Nwom15 http://www.cheapassignmenthelp.co.uk/

Full length article
“Understanding a fury in your words”: The effects of posting and
viewing electronic negative word-of-mouth on purchase behaviors
Su Jung Kim a, *
, Rebecca Jen-Hui Wang b
, Ewa Maslowska c
, Edward C. Malthouse c, d
a
Greenlee School of Journalism and Communication, Iowa State University, 116 Hamilton Hall, Ames, IA 50011, USA
b
Kellogg School of Management, Northwestern University, 2001 Sheridan Road, Evanston, IL 60208, USA
c
Medill IMC Spiegel Research Center, Medill School of Journalism, Media and Integrated Marketing Communication, Northwestern University, 1845
Sheridan Road, Evanston, IL 60208, USA
d
McCormick School of Engineering, Northwestern University, 2145 Sheridan Road, Evanston, IL 60208, USA
a r t i c l e i n f o
Article history:
Received 29 May 2015
Received in revised form
14 August 2015
Accepted 18 August 2015
Available online xxx
Keywords:
Word-of-mouth (WOM)
Complaint
Company usefulness
Company apology
Webcare
Propensity score matching
a b s t r a c t
Marketing scholars and practitioners have long recognized that the power of electronic negative word-
of-mouth (e-NWOM) can influence brand revenues and firm performance, but most previous studies
have only examined the effect of viewing. This study is one of the initial attempts to test the effects of e-
NWOM on both posters and viewers. We also test the moderating effects of company usefulness and
company apology in a separate study. Using an observational dataset that contains NWOM viewing and
posting records and customers' purchase transactions from a real company, Study 1 finds that viewing e-
NWOM has a negative effect on subsequent purchases, whereas posting e-NWOM has a positive inter-
action effect with company usefulness. Study 2 shows that a company's public apology has a positive
effect on viewers, but not posters. We conclude with the theoretical, methodological, and managerial
implications of e-NWOM and webcare research.
© 2015 Elsevier Ltd. All rights reserved.
1. Introduction
The study of word-of-mouth (WOM) is not new (Richins & Root-
Shaffer, 1988; Verhagen, Nauta, & Feldberg, 2013; Westbrook,
1987), but recently it has attracted much attention from scholars
and practitioners with the advancement of Web 2.0 and subsequent
development of social media platforms. Despite the recognition
that WOM can influence brand awareness, attitude, or purchase
decision, traditional WOM research has faced challenges in tracking
WOM due to its interpersonal and ephemeral nature. In the era of
Web 2.0, consumers are able to voice opinions about a brand on
various digital media platforms such as brand websites, personal
blogs, social media, or third-party review sites where their com-
ments are read by other consumers in real time (Pitt, Berthon,
Watson, & Zinkhan, 2002). The same technological environment
allows scholars to collect and analyze the actual WOM data from
these media channels and estimate their impact on consumers who
participate in WOM activities.
The proliferation of electronic WOM (e-WOM) has transformed
the ways in which marketing communication has traditionally
operated as one-way communication from companies to con-
sumers via mass communication channels (Campbell, Pitt, Parent,
& Berthon, 2011). Consumers no longer solely rely on ad mes-
sages to obtain brand information and make a purchase decision
(Edelman, 2010). User-generated content, e-WOM, or online con-
versations between companies and consumers can all influence
consumer behavior due to easy accessibility of these messages. A
Nielsen report (2013) finds that consumer opinions posted online
are trusted more than ads delivered via mass media. More than two
thirds of consumers also report that they take action after reading
other consumers' e-WOM messages, showing its potential as a new
form of social influence that impacts consumer trust and behavior.
E-WOM poses new challenges to companies since they cannot
control its creation and dissemination. In particular, the power of e-
WOM is amplified when consumers share negative feelings and
thoughts publicly online after experiencing dissatisfaction with a
product or service. Many anecdotes have shown that a brand's image
can be tainted by a single negative piece of user-generated content
that gets propagated (e.g., “United Breaks the Guitars”). Dissatisfied
customers express electronic negative WOM (e-NWOM) on com-
pany websites, online review sites, third-party complaint sites or
* Corresponding author.
E-mail addresses: sjkim@iastate.edu (S. Kim), r-wang@kellogg.northwestern.edu
(R.J.-H. Wang), ewa.maslowska@northwestern.edu (E. Maslowska), ecm@
northwestern.edu (E.C. Malthouse).
Contents lists available at ScienceDirect
Computers in Human Behavior
journal homepage: www.elsevier.com/locate/comphumbeh
http://dx.doi.org/10.1016/j.chb.2015.08.015
0747-5632/© 2015 Elsevier Ltd. All rights reserved.
Computers in Human Behavior 54 (2016) 511e521

social media channels. Research has shown that unfavorable mes-
sages posted on these online outlets can negatively influence con-
sumers' attitudes and behaviors (Bickart & Schindler, 2001;
Chevalier & Mayzlin, 2006; Davis & Khazanchi, 2008; Doh &
Hwang, 2009; Liu, 2006; Park & Lee, 2009; Reichheld, 2003).
On the flip side, other case studies have shown that prompt
responses to e-NWOM crises can stimulate advocacy and
contribute to regaining trust and satisfaction from customers (e.g.,
the Customer Bill of Rights by JetBlue or the Pizza Turnaround by
Domino's). Online feedback mechanisms serve as an effective and
cost-efficient way for companies to rebuild their online reputation
during such crises. By actively searching for consumers' complaints
and providing a remedy for a cause of dissatisfaction, firms can
prevent the subsequent spread of e-NWOM or even turn negative
sentiments to positive ones, as shown in recent literature on
webcare (Lee & Song, 2010; van Noort & Willemsen, 2011).
This paper concerns the power of e-NWOM and analyzes the
effect of e-NWOM on purchase decisions. At the same time, it pays
attention to the moderating role of the relationship between con-
sumers and firms when e-NWOM exerts influence on consumers.
This study makes three substantive contributions by addressing
limitations of previous research on e-NWOM. First, most studies on
e-NWOM have focused on the effect of reception to e-NWOM, not
the expression of it (Verhagen et al., 2013). Recognizing the dearth
of research on the influence of e-WOM messages on their senders,
this study examines the effects of e-NWOM on its senders and re-
ceivers separately. In addition, we identify possible moderators of
e-NWOM effects, both of which are driven from the perception
about or the reaction of the given firm that creates the e-NWOM
incident. We include these moderators in two separate studies.
Second, despite the availability of WOM data, few organizations can
link it to individual customers' purchases or vice versa, so it is
challenging for them to establish a direct link between e-NWOM
behavior and purchase behavior. For example, researchers can
easily access e-WOM data on Facebook or Twitter, but have no way
of linking these data with purchase transaction data. Retailers and
service providers have extensive records on customer purchases,
but cannot easily match them to social media data. In Study 1, we
use a single-source dataset that combines e-NWOM data from a
real company's online community with actual purchase data of the
same customers who engaged in e-NWOM activities. This unique
observational dataset allows us to estimate the financial impact of
e-NWOM empirically on the customer level. Third, there are only a
few studies that test the effects of webcare in response to e-NWOM.
The question of how e-NWOM posters and viewers might react
differently to company's corrective action has not been thoroughly
investigated. In Study 2, we conduct an experiment to test the effect
of company apology to e-NWOM senders and receivers using a
similar negative incident of Study 1 (i.e., a policy change) in a
different industry context.
In the following sections, we review the previous research on e-
NWOM, focusing on online complaining behavior and suggest a
conceptual model that explains the effects of e-NWOM on its
senders and receivers. In addition to the differential effects on
senders and receivers, we include two moderators of e-NWOM ef-
fects. Next, we present our two studies, empirical findings, and their
respective conclusions. Finally, we summarize contributions and
limitations of our study, followed by suggestions for future research.
2. Literature review
2.1. E-WOM, valence, and online public complaining behavior
E-WOM is defined as “any positive or negative statement made
by potential, actual, or former customers about a product or
company, which is made available to a multitude of people and
institutions via the Internet” (Hennig-Thurau, Gwinner, Walsh, &
Gremler, 2004, p. 39). E-WOM is distinguished from traditional
WOM in that it is (1) mediated in the form of reading or writing, (2)
presented in public forums for other consumers or companies to
observe, and (3) electronically stored and can be searched for future
use (Andreassen & Streukens, 2009).
Previous research on e-WOM has found that it influences brand
awareness (Davis & Khazanchi, 2008), brand attitude (Doh &
Hwang, 2009), purchase intention (Bickart & Schindler, 2001;
Park & Lee, 2009), product sales (Chevalier & Mayzlin, 2006;
Davis & Khazanchi, 2008; Liu, 2006), as well as revenue growth
(Reichheld, 2003). Depending on its valence, WOM can affect
customer loyalty or firm revenues either negatively or positively
(Dellarocas, Awad, & Zhang, 2004; East, Hammond, & Lomax, 2008;
Liu, 2006). For instance, positive WOM (PWOM) enhances expected
quality and brand attitude, and leads to recommendation for
product purchases, whereas NWOM elicits product denigration,
rumor, private complaining, and ultimately diminishes purchase
intentions and sales (Chevalier & Mayzlin, 2006; Huang & Chen,
2006; Mizerski, 1982). Literature also suggests that the impact of
NWOM on decreasing sales is greater than the impact of PWOM on
increasing sales (Mittal, Ross, & Baldasare, 1998; Park & Lee, 2009).
Another stream of research that draws special attention to
NWOM is consumer complaint behavior (CCB). Supplementing
Hirschman's (1970) classic CCB model of exit (e.g., switching
brands), voice (e.g., making a complaint to the seller), and loyalty
(e.g., continuing to purchase from a dissatisfying seller), recent
studies incorporate NWOM as a new type of CCB (Goetzinger, 2007;
Singh, 1990). They also distinguish between CCB that occurs in a
private setting where customers tell others about unsatisfactory
experiences (e.g., traditional NWOM), and a public setting where
customers express NWOM to broader audiences (e.g., e-NWOM). In
particular, studies have shown that online public complaining
behavior has an aggravating effect on firm performance (Gregoire,
Tripp, Legoux, 2009; Lee Song, 2010).
Scholars have emphasized the need to categorize those who
exhibit complaint behaviors (Singh, 1990) or their response styles
(Schoefer Diamantopoulos, 2009) to better handle complaint
situations. In an online context, Lee and Song (2010) classify con-
sumers who display online complaint behaviors into complainers,
repliers, and observers. Despite a vast number of studies on ob-
servers, little research has examined how making online com-
plaints affects the attitudes or behaviors of complainers (i.e.,
customers who create e-NWOM messages). Verhagen et al. (2013)
point out the dearth of “sender-oriented” studies in e-WOM
research. This study extends this line of research by separating the
effects of e-NWOM on senders (i.e., e-NWOM posters) and receivers
(i.e., e-NWOM viewers). We also suggest two potential moderators
of e-NWOM effects that are relevant to relationship management
prior to or during the e-NWOM incident. Fig. 1 provides our con-
ceptual framework that explains the effect of e-NWOM posting and
viewing behavior as well as moderators identified in previous
research, which we elaborate in the following sections.
2.2. Differentiating the effect of posting and viewing e-NWOM:
Posting e-NWOM
According to social sharing of emotion theory (Rime, 2009),
humans have a natural tendency to share emotional experiences
with others. The social environment in which humans live moti-
vates them to express emotions to people around them in order to
seek help and support, vent, bond, or get validation. The same
tendency can be observed in the context of negative consumption
experiences. Research on consumption emotions has shown that
S.J. Kim et al. / Computers in Human Behavior 54 (2016) 511e521512

negative emotions triggered by unsatisfactory consumption expe-
riences increase the likelihood to produce e-NWOM (Ladhari, 2007;
Maute Dubes, 1999; Riegner, 2007; S€oderlund Rosengren,
2007). Consumers engage in expressing NWOM for various rea-
sons, including: (1) preventing others from experiencing the same
problem that they had encountered, (2) seeking advice on how to
solve their problems, (3) venting their anger through NWOM as a
way of reducing cognitive dissonance, or (4) retaliating against the
offering company (Hennig-Thurau et al., 2004; Sundaram, Mitra,
Webster, 1998).
The question then arises as to what happens to those who share
their negative experience publicly on the Internet. Regarding the
effect of posting, we draw from cognitive dissonance theory
(Festinger, 1957). Cognitive dissonance occurs when people are
confronted with inconsistent attitudes or beliefs. The negative
intrapersonal state thereby motivates them to reduce the aversive
psychological state. Existing customers may experience cognitive
dissonance when companies fail to provide a product or a service
that meets their expectations, or when they see concerns sur-
rounding the general business conduct of ﬁrms (Andreassen
Streukens, 2009). One way to reduce such dissonance is to openly
deliberate a negative experience and announce how they might
behave in response to companies' failings. However, publicly
revealing their negative feelings can bring about two opposite con-
sequences. As presented in Fig. 1, we identify two competing theo-
retical explanations e self-prophecy and catharsis through venting e
that may impact customers' subsequent purchase behavior.
The self-prophecy approach (Sherman, 1980; Spangenberg
Giese, 1997) posits that engaging in pre-behavioral cognitive
work of stating a sequence of behavioral intentions leads people to
become committed to what they had stated. Thus, after customers
elaborate on their negative feelings and thoughts about a company
by posting e-NWOM, their negative attitudes and opinions serve as
a cognitive frame for their future purchases from the company
(Nyer Gopinath, 2005). Prior studies provide additional support
that creating a message results in a deeper understanding and long-
term recollection of the subject matter (Chi, De Leeuw, Chiu,
Lavancher, 1994; Nekmat, 2012), which reinforces the pre-
existing attitudes of the message creator (Prislin et al., 2011). The
same conclusion has been reported about revenge (del Río-Lanza,
Vazquez-Casielles, Díaz-Martín, 2009). Customers who create
e-NWOM go through a process of self-prophecy by elaborating
their unsatisfying experience with a company and thereby
strengthening their negative attitudes and opinions, which de-
creases their future purchases from the given company, as sug-
gested using a minus sign in Fig. 1.
On the contrary, the venting approach provides a competing
explanation of the posting effect in that it focuses on the power of
emotional release through venting (Barclay Skarlicki, 2009;
Blodgett, Hill, Tax, 1997; Hennig-Thurau et al., 2004). Creating
messages in a stressful situation has been shown to reduce the
intense emotion of the message creator in various contexts
(Pennebaker, 1997). Barclay and Skarlicki (2009) show that venting
about workplace injustice reduces intentions to retaliate and in-
creases the levels of psychological well-being and personal reso-
lution. Nyer and Gopinath (2005) ﬁnd that the emotional release
from complaining behavior reduces customer dissatisfaction. Those
in favor of venting maintain that expressing e-NWOM alleviates the
negative reactions of those who share their unsatisfying experience
with a company because of the catharsis that the venting behavior
provides (Berger, 2014). The expected positive impact of venting is
suggested using a plus sign in Fig. 1.
Fig. 1. Conceptual framework.
S.J. Kim et al. / Computers in Human Behavior 54 (2016) 511e521 513

2.3. The moderating effect of company usefulness on posting e-
NWOM
The aforementioned effects of posting e-NWOM can be
moderated by the nature of the relationship between the firm and
the poster (Chiou, Hsu, Hsieh, 2013; Gregoire et al., 2009). Ac-
cording to Aggarwal (2004), there are two types of relationships
between brands and their consumers, which have their respective
norms of behavior: First, exchange relationships are based on the
assumption that those who provide benefits to the other party
expect to receive something in return. On the contrary, in
communal relationships the motivation is altruistic. People provide
benefits because they care for the other party, not because they
expect something in return. We expect that posters who have an
existing relationship with a firm and perceive the firm as useful are
more likely to engage in e-NWOM posting behavior because they
see the value of maintaining the relationship with the brand and
want to help the company (Hennig-Thurau et al., 2004; Sundaram
et al., 1998; Verhagen et al., 2013). Given that those who are vocal
about a brand already have a higher level of brand commitment
(Kim, Sung, Kang, 2014), consumers who participate in a firm's
online community maintain a communal relationship with the
firm. When an e-NWOM crisis happens, they post about their
negative experience to make the firm aware of an issue so that it
can handle it promptly and properly.
In this vein, Verhagen et al. (2013) test the moderating effect of
company usefulness on intentions to switch and repatronage. They
hypothesize a positive interaction effect between posting e-NWOM
and company usefulness on intention to repatronage and a negative
interaction effect on intention to switch to a competing brand. They
fail to detect a significant effect of company usefulness presumably
due to the context of their study where the high levels of empow-
erment towards the company and social connection among forum
members do not allow a customer with a desire to help the company
to speak out and post information that may seem empathetic to the
firm. In addition, we argue that the operationalization of company
usefulness in the study may not be suitable to measure the level of
company usefulness. In their study, company usefulness is oper-
ationalized as the extent to which consumers agree that their
messages contribute to the development, improvement, effective-
ness, and operation of the company. This captures the intention to
help the company, but does not indicate whether consumers see the
company and their relationship as useful.
In this study, we define company usefulness as the extent to
which consumers perceive a company's product or service provides
value to them. Instead of measuring intentions to help the company
as an indicator of company usefulness, we operationalize it as
whether a consumer had an opportunity to assess the value of the
company. If such an opportunity motivates a consumer to cogni-
tively process the benefits of the firm, this will result in a positive
moderating effect on e-NWOM posters, which is proposed using a
plus sign in Fig. 1. Study 1 tests the moderating role of company
usefulness by using a real company's community website where
visitors are existing customers of the company. Thus, customers
who maintain a good relationship with the company and desire to
help it can post e-NWOM without the fear of going against the
community norm. In other words, if a customer posts NWOM and
experience the value of a product or a service and perceive its
usefulness after he or she posts, we expect that his or her subse-
quent purchases will increase.
H1. Company usefulness has a positive moderating effect on the
relationship between posting e-NWOM and subsequent purchases
because it amplifies a positive effect of venting and buffers a
negative effect of self-prophecy.
2.4. Differentiating the effect of posting and viewing e-NWOM:
Viewing e-NWOM
As mentioned earlier, Lee and Song (2010) classify consumers
who engage in online complaint behavior into complainers, re-
pliers, and observers. Observers, who only read other consumers'
negative posts, evaluate the given brand by perusing negative in-
formation on the site where complaints are posted. They are less
likely to take any action until they recognize that there is an issue
relevant to them. Attribution theory provides an explanatory
framework for these recipients of e-NWOM messages. According to
this theory, causal analysis is inevitable in an individual's need to
understand social events and decide which actions to take (Keller,
2007). In our context, viewers (i.e., observers) are motivated to
process e-NWOM messages in order to make judgments about the
causes that trigger the creation of e-NWOM by other consumers.
Depending on how they attribute the causedwhether it is the
company, the poster, or other circumstancesdsuch attributions
influence subsequent actions they take toward the company.
Laczniak, DeCarlo, and Ramaswami (2001) specified three types
of information that consumers use to make causal attributions: (1)
consensus (i.e., the extent to which other consumers agree with the
negative views of the poster), (2) distinctiveness (i.e., whether the
negative information is associated with a particular brand or with
other brands), (3) consistency (i.e., the extent to which a poster is
stable in his position across time and situations). Consumers blame
the company when the information they receive is high on all three
dimensions, and the communicator when the dimensions are low,
concluding that the problem is unique to the complainer.
Both Study 1 and 2 examine NWOM engendered due to firm
policy changes that influence all the existing customers. The con-
texts in both studies present the case in which all three dimensions
of attribution are considered high. It is high consensus because the
policy change impacts all customers of the company, which elicit
similar negative reactions on the community site. It is high
distinctiveness since the website is exclusively for customers of the
given brand. It is high consistency because the tone of the posts was
uniformly negative for those who posted multiple times. In sum, we
expect consumers to attribute the cause of e-NWOM to the com-
pany and blame its decision to change the company policy. We also
expect high consensus and consistency conditions to create a sense
of social norm (Lee, Park, Han, 2008). Thus, viewers will conform
to the social norm of the community site and form negative atti-
tudes toward the firm, which will negatively affect future purchase
decisions (Duan, Gu, Whinston, 2008). Accordingly, we present a
minus sign in Fig. 1 and formally propose the following hypothesis.
H2. An e-NWOM viewer decreases his or her subsequent pur-
chases after he or she reads e-NWOM in high levels of consensus,
distinctiveness, and consistency.
3. Study 1
3.1. Data collection
The data for Study 1 come from a large coalition loyalty program
that maintained an online community forum. Members earn points
for purchases at sponsors in different product categories including
groceries, gas, pharmacy, and credit card. They can then use the
points to redeem a variety of rewards such as gift cards,
merchandise, and travel. In November 2011, the program
announced a policy changedearned points would expire after five
yearsdthat affected all members and triggered them to post or
read NWOM on their community site and other social media sites.
For this study, the company provided point accrual and redemption

records of 76 posters who created 116 total messages, 681 viewers,
and a control group of 10,000 customers who neither posted nor
viewed. When members make a purchase at a sponsor location and
earn points, the loyalty program receives a payment from the
sponsor. Thus, point accumulation is directly related to the loyalty
program's revenues. The policy change is an exogenous shock that
prompts customers to post or view messages on the online forum,
and the resulting data offer a unique opportunity to measure the
effect of both posting and viewing NWOM on actual purchase be-
haviors. The study period is 15 weeks. The four weeks prior to the
policy change announcement establish existing purchase charac-
teristics prior to posting or viewing. The 11 weeks after the
announcement are used to evaluate the behaviors after posting or
viewing, relative to control customers who neither posted nor
viewed.
3.2. Variables
Our independent variables are posting and viewing e-NWOM.
Posters are defined as those who posted at least once during the 11-
week period after the announcement. Viewers are those who read
any of the messages at least once during the 11-week period. The
moderating variable, company usefulness, is whether a member
experienced the brand value after posting or viewing. We oper-
ationalize company usefulness with a binary variable that indicates
whether a member redeemed a reward after the policy change.
Members accumulate points with the goal of redeeming them for
various rewards. Thus, redemption reminds them of the usefulness
of the company. Our dependent variable is point accumulation,
which is proportional to the revenues received from the sponsors
where customers earn points. We aggregate point transactions to
the weekly level because point accrual is somewhat periodic on a
weekly cycle, with two of the major sponsor categories being gro-
cery and gas.
3.3. Propensity score model
Before testing our hypotheses with regressions models, we
address the possibility of selection bias, where posters or viewers
are different than controls prior to the announcement. Such dif-
ferences could affect members' forum participation as well as their
subsequent purchase behaviors. To reduce the possibility of such
confounding, we employ a propensity score model (Rosenbaum
Rubin, 1983; Rubin, 1997) to identify matched controls. Matching
is done by first estimating the probability Pi that member i self-
selects into the treatment group (posting or viewing) using only
on information that was available prior to the announcement, and
then by finding control-group members with similar values of Pi.
We use the following covariates in the model: vector mi records the
points accrued in bank services, grocery, pharmacy, retail, or other
categories, vector ri records points used to redeem gift and cash
certificates, goods, or other categories of rewards, and ni is the
number of sponsors the customer purchased from. To obtain each
customer's propensity score we use a binary logistic regression:
ln

Pi
1ÀPi

¼ l0 þl1lnðmi þ1Þ þl2lnðri þ1Þþl3lnðni þ1Þþεi;
(1)
The logarithmic transformation is used because the predictors
may be skewed to the right with outliers. After obtaining the pro-
pensity scores, we employ 5:1 matching with the nearest neighbor
algorithm to select a group of control customers that resembles the
posters/viewers before the announcement. After matching, we
verify that covariate balance is improved by comparing
quantileequantile plots of all predictor variables before and after
matching. The propensity score distributions are more similar with
those of the matched control groups than the full control sample.
Final sample consists of 76 posters, 681 viewers, and 3784 matched
controls.
After finding matches, we estimate a regression model to test
the hypotheses:
lnðyit þ 1Þ ¼ða0 þ a0Þ þ b1xit þ b2vit þ b3wit þ b4xit Â wit
þ b5vit Â wit þ b6qi þ b7ki þ b8 lnðst þ 1Þ
þ b9 lnðmi þ 1Þ þ b10 lnðri þ 1Þ þ b11ðni þ 1Þ
þ b12 bpi þ eit;
(2)
where yit is the points accrued by customer i from purchases made in
week t, which is directly related to the company's revenue from the
issuing sponsor. Values of t ranging from one to four represent the
weeks prior to the announcement of the policy change, and five to
fifteen for the weeks after the announcement. Binary variables vit
and xit indicate whether customer i viewed or posted during or after
week t, respectively. We control for the main effect of redeeming
rewards after the policy change announcement (wit). We also
include interaction effects between posting and redeeming rewards
after the policy change (xit Â wit), and viewing and redeeming re-
wards after the policy change (vit Â wit). To test H1, we observe the
direction and magnitude of the coefficient for the interaction be-
tween posting and experiencing the brand benefit, i.e., b4. To test H2,
we observe the sign and magnitude of b2. We account for customer
heterogeneity by including covariates measuring the preexisting
characteristics used in the propensity score model. To account for
any remaining unobserved heterogeneity, we include the random
intercept (ai). We also control for the fixed effects of whether a
subject is a poster or viewer, denoted by qi and ki, respectively, to
account for a priori heterogeneity amongst posters, viewers, and
control customers. Lastly, to account for seasonality, we include the
weekly average st of accrued points from all 10,000 control cus-
tomers who have neither posted nor viewed the online forum.
3.4. Results of Study 1
First, we estimate the propensity score model, which predicts the
likelihoodthatamemberposted orviewede-NWOMmessagesonthe
forum. The estimates suggest that posting is associated with accruing
morepoints fromfood(bb ¼ 0.06,p.05),retail(bb ¼ 0.19,p.001),and
other (bb ¼ 0.41, p .001) categories in the pre-announcement period.
They are also more likely to redeem rewards (bb ¼ 0.07, p .05). The
results suggest that forum visitors are more engaged with the firm to
beginwith.Thisfindingalsoconfirmsthatmatchinge-NWOMposters
and viewers with a control group that exhibits similar a priori be-
haviors is necessary to reduce selection bias.
The regression specified in Equation (2) quantifies the effect of
our main independent variables, as shown in Table 1. We test H1 by
observing the coefficient estimate of the interaction between
posting and redeeming. The regression estimate of the interaction
is positive and significant on subsequent spending (bb ¼ 0.31,
p .05), supporting H1. The interaction effect is illustrated in Fig. 2.
The estimate for viewing behavior is negative and significant,
suggesting that viewing e-NWOM decreases subsequent spending
by 12% (p .001). Thus, H2 is supported.
3.5. Discussion of Study 1
Our first hypothesis tests whether the effect of posting e-NWOM
is moderated by company usefulness (i.e., redemption behavior).

We found a positive interaction between posting e-NWOM and
point redemption. Our analysis shows that spending level is
increased by 37% (e0.314
¼ 1.37) for those who posted e-NWOM and
also experienced the benefit of the company through point
redemption. The finding suggests that for e-NWOM posters, it is
important that they are given an opportunity to recognize the
usefulness of the company by experiencing the brand value
through redemption after posting e-NWOM.
Our second hypothesis tests the effect of viewing e-NWOM
messages high on consensus, distinctiveness, and consistency. We
found that viewers decrease their spending level after reading e-
NWOM messages. The finding is consistent with previous research
that showed a negative influence of e-NWOM on its readers. Upon
reading uniformly negative posts, viewers are more likely to attri-
bute the e-NWOM incident to be the company, which is likely to
create negative attitudes toward the firm and reduce their subse-
quent spending.
While posting and viewing e-NWOM have different effects on
future behaviors, it seems that there is a single company response
that could address both and create a situation where both cus-
tomers and the company benefit. Since point redemption reminds
posters of the brand's usefulness, to mitigate negative effects
brought forth by the policy change, the company could have offered
a promotion/incentive to encourage redemption. Moreover, since
redemption increases future point accumulation, it is possible that
revenues due to the change in future point accumulation will
exceed the cost of the promotion and the reward, which would
create a net positive change in customer value from the firm's point
of view, and increase the value consumers derive from their rela-
tionship with the firm. This is an example of value fusion (Lariviere
et al., 2013), where both parties derive value.
For viewers, it is important that the company provides further
information and shows concern for the complaints of other cus-
tomers. Such interventions show how attentive the company is to
customer feedback (Allsop, Bassett, Hoskins, 2007). Since the
company did not make any efforts to respond to customers'
complaints posted on its community website or social media
channels, we posit that the company missed an opportunity to
address the frustration openly on the community forum. A public
notice of the promotion/incentive mentioned above would have
been viewed by all, and could have allayed the effects of the e-
NWOM by establishing the perception that the company cares
about its customers' anger and makes an effort to provide value to
them. Such a company intervention could have also motivated
posters to adjust their position on the e-NWOM incident and write
a more positive message, which would lower the levels of
consensus and consistency of e-NWOM messages and ultimately
change the attribution among viewers. To examine the issue of
Table 1
Model estimates e effects of viewing, posting, and reward redemption.
Model
Independent variable Points accumulated (logged)
Intercept À1.204 (0.128)***
Treatment variables
Posting À0.227 (0.105)*
Redeeming following policy change 0.137 (0.03)***
Posting Â redeeming following policy change 0.314 (0.160)*
Viewing À0.116 (0.034)***
Viewing Â redeeming following policy change À0.041 (0.058)
Fixed effects
Is a poster 0.137 (0.092)
Is a viewer 0.074 (0.029)**
Behaviors before policy change
ln(Bank Points Accumulated þ 1) 0.145 (0.008)***
ln(Food Points Accumulated þ 1) 0.289 (0.006)***
ln(Retail Points Accumulated þ 1) 0.080 (0.0103)***
ln(Gas Points Accumulated þ 1) 0.249 (0.008)***
ln(Other/Misc. Points Accumulated þ 1) 0.094 (0.015)***
ln(Number of Sponsors Purchased From þ 1) 0.084 (0.039)*
ln(Points Used to Redeem Travel þ 1) 0.044 (0.018)**
ln(Points Used to Redeem Cash/Certificates þ 1) 0.041 (0.007)***
ln(Points Used to Redeem Goods/Other þ 1) 0.045 (0.008)***
Other control variables
Seasonality 0.490 (0.037)***
Propensity Score 0.406 (0.295)
Random Intercept Covariance 0.217 (0.466)
Note. Standard errors are presented in parentheses. *
p .05, **
p .01, ***
p .001.
-0.2-0.10.00.10.2
Posting
EstimatedMarginalMeans
No Yes
Redeeming
Yes
No
Fig. 2. Interaction effect between posting and redeeming (study 1).

company intervention further, Study 2 tests the effect of an apology
on e-NWOM posters and viewers using a randomized controlled
experiment.
4. Study 2
4.1. The moderating effect of company apology
Two questions emerge from the findings of Study 1: (1) Will we
see the differential effects of posting and viewing in another in-
dustry? (2) How would have the results changed if the company
had responded to NWOM on the community forum to remind its
customers of its value and offer an apology? In Fig. 1, we identified
company compensation as the second possible moderator of the
effects of e-NWOM behavior. Study 2 addresses the aforementioned
questions.
A growing body of literature has shown that when a company
handles complaints properly, its reputation is not damaged, and it
can even potentially benefit by regaining customer satisfaction and
loyalty (Buttle Burton, 2002; Lee Song, 2010; Willemsen,
Neijens, Bronner, 2013). Previous research has not, however,
investigated the differential effect of webcare on NWOM viewers
and posters. We expect that in the presence of the company's
webcare effort, viewers observing the dialog between complainers
and the company will consider the firm's effort to offer a response
as credible and beneficial (Breitsohl, Khammash, Griffiths, 2010).
We also expect that a public apology on a company's community
website will create the perception that the company is sincere and
respects the communal relationship with its customers, which, in
turn, will increase future purchases for both viewers and posters
(Gregoire et al., 2009).
H3. A company apology has a positive moderating effect on the
relationship between posting e-NWOM and behavioral intentions,
i.e., the apology will make posters less likely to conduct negative
behaviors, such as quitting.
H4. A company apology has a positive moderating effect on the
relationship between viewing e-NWOM and behavioral intentions,
i.e., the apology will make viewers less likely to conduct negative
behaviors, such as quitting.
4.2. Participants and design
To test our predictions, we conducted an experiment with a 2
(viewer vs. poster) Â 2 (apology vs. no apology) þ 1 (control group)
between-subject design. The experiment was conducted among
graduate students (n ¼ 127, 74.8% female, Mage ¼ 25.09,
SDage ¼ 6.12) at an urban private university in the United States.
Participants represented 16 countries with the majority coming
from the U.S. (33.9%) and China (44.9%). Students were not
compensated for their participation.
4.3. Procedure
Participants were emailed a request to take part in an online
study. At the beginning of the experiment, they were asked to give
informed consent. Next, they were exposed to a policy change
scenario. After reading the scenario, participants were randomly
assigned to one of the three conditions: control (n ¼ 29), viewer
(n ¼ 51), and poster (n ¼ 47). Participants in the control condition
were asked to complete the survey. Respondents in the viewer and
poster conditions were asked to read or post comments about the
policy change. Next, they were randomly exposed to either the
apology condition (nviewer ¼ 24, nposter ¼ 22), in which they were
asked to read the company apology or not shown the apology
(nviewer ¼ 27, nposter ¼ 25). Finally, all participants were debriefed
and thanked for their participation. To make sure that the partici-
pants read the scenario and reviews, a timer was used. On average,
participants were exposed to the scenario for about 53 s
(Mdn ¼ 30.28, M ¼ 53.09, SD ¼ 86.46), to the reviews for about
298 s (Mdn ¼ 82.93, M ¼ 297.88, SD ¼ 1298.33), and to the apology
for 440 s (Mdn ¼ 49.18, M ¼ 440.45, SD ¼ 2057.29). Participants in
different conditions did not differ with respect to the time spent
reading the scenario, F2,124 ¼ 0.543, p ¼ .582 (review condition),
t125 ¼ 0.985, p ¼ .327 (apology condition). Also, the two apology
conditions did not differ with respect to the time spent on reading
the reviews, t49 ¼ À1.120, p ¼ .268. Finally, viewers and posters
exposed to the apology condition did not differ in the time spent on
reading the apology, t44 ¼ 1.206, p ¼ .234.
4.4. Stimulus materials
In all conditions the scenario read as follows: “Imagine that you
received an e-mail from your current mobile service provider,
American Telecom, stating that your plan conditions were going to
change. You have a limit of 500MB of data per month. The previous
policy was to slow down Internet speeds after exceeding the limit,
but not charge anything. The email informed you that in the future
you will pay a penalty of $15 if you exceed the data limit, and $1/MB
that you use beyond 500MB.” After reading the scenario, partici-
pants in the viewer condition were requested to imagine that they
visited the company's community site and discovered other cus-
tomers' NWOM. Participants in the poster condition were asked to
write a comment on the community site expressing their thoughts
and feelings.
Participants in the viewer condition were then exposed to four
negative reviews about American Telecom (Appendix A, hereafter
called “A.T.”). To increase external validity, the posts were collected
from real review sites and modified to reflect the context of Study 2.
As with Study 1, the NWOM condition in Study 2 was of high
consensus, distinctiveness, and consistency. Next, depending on
the apology condition, participants were either asked to read a post
from a company apologizing for the situation (Appendix B) or to
complete the survey. The post was based on companies' real re-
actions to negative reviews and contained an apology and an
explanation of steps the company is going to introduce to minimize
the risk of overages. To be consistent with the context of Study 1,
which tests for company usefulness, the apology condition also
included a paragraph that explains the benefits that A.T. provides to
its customers.
4.5. Measures
To make sure that our policy change scenario generated a
negative e-WOM incident, we asked respondents to answer three
questions measuring their level of anger and regret based on
Tykocinski and Pittman (2001), using a 7-point Likert scale
(1 ¼ completely disagree; 7 ¼ completely agree). The order of the
items was randomized. To confirm that the manipulation of the
reviews was successful, participants were asked to rate the nega-
tivity of the four reviews they had read or the review they had
written on a scale from 1 (very negative) to 7 (very positive). The
apology manipulation was checked with one item measured on a 7-
point Likert scale (1 ¼ completely disagree; 7 ¼ completely agree).
Negative behavioral intentions were measured with a 7-point Lik-
ert scale (1 ¼ not at all likely; 7 ¼ extremely likely) of how likely
they were to 1) search for alternative providers, 2) stay with A.T.
(reversed), 3) quit A.T., 4) recommend A.T. to a friend (reversed).
Items were presented in a random order (Cronbach's a ¼ .71).

Finally, information on gender, age, education level and country of
origin was collected.
4.6. Results of Study 2
4.6.1. Manipulation and randomization checks
Participants in the three conditions did not differ with respect to
gender (c2
(2) ¼ 2.031, p ¼ .362), education (c2
(8) ¼ 9.295, p ¼ .318),
country of origin (c2
(30) ¼ 33.686, p ¼ .294), or age (F2,121 ¼ 1.554,
p ¼ .216). We assessed if our manipulations worked. First, the
scenario successfully evoked negative feelings: Participants felt
angry at themselves for choosing A.T. (M ¼ 3.97, SD ¼ 1.74),
regretted having decided to choose A.T. (M ¼ 5.45, SD ¼ 1.33), and
felt angry at A.T. (M ¼ 5.97, SD ¼ 1.15). Second, the four negative
reviews in the stimulus material were perceived as negative
(M1 ¼ 1.86, SD1 ¼ 1.18; M2 ¼ 2.02, SD2 ¼ 1.14; M3 ¼ 2.02, SD3 ¼ 1.16;
M4 ¼ 2.02, SD4 ¼ 1.14). We also asked posters to evaluate their own
reviews, which were perceived as negative (M ¼ 2.47, SD ¼ 1.16).
There were no differences in how negative the reviews were
perceived, i.e., reviews posted by others were perceived equally
negative as reviews posted by the participants, F(3, 94) ¼ 1.75,
p ¼ .162. Finally, participants reported that the company apology
used the right tone of voice (M ¼ 4.52, SD ¼ 1.57).
4.6.2. Hypotheses testing
We used a two-way ANOVA to test our hypotheses. Because the
3 Â 2 design had a missing cell, Type IV sums of squares were used,
which compare a given cell with averages of other cells and are
preferred when there are missing treatment combinations
(Milliken Johnson, 2009). The interaction between condition and
company apology on negative behavioral intention was borderline
significant (F1,122 ¼ 3.109, p ¼ .040, h2
¼ 0.03) using one-sided
tested due to the directional nature of our hypotheses (Fig. 3).
Moreover, the main effects of condition, F2,122 ¼ 4.750, p ¼ .005,
h2
¼ 0.07, and company apology, F1,122 ¼ 8.369, p ¼ .003, h2
¼ 0.06,
on intention were also significant (one-sided).
To disentangle the interaction effect, simple effects tests with
least significant differences were applied. Mean scores for all var-
iables are included in Table 2 and significant differences are shown
based on multiple comparisons (p .05). Although the effect of
apology on posters was in the right direction, posters in the no
apology condition did not differ significantly from posters in the
apology condition (Mdiff ¼ À0.211, SE ¼ 0.27, F1,122 ¼ 0.613, p ¼ .435,
h2
¼ 0.01), leading us to reject H3. The hypothesized positive effect
of apology on viewers (H4) was confirmed by a significant differ-
ence between viewers in the no apology condition and those in the
apology condition (Mdiff ¼ À0.869, SE ¼ 0.26, F1,122 ¼ 11.304,
p ¼ .001, h2
¼ 0.09). When the apology was offered, viewers scored
lower on negative behavioral intentions than posters
(Mdiff ¼ À0.732, SE ¼ 0.27, F1,122 ¼ 7.242, p ¼ .008, h2
¼ 0.06). When
no apology was present, the differences between conditions were
not significant. Summarizing, the apology affected only viewers of
e-NWOM: viewers who read the public company apology
decreased their intention to quit and increased their intention to
stay and recommend the company.
4.7. Discussion of Study 2
Hypothesis 3 predicts that a company apology has a positive
moderating effect on posters, but we cannot confirm it. Posters who
were offered an apology did not differ from those who were not.
Posters may have just wanted to experience catharsis by releasing
negative emotions without expecting that the company would
respond to their complaints. This explanation not only confirms
venting as a mechanism for resolving cognitive dissonance, but also
is consistent with the finding that webcare satisfaction is negatively
affected by venting (Willemsen et al., 2013). Also, an apology from
the company may be associated with guilt, which people consider a
negative sign (Blodgett et al., 1997). Finally, it is possible that an
apology was not enough to evoke fairness perception or to trigger
elaboration on company usefulness. Cognitive appraisal theory
predicts that people's evaluation of a situation triggers emotions (del
Río-Lanza et al., 2009). If the apology was not perceived as a remedy
for the situation, it failed to elicit positive emotions about the firm.
Hypothesis 4 predicts that viewers who receive the company
apology show more positive behavioral intentions. Our results
confirm Hypothesis 4: Viewers who saw that the complaints had
been addressed by the company increased their positive behavioral
intentions. This is consistent with van Noort and Willemsen (2011),
showing that webcare can diminish the negative effects of NWOM.
This also suggests that for e-NWOM viewers, the additional infor-
mation provided by the company is used for the viewers' attribu-
tion process. Company apology changes the level of consensus in
what viewers read on the online forum, which then influences the
causal attribution of viewers.
5. General discussion
This study examines the effect of posting and viewing e-NWOM
on purchase behaviors, and the moderating role of company use-
fulness and company apology. A series of two studies showed that
(1) posting e-NWOM has a positive effect when posters are
reminded of the firm's usefulness after engaging in posting
behavior, (2) viewing e-NWOM that consists of uniformly negative
opinions decreases subsequent purchases, and (3) a company
4.44.64.85.05.25.45.6
Conditions
EstimatedMarginalMeans
Control Viewer Poster
Apology
No
Yes
Fig. 3. Interaction effect between condition and company apology on negative
behavioral intention (study 2).
Table 2
Mean differences on intention.
Conditions No apology Apology
M SD N M SD N
Control 5.69a 0.95 29 e e e
Viewer 5.30a 0.95 27 4.43b 0.73 24
Poster 5.37a 1.04 25 5.16a 0.89 22
Total 5.46a 0.98 81 4.78a 0.88 46
Note. Means with different subscripts (a, b) differ significantly at p .05.

responsedreminding customers of the value provided by the firm
and apologizing for the policy changedincreases behavioral
intention among e-NWOM viewers.
The effect of posting e-NWOM is moderated by whether posters
are aware of the usefulness of the brand through point redemption.
The findings suggest the importance of relationship marketing in
the emergence of e-NWOM crisis. The existing relationship be-
tween customers and the company, especially when customers
perceive the company to be useful to them, works as a positive
moderator that mitigates the negative sentiments from elaborating
what they have experienced (i.e., self-prophecy). In addition, the
positive impact of posting through venting (Nyer Gopinath,
2005) can be escalated when customers are placed in a situation
where they are reminded of how the brand creates value.
In addition to the positive impact of venting, the specific context
of e-NWOM (i.e., online public forum) may provide another
explanation for why expression of e-NWOM may exert a positive
influence on posters. According to Liu and Shrum (2002), a higher
level of interactivity helps users to gain a greater sense of infor-
mation control, which results in a more positive mood and attitude
toward the website and user satisfaction. The interface of an online
discussion board (i.e., message threads) also gives a sense that
people are engaged in an exchange of dialog with other partici-
pants, and the exchange directly leads to a higher sense of user
engagement and positive attitude toward the website and the
brand (Sundar, Bellur, Oh, Xu, Jia, 2013). Given that our study
used a firm's community website that provides two-way commu-
nication among consumers and/or between the firm and con-
sumers in a message-thread style, this situational factor may have
contributed to a positive outcome from e-NWOM posters by
providing them with a sense of agency and engagement.
For e-NWOM viewers, we find that merely viewing without
participating in the online discussion has a negative influence on
future purchases. Without releasing negative feelings and thoughts
in the form of writing, e-NWOM viewers are most likely to obtain
negative information and absorb negative sentiments about the
company. In line with attribution theory, viewers who use the
uniformly negative information to assess causal attributions are
more likely to conform to the group opinion and blame the firm to
be the culprit of the complaints, which results in decreasing pur-
chases from the company (Laczniak et al., 2001).
We also examine whether the offering of a company apology
reminding customers about the value of the company alleviates
negative purchase intentions of posters and viewers. The findings
indicate that the apology works for both viewers and posters,
although the effect on posters was not statistically significant. The
differences between posters and viewers could be explained by
different levels of issue involvement in e-NWOM processing
(Breitsohl et al., 2010). Because viewers are not as emotionally
involved as complainers, they are more likely to engage in shallow
cognitive processing. On the other hand, posters, who engage
themselves in the situation, have higher levels of cognitive and
emotional involvement, which will lead them to deeper cognitive
processing. As argued by the elaboration likelihood model (Petty
Cacioppo, 1986), individuals who engage in deep processing pay
more attention to the content of the message and arguments pre-
sented in it, while individuals following the peripheral route of
processing pay more attention to heuristic cues. We posit that the
company apology was accepted by viewers based on the heuristic:
the presence of company apology itself is a sign of webcare and
enough for viewers to rebuild trust. Posters, however, need more
than a mere presence of company apology and may need a more
concrete reminder of how the company provides value or some
type of compensation that serves as a remedy for their high level of
negative involvement with the issue in hand.
The differential effect of webcare on posters and viewers may
also be explained by social learning theory (Bandura, 1977). As
Schamari and Schaefers (2015) discuss, consumers learn not only
from direct experience, but also through observing others and
imitating their behaviors (Blazevic et al., 2013; Libai et al., 2010).
Hence, when a consumer observes the dialog between posters and
the company, she learns that the company makes an effort to
address problems consumers encounter. Viewers may perceive the
firm's response as having higher complaint utility, because the
firm's identity and expertise are evident (Breitsohl et al., 2010).
Hence, viewers may perceive company response as more credible
than the original complaint, and show positive responses towards
the company.
6. Contributions and practical implications
This study offers substantive contributions. First, we advance
our understanding of the effects of e-NWOM by distinguishing
posting and viewing behavior. We test how perceptions about
company usefulness and a company apology moderate the effects
of engaging in e-NWOM behavior. Given the dearth of research on
those who create messages, this study provides a comprehensive
picture of the effects of producing and consuming it. Second, we
provide evidence showing how e-NWOM impacts a firm's revenue
by linking online social media data with customers' actual purchase
data (Study 1). Methodologically, we use propensity score matching
to reduce potential selection bias in actual social media and pur-
chase data, which is underrepresented in advertising literature. Our
paper can serve as an example of how field data can be used to
advance our understanding of advertising research. Third, we
contribute to the growing literature on webcare by testing the ef-
fect of a company apology. Lastly, Study 2 provides a more inter-
nally valid test of the relationships between e-NWOM activities, the
firm's response, and consumer behaviors by conducting a ran-
domized controlled experiment.
Managerially, our findings suggest that companies should
customize strategies in response to e-NWOM messages, since the
effects of posting and viewing differ. For viewers, it is important to
address the dominant opinions by being transparent and keeping
them informed so that they perceive the brand as being beneficial
and useful. Such an intervention will show that the company is
timely in handling complaints and attentive to its customers
(Allsop et al., 2007). For posters, a company should encourage
customers to participate in their WOM activities. At the same time,
it should also provide an opportunity that helps posters be engaged
with the brand and reminds them of the brand's benefits. Brand
managers should carefully monitor the sentiments expressed in e-
NWOM messages and provide reactive interventions.
This study has limitations and raises questions for future
research. Our data does not allow us to study the long-term effect of
e-NWOM. Thus, we are not be able to answer the question of how
long the observed effect will hold and what company interventions
are appropriate across the time span. Second, we used a similar e-
NWOM incident that is specific to the members of the firms
involved and is of high consensus. Future research should investi-
gate e-NWOM generated in different configurations of consensus,
distinctiveness, and consistency to see how these conditions
change the effect of posting and viewing.
Acknowledgment
We appreciate support from Northwestern University's Spiegel
Research Center on Digital and Database Marketing. We also thank
the Center's Executive Director Tom Collinger for sharing his in-
sights on earlier versions of this manuscript.

Appendix A
Reviews
Appendix B
Apology
References
Aggarwal, P. (2004). The effects of brand relationship norms on consumer attitudes
and behavior. Journal of Consumer Research, 31(1), 87e101. http://dx.doi.org/
10.1086/383426.
Allsop, D. T., Bassett, B. R., Hoskins, J. A. (2007). Word-of-mouth research: prin-
ciples and applications. Journal of Advertising Research, 47(4), 398e411.
Andreassen, T. W., Streukens, S. (2009). Service innovation and electronic word-
of-mouth: is it worth listening to? Managing Service Quality, 19(3) http://
dx.doi.org/10.1108/09604520910955294, 249-249-265.
Bandura, A. (1977). Social learning theory. Englewood Cliffs, N.J.: Prentice Hall.
Barclay, L. J., Skarlicki, D. P. (2009). Healing the wounds of organizational injus-
tice: examining the benefits of expressing writing. Journal of Applied Psychology,
94(2), 511e523. http://dx.doi.org/10.1037/a0013451.
Berger, J. (2014). Word of mouth and interpersonal communication: a review and
directions for future research. Journal of Consumer Psychology, 24(4), 586e607.
http://dx.doi.org/10.1016/j.jcps.2014.05.002.
Bickart, B., Schindler, R. M. (2001). Internet forums as influential sources of
consumer information. Journal of Interactive Marketing, 15(3), 31e40. http://
dx.doi.org/10.1002/dir.1014.
Blazevic, V., Hammedi, W., Garnefeld, I., Rust, R. T., Keiningham, T.,
Andreassen, T. W., et al. (2013). Beyond traditional word-of-mouth: an
expanded model of customer-driven influence. Journal of Service Management,
24(3), 294e313. http://dx.doi.org/10.1108/09564231311327003.
Blodgett, J. G., Hill, D. J., Tax, S. S. (1997). The effects of distributive, procedural,
and interactional justice on postcomplaint behavior. Journal of Retailing, 73(2),
185e210. http://dx.doi.org/10.1016/S0022-4359(97)90003-8.
Breitsohl, J., Khammash, M., Griffiths, G. (2010). E-business complaint manage-
ment: perceptions and perspectives of online credibility. Journal of Enterprise
Information Management, 23(5), 653e660. http://dx.doi.org/10.1108/
17410391011083083.
Buttle, F., Burton, J. (2002). Does service failure influence customer loyalty?
Journal of Consumer Behaviour, 1(3), 217e227. http://dx.doi.org/10.1002/cb.67.
Campbell, C., Pitt, L. F., Parent, M., Berthon, P. R. (2011). Understanding consumer
conversations around ads in a Web 2.0 World. Journal of Advertising, 40(1),
87e102.
Chevalier, J. A., Mayzlin, D. (2006). The effect of word of mouth on sales: online
book reviews. Journal of Marketing Research, 43(3), 345e354. http://dx.doi.org/
10.1509/jmkr.43.3.345.
Chi, M. T. H., De Leeuw, N., Chiu, M.-H., Lavancher, C. (1994). Eliciting self-
explanations improves understanding. Cognitive Science, 18(3), 439e477.
http://dx.doi.org/10.1207/s15516709cog1803_3.
Chiou, J.-S., Hsu, A. C.-F., Hsieh, C.-H. (2013). How negative online information
affects consumers' brand evaluation. Online Information Review, 37(6), 910e926.
http://dx.doi.org/10.1108/OIR-02-2012-0014.
Davis, A., Khazanchi, D. (2008). An empirical study of online word of mouth as a
predictor for multi-product category e-commerce sales. Electronic Markets,
18(2), 130e141. http://dx.doi.org/10.1080/10196780802044776.
Dellarocas, C., Awad, N., Zhang, X. (2004). Exploring the value of online reviews to
organizations: implications for revenue forecasting and planning. In Proc. 24th
International Conference on Information Systems.
Doh, S.-J., Hwang, J.-S. (2009). How consumers evaluate eWOM (electronic word-
of-mouth) messages. CyberPsychology Behavior, 12(2), 193e197. http://
dx.doi.org/10.1089/cpb.2008.0109.
Duan, W., Gu, B., Whinston, A. B. (2008). The dynamics of online word-of-mouth
and product sales: an empirical investigation of the movie industry. Journal of
Retailing, 84(2), 233e242. http://dx.doi.org/10.1016/j.jretai.2008.04.005.
East, R., Hammond, K., Lomax, W. (2008). Measuring the impact of positive and
negative word of mouth on brand purchase probability. International Journal of
Research in Marketing, 25(3), 215e224. http://dx.doi.org/10.1016/
j.ijresmar.2008.04.001.
Edelman, D. C. (2010). Branding in the digital age. Harvard Business Review, 88(12),
62e69.
Festinger, L. (1957). A theory of cognitive dissonance. Evanston, IL: Row.
Goetzinger, L. M. (2007). Consumer complaint behavior: Studies on behavioral di-
mensions and the impact of the Internet (Consumer Sciences and Retailing Ph.
D.). United States e Indiana: Consumer Sciences and Retailing. Retrieved from
http://turing.library.northwestern.edu/login?url¼http://search.proquest.com/?
url¼http://search.proquest.com/docview/304841878?accountid¼12861 (Pro-
Quest database).
Gregoire, Y., Tripp, T. M., Legoux, R. (2009). When customer love turns into lasting
hate: the effects of relationship strength and time on customer revenge and
avoidance. Journal of Marketing, 73(6), 18e32. http://dx.doi.org/10.1509/
jmkg.73.6.18.
Hennig-Thurau, T., Gwinner, K. P., Walsh, G., Gremler, D. D. (2004). Electronic
word-of-mouth via consumer-opinion platforms: what motivates consumers to
articulate themselves on the Internet? Journal of Interactive Marketing, 18(1),
38e52. http://dx.doi.org/10.1002/dir.10073.
Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to decline in firms, orga-
nizations, and states. Cambridge, Mass: Harvard University Press.
Huang, J.-H., Chen, Y.-F. (2006). Herding in online product choice. Psychology and
Marketing, 23(5), 413e428. http://dx.doi.org/10.1002/mar.20119.
Keller, E. (2007). Unleashing the power of word of mouth: creating brand advocacy
to drive growth. Journal of Advertising Research, 47(4), 448e452.
I just received my bill from you, and it was a surprise. $15 data penalty plus another $20 for going over my limit. THIS IS SO UNFAIR!!! You are just greedy!
American Telecom just hit me with a HUGE overage charge. I called to complain and they told me I have to pay, and there will be a late fee if I don't pay the overage charge
on time. There was absolute no effort by the supervisor to preserve goodwill. If this is not illegal, then it is at least unethical.
I went over my data by a measly 2 MB and American Telecom charged me $17! Can you believe it? Wow, how stingy and deceitfuldbeware!!!
I wish I could give a half of a star … I am so unhappy with them I can't begin to describe it. I went to speak with a manager who sat and said there is absolutely nothing he
can do to help me out, just that their policy had suddenly changed. If they had told me when I was getting my phone with them, I would've totally been fine, but to just
inform us about the changes like that, it is unacceptable. They also didn't seem like you wanted to help in any way which just tells me that this company does not value
their customers at all.
We are sorry that you are angry about our data-overage policy change. Our goal has always been to provide the best cell phone service in the industry. We offer less
expensive plans with low data limits for customers who do not plan to use much data, and other plans for customers who use their mobile phone for dataeintensive
activities, such as music and video streaming. We want customers to have a plan that is right for their needs.
In the future, we will send you SMS messages when you are nearing your data limit or when you are about to exceed your limit, and ask whether you wish to exceed your
limit and pay the fee.
American Telecom offers many benefits that other carriers do not have. Only our 4G network is 100% 4G LTE the gold standard of wireless technology. Available in over 500
cities, American Telecom 4G LTE covers almost 97% of the U.S. population. Moreover, American Telecom makes it easier to buy a new device, with a low upfront cost and
affordable monthly payments.

Kim, E., Sung, Y., Kang, H. (2014). Brand followers' retweeting behavior on
Twitter: how brand relationships influence brand electronic word-of-mouth.
Computers in Human Behavior, 37, 18e25. http://dx.doi.org/10.1016/
j.chb.2014.04.020.
Laczniak, R. N., DeCarlo, T. E., Ramaswami, S. N. (2001). Consumers' responses to
negative word-of-mouth communication: an attribution theory perspective.
Journal of Consumer Psychology, 11(1), 57e73.
Ladhari, R. (2007). The effect of consumption emotions on satisfaction and word-of-
mouth communications. Psychology Marketing, 24(12), 1085e1108.
Lariviere, B., Joosten, H., Malthouse, E. C., Birgelen, M. v., Aksoy, P., Kunz, W. H., et al.
(2013). Value fusion: the blending of consumer and firm value in the distinct
context of mobile technologies and social media. Journal of Service Management,
24(3), 268e293. http://dx.doi.org/10.1108/09564231311326996.
Lee, J., Park, D.-H., Han, I. (2008). The effect of negative online consumer reviews
on product attitude: an information processing view. Electronic Commerce
Research and Applications, 7(3), 341e352. http://dx.doi.org/10.1016/
j.elerap.2007.05.004.
Lee, Y. L., Song, S. (2010). An empirical investigation of electronic word-of-mouth:
informational motive and corporate response strategy. Computers in Human
Behavior, 26(5), 1073e1080. http://dx.doi.org/10.1016/j.chb.2010.03.009.
Libai, B., Bolton, R., Bügel, M. S., de Ruyter, K., G€otz, O., Risselada, H., et al. (2010).
Customer-to-customer interactions: broadening the scope of word of mouth
research. Journal of Service Research, 13(3), 267e282. http://dx.doi.org/10.1177/
1094670510375600.
Liu, Y. (2006). Word of mouth for movies: its dynamics and impact on box office
revenue. Journal of Marketing, 70(3), 74e89. http://dx.doi.org/10.2307/
30162102.
Liu, Y., Shrum, L. J. (2002). What is interactivity and is it always such a good thing?
implications of definition, person, and situation for the influence of interactivity
on advertising effectiveness. Journal of Advertising, 31(4), 53e64.
Maute, M. F., Dubes, L. (1999). Patterns of emotional responses and behavioural
consequences of dissatisfaction. Applied Psychology, 48(3), 349e366. http://
dx.doi.org/10.1111/j.1464-0597.1999.tb00006.x.
Milliken, G. A., Johnson, D. E. (2009). Analysis of messy data volume 1: designed
experiments (2nd ed.). Boca Raton, FL: CRC Press.
Mittal, V., Ross, W. T., Jr., Baldasare, P. M. (1998). The asymmetric impact of
negative and positive attribute-level performance on overall satisfaction and
repurchase intentions. Journal of Marketing, 62(1), 33e47.
Mizerski, R. W. (1982). An attribution explanation of the disproportionate influence
of unfavorable information. Journal of Consumer Research, 9(3), 301e310. http://
dx.doi.org/10.2307/2488625.
Nekmat, E. (2012). Message expression effects in online social communication.
Journal of Broadcasting Electronic Media, 56(2), 203e224. http://dx.doi.org/
10.1080/08838151.2012.678513.
van Noort, G., Willemsen, L. M. (2011). Online damage control: the effects of
proactive versus reactive webcare interventions in consumer-generated and
brand-generated platforms. Journal of Interactive Marketing, 26(3), 131e140.
http://dx.doi.org/10.1016/j.intmar.2011.07.001.
Nyer, P. U., Gopinath, M. (2005). Effects of complaining versus negative word of
mouth on subsequent changes in satisfaction: the role of public commitment.
Psychology Marketing, 22(12), 937e953. http://dx.doi.org/10.1002/mar.20092.
Park, C., Lee, T. M. (2009). Information direction, website reputation and eWOM
effect: a moderating role of product type. Journal of Business Research, 62(1),
61e67. http://dx.doi.org/10.1016/j.jbusres.2007.11.017.
Pennebaker, J. W. (1997). Writing about emotional experiences as a therapeutic
process. Psychological Science, 8(3), 162e166. http://dx.doi.org/10.1111/j.1467-
9280.1997.tb00403.x.
Petty, R. E., Cacioppo, J. T. (1986). Communication and persuasion: Central and
peripheral routes to attitude change. New York: Springer-Verlag.
Pitt, L. F., Berthon, P. R., Watson, R. T., Zinkhan, G. M. (2002). The Internet and the
birth of real consumer power. Business Horizons, 45(4), 7e14. http://dx.doi.org/
10.1016/s0007-6813(02)00220-3.
Prislin, R., Boyle, S. M., Davenport, C., Farley, A., Jacobs, E., Michalak, J., et al. (2011).
On being influenced while trying to persuade: the feedback effect of persuasion
outcomes on the persuader. Social Psychological and Personality Science, 2(1),
51e58. http://dx.doi.org/10.1177/1948550610377238.
Reichheld, F. F. (2003). The one number you need to grow. Harvard Business Review,
81(12), 46e54.
Richins, M. L., Root-Shaffer, T. (1988). The role of involvement and opinion
leadership in consumer word-of-mouth: an implicit model made explicit. Ad-
vances in Consumer Research, 15(1), 32e36.
Riegner, C. (2007). Word of mouth on the web: the impact of Web 2.0 on consumer
purchase decisions. Journal of Advertising Research, 47(4), 436e447.
Rime, B. (2009). Emotion elicits the social sharing of emotion: theory and empirical
review. Emotion Review, 1(1), 60e85. http://dx.doi.org/10.1177/
1754073908097189.
del Río-Lanza, A. B., Vazquez-Casielles, R., Díaz-Martín, A. M. (2009). Satisfaction
with service recovery: perceived justice and emotional responses. Journal of
Business Research, 62(8), 775e781. http://dx.doi.org/10.1016/
j.jbusres.2008.09.015.
Rosenbaum, P. R., Rubin, D. B. (1983). The central role of the propensity score in
observational studies for causal effects. Biometrika, 70(1), 41e55. http://
dx.doi.org/10.1093/biomet/70.1.41.
Rubin, D. B. (1997). Estimating causal effects from large data sets using propensity
scores. Annals of Internal Medicine, 127(8_Part_2), 757e763. http://dx.doi.org/
10.7326/0003-4819-127-8_Part_2-199710151-00064.
Schamari, J., Schaefers, T. (2015). Leaving the home turf: how brands can use
webcare on consumer-generated platforms to increase positive consumer
engagement. Journal of Interactive Marketing, 30, 20e33. http://dx.doi.org/
10.1016/j.intmar.2014.12.001.
Schoefer, K., Diamantopoulos, A. (2009). A typology of consumers' emotional
response styles during service recovery encounters. British Journal of Manage-
ment, 20(3), 292e308. http://dx.doi.org/10.1111/j.1467-8551.2008.00589.x.
Sherman, S. J. (1980). On the self-erasing nature of errors of prediction. Journal of
Personality and Social Psychology, 39(2), 211e221. http://dx.doi.org/10.1037/
0022-3514.39.2.211.
Singh, J. (1990). A typology of consumer dissatisfaction response styles. Journal of
Retailing, 66(1), 57e99.
S€oderlund, M., Rosengren, S. (2007). Receiving word-of-mouth from the service
customer: an emotion-based effectiveness assessment. Journal of Retailing and
Consumer Services, 14(2), 123e136. http://dx.doi.org/10.1016/
j.jretconser.2006.10.001.
Spangenberg, E. R., Giese, J. L. (1997). An exploratory study of word-of-mouth
communication in a hierarchy of effects context. Communication Research Re-
ports, 14(1), 88e96.
Sundaram, D. S., Mitra, K., Webster, C. (1998). Word-of-mouth communications: a
motivational analysis. Advances in Consumer Research, 25(1), 527e531.
Sundar, S. S., Bellur, S., Oh, J., Xu, Q., Jia, H. (2013). User experience of on-screen
interaction techniques: an experimental investigation of clicking, sliding,
zooming, hovering, dragging, and flipping. HumaneComputer Interaction, 29(2),
109e152. http://dx.doi.org/10.1080/07370024.2013.789347.
The Nielsen Company. (2013). Global trust in advertising and brand messages. http://
www.nielsen.com/us/en/newswire/2013/under-the-influence-consumer-trust-
in-advertising.html.
Tykocinski, O. E., Pittman, T. S. (2001). Product aversion following a missed op-
portunity: price contrast or avoidance of anticipated regret? Basic and Applied
Social Psychology, 23(3), 149e156. http://dx.doi.org/10.1207/
S15324834BASP2303.
Verhagen, T., Nauta, A., Feldberg, F. (2013). Negative online word-of-mouth:
behavioral indicator or emotional release? Computers in Human Behavior,
29(4), 1430e1440. http://dx.doi.org/10.1016/j.chb.2013.01.043.
Westbrook, R. A. (1987). Product/consumption- based affective responses and
postpurchase processes. Journal of Marketing Research, 24(3), 258e270.
Willemsen, L., Neijens, P. C., Bronner, F. A. (2013). Webcare as customer rela-
tionship and reputation management? motives for negative electronic word of
mouth and their effect on webcare receptiveness. In S. Rosengren, M. Dahlen,
S. Okazaki (Eds.), Advances in advertising research (vol. IV, pp. 55e69). Springer
Fachmedien Wiesbaden.

Nwom15 http://www.cheapassignmenthelp.co.uk/

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Nwom15 http://www.cheapassignmenthelp.co.uk/

Similar to Nwom15 http://www.cheapassignmenthelp.co.uk/ (20)

More from Assignment Help

More from Assignment Help (20)

Recently uploaded

Recently uploaded (20)

Nwom15 http://www.cheapassignmenthelp.co.uk/