validity) and the lack of relations of each item with factors they were not supposed to measure (their discriminant validity) while accounting for errors of measurement.
Items were forced to load on their respective
factors and they were not allowed to cross load. The results
of the confirmatory factor analysis indicated that the model
offered good fit to the data (
= 1169.6, df = 532; CFI = .98; TLI = .98; RMSEA = .062). The
reliability and validity of each of the eight factors was assessed
using Fornell and Larcker's (1981) criteria. All factors
were shown to have a construct reliability of at least .74,
as construct reliability estimates ranged from .74 (saw on TV)
to .96 (price consciousness). Likewise, discriminant validity
was supported as the average variance extracted (AVE) for each
factor exceeded the shared variance between it and the other
factors. Scale statistics and detailed results of the confirmatory
factor analysis including means, standard deviations, construct
reliabilities, average variances extracted, and correlations
are presented in Table 2.
Table 2. Scale Statistics and Overall Model Fit from the Confirmatory Factor Analysis of the 35-Item Scale
Because convergent validity was not demonstrated for two of the factors, a total of three items that did not load above .6 on their intended factor were eliminated from the to get information and the ease of use factors, and the model was estimated a second time using AMOS with a total of 32 items. When items do not correlate strongly with their intended factor, more error variance than unique variance becomes associated with a factor. With the three items eliminated, the amount of error variance associated with the to get information and the ease of use factors was reduced, and their average variances extracted exceeded .50, which demonstrates convergent validity according to Fornell and Larcker's (1981) generally accepted criteria. The results of the second confirmatory factor analysis can be seen in Table 3. Finally, an exploratory factor analysis of the 32 items was performed as before. Table 4 presents the 32 remaining scale items and the factor loadings from the EFA pattern matrix from this analysis.
Table 3. Scale Statistics and Overall Model Fit from the Confirmatory Factor Analysis of the 32-Item Scale
Table 4. Factor Structure of the Motives for Seeking Opinions Online Scale
Criterion-Related Validity
The scores on the eight measures of motives for seeking eWOM were summed so that higher scores indicated greater levels of the construct. These scores were then correlated with two criterion measures that were computed from self-reports of buying activity by the survey participants. Our presumption was that higher levels of a motive to seek opinions online would be positively related to amount of online purchasing; the more people buy online, the more they seek the opinions of other. Two separate measures were employed in order to enhance the reliability of the construct measure. The first measure consisted of three statements labeled "Online Buying." These were: "I use the web to purchase products/services online," "It is common for me to purchase products/services online," and "I buy a lot of product/services online compared to others." A five-point Likert response format was provided where 1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, and 5 = strongly agree. A common factor analysis showed that these three items formed a unidimensional scale. The items were summed so that higher scores indicated higher levels of online buying. Internal consistency (alpha = .87) was adequate for the purposes of the study. The second measure of online buying also consisted of three items labeled "Amount of Online Purchasing." The first was a statement, "How often do you purchase anything online?" with a five-point response format where 1 = never, 2 = rarely, 3 = seldom, 4 = sometimes, and 5 = frequently. The second statement asked respondents to note the frequency with which they purchased online using a seven-point response format in increments from "never" to "daily." The third statement asked respondents to report the amount of money they spent buying products online in a typical month. The six response categories ranged from $0 - $99 to $500+. Because these three items used different response formats, they were combined by using principal components analysis to assess that only one principal component was needed to account for the majority of the variance in their covariance matrix and to compute factor scores based on this component. The correlation between the scores on the first summed three-item scale and the factor scores for the second three items was .72, supporting the convergent validity of these two measures.
The correlations between the measures of the eight motives for seeking opinions online and the two measures of online buying appear in Table 6. The presumption was that consumers who purchased more online would also seek opinions online was generally supported. Six of the motive measures, risk, price, ease of use, accident, information, and TV, were positively related to both measures of online buying. This suggests that these motives are positively related to online buying, which one would suspect; consumers who are motivated to seek opinions online are likely to buy online, indicating that they have greater experience than other consumers.
The aim of the second quantitative study was to assess the possible influence of two prominent types of response bias on the motives to seek opinions online scale: social desirability and acquiescence (Baumgartner and Steenkamp 2001). Social desirability is the "tendency to give answers that make the respondent look good" (Paulhus 1991, p. 17). "Acquiescence is the tendency to agree rather than disagree with propositions in general" (Paulhus 1991, p. 46). Measures that are influenced by social desirability and/or acquiescence confound that measure's content with these response biases and might be difficult to interpret or are correlated with other measures similarly affected. Because the behaviors described by the motives for seeking opinions online scale could be interpreted as representing socially good actions and attitudes, we felt it was important to assess its possible relationship with a measure of social desirability. Similarly, because all the scale items are worded in the positive direction, we felt it was important to assess the possible influence of acquiescence or "yea-saying" on the responses. Moreover, we wanted to assess the possible relationships between the motives for seeking opinions online scales with two new criteria. Study one showed that several of the scales were correlated with measures of buying or purchasing online. In the second study, we measured two additional criteria, the perceived importance of getting opinions online and future intentions to get opinions online.
Sample
The data from study two came from a convenience sample of students in marketing similar to those in study one. There were 109 in total, 67 (61.5%) women and 42 (38.5%) men. Their ages ranged from 19 years to 32 years, with a mean age of 21.3 (SD = 1.5) years.
Measures
The questionnaire contained the 32 items from the motivations for seeking opinions online scale constructed in study one. The means, standard deviations, and correlations between the eight variables can be seen in Table 5. Also included were measures of (1) social desirability response set, the short-form Crowne-Marlow scale from Strahan and Gerbasi (1972); (2) a measure of acquiescence or yeasaying (Couch and Keniston 1960); (3) a single item asking "Compared to advertising, how important was the other consumers' information?" with a 7-point semantic differential response format using endpoints labeled "Less important than advertising" and "More important than advertising"; and (4) a behavioral intentions measure reading "How likely are you to use online information from other consumers again?" with a 7-point semantic differential response format using endpoints labeled "Very likely" and "Not likely at all."
Table 5. Correlation Table for Study Two
The Crowne-Marlowe items were summed so that
higher scores reflected a greater tendency to respond in a socially
desirable manner, and the acquiescence items were summed so
that higher scores indicated a greater tendency to agree with
items regardless of their content. Disappointingly, the internal
consistency estimates (coefficient alpha) of these two scales
were low: social desirability = .64 and acquiescence = .59,
so the findings should be viewed with caution.
The means of the single item indicators of relative importance and future intentions (mean importance = 4.82, mean intentions = 4.81), their negative skewness (importance = -.685, intentions = -.467), and the distributions of their scores showed that most respondents felt that eWOM was more important than advertising and that they intended to continue to consult with other consumers online in the future. This testifies to the importance of eWOM in ecommerce.
Correlations between the scores on the eight measures of motives to seek opinions online and social desirability, acquiescence, importance, and intentions appear in Table 6. The direction and size of the correlations between the motive measures with social desirability and acquiescence suggest that they are largely free from both response set confounds. Only 2 of the 16 correlations were statistically significant (p < .05) and these were not large.
Table 6. Correlations Between Motives for Seeking Opinions and Other Variables
The correlations of the motive measures with relative importance and intentions revealed several positive, significant relationships. The movies of risk, price, easy, accident, TV, and information, were correlated with at least one of the two criteria, suggesting that consumers more highly motivated to seek opinions online find their information more important than advertising and intend to do it more in future than do their less motivated counterparts. A final analysis repeated these correlations as partial correlations with the effect of social desirability and acquiescence controlled and showed exactly the same results, confirming the freedom of the new measures from these response set influences.
Discussion
The results of our study suggest that consumers seek the opinions of others online for a variety of reasons. To uncover these, previous research and a qualitative study of responses to open-ended questions were used to generate a set of scale items, and two quantitative studies were performed. The result was a multidimensional scale containing eight subscales. The motivational factors range from basic utilitarian motives such as to get information to a more hedonic motive such as it's cool. Furthermore, it is evident that some of the factors seem more deliberate and planned, while other motivations are more spontaneous in nature. For example, the perceived risk factor contains items that suggest consumers go online to seek information as part of a consumer search process that rigorously screens and compares products and services online. This factor is much different than the accidentally factor which portrays consumers as going online to seek information in more of a spontaneous manner. Overall, these motives appear to be similar to the motivations influencing consumers to seek opinions offline and are consistent with Bailey's (2005) limited study.
A recent study described by Parker (2005) reports that eWOM (both giving and seeking) is pervasive and growing. Consumers in this study were approximately 16% more likely to be influenced by eWOM than by traditional advertising media (radio, TV, and newspapers). If it is true that, in the words of Mike Nazzaro, CEO of Intelliseek, the source of the study, "we're really on the front end of what will be a fundamental shift in the way consumers get and receive information" (Parker 2005), then the study of consumer motivations to seek eWOM is quite important. Matched with Hennig-Thurau et al.'s (2004) scale measuring motivations for giving opinions online, our scale can be used to explore the reasons consumers participate in eWOM.
Consumer and communication researchers will want to formulate models of eWOM in order to understand its causes, processes, and effects. The large body of theory and research on social communication will provide much of the background for these efforts. Our scale can contribute to this effort by giving these researchers a standard instrument to measure motives for seeking opinions online that can be used to test portions of these models. For example, the pioneering work of Birnie and Horvath (2002) uses social network theory to understand why and how much people use the Internet for social communication. Our scale mirrors this effort in the specific domain of online consumer behavior. The recent study by Bart et al. (2005) highlights the moderating role of purchase context on consumer behavior online. A future study could examine how consumers' motivations for online opinion seeking vary across online medium and purchase contexts. As more and more people grow up immersed in online communications, online communication medium evolve, and types of purchases and exchanges that are made online change, the manner in which eWOM is conceptualized and measured must also change.
The variability in the motivations for online opinion seeking identified in this study suggests that managers must consider a broad range of reasons that lead consumers to engage in eWOM. One area of study that could yield valuable information would be to apply the scale to a variety of types of consumers and product categories. Some consumers are likely to rely on eWOM more than others; knowing what motives this behavior would be useful for influencing their online opinion seeking behavior. Consumers are not likely to be homogenous in terms of seeking eWOM. They might be segmented into distinct groups based on their different motives. Managers across different product fields could use this scale to identify the factors that are most influential in leading their specific consumers to seek eWOM. Are consumers motivated to seek opinions online because of utilitarian reasons such as finding lower prices or for hedonic reasons such as because it's cool? With this knowledge managers could devise strategies that could motivate more of their consumers to engage in eWOM. For example, consumers could be encouraged using other media to seek out specific web sites by appealing to their dominant motives for seeking eWOM.
It is important to recognize the limitations of this study. The first limitation is that the qualitative portion of this study, its starting point, was performed using a student sample. If a different demographic group were used as a starting point, it is possible that the results could have been different. We may not have uncovered all the motives consumers have for seeking eWOM. The non-probability samples in the quantitative studies also limit the generalizability of the results with regard to point and interval estimates. However, in the quantitative portion of the study many non-students subjects were surveyed, giving the data desirable variation. Another limitation lies in the limited number of variables examined in relation to the motives. Future studies should expand this research by studying a variety of additional constructs.
References
Armstrong, Arthur and John Hagel III (1995), "Real Profits from Virtual Communities," The McKinsey Quarterly, 3, 127-141.
--- and --- (1996), "The Real Value of On-line Communities," Harvard Business Review, 74 (May-June), 134-141.
Arndt, Johan (1967), "Role of Product-Related Conversations in the Diffusion of a New Product," Journal of Marketing Research, 4 (August), 291-295.
Bailey, Ainsworth Anthony (2005), "Consumer Awareness and Use of Product Review Websites," Journal of Interactive Advertising, 6 (1) <http://www.jiad.org/vol6/no1/bailey/index.htm>.
Balasubramanian, Sridhar and Vijay Mahajan (2001), "The Economic Leverage of the Virtual Community," International Journal of Electronic Commerce, 5 (3), 103-138.
Bart, Yakov, Venkatesh Shankar, Fareena Sultan, and Glen L. Urban (2005), "Are the Drives and Roles of Online Trust the Same for All Web Sites and Consumers? A Large-Scale Exploratory Empirical Study," Journal of Marketing, 69 (4), 133-152.
Baumgartner, Hans and Jan-Benedict Steenkamp (2001), "Response Styles in Marketing Research: A Cross-National Investigation," Journal of Marketing Research, 38 (2), 143-156.
Bell, David R. and Sangyoung Song (2004), "Social Contagation and Trial on the Internet: Evidence from Online Grocery Retailing," Working Paper: The Wharton School, University of Pennsylvania.
Birnie, Sarah A. and Peter Horvath (2002), "Psychological Predictors of Internet Social Communication," Journal of Computer-Mediated Communication, 7 (4).
Bitner, Mary Jo, Bernard H. Booms, and Mary Stanfield Tetreault (1990), "The Service Encounter: Diagnosing the Favorable and Unfavorable Incidents," Journal of Marketing, 54 (January), 71-84.
Breakenridge, Deirdre (2001), Cyberbranding, New York: Prentice Hall.
Couch, Arthur and Kenneth Keniston (1960), "Yeasayers and Naysayers: Agreeing Response Set as a Personality Variable," Journal of Abnormal and Social Psychology, 60 (2), 151-174.
De Bruyn, Arnaud and Gary L. Lilien (2004), "A Multi-Stage Model of Word of Mouth through Electronic Referrals," eBusiness Research Center Working Paper, Penn State University.
DeVellis, Robert F. (1991), Scale Development: Theory and Applications, Newbury Park, CA: Sage Publications.
Dichter, E. (1966), "How Word-of-Mouth Advertising Works," Harvard Business Review (November), 131-144.
Feather, Frank (2000), future consumer.com, Toronto: Warwick Publishing.
Fornell, Claes and David F. Larcker (1981), "Evaluating Structural Equation Models with Unobservable Variables and Measurement Error," Journal of Marketing Research, 18 (February), 39-50.
Gelb, Betsy D. and Suresh Sundaram (2002), "Adapting to ‘Word of Mouse'," Business Horizons, 45 (4), 21-25.
Gladwell, Malcolm (2000), The Tipping Point, Boston: Little, Brown and Co.
Godes, David and Dina Mayzlin (2004), "Using Online Conversations to Study Word-of-Mouth Communications," Marketing Science, 23 (4), 545-560.
Goldsmith, Ronald E. (2006), "Electronic Word-of-Mouth," in Encyclopedia of E-Commerce, E-Government and Mobile Commerce, Mehdi Khosrow-Pour, Ed., Hershey, PA: Idea Group Publishing, forthcoming.
Guadagno, Rosanna E. and Robert B. Cialdini (2004), "Online Persuasion and Compliance: Social Influence on the Internet and Beyond," in The Social Net: The Social Psychology of the Internet, Y. Amichai-Hamburger, Ed., Oxford, UK: Oxford University Press, 91-113.
Hawkins, Del I., Roger J. Best, and Kenneth A. Coney (2004), Consumer Behavior: Building Marketing Strategy, 9th edition, Boston: McGraw Hill-Irwin.
Haywood, K. Michael (1989), "Managing Word of Mouth Communications," The Journal of Services Marketing, 3 (2), 55-67.
Hennig-Thurau, Thorsten, Kevin P. Gwinner, Gianfranco
Walsh, and Dwayne D. Gremler (2004), "Electronic Word-of-Mouth
via Consumer-Opinion Platforms: What Motivates Consumers to
Articulate Themselves on the Internet?" Journal of
Interactive Marketing, 18 (1), 38-52.
Hughes, Mark (2005), Buzz Marketing: Get People to Talk
About Your Stuff, New York: Portfolio.
Humphries, Stephen (2004), "Heard it through the Web grapevine," in The Christian Science Monitor, October 8.
Katz, Elihu and Paul F. Lazarsfeld (1955), Personal Influence: The Part Played by People in the Flow of Mass Communication, New York: The Free Press.
Keller, Edward B. and Jonathan L. Berry (2003), The Influentials: One American in Ten Tells the Other Nine How to Vote, Where to Eat, and What to Buy, New York: Free Press.
Kiecker, Pamela and Deborah L. Cowles (2001), "Interpersonal Communication and Personal Influence on the Internet: A Framework for Examining Online Word-of-Mouth," Journal of Euromarketing, 11 (2), 71-88.
Lazarsfeld, Paul F., Bernard Berelson, and Hazel Gaudet (1944), The People's Choice, New York: Duell, Sloan, and Pearce.
Lindgreen, Adam and Joelle Vanhamme (2005), "Viral Marketing: The Use of Surprise," in Advances in Electronic Marketing, Irvine Clarke and Theresa B. Flaherty, Eds., Hershey, PA: Idea Group Publishing, 122-138.
Negroponte, Nicholas and Pattie Maes (1996), "Electronic Word of Mouth," Wired, 4.10 (October).
Netemeyer, Richard G., William O. Bearden, and Subash Sharma (2003), Scaling Procedures: Issues and Applications, Thousand Oaks, CA: Sage Publications.
Osenton, Tom (2002), Customer Share Marketing, Upper Saddle River, NJ: Prentice Hall.
Parker, Pamela (2005), "Reach 'Ad Skippers' Via Online Word-of-Mouth," <http://www.clickz.com/news/article.php/3551611> (accessed on 2/15/2006).
Paulhus, Delroy L. (1991), "Measurement and Control of Response Bias," in Measures of Personality and Social Psychological Attitudes, John P. Robinson, Phillip R. Shaver, and Lawrence S. Wrightsman, Eds., San Diego: Academic Press, Inc., 17-59.
Pett, Marjorie A., Nancy R. Lackey, and John J. Sullivan (2003), Making Sense of Factor Analysis, Thousand Oaks, CA: Sage Publications.
Rayport, Jeffrey F. and John J. Sviokla (1994), "Managing in the Marketspace," Harvard Business Review, 72 (November-December), 141-150.
Reis, Al and Laura Reis (2002), The End of Advertising and the Rise of PR, New York: Harper Business.
Rosen, Emanuel (2000), The Anatomy of Buzz: How to Create Word of Mouth Marketing, New York: Doubleday.
Senecal, Sylvain and Jacques Nantel (2001), "Online Interpersonal Influence: A Framework," Working Paper: RBC Financial Groups Chair E-Commerce, HEC Montreal, University of Montreal.
--- and --- (2004), "The Influence of Online Product Recommendations on Consumers' Online Choices," Journal of Retailing, 80, 159-169.
Serwer, A. (2003) "The Hole Story: How Krispy Kreme Became the Hottest Brand in America", Fortune, 148 (1), 52-58.
Smith, Donnavieve, Satya Menon, and K. Sivakumar (2003), "Trust Me, Would I Steer you Wrong? The Influence of Peer Recommendations within Virtual Communities," in Proceedings, Academy of Marketing Science Annual Conference.
Spector, P.E. (1992), Summated Rating Scale Construction: An Introduction, Newbury Park, CA: Sage.
Spiggle, Susan (1994), "Analysis and Interpretation of Qualitative Data in Consumer Research," Journal of Consumer Research, 21 (December), 491-503.
Strahan, Robert and Kathleen C. Gerbasi (1972), "Short, Homogeneous Versions of the Marlow-Crowne Social Desirability Scale," Journal of Clinical Psychology, 28 (2), 191-193.
Streisand, Betsy (1999), "The Price of Good Buzz," U.S. News & World Report, September 27, 48.
Tax, Stephen S., Stephen W. Brown, and Murali Chandrashekaran (1998), "Customer Evaluations of Service Complaint Experiences: Implications for Relationship Marketing," Journal of Marketing, 60 (April), 60-76.
Walsh, Gianfranco, Vincent-Wayne Mitchell, and Klaus-Peter Wiedmann (2004), "Segmenting The Online eMaven Information Disseminator," Proceedings of the American Marketing Association, 3.
Weimann, Gabriel (1994), The Influentials, Albany: State University of New York Press.
Weiss, Michael J. (2003), "To Be About To Be," American Demographics, 25 (7), 28-36.
Whyte, William H., Jr. (1954), "The Web of Word of Mouth," Fortune, 50 (November), 140.
Wind, Yoram, Vijay Mahajan, and Robert E. Gunther (2002), Convergence Marketing: Strategies for Reaching the New Hybrid Consumer, Upper Saddle River, NJ: Prentice Hall.
About the Authors
Ronald E. Goldsmith (Ph.D., University of Alabama) is the Richard M. Baker Professor of Marketing at Florida State University. His research interests are consumer innovativeness and scaling in marketing and consumer research. He has published over 200 articles and papers in a variety of marketing and consumer journals and conferences.
David Horowitz is a doctoral student at Florida State University. His specialty is qualitative research methods applied to consumer behavior. He has published in the Journal of the Academy of Marketing Science.
