Journal of Business Research 66 (2013) 801–808

Contents lists available at ScienceDirect

Journal of Business Research

Modeling word-of-mouth usage

William C. Martin a,⁎, Jason E. Lueg b, 1
a
Department of Marketing, College of Business and Public Administration, University of North Dakota, P.O. Box 8366, Grand Forks, ND 58202-8366, USA
b
Department of Marketing, College of Business, Mississippi State University, P.O. Box 9582, Mississippi State, MS 39762, USA

a r t i c l e i n f o a b s t r a c t

Article history: Despite the recognized influence of word-of-mouth (WOM) in the consumer decision making process, research
Received 1 March 2010 investigating how listeners of WOM use this communication is limited. In this paper, the authors present a model
Received in revised form 1 November 2010 which integrates factors influencing listeners' usage of WOM (WOMU) and the consequences of WOMU in
Accepted 1 May 2011
listeners' purchase decisions. Empirical testing of the model indicates that characteristics of both the WOM
Available online 7 July 2011
speaker (trustworthiness, experience, and evidence) and the WOM listener (self-perceived knowledge and
Keywords:
purchase involvement) affect WOMU, as well as whether the WOM is face-to-face or online. The results also show
Interpersonal communication that WOMU strongly relates to attitude toward the recommended product. Implications for retailers and
Social influence marketing researchers are given along with directions for future research.
Adoption © 2011 Elsevier Inc. All rights reserved.
Diffusion

1. Introduction
Word-of-mouth (WOM) has been an important research topic in
marketing literature for over fifty years. Early research indicates that
WOM can have a substantial impact in the marketplace (Whyte, 1954),
especially regarding product adoption and the diffusion process (e.g.,
Arndt, 1967; Brooks, 1957; Coleman, Katz, & Menzel, 1966; Czepiel,
1974; Engel, Kegerreis, & Blackwell, 1969; Holmes & Lett, 1977;
Liu, 2006; Martilla, 1971; Sheth, 1971). WOM can be considerably
more effective than conventional marketing approaches in influencing
consumers' behavior (Katz & Lazarsfeld, 1955; Trusov, Bucklin, &
Pauwels, 2009) and is a very dynamic form of interpersonal commu-
nication which goes far beyond the mere exchange of commercial
information (Kozinets, de Valck, Wojnicki, & Wilner, 2010). Constructs
related to WOM, such as individuals' intention to engage in WOM, are
among the most common dependent variables of interest in marketing
research in recent years (e.g., Blodgett, Granbois, & Walters, 1993;
Bowman & Narayandas, 2001; de Matos & Rossi, 2008; Hartline & Jones,
1996; Maxham, 2001). Further, the rise of electronic word-of-mouth
(eWOM) also garners significant attention as this form of communica-
tion is much more visible and accessible than traditional WOM
(Breazeale, 2009; Godes & Mayzlin, 2004; Phelps, Lewis, Mobilio,
Perry, & Raman, 2004) and may influence hundreds or thousands of
potential customers.
Most existing WOM research investigates how this communica-
tion operates from the perspective of the WOM source or speaker
⁎ Corresponding author. Tel.: + 1 701 777 3929; fax: +1 701 777 2225.
E-mail addresses: wmartin@business.und.edu (W.C. Martin),
jlueg@cobilan.msstate.edu (J.E. Lueg).
1
Tel.: + 1 662 325 7011; fax: +1 662 325 7012.
(Brown & Reingen, 1987; Reingen & Kernan, 1986; Richins &
Root-Shaffer, 1988; Sundaram, Kaushik, & Webster, 1998), the effects
of WOM on the listener (Bone 1995; Herr, Kardes, & Kim, 1991), and
how to generate positive WOM (Higie, Feick, & Price, 1987; Maxham,
2001) while avoiding negative WOM (DeCarlo, Laczniak, Motley, &
Ramaswami, 2007; Richins, 1983). Despite this wealth of research,
there is little research concerning how listeners make use of the
information a source of WOM provides in purchase decisions.
While some researchers assume that positive WOM will invariably
lead to action, such as product purchase, on the part of the listener
(e.g., Maxham, 2001; Reichheld & Sasser, 1990), research indicates
that listeners do not make use of all WOM communications they
receive, but rather that listeners evaluate the value of such commu-
nication according to a number of factors (Dichter, 1966). Currently, as
noted in other research (Sweeney, Soutar, & Mazzarol, 2008), there is a
gap in the marketing literature with regard to understanding what
occurs after an individual provides WOM to a listener and whether and
how that WOM will influence the listener's consumer behavior.
Research in this area can benefit marketers' understanding as to when
WOM is likely to be more influential, as well as how to best link
referrals to customer lifetime value. As such, the purpose of this
research is to develop a framework of the processes occurring between
a positive WOM recommendation and the listener's purchase in-
tentions. Specifically, this study examines what characteristics of the
source of the positive WOM and what characteristics of the listener
influence this process, as well as whether WOM usage operates
differently if the communication is face-to-face or online. Since
research indicates that positive and negative WOM operate differently
at the speaker level, listeners of WOM may respond differently to
positive and negative WOM. Thus, positive WOM is the focus of the
model presented herein.

0148-2963/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jbusres.2011.06.004
802 W.C. Martin, J.E. Lueg / Journal of Business Research 66 (2013) 801–808
The research question addressed here is how listeners use WOM
from a speaker. This study makes three important contributions to the
marketing literature. First, a greater understanding of the role of
WOM communication from the listener's perspective has implications
for retailers and brand and/or product managers attempting to best
target and reach potential consumers. Second, the empirical exam-
ination of multiple factors related to WOM communication concur-
rently allows researchers to better understand the relative impact of
each on WOM listeners. Third, since WOM communication serves a
significant role in consumer socialization, the processes of adoption
and diffusion of innovations, and how market mavens and opinion
leaders spread their influence in the marketplace, the findings of this
study can help marketing researchers better understand the mech-
anisms underlying these factors.
2. Literature review and hypotheses
WOM refers to interpersonal, informal communication about prod-
ucts, which can take the form of goods or services (Godes & Mayzlin,
2004; Liu, 2006; Richins & Root-Shaffer, 1988). Though research
regarding WOM consistently finds that its power in influencing listeners
is great (e.g., Arndt, 1967; Bone, 1995; Dichter, 1966; Sheth, 1971),
knowledge regarding how listeners utilize the information a WOM
source provides is limited. Understanding how consumers make use of
WOM is critical for understanding the link between WOM and the
listener's desire to act on that WOM (e.g. purchase intentions). The
extent to which listeners utilize WOM information from a speaker in
their decision making processes is herein referred to as WOM usage
(WOMU). In a step toward better understanding WOMU, a model of
WOMU is shown in Fig. 1. A discussion of the antecedents and con-
sequences of WOMU, as well as moderators, follows.
2.1. Antecedents of WOMU
While many variables may impact WOMU, the following are those
that have been most prominently and frequently cited in the
interpersonal communication and marketing literature as influencing
listeners of WOM. A thorough review of the WOM literature is beyond
the scope of this study; Kozinets et al. (2010), Duhan, Johnson, Wilcox,
and Harrell (1997), and Breazeale (2009) provide such reviews.
Source Trustworthiness
Source Experience
Positive WOMU
Source Expertise
Evidence
Purchase Involvement Self-Perceived
Knowledge
Fig. 1. Model of positive WOM Usage.
2.2. Source trustworthiness
Source trustworthiness refers to the extent to which an individual's
statements are believed to be genuine (Pornpitakpan, 2004). An integral,
perhaps most vital, element of WOM is that its source can be trusted or
viewed as impartial. This implies that the source does not have a vested
interest in the listener's behavioral response to the communication.
Attribution theory affirms that individuals make causal attributions in
order to understand and account for events (Kelly, 1973). If a listener
ascribes the reason for a product recommendation, for instance, to be the
source's self-serving motives and not for the good of the listener, then the
recommendation is likely to be heavily discounted. As such, marketers
cannot typically serve as sources of WOM since their motive for
communication is usually perceived to be focused on profit, not
consumers' best interests (Bone, 1995). Effective WOM usually takes
place when the speaker is not concerned with whether the listener
engages in a specific behavior as a result of the communication. For
instance, in studying the effects of WOM regarding a new food product,
Arndt finds that “comments tended to reflect exchanges of opinion more
than attempts to control the purchasing actions of the receivers of word of
mouth” (1967, p. 295). Dichter (1966) states that listeners are seriously
concerned with whether they can trust the speaker's comments.
H1. Source trustworthiness is positively related to WOMU.
2.3. Source experience
Source experience refers to the level of a source's familiarity with a
specific topic that comes from actual usage, as perceived by WOM
listeners (Braunsberger & Munch, 1998). Individuals who have engaged
in product trial, for instance, have more experience with the product
than others. Information processing theory affirms that the experience of
information sources is particularly relevant for consumers in making
decisions (Bettman, 1979). In studying consumers' patronage of
automobile diagnostic centers, Engel et al. (1969) find that early patrons
were frequently solicited for information regarding their experience.
Source experience can be important even in low risk purchase decisions.
For instance, a study by Sheth (1971) regarding the adoption of
innovative products reveals that 48% of those who purchased stainless
steel blades used for shaving said their purchase was influenced by a
recommendation from a source who had experience with the blades.
H2. Source experience is positively related to WOMU.
Attitude Toward Product Purchase Intentions
WOM Format
 W.C. Martin, J.E. Lueg / Journal of Business Research 66 (2013) 801–808
2.4. Source expertise
Source expertise is the extent to which listeners of WOM believe a
source to have a high degree of skill in or knowledge of a particular
subject (Braunsberger & Munch, 1998). Those WOM sources who are
trained or possess significant information in an area have greater
expertise than others. Thus, while source experience is dependent on
the source's actual use of the recommended product, source expertise
is dependent on the extent of information and proficiency in the
product category held by the source. Information sources with greater
expertise have greater influence over consumers than do other sources
(Sweeney et al., 2008; Woodside & Davenport, 1974). The expertise of
a WOM source is especially important in the product adoption process,
where consumers are often concerned with minimizing the risk
associated with the purchase of unfamiliar, untried products. For
instance, Leonard-Barton (1985) finds that the opinions of experts
regarding new dental technology significantly influence the attitudes
and adoption decisions of dentists.
H3. Source expertise is positively related to WOMU.
2.5. Evidence
In this context, the evidence of a source refers to the extent to which
a listener of WOM believes that a source can demonstrate the validity of
his or her product related information. For instance, individuals who can
readily show others that their claims regarding a product are legitimate
have greater evidence than those who cannot corroborate their claims
(Reinard 1988). Though seldom studied in the marketing literature, the
evidence provided by WOM sources may be a critical factor in whether a
listener will attend to WOM (Reinard 1988). Indeed, “where the speaker
has at his disposable perceivable proof of a product's efficacy, this fact
may override all other factors of personal and product relations”
(Dichter 1966, p.154).
H4. Evidence of WOM claims is positively related to WOMU.
2.6. Consequences of WOMU
2.6.1. Attitude toward the product
While an in-depth discussion of the attitude literature is beyond
the scope of this paper and the authors refer the reader to Eagly and
Chaiken (1993) and Cialdini, Petty, & Cacioppo (1981) for reviews of
this topic, following is a brief discussion of some of the foundations of
attitude formation. One means put forth for explaining how in-
dividuals' attitudes are formed is expectancy-value models (Eagly &
Chaiken 1993). The attitude toward the object model, a widely
utilized expectancy-value model, affirms that attitudes are a function
of individuals' beliefs that an object possesses specific attributes and
their evaluations of those attributes (Fishbein 1963). Applying
this model to WOMU, when a listener utilizes positive information
regarding a product from a WOM source, the listeners' beliefs that
product purchase will lead to some positive outcome should improve.
This is likely to result in an improved attitude toward the product.
Indeed, research shows that positive WOM results in improved brand
attitudes of listeners (Sundaram & Webster, 1999). As put forth here,
positive WOM does not directly affect the product-related attitudes of
listeners, but rather the listener's use of the information provided by a
positive WOM source is the underlying factor leading to favorable
attitudes toward the recommended product.
H5. WOMU is positively related to attitude toward the product.
2.7. Moderators of WOMU relationships
2.7.1. Purchase involvement
The level of purchase involvement an individual has with a par-
ticular product can alter how he or she makes use of potential WOM
803
sources. When purchase involvement is low, individuals are not as
likely to seek out or attend to information regarding product quality,
features, tradeoffs, or other product aspects. Rather, individuals tend to
make choices regarding low involvement products on the basis of
peripheral cues regarding the product (Petty & Cacioppo, 1986).
However, when individuals' purchase involvement is high, they are
more likely to give consideration to and place importance on the above
referenced factors relating to WOMU. In purchasing a home, for
example, individuals are likely to rely on WOM sources that they
perceive to have knowledge of a particular neighborhood or school
district and give greater consideration to the recommendations of
their friends, whom they trust, than to a real estate agent who has a
vested interest in the individual's purchase decision. In addition, when
individuals are highly involved with a purchase decision, they are
more likely to use recommendations from others in their purchase
decision.
H6. When purchase involvement is high, the relationships between
(a) source trustworthiness, (b) source experience, (c) source
expertise, and (d) evidence with WOMU are stronger.
H7. When purchase involvement is high, the relationship between
WOMU and attitude toward the product is stronger.
2.7.2. Self-perceived knowledge
Self-perceived knowledge refers to “the extent of experience and
familiarity that one has with a product (good or service)” (Duhan et al.,
1997, p. 286). When listeners have prior knowledge of the product
concerned, they are more likely to form their opinions and beliefs on the
basis of their own knowledge, rather than on the basis of information
provided by WOM sources. While WOM communication can strongly
influence product judgments, for instance, this relationship is weaker
when the listener has prior opinions of the topic from memory (Herr
et al., 1991).
Further, a major motivation for individuals to seek out and make
use of WOM communication is to reduce their perceived level of risk
when purchasing products for which they lack familiarity (Arndt,
1967). Consumers' brand attitude ratings and purchase intentions are
more strongly affected by positive WOM when the brand concerned is
unfamiliar to the listener, as opposed to when the brand is familiar
(Sundaram & Webster, 1999). As individuals place greater emphasis
on the value of information related to products for which they are
unfamiliar, the antecedents of WOMU gain importance in influencing
WOMU when their product knowledge is low. Conversely, individuals
who are already familiar with the product concerned in WOM com-
munication are apt to form their opinions and beliefs with sub-
stantially less regard to the recommendations of WOM sources and,
hence, are less liable to utilize a WOM source.
H8. When self-perceived knowledge is high, the relationships
between (a) source trustworthiness, (b) source experience, (c) source
expertise, and (d) evidence and WOMU are weaker.
H9. When self-perceived knowledge is high, the relationship between
WOMU is weaker.
2.7.3. WOM format
Two forms of evidence that WOM speakers can provide to listeners
exist: statistical and narrative evidence (Reinard 1988). Statistical
evidence refers to factual information which is independent of the
speaker, whereas narrative evidence refers to first-hand accounts of
personal experiences. In face-to-face WOM situations, WOM speakers
can often provide both statistical as well as narrative evidence of their
recommendations, of which listeners are apt to at least be implicitly
aware. However, when WOM recommendations occur online, speakers
cannot typically provide statistical evidence. Links to company websites
804 W.C. Martin, J.E. Lueg / Journal of Business Research 66 (2013) 801–808
or other online references might be viewed as biased and not highly
trusted. Since less evidence can be provided online, listeners are apt to
discount the importance of evidence with regard to eWOM. However,
with regard to eWOM, the trustworthiness of the WOM source is the
lynchpin influencing whether a listener will use WOM information in
their purchase decision. In this situation, listeners' trust of the source
may be the only reason they use information from that source.
H10. When WOM is provided via face-to-face interaction, the relation-
ship between evidence and WOMU is stronger.
H11. When WOM is provided via online interaction, the relationship
between source trustworthiness and WOMU is stronger.
3. Method
This study utilizes cross-sectional data analyzed via structural
equations modeling (SEM) as the approach to test the hypotheses.
While an experiment might yield better internal validity, the large
sample sizes needed to investigate the constructs of interest with SEM
(Michon & Chebat, 2008) would be excessive. Further, the critical
incident technique (Bitner, Booms, & Mohr, 1994; Bitner, Booms, &
Tetreault, 1990) was the basis for the survey. Respondents were first
provided with a definition of WOM and are then asked to think back to
the last time someone made a positive recommendation to them
concerning a product. Fifty-four percent of responses concerned the
last face-to-face recommendation made to respondents and forty-six
percent concerned the last online recommendation made to them.
Data were collected using the student referral method (Babin,
Hardesty, & Suter, 2003). Students in marketing courses received extra
credit for participation in the survey. These students were also allowed
to recruit up to four other individuals to participate in the survey,
resulting in the collection of data from 589 individuals. Approximately
10% of respondents were randomly contacted in order to verify the
authenticity of the responses. Six responses had greater than 10%
missing data and were eliminated from further analysis, as were
thirty-seven responses which appeared to display high levels of
acquiescence bias. Thus, the final sample size is 546. Mean respondent
age is 32 and 57% of respondents are female. Self-reported household
income by categories is: 37% in the less than $20,000 category, 13% in the
$20,000 to $39,999 category, 10% in the $40,000 to $59,999 category, 9%
in the $60,000 to $79,999 category, 7% in the $80,000 to $99,999
category, 19% in the $100,000 and over category, and 5% did not respond
with regard to income. Neither gender, nor age, nor income has a
significant effect on any of the variables in the analysis. Participants also
reported what type of product was recommended to them; a wide
variety of products were cited, including electronics (18%), clothing and
shoes (15%), cosmetics (14%), food and beverages (8%), sporting goods
(7%), automobiles and motorcycles (4%), services (4%), games and toys
(3%), books (3%), music and movies (2%), tools (2%), and others (20%).
Source trustworthiness was measured using Pornpitakpan's
(2004) four-item, seven-point semantic differential response scale.
Source experience and source expertise were measured using revised
versions of Braunsberger and Munch's scales (1998); both are three-
item scales which utilize Likert-type items anchored by “Strongly
Disagree” (1) and “Strongly Agree” (7). Attitude toward the product
was measured with a four-item, seven-point semantic differential
response scale (Iyer, 1988). Purchase intentions were measured with a
four-item scale using Likert-type items anchored by “Strongly
Disagree” (1) and “Strongly Agree” (7) (Putrev & Lord, 1994). Purchase
involvement was measured with a six-item, seven-point semantic
differential response scale (Zaichkowsky, 1985). Self-perceived
knowledge was measured using a two-item scale adapted from a
similar measure by Salwen & Dupagne (2001) which utilizes a Likert-
type response scale anchored by “Strongly Disagree” (1) and “Strongly
Agree” (7). No scales existed for either evidence or WOMU. Evidence
was measured by three original items (e.g. “There is proof that this
person's claim regarding this product is authentic.”) and WOMU was
measured with six original items (e.g. “I will rely on the advice given to
me by this person when making my choice.”); each utilizes Likert-type
items anchored by “Strongly Disagree” (1) and “Strongly Agree” (7).
4. Results
In the first stage of measure validation, a principal component
analysis of all nine measures indicated that nine factors have an
eigenvalue greater than one. These factors account for 81.87% of the
variance in the items. After these factors were rotated with Promax
rotation, all items loaded strongly (N.70) on their respective factors
with no significant cross-loadings (b.35).
Next, a confirmatory factor analysis (CFA) using maximum likelihood
estimation of the covariance matrix was conducted. Table 1 contains the
standardized construct correlation matrix of this CFA. In Table 1, the
diagonal values represent the AVE for each construct, values below the
diagonal are correlation estimates, and values above the diagonal are
squared correlation estimates; t-values of correlations are italicized. The fit
of the measurement model is good (χ2 =1222.19, df=492, pb .001;
χ2/df = 2.48; RMSEA = .052; CFI = .96; NNFI = .96). All items have
standardized loadings greater than .70 and average variance extracted
greater than .50, providing evidence of the measures' convergent validity
(Hair, Black, Babin, Anderson, & Tatham, 2006). As shown in Table 1, the
AVE for each pair of constructs is greater than their squared correlation
coefficient, providing evidence of the measures' discriminant validity
(Fornell & Larcker, 1981). All of the measures display excellent reliability
as evidenced by high coefficient alphas (N.80) and mean inter-item
correlations (N.40) (Netemeyer, Bearden, & Sharma, 2003).
Hypotheses one through six were tested by examining the structural
model shown in Fig. 2. This model has good fit to the data (χ2 = 820.44,
df= 287, p b .001; χ2/df = 2.86; RMSEA = .058; CFI= .96; NNFI = .96).
Squared multiple correlations for WOMU, attitude toward the product,
and purchase intentions are .41, .29, and .33, respectively. As hy-
pothesized, trustworthiness (p b .001), source experience (p= .001),
and evidence (p= .003) all have significant positive relationships with
WOMU, supporting H1, H2, and H4, respectively. Though the relation-
ship between source expertise and WOMU is in the hypothesized
direction, this relationship is not significant (p= .119); the data do not
support H3. WOMU is positively related to attitude toward the product
(pb .001), which in turn is positively related to purchase intentions
(pb .001). The results support H5.
The moderating roles of purchase involvement, self-perceived
knowledge, and WOM format were then investigated via multigroup
analyses in SEM. Groups were created for purchase involvement
and self-perceived knowledge via a median split. The results of the
moderation analyses are shown in Table 2.
In examining purchase involvement, a more restricted baseline
model, which constrained all of the relationships to be equal among
both the low and high involvement groups, was investigated. This
baseline model has acceptable fit to the data (χ 2 = 1861.63, df = 638,
p b .001; χ 2/df = 2.92; RMSEA = .084; CFI = .91; NNFI = .91). The
multigroup analysis reveals that purchase involvement does not
moderate the relationships between source trustworthiness, source
experience, source expertise, or evidence with WOMU. The data do not
support H6. Purchase involvement significantly moderates the
relationship between WOMU and attitude toward the product, though
this relationship is stronger in the low purchase involvement group
than in the high purchase involvement group, contradictory to H7.
Next, the moderating role of self-perceived knowledge was exam-
ined. The baseline model has acceptable fit to the data (χ2 = 1840.27,
df= 638, p b .001; χ2/df = 2.88; RMSEA = .083; CFI= .93; NNFI = .92).
The multigroup analyses indicate that when self-perceived knowledge
is low, the relationships between source trustworthiness, source
experience, and evidence with WOMU are significantly stronger.
 W.C. Martin, J.E. Lueg / Journal of Business Research 66 (2013) 801–808 805

Table 1
Standardized construct correlation matrix.

Source Source Source Evidence WOMU Attitude toward Purchase Purchase Self-perceived
trust. experience expertise product intentions involvement knowledge

Source trust .70 .10 .04 .25 .29 .40 .16 .14 .02
Source experience .32 .80 .40 .34 .20 .12 .18 .18 .03
6.27
Source expertise .19 .63 .71 .23 .12 .07 .09 .10 .08
3.76 10.72
Evidence .50 .58 .48 .76 .25 .26 .22 .09 .01
9.16 10.42 8.86
WOMU .54 .45 .34 .50 .79 .26 .20 .12 .01
9.51 8.57 6.79 9.33
Attitude toward product .63 .35 .26 .51 .51 .72 .31 .26 .05
10.42 7.00 5.22 9.44 9.33
Purchase intentions .40 .42 .30 .47 .45 .56 .86 .18 .04
7.77 8.29 6.18 9.11 8.92 10.29
Purchase involvement .38 .27 .31 .30 .34 .51 .42 .78 .16
7.30 5.53 6.25 6.18 6.92 9.29 8.46
Self-perceived knowledge .15 .18 .28 .12 .12 .23 .21 .40 .93
3.21 3.91 5.85 2.63 2.72 4.96 4.60 8.03

There is modest support for the strengthening of the relationship
between source expertise and WOMU when self-perceived knowledge
is low. The findings support H8a, H8b, H8d, and modestly support H8c.
Also, the relationship between WOMU and attitude toward the product
is stronger when self-perceived knowledge is low, supporting H9.
Last, the moderating role of WOM format was examined. The baseline
model fits the data fairly well (χ 2 = 1289.99, df = 638, p b .001;
χ 2/df = 2.02; RMSEA =.061; CFI= .95; NNFI= .95). The multigroup
analysis reveals that the relationship between evidence and WOMU is
stronger when the WOM occurs via face-to-face interaction as opposed to
online interaction, supporting H10. However, no significant difference in
the strength of the relationship between trust and WOMU across the two
WOM formats exists; the data do not support H11. Though hypotheses
were not made concerning the moderating role of WOM format on other
relationships in the conceptual model, these were investigated for
exploratory purposes. This investigation reveals that WOM format does
not moderate the relationship between source experience or source
expertise with WOMU, nor does WOM format moderate the relationship
between WOMU and attitude toward the product.
5. Discussion and implications
The majority of research regarding WOM investigates this form of
interpersonal communication from the standpoint of the speaker,
especially with regard to identifying which factors influence the
extent to which an individual will engage in WOM. The findings of this
study are a step toward bridging the current gap in understanding
what transpires after an individual has engaged in positive WOM.
Specifically, this research provides a framework detailing the means
by which listeners determine whether to utilize positive WOM from a
source and how their use of that positive WOM (a construct termed
WOMU) influences their purchase decision. This study investigates
how factors related to the WOM source, as well as those related to the
listener, affect WOMU.
First, unlike previous research which assumes that positive WOM
directly influences listeners' attitude toward the product and their
purchase intentions, these findings indicate that several characteristics
of the positive WOM source influence listeners' use of positive WOM in
their purchase decision (WOMU). A robust finding from this study is

Source Trustworthiness

.40*

Source Experience .18*

.54* .57*
Positive WOMU Attitude Toward Product Purchase Intentions

.08
Source Expertise

.16*

Evidence

*p-value < .01

Fig. 2. Results of model of positive WOM Usage with standardized path estimates.
806 W.C. Martin, J.E. Lueg / Journal of Business Research 66 (2013) 801–808

Table 2
Results of moderation analyses.

Purchase involvement Standardized β χ2difference

Low purchase involvement (mean = 4.26, n = 252) High purchase involvement (mean = 6.29, n = 294)

Source trustworthiness → WOMU .38⁎⁎⁎ .34⁎⁎⁎ .16

Source experience → WOMU .19⁎⁎⁎ .19⁎⁎⁎ .17
Source expertise → WOMU .04 .05 .22
Evidence → WOMU .19⁎⁎⁎ .14⁎⁎ 1.25
WOMU → Attitude toward product .65⁎⁎⁎ .25⁎⁎⁎ 11.46⁎⁎⁎

Self-perceived knowledge

Low self-perceived knowledge (mean = 3.06, n = 225) High self-perceived knowledge (mean = 5.83, n = 321)

Source trustworthiness → WOMU .49⁎⁎⁎ .30⁎⁎⁎ 6.47⁎⁎

Source experience → WOMU .25⁎⁎⁎ .11⁎ 10.49⁎⁎⁎
Source expertise → WOMU .13⁎ .04 3.18⁎
Evidence → WOMU .27⁎⁎⁎ .08 13.66⁎⁎⁎
WOMU → attitude toward product .69⁎⁎⁎ .37⁎⁎⁎ 4.74⁎⁎

WOM format

Face-to-face (n = 297) Online (n = 249)

Source trustworthiness → WOMU .43⁎⁎⁎ .36⁎⁎⁎ .33

Source experience → WOMU .17⁎⁎⁎ .17⁎⁎ .10
Source expertise → WOMU .10 .07 .33
Evidence → WOMU .23⁎⁎⁎ .09 4.84⁎⁎
WOMU → attitude toward product .57⁎⁎⁎ .49⁎⁎⁎ 1.93
⁎⁎⁎ p-value b .01.
⁎⁎ p-value b .05.
⁎ p-value b .10.

that WOMU strongly influences listeners' attitude toward the product;
when positive WOM is provided to consumers, WOMU accounts for
more than one-fourth of the variance in their attitude toward the
recommended product. The effect of WOMU on listeners' attitude
toward the product is stronger when listeners' purchase involvement is
low and when their self-perceived knowledge of the recommended
product is low. Those who are not highly involved with a particular
purchase are more likely to attend to a WOM source than those who are
involved with the purchase. This finding suggests that WOM regarding
retailers of products which typically elicit low levels of purchase
involvement is likely to have a greater impact on listeners' purchase
decisions than other WOM. Also, information provided by a WOM
source has a greater effect on individuals’ attitude toward the product
when they feel that they have little knowledge regarding the product.
WOM involving retailers selling truly new products is likely to be more
influential than WOM regarding existing products. This finding supports
the assertion of previous research that WOM has a strong influence on
the processes of new product adoption and diffusion.
Second, among the four variables hypothesized to influence
WOMU, the trustworthiness of the WOM source is by far the strongest.
The degree to which listeners trust a WOM source has a stronger effect
on their use of that WOM than the source's experience with the
product, knowledge or skill in the product category, or evidence in
support of his or her recommendation. Further, the trustworthiness of
the WOM source has a greater impact on WOMU when listeners
perceive that they have little knowledge of the product. As early
research regarding WOM notes, listeners do indeed appear to be very
concerned with whether a WOM source truly has their interests in
mind and does not have an ulterior motive in making the recommen-
dation (Arndt, 1967; Dichter, 1966). This finding implies that WOM
recommendations which retailers attempt to employ in their
promotional efforts (e.g. Sussan, Gould, & Weisfeld-Spolter, 2006)
are likely to be discounted heavily when compared to traditional
WOM. While some retailers have recently attempted to encourage
WOM by using employees to spread eWOM, this result indicates that
since consumers tend not to have high levels of trust for marketers,
such efforts are unlikely to have a substantial direct influence on
consumers' purchase decisions. However, such efforts may be fruitful
for retailers that are highly trusted by consumers, such as the Mayo
clinic, which employs individuals to spread eWOM regarding the
organization (Levy, 2009). Most retailers will have the best chance for
success with eWOM if any affiliation with the retailer is not revealed,
though such an action raises ethical issues. In addition, retailers who
reward consumers for engaging in WOM (e.g. Ryu & Feick, 2007) may
be unintentionally compromising the trustworthiness of WOM
speakers. If listeners perceive that a WOM source is recommending a
product for his or her personal gain, the trustworthiness of that source
and listeners' intention to use their information may be seriously
impeded.
Third, the experience of the WOM source with the product also has a
positive influence on WOMU. Surprisingly, the expertise of the source,
referring to the source's skill and training in the product category, does
not lead to greater WOMU. While the two are significantly correlated
with each other, listeners of WOM appear to place substantially greater
weight on the source's actual experience with the product than on his or
her competence in the product category. Listeners of WOM seem to be
more likely to use information from an individual who is familiar with
the specific product being recommended than from a noted authority in
the area. This finding also relates to the comparative influence of opinion
leaders and market mavens. While both consumer types can exert
influence over listeners (e.g. Higie et al., 1987; King & Summers, 1970),
the domain specificity of opinion leaders' experience in a specific
product category suggests that their WOM regarding that product
category has a greater impact on listeners than that of market mavens,
who likely possess more general knowledge. Also, source experience has
a stronger effect on WOMU when listeners' self-perceived knowledge is
low. If listeners feel they already possess substantial information
concerning the product, the experience of the WOM source is less
valuable to them.
Fourth, the evidence that WOM sources can provide to substantiate
their product recommendations also influences listeners' WOMU.
However, this influence is only present when listeners' self-perceived
 W.C. Martin, J.E. Lueg / Journal of Business Research 66 (2013) 801–808
knowledge of the product is low. When listeners of WOM have a high
degree of self-perceived knowledge, they seem unconcerned with the
ability of sources to demonstrate the legitimacy of their recommenda-
tions. Consequently, if most of the consumers in a retailer's target
market feel they have little knowledge regarding a product, such as if the
product is new, the retailer would do well to provide its customers with
some means of demonstrable evidence of their positive experiences
with the retailer's product so as to actively promote WOMU when WOM
occurs. When a retailer's target market has considerable knowledge of
the product, however, such efforts seem likely to be ineffective.
Lastly, while evidence has a stronger effect on WOMU in face-to-face
WOM situations, as compared to eWOM, this study reveals no other
significant differences between these two forms of WOM. Overall, the
process by which listeners determine whether to utilize WOM does not
seem to differ substantially between face-to-face or online WOM.
However, this finding does not discount that eWOM may be more
susceptible to skepticism.
6. Limitations and future research
By using the critical incident technique, data concerning a wide
variety of actual product recommendations were obtained; however,
cross-sectional data cannot be used to provide evidence of causality.
In future research efforts, the experimental manipulation of the
antecedents of WOMU identified in the current study can potentially
provide this evidence. These manipulations could also help to identify if
differences exist between different product categories (e.g., electronics,
clothing, music, etc.), and retail formats (e.g., brick-and-mortar,
bricks-and-clicks, e-tailers). Consideration should be given to how
eWOM usage might be composed of artifacts unique to eWOM. For
instance, the type of eWOM conducted (e.g., email, online forum) may
well influence WOMU (e.g. Kozinets et al., 2010).
While the antecedents of WOMU accounted for 41% of the variance
in this construct, the majority of the variance remains unaccounted for,
implying that constructs not included in the model (e.g., source's social
status, source's attractiveness, tie strength) also influence WOMU.
Seeking out these constructs should be an objective of future research.
Future studies might consider the effects of cross-cultural differences
such as power distance and collectivism on WOMU. Further, future
research should examine how negative WOM is utilized by listeners.
References
Arndt J. Role of product-related conversations in the diffusion of a new product. Journal
of Marketing Research 1967;4:291–5. August.
Babin BJ, Hardesty DM, Suter TA. Color and shopping intentions: the intervening effect
of price fairness and perceived affect. Journal of Business Research 2003;56:
541–51. July.
Bettman JR. An Information Processing Theory of Consumer Choice. Reading:
Addison-Wesley; 1979.
Bitner MJ, Booms BH, Mohr LA. Critical service encounters: the employee's viewpoint.
Journal of Marketing 1994;58:95–106. October.
Bitner MJ, Booms BH, Tetreault MS. The service encounter: diagnosing favorable and
unfavorable incidents. Journal of Marketing 1990;54:71–84. January.
Blodgett JG, Granbois DH, Walters RG. The effects of perceived justice on complainants'
negative word-of-mouth behavior and repatronage intentions. Journal of Retailing
1993;69(4):399–426.
Bone PF. Word-of-mouth effects on short-term and long-term product judgments.
Journal of Business Research 1995;32(3):213–23.
Bowman D, Narayandas D. Managing customer-initiated contacts with manufacturers:
the impact on share of category requirements and word-of-mouth behavior.
Journal of Marketing Research 2001;38:281–97. August.
Braunsberger K, Munch JM. Source expertise versus experience effects in hospital
advertising. Journal of Services Marketing 1998;12(1):23–38.
Breazeale M. Word of mouse: an assessment of electronic word-of-mouth research.
International Journal of Marketing Research 2009;51(3):297–318.
Brooks RC. ‘Word-of-mouth’ advertising in selling new products. Journal of Marketing
1957;22(2):154–61.
Brown JJ, Reingen PH. Social ties and word-of-mouth referral behavior. Journal of
Consumer Research 1987;14:350–62. December.
Cialdini RB, Petty RE, Cacioppo JT. Attitude and attitude change. Annual Review of
Psychology 1981;32:357–404.
807
Coleman J, Katz E, Menzel H. Medical Innovation. A Diffusion Study. New York: Bobbs-Merrill;
1966.
Czepiel JA. Word-of-mouth processes in the diffusion of a major technological
innovation. Journal of Marketing Research 1974;11:172–80. May.
de Matos CA, Rossi CAV. Word-of-mouth communications in marketing: a meta-analytic
review of the antecedents and moderators. Journal of the Academy of Marketing
Science 2008;36(4):578–96.
DeCarlo TE, Laczniak RN, Motley CM, Ramaswami S. Influence of image and familiarity
on consumer response to negative word-of-mouth communication about retail
entities. Journal of Marketing Theory and Practice 2007;15(1):41–51.
Dichter E. How word-of-mouth advertising works. Harvard Business Review 1966;44:
147–57. November-December.
Duhan DF, Johnson SD, Wilcox JB, Harrell GD. Influences on consumer use of word-of-
mouth recommendation sources. Journal of the Academy of Marketing Science
1997;25(4):283–95.
Eagly AH, Chaiken S. The Psychology of Attitudes. Orlando, FL: Harcourt Brace
Jovanovich College Publishers; 1993.
Engel JF, Kegerreis RJ, Blackwell RD. Word-of-mouth communication by the innovator.
Journal of Marketing 1969;33:15–9. July.
Fishbein M. An investigation of the relationships between beliefs about an object and
the attitude toward that object. Human Relations 1963;16:233–40.
Fornell C, Larcker DF. Evaluating structural equation models with unobservable variables
and measurement error. Journal of Marketing Research 1981;18:39–50. February.
Godes D, Mayzlin D. Using online conversations to study word-of-mouth communication.
Marketing Science 2004;23(4):545–60.
Hair JF, Black WC, Babin BJ, Anderson RE, Tatham RL. Multivariate Data Analysis. 6th ed.
Upper Saddle River: Pearson Prentice Hall; 2006.
Hartline MD, Jones KC. Employee performance cues in a hotel service environment:
influence on perceived service quality, value, and word-of-mouth intentions.
Journal of Business Research 1996;35(3):207–15.
Herr PM, Kardes FR, Kim J. Effects of word-of-mouth and product attribute information
on persuasion: an accessibility-diagnosticity perspective. Journal of Consumer
Research 1991;17:454–62. March.
Higie RA, Feick LF, Price LL. Types and amount of word-of-mouth communications about
retailers. Journal of Retailing 1987;63(3):260–78.
Holmes JH, Lett JD. Product sampling and word of mouth. Journal of Advertising
Research 1977;17(5):35–40.
Iyer ES. The influence of verbal content and relative newness on the effectiveness of
comparative advertising. Journal of Advertising 1988;17(3):15–21.
Katz E, Lazarsfeld PF. Personal Influence. Glencoe, IL: Free Press; 1955.
Kelly HH. The process of causal attribution. The American Psychologist 1973;28:107–28.
King CW, Summers JO. Overlap of opinion leadership across consumer product
categories. Journal of Marketing Research 1970;7(1):43–50.
Kozinets RV, de Valck K, Wojnicki AC, Wilner SJS. Networked narratives: understanding
word-of-mouth marketing in online communities. Journal of Marketing 2010;74:
71–89. March.
Leonard-Barton D. Experts as negative opinion leaders in the diffusion of a
technological innovation. Journal of Consumer Research 1985;11:914–26. March.
Levy P. Get the word out. Marketing News 2009;43(5):16–8.
Liu Y. Word of mouth for movies: its dynamics and impact on box office revenue.
Journal of Marketing 2006;70:74–89. July.
Martilla JA. Word-of-mouth communication in the industrial adoption process. Journal
of Marketing Research 1971;8:173–8. May.
Maxham JG. Service recovery's influence on consumer satisfaction, positive word-of-
mouth, and purchase intentions. Journal of Business Research 2001;54(1):11–24.
Michon R, Chebat JC. Breaking open the consumer behavior black box: SEM and retail
atmospheric manipulations. Journal of Marketing Theory & Practice 2008;16:
299–307. Fall.
Netemeyer RG, Bearden WO, Sharma S. Scaling Procedures: Issues and Applications.
Thousand Oaks, CA: Sage Publications, Inc.; 2003.
Petty RE, Cacioppo JT. The elaboration likelihood model of persuasion. In: Leonard
Berkowitz, editor. Advances in Experimental Social Psychology, Vol. 19. New York:
Academic Press; 1986. p. 123–205.
Phelps JE, Lewis R, Mobilio L, Perry D, Raman N. Viral marketing or electronic
word-of-mouth advertising: examining consumer responses and motivations to
pass along email. Journal of Advertising Research 2004;44:333–48. December.
Pornpitakpan C. The effect of Americans' adaption to Malaysians on perceived
trustworthiness. Journal of International Consumer Marketing 2004;16(3):7–23.
Putrev S, Lord KR. Comparative and noncomparative advertising: attitudinal effects under
cognitive and affective involvement conditions. Journal of Advertising 1994;23:77–91.
June.
Reichheld FF, Sasser Jr WE. Zero defections: quality comes to services. Harvard Business
Review 1990:105–11. September-October.
Reinard JC. The empirical study of the persuasive effects of evidence: the status after
fifty years of research. Human Communications Research 1988;15(1):3–59.
Reingen PH, Kernan JB. Analysis of referral networks in marketing: methods and
illustration. Journal of Marketing Research 1986;23:370–8. November.
Richins M. Negative word-of-mouth by dissatisfied customers: a pilot study. Journal of
Marketing 1983;47:68–78. Winter.
Richins M, Root-Shaffer T. The role of involvement and opinion leadership in consumer
word-of-mouth: an implicit model made explicit. Advances in Consumer Research
1988;15(1):32–6.
Ryu G, Feick L. A penny for your thoughts: referral reward programs and referral
likelihood. Journal of Marketing 2007;71:84–94. January.
Salwen MB, Dupagne M. Third-person perception of television violence: the role of
self-perceived knowledge. Mediapsychology 2001;3:211–36.
808 W.C. Martin, J.E. Lueg / Journal of Business Research 66 (2013) 801–808
Sheth JN. Word-of-mouth in low-risk innovations. Journal of Advertising Research
1971;11(3):15–8.
Sundaram DS, Kaushik M, Webster C. Word-of-mouth communications: a motivational
analysis. Advances in Consumer Research 1998;25(1):527–31.
Sundaram DS, Webster C. The role of brand familiarity on the impact of word-of-mouth
communication on brand evaluations. Advances in Consumer Research 1999;26(1):
664–70.
Sussan F, Gould S, Weisfeld-Spolter S. Location, location, location: the relative roles of
virtual location, online word-of-mouth (ewom) and advertising in the new product
adoption process. Advances in Consumer Research 2006;33(1):649–50.
Sweeney JC, Soutar GN, Mazzarol T. Factors influencing word of mouth effectiveness:
receiver perspectives. European Journal of Marketing 2008;42(3/4):344–64.
Trusov M, Bucklin RE, Pauwels K. Effects of word-of-mouth versus traditional
marketing: findings from an internet social networking site. Journal of Marketing
2009;73:90–102. September.
Whyte Jr WH. The web of word of mouth. Fortune 1954;50:140–3. November.
Woodside AG, Davenport Jr W. The effect of salesman similarity and expertise on consumer
purchasing behavior. Journal of Marketing Research 1974;11:198–202. May.
Zaichkowsky JL. Measuring the involvement construct. Journal of Consumer Research
1985;12:341–52. December.