Virtually present others and their influence on

complainants’ follow-ups and firm response

Rosa E. Rios
College of Business, Victoria University, Melbourne, Australia
Hernan E. Riquelme
La Trobe Business School, La Trobe University, Melbourne, Australia, and
Alessandro Comai
Graduate School of International Management, International University of Japan, Minami Uonuma Shi, Japan

Abstract
Purpose – The purpose of this empirical research is to investigate the influence of interactive virtually present others (VPOs) on a firm response and
customers’ reiterated complaints (follow-ups) during service failures. This research follows up on previous experimental studies that suggest VPOs
affect the focal customers’ complaint intention.
Design/methodology/approach – More than 16,000 posts (of complaints and complainants’ follow-ups) on 13 airline Facebook pages were
analyzed using partial least squares.
Findings – This empirical study found that customers’ complaints are attended to the extent the complaints are followed up with more comments as
supported by VPOs suggesting a contagion effect. Besides, it appears, the squeaky wheel is the one that gets the grease. The interactive virtual
presence of others does not have an effect on a firm response toward them, despite their support to complainants and calling the airline to act.
Originality/value – To date, previous experimental studies have only featured a single VPO in potential complainants’ responses; however, in this
empirical study the authors take into consideration the interaction of several VPOs in the service failure. Also, the focus is on the influence of VPOs
on customers’ complaints rather than the other way around.
Keywords Social media complaints, Interactive virtual presence, Virtual presence influence, Firm response
Paper type Research paper

Introduction (Armstrong et al., 2021b) and the effect of organizational

complaint responses on active observers rather than simply the
Service firms may inevitably suffer service failures, and they are complainant (Armstrong et al., 2021a). More broadly, by
considerably more damaging in the new “Service Encounter studying the effect of VPO on complainants and firm
2.0” environment (Lariviere et al., 2017) where customers take responsiveness, this investigation addresses the call for research
increasingly active roles and interact with other customers on the interdependent nature of customer engagement and
beyond the physical constraints of space. When facing service relationships in the multi-actor service ecosystem (Sharma,
failures, companies not only do deal with the complaints and et al., 2020).
rage in social media of directly affected customers but also the In response to these calls, this research digs data from airline
behavior of other customers online. Perhaps this is one reason service failures posted on their Facebook pages and analyzes the
for firms’ reluctance to handle complaints in social media communication exchanged between customers’ complaints,
(Schaefers and Schamari, 2016). their reiteration (follow-up) of the complaints, online
As of today, the effects of social media complaints and bystanders and firms.
recovery on third-party online bystanders have received limited The little research, in the online context and especially in
attention (Hutzinger and Weitzl, 2021). Within this stream of social media, investigates the impact of service failures on
research, not much is known about the effect of what effect online bystanders such as Schaefers and Schamari (2016),
active virtually present others (VPO) have on customers’ Hogreve et al. (2019) Hutzinger and Weitzl, (2021), Huang
reiteration (follow-up) of the complaints and firm and Ha (2020) and Armstrong et al. (2021a). Our study focuses
responsiveness. This research responds to recent calls to on the effect of active VPOs on directly affected customers and
consider the effect of other consumers and their comments on the firms. We are interested in those online bystanders that
the complaints of others in Facebook business pages interact with complainants and firms. Bystanders’ perceptions
and interactions with complainants and firms are important
The current issue and full text archive of this journal is available on Emerald because their perceptions of firms’ service recovery can shape
Insight at: https://www.emerald.com/insight/0887-6045.htm their behavior toward firms (Purnawirawan et al., 2015),

Received 22 August 2022

Journal of Services Marketing Revised 9 January 2023
© Emerald Publishing Limited [ISSN 0887-6045] 30 January 2023
[DOI 10.1108/JSM-08-2022-0276] Accepted 22 March 2023
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

inflicting additional damage to firms in social media if service effect of the engagement of VPOs with the complainant as
recovery is perceived as negative, and they can exert additional endorsers of the complaints. Previous experimental study
influence on complainants to persevere in the complaint. (Armstrong et al., 2021a) has only featured a single VPO in
We use the terms online bystanders, or VPOs potential complainants’ responses, however, in this empirical
interchangeably and define them as: “consumers who are in a study we take into consideration the interaction of several
public online environment and can observe and interact with VPOs in the service failure. Finally, this research responds to
complainants who are in dialogue with the firm” (Armstrong recent calls on the effect of organizational complaint responses
et al., 2021a, p. 34). on active observers rather than simply the complainant
We wonder if the reaction to complaints by VPOs, based on (Armstrong et al., 2021a).
actual in situ data in Facebook does have an effect on
companies’ response and customers following up on their Overview of the influence of virtually present
complaints. According to the social influence theory (Kelman, others on complaints and firm responses in social
1958), online bystanders should not be influenced by the action
media
of unrelated individuals (nonreferent others) online.
The research questions that lead our study are: To what In an era of abundance of consumer generated content, it may
extent do customer’s complaint follow-up affect the airline not come as a surprise that consumers engage with one another
responsiveness? Do VPOs influence customers’ follow-up posts on social media to inform companies about their failures, to
on Facebook? Do VPOs’ engagement with customers’ elicit companies’ prompt responses, and to achieve
complaints and firm responses to complaints influence airlines compensation (Hennig-Thurau et al., 2004; Xu et al., 2016). A
to respond to them? report on Czech Facebook users confirms that these individuals
This research contributes to the conversation on the are fully aware that complaining publicly via Facebook can get a
influence of customers copresence (Colm et al., 2017) in social company in serious trouble (St
ríteský and Stranska, 2015). For
media specifically complaints in airline Facebook pages, mostly companies, the complaints made public on social media pose
ignored in empirical studies. Previous research on the influence not only a threat of losing current clients if the company does
of consumer social presence in service settings has been based not respond satisfactorily but also, the complaints have the
on physical social presence (He et al., 2012) rather than virtual potential to affect bystanders (Hogreve et al., 2019). These
presence (Chen et al., 2014). Moreover, online studies that negative comments by the author of a complaint signal
have investigated social presence have done so by examining important information to online bystanders affecting their
comments and reviews by third parties, especially in the attitudes toward the companies under scrutiny (Hutzinger and
hospitality industry (Sreejesh et al., 2019; Wan et al., 2011). Weitzl, 2021). From online reviews, we know that negative
These studies do not properly represent the social presence reviews are more useful to readers than positive ones, therefore
effect; they have been static (posts, that are notably positive, are companies need to mitigate these comments (Sen and Lerman,
left on review pages), with low social interactivity among 2007). On the other hand, virtual presence, whether interactive
reviewers, and they are product centric, i.e. reviews of products or passive, is critically important when companies try to resolve
(Ma et al., 2015). However, this study investigates VPOs’ service failures because the positive effects of successful
comments in a dynamic mode as they interact with the recovery in social media will be significantly higher due to such
complainants in Facebook (a user-centric platform). In virtual presence (Chen et al., 2014).
addition, most of the research testing the effect of complaint A variety of studies have examined the role of peer
handling or service recovery on online bystanders on social relationships and their influence on deviant behaviors in a wide
media has been experimental. This in no way means to deny spectrum of fields such property crimes, computer crime like
experiments are not empirical. Although respondents may digital piracy, computer hacking and password-guessing
genuinely report the scenarios are realistic, and are assumed to (Miller and Morris, 2014). Some evidence supports the idea
experience the failure, their responses on measurement scales that virtual peer interactions have a substantive impact on an
do not reflect the actual experience but consumer perceptions. individual’s behavior. In a series of experimental designs,
In this regard, some suggestions have been provided to increase Schaefers and Schamari (2016) compared mere virtual
the realism of experiments such as the use of 3D animation, presence (VP) (i.e. read comments only) with interactive VP
avatars and use of video vignettes to make respondents visualize (i.e. others comment on the service provision). Interactive VP
a real consumption experience (Gelbrich and Roschk, 2011). was found beneficial when the interaction was of a positive
Our empirical study contributes in three ways. Firstly, it valence and when the company resolved the complaint and it
enriches the literature on the effect of complaints on VPOs on led to attitude polarization. Using multiple experiments,
social media. There is not much investigation on the influence Hogreve et al. (2019) found that service recovery transparency
of social media and the presence of active online bystanders on (i.e. the complaints and recovery service are visible to the
firms’ response to complaints (Gunarathne et al., 2017) nor the public) can create positive quality signals and improve trust
influence of VPOs on complainants’ behavior and firm perceptions, thereby enhancing virtual presence of others’
response (Armstrong et al., 2021b). More broadly, this study intentions toward the provider. When firms handle the
contributes to the call on the role of the “other” actors in the complaints opaquely (i.e. they ask the complainant to discuss
Service Encounter 2.0 context. Second, our study the matter privately), it creates a negative effect on the virtual
complements experimental studies by analyzing actual presence of others, especially if the service recovery is
reactions (reported experiences in Facebook) to a service unsuccessful. The results are consistent with the theoretical
failure. Furthermore, this empirical study sheds light on the framework suggesting that engaging in proactive transparency
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

through self-disclosures, companies can prevent social heterogeneous set of constituents than in the traditional media
disapproval from generating (Wang et al., 2021). era. These mechanisms of social media influence firms’
A company response to a complaint also influences online response to social disapproval, that is, “constituents general
observers on social media. Companies that use a high enmity toward a firm that reflects their intuitive negative
conversational human voice (i.e. a more natural, close and perceptions” (Wang et al., 2021, p. 276). According to the
human style of online communication (Barcelos et al., 2018) in model, companies have the options of responding widely to all
their complain handling affect observers’ perceptions of how constituents or pursuing a “reactive reticence strategy” that
fair the applied procedure and outcome of the complaint consists in limiting the scope of the response, or in not
handling are (Javornik et al., 2020). An informal (human) tone responding at all.
(i.e. language that occurs in casual conversation rather than in Social media technologies afford consumers to publish
written language) can decrease the perception of procedural content (e.g. commentaries, complaints and preferences) in a
and distributive justice, whereas a conversational human voice costless manner, with less concern for censorship (Hermida
translates into more positive interactional justice perceptions. et al., 2012), as frequently as they want (persistence), and such
In relation to the length of the company communication, while content is visible to others such that VPOs can react
no direct effects on justice perception are observed, it does (metavoicing) online; a single complaint is amplified and the
affect the observers’ satisfaction with the complaint handling firm response is open to scrutiny (Sun et al., 2021). These
through procedural justice (Javornik et al., 2020). affordances can accelerate the spread of complaints (Wang
Observers’ perceptions are also influenced by how warm the et al., 2021). For example, David Dao’s screaming while he was
company response is perceived. Warmth-oriented responses to being dragged out of the United Airlines flight in 2017 went
online complaints make a stronger impression and are viral and millions saw the videos posted in social media in
perceived as more sincere than competence-oriented matters of hours (Bever, 2019).
responses. This perception is moderated by the observers’ An earlier study notes that 46% of individuals have
relationship orientation, i.e. individual sensitivity toward the complained in Twitter and Facebook (Istanbulluoglu, 2017)
needs of others in social relationships (p. 624). Those with a and, despite the fact that many passengers do not receive a
communal (versus an exchange) relationship orientation response from an airline on Twitter, they still indicated they
considered warmth-oriented responses to be more relevant and would continue using Twitter to complain, presumably, to
more sincere than competence-oriented responses (Huang and publicly punish the brand (St
ríteský and Stranska, 2015). An
Ha, 2020). A personalized company response style to a exceptional example of persistence is Canadian musician Dave
complaint (compared to an automatic response) in Twitter also Carroll’s “United breaks guitars” video who named and
corroborates the influence of the style of company response. shamed the airline on YouTube and finally received
Personalized responses led to more favorable perceived brand compensation after a lot of attention in the public sphere
image, purchase intentions and eWOM than automatic (Negroni, 2009).
company responses. Furthermore, when the complainant was As comments from complaints accumulate and persist on
an influencer (compared to an ordinary Twitter user), a social media, firms are more inclined to respond for fear these
personalized response did not immunize the company from a complaints may become viral (Lopez-Lopez et al., 2021).
decline in purchase intention, brand image or intention to Recent literature suggests (Tellis et al., 2019) that emotional
engage in eWOM (Lopez-Lopez et al., 2021). Emotions also content and popular customers or actor prominence (Wang
play a role in observers’ perceptions. A personalized response et al., 2021), those whose comments resonate more among
generates positive emotions that lead to higher perceived brand other publics, or have more followers, are more likely to receive
image, purchase intention and eWOM when the complainant responses from companies in general than others with less
was an ordinary Twitter user. However, when the complainant followers.
was an influencer, the results were opposite damaging all three Previous studies on organizational responses to
marketing outcomes mentioned above (Lopez-Lopez et al., complaints have identified six different dimensions:
2021, p. 675).
timeliness, facilitation, redress, apology, credibility and
A tabular summary of the relevant literature reviewed is
attentiveness (Davidow, 2003). Firms that do respond to
provided in the appendix.
complaints promptly influence positively complainants’
We the previous literature in the background, we next
experience evaluations and their interactions with VPOs
develop the hypotheses in this study.
(Bacile et al., 2020), and influence the attitude toward the
firm (Armstrong et al., 2021b). Moreover, those firms that
Hypothesis formulation personalize the responses and provide evidence of redress
Our first hypothesis tests whether customers’ complaints and lead to more positive observers’ perceptions of complaint
their reiteration of the complaints (follow-up) on the firms’ handling (Javornik et al., 2020) and increase the likelihood
Facebook pages put organizational accountability pressures to of observers to complain (Armstrong et al., 2021b).
respond to the claimants. Because complaints can be categorized as social disapproval
We draw on the model of social disapproval (Wang et al., of a firm practice, in our research, the practice of cancelling
2021) to justify why airlines will respond to accountability flights without redress at the time of the announcements, the
pressures. The model posits that, in the new social media era, complaints are emotionally laden, and they were populated
greater velocity, emotionality and communality, expose firms quickly by a large number of constituents on the Facebook
to faster, more emotional evaluations from a more company page, we predict the first hypothesis:
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

H1. As the volume of follow-ups by complainants increases, H2. The more virtually present others endorse complainants
the accountability pressures on organizations to respond and call the firm to act, in the Facebook business page,
to the complaints expand. the more firms will respond to them.

Service encounters are usually embedded in a complex web of Research evidence, based on social learning theory, has found
spectators and participants, more so in an online environment, that observers of online public communication between firms
elsewhere named “Service Encounter 2.0” (Lariviere et al., and customers are influenced by their peers’ comments and
2017). In a review of the literature on third parties’ roles behaviors online (Miller and Morris, 2014; van Vaerenbergh
between the interaction of customers and service providers, et al., 2013). The coparticipation that emanates from the
Abboud et al. (2021) identified five roles nontechnological interaction between complainant and observers create group-
parties can adopt, namely, bystander, connector, endorser, based emotions that influence the firm recovery efforts (Maher
balancer and partner. The bystander can play a passive role or and Sobh, 2014). An experimental study conducted by
an active one. In the latter case, the third-party bystander can Facebook suggested that emotional contagion occurs online
act so by observing the nonresponse of a service provider to a even in absence of nonverbal cues typical of in-person
complaint. It is of interest, in this research, the role of the interactions (Ferrara and Yang, 2015). Other experimental
endorser, that is, a third party that provides informational, study confirms that complaint behavior of others increases focal
emotional or practical resources to the customer or service customers’ complaint intention because watching others
provider (Abboud et al., 2021). Several studies have complain causes customers to feel more angry (Du et al., 2014)
demonstrated this role when third parties give advice to a and give the complainant greater perceived power (Armstrong
customer (Becker and Pizzuti, 2017), react to complaints et al., 2021a). There are several examples that demonstrate how
(Gregoire et al., 2015) or share positive or negative consumers express their displeasure and anger on social media
consumption experiences during an online exchange (Lopez- and influence others to boycott firms (Makarem and Jae,
Lopez et al., 2021). 2016). A study in the telecommunication sector found that
Research on complaints on social media indicates complaints customers that are in a negative state are more likely to
go beyond the customer directly involved in the service failure complain when they hear more positive messages from referent
and it reaches “silent observers” (Armstrong et al., 2021a). others (Ma et al., 2015).
Hence, we hypothesize that as more and more social media Some of these referent others act as advocates of the
bystanders engage in the original complaints, and call the firm complainant, that is, they engage positively with the
to act, the more pressure for firms to respond. The extant complainant by indicating “likes,” making positive comments
literature advanced two reasons: One is that, these online on the complainant’s diatribe and commenting against the
bystanders, by observing the public interaction between the company. These Listeners (mere virtual presence) or Helpers
company and the customer, may expect a similar treatment in the online customer to customer interaction (Xu et al., 2016)
(Abney et al., 2017) after all, the presence of complaints of can influence a target consumer’s brand evaluations and
others is a visual cue that influences a consumer’s decision to purchase intentions (Naylor et al., 2012). In contrast, some
complain on a Facebook business page (Armstrong et al., VPOs, classified as Ironists (Xu et al., 2016), may engage in a
2021b). Two, the bystanders on social media could be potential negative way with the complainant, making positive remarks
consumers, or previous customers, that may not have had a bad about previous experiences with the company, trivializing or
experience yet, therefore a firm response will signal to them the doubting the complaint, or making negative comments on a
airline does care for passengers. complainant’s comment page.
Recent experimental investigations of people observing Hutzinger and Weitzl (2021), through a series of experiment
complaints in social media found that a firm response can signal and scenarios, demonstrated the importance of the interaction
cues for observers of how a complaint is handled (Armstrong between VPOs and the complainants’ comments on the brand
et al., 2021b), affect their perception of fairness in the attitude of online bystanders. When the VPOs engage in
communication exchange (Javornik et al., 2020), elicit trust and positive comments supporting the company, and the recovery
improve bystanders’ word of mouth (WOM). And when the response is initiated by the company marketer vouching, the
service recovery is unsuccessful, the signaling to online positive comments of VPOs have no incidence on bystanders’
bystanders has negative implications on WOM, overall brand attitude. However, negative comments of VPOs have a
satisfaction (van Vaerenbergh et al., 2013) and purchase clear negative influence on online bystanders’ brand attitude
intentions (Hogreve et al., 2017; van Vaerenbergh et al., 2013). that are highly susceptible to normative influence, even when
Furthermore, we believe firms will respond to active VPOs the initial recovery response came from the firm marketer.
because the presence of supportive VPOs contributes to When the initial recovery response is consumer vouching, the
creating a collective public outrage pressuring needed change negative comments from VPO have no effect, while positive
and holding companies accountable for their actions comments of VPO have a positive impact on bystanders’ brand
(Champoux et al., 2012). attitude. Other studies (Fisk et al., 2010; Nicholls, 2010; Xu
An alternative view suggests that firms may not want to et al., 2016) that have evaluated the interaction between
respond widely to complaints to avoid opportunities for complainants and other online bystanders, note the disrupting
disgruntled constituents to reengage with the story and effect that dysfunctional behavior enacted by an online
intensify their emotions, which can spread social disapproval bystander can create on the service experience for
(Wang et al., 2021, p 282). complainants, customers, employees and firms. The complaint
Therefore, we predict the second hypothesis: contagion effect has been found to be stronger for customers
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

with high social identification with others, high perceived Methodology

credibility and weak prior relationship of complaining
Sample and data collection
customers with the service provider (Chen et al., 2020).
The data for this study consist primarily of Facebook posts
Supportive comments to the complainants made by VPOs
available during the period March 8–9, 2020, concerning the
also influence others to make observable complaints, increase
airlines’ own announcements such as cancellations due to
the likelihood to complain on the company Facebook page and
COVID-19 restrictions, complaints by passengers, carriers’
reduce their likelihood to make a nonobservable complaint via
responses, customer responses and online bystanders’
Facebook Messenger (Armstrong et al., 2021a). However, the
comments on an author’s complaint/comment. The reason to
likelihood of making a complaint by either supportive or
concentrate on Facebook is that this platform is one of the most
nonsupportive VPOs is influenced by the perception of social
preferred by customers to voice their issues (Einwiller and
risk (i.e. the probability of being scrutinized publicly) and the
Steilen, 2015) and Facebook business page allows for
intervention of an employee with the comments (idem). Hence,
comments to be posted by consumers who view and often
we posit:
interact with, the online complaint process (Armstrong et al.,
H3. The more virtually present others engage with the 2021a). The reason for concentrating on this one industry is
complainants, endorsing the complaint, the more follow- that it received a disproportionate number of complaints due to
ups by complainants. issues with refunds for unused travel. This allowed the
researchers to examine, within a short period, a large volume of
Figure 1 illustrates the structural model to be tested. The data on complaints, bystander reactions to complaints and firm
model hypothesizes that H1 the more complainants persist responses. The selection of the two days, namely, March 8 and
in following up with their complaints on the social media 9 were chosen because on these days, airlines posted several
channel, the more likely an airline will reply to passengers’ announcements on Facebook about flight cancellations due to
complaints; H2 predicts the more VPOs engage with the the flight restrictions imposed by European governments. This
passengers’ complaints and call the airline to take action, was the start of complaints on Facebook therefore it was
the more likely VPOs will receive a response from the relatively easy to observe the history of stimuli (thread of
airline; also, H3 predicts the more VPOs engage with interactions) that complainants, online bystanders and firms
passengers’ complaints, the more complaining authors will were exposed to and the responses they produced. The two
post follow-up comments. Finally, the airline type is days provided 16,810 comments which include 6,409
considered a control variable as budget airlines are less complaints, 2,348 complainants’ follow-ups, 6,017 online
likely to reply than flag airliners. bystanders’ engagements and 2,036 airlines’ responses.

Figure 1 Structural model of the influence of VPOs

Number of “thumb Number of “Like”

Number of + up” to complaint of the complaint
posted
comments

H3 VPOs Control Variable:

Claimant’s engagements Type of airline
follow-ups
H2

H1
Airline
Response

Number of Interacons between

answers of the claimant and the
airline airline

Source: Own work

 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

Although two days may seem a short period to gather data, short phrases, e.g. “feeling happy” (Shawky et al., 2020), and in
most customers normally receive responses to their complaints some other social media sites engagement is symbolically
within 1–3 h and a very few receive a final message more than manifested in “thumbs-up,” “likes,” commenting and sharing
48 h after complaining (Istanbulluoglu, 2017). (Khan, 2017; DiStaso et al., 2015). In the case of the airline
We qualified and quantified each message thread from the response, we used two observable variables, namely, whether
VPO as to who it was directed: to the complainant, the firm or the airline replied to the complaint or not and the number of
other online bystanders. The firm response was also classified interactions between the airline and the complainant. This
as directed to the complainant or to the active online bystander second measure served the purpose of validating the firm
that connected with the firm. For example, for a complaint to response because some responses suggested an automatic
Lufthansa: “I have been trying to reach your Johannesburg response.
offices for the past 2 days. I’ve held on for longer than 25 min
with no answer! They do not reply to my emails either. I need to Multivariate analysis tool
know the status on our refund for our flights to Milan please! To test the hypotheses, the SmartPLS software version 2 M3
We cancelled on the 25 February – and have not received was used and ADANCO software to test discriminant validity
anything as of yet. . .. . ..” Reply from Lufthansa: “Hi (name (heterotrait-monotrait [HTMT] criterion). The use of partial
omitted), please note that it may take a while until refunded
least squares-structural equation modeling (PLS-SEM) was
amounts are visible on your statement. The agents in the
preferred to the traditional covariance-based (CB-SEM)
Service Center can verify the status of your refund for you –
because of the former’s ability to estimate complex models, its
however, please note that they are facing a high call volume at
emphasis on rigorous prediction perspective and the
the moment, so longer waiting times are expected. I can only
achievement of high levels of statistical power (Sarstedt et al.,
advise you to keep on trying to contact them in short, regular
2022). The following standard bootstrapping settings were
intervals. Thank you./(named erased).” Follow-up example:
used to determine T-values and confidence intervals, PLS
“Lufthansa Thank you Name omitted! I will wait. Please can you
assist with the status of the 2 following flights for Sunday 15/03/ algorithm – settings, sign changes – no sign changes, 4,000
2020 LX2813 and LX4370 as we fly out then to Malta.” samples of 300 cases. The standard PLS procedure was
Example of a complainant of Ryanair: “Bloody ridiculous – adopted: path weighting scheme, Mean 0, Var. 1, max.
you aren’t flying to my destination because the country is on iterations 4,000, abort criterion 1.0
105 and initial weights
complete lockdown. Yet you still won’t deem the flight 1.0. In evaluating and reporting the results, guidelines for PLS-
cancelled for us customers to get our monies refunded. I have SEM suggested by Chin (2010) and Hair et al. (2012) were
tried calling, I’ve tried emailing, I’m tried Facebook[. . .]See followed.
More” (No response from Ryanair). Example from an active
online bystander: “Same for me. How can the reservation Results
number be wrong when you copy and paste. Does Ryanair not Descriptive analysis
want to refund us OUR money. It’s getting ridiculous now.” Airlines responsiveness: In general, airlines answered less than
To have a comprehensive coverage of European airlines, we
one-fourth (24.24%) of the total complaints received. This is in
chose 13 top flag and low-cost companies for this study (similar
line with an annual survey from A T Kearney that found 70% of
to other studies, e.g. Chang and Hung, 2013; Mason, 2001),
complaints are going unanswered (Austin, 2012), and is
namely, Alitalia, Air Dolomiti, Air France, Austrian Airline,
surprising given the evidence that ignoring complaints on social
British Airways, EasyJet Finnair, KLM, Iberia, Lufthansa,
media make consumers less likely to buy from such companies
Swiss International Air Lines, Vueling and Ryanair. The
(Armstrong et al., 2021b). The difference between brands and
collection was done between March 22 and 23, 2020 using a
python script. Although personal information is available on type of airline is remarkable. For instance, AirDolomiti stands
Facebook (Chu, 2011), and is voluntarily disclosed by out for being the one that answered the highest number of
consumers with the interest of engaging online (Lee et al., complaints with a response rate of 70.83%; second, Vueling
2018), we chose not to collect any personal information and attended 68.79% of its complaints and Lufthansa 62.17%. On
replaced names with codes. Each active VPO was identified in the other hand, all other brands did not respond to more than
each message thread and his/her message was classified as to 35% of the complaints, leaving a very large number of
who it was directed: to the complainant, the firm or other unattended complaints. On the extreme, Ryanair (a budget
online bystanders. The firm response was also classified as airline) attended only to 0.65% of complaints. Please see
directed to the complainant or to the active online bystander Figure 2 for details.
that connected with the firm. Customers’ complaints and their follow-up: Not every customer
followed up on his/her complaint. Indeed, 69.11% of the total
Variables complainants did not make any follow-up at all. 19.39% of the
Drawing on the threads of interaction in each airline Facebook authors made just one follow-up commentary and only 5.83%
pages, we identified seven observable variables (described in made two follow-ups. In addition, descriptive analysis shows
Table 1) to reflect on two latent constructs, namely, airline that customers that did not make any follow-up in flag airlines
response, and VPOs engagement. The observable variables is a little bit higher than low-cost airlines although the
reflecting engagement used in this study are not uncommon in difference between flag airline and low-cost is quite modest
social media. For example, Facebook is continuously adding resulting in 67.49% and 74.10% for low-cost and flag airlines,
new and easier ways for actors to interact through emojis, GIFs, respectively. Presumably, customers using a low-cost airline are
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

Table 1 Description of the variables

Construct Observable variable: description
Customer’s complaint In general, is a piece of text that contains dissatisfaction with a service or lack of it (e.g. flight cancellation). It
includes negative communications to the responsible company. Specifically, we followed Decock and Depraetere’s
(2018) definition, to distinguish the initial complaint from the follow-up comments. A complaint consists of four
components which (A) the event about which the customer is complaining. (B) the expression by the complaining
customer of some degree of dissatisfaction with respect to the product/service. (C) identifies the person or
institution that is considered by the customer to be responsible. Finally, component D concerns the customer’s wish
for the offence to be remedied
Example of complaint
Hello, Lufthansa, the French number of Lufthansa isn’t working. And I’m stuck in Bali. A solution maybe?
The action from Lufthansa resulted not satisfactory for the complainant
Customer’s complaint follow-up “It is a content published by the complainants after they have posted a complaint. It may be a response to the
airline or an insisting note requesting a response from the airline
Example of follow-up from complainant”
Lufthansa the French number is disconnected. Why you don’t understand that?
Engagement of the interactive Engagement is a piece of content published by online (active) bystanders to comment, cite similar experience, or
virtually present other (VPO) provide an opinion regarding a complaint. These comments represent strong (active) form of virtual presence. For
example, an online bystander commented to a complainant in Ryanair: “Maybe this petition is worth signing.”
The complainant replies to the online bystander: “ I honestly don’t think they care enough to be honest they are
useless and no motive up at all explaining why it’s happening just ignoring people.” The engagement can also be
in the form of a symbol (tick on “like” or ‘Thumb up”). This is a form of shallow engagement in the virtual
presence
Number of likes of the complaint:
The number is calculated by summing the “Likes” to an author’s complaint by online bystanders. This represents a
shallow virtual presence
Number of likes of the customer’s complaint follow-up:
The number is calculated by summing the “Like” to an author’s follow-up by online bystanders. This represents a
shallow virtual presence
Airline response Interactions between the complainant and the airline:
It is calculated using a pair of comments between Airline and Author of the complaint where first, the Airline
answers the Author of the complaint and then the Author follows up by answering to the airline response. If there
is a pair, in this order, we count it as one interaction
Example of airline response:
Hello E (name disguised), if you are unable to get through, please try calling in short, regular intervals. You may
also contact the colleagues for a rebooking after departure, if your flight has been cancelled. Here’s information on
our rebooking policies due to the current situation: http://bit.ly/38qPnOM./L(name disguised)
Number of comments of airline:
Numerical value that results from observing if the airline replied to the claimant. It takes the value of 0 when the
airline has not replied the claimants’ complaints and 1 if it has
Interactions between the VPOs and the airline:
Numerical value that results from counting the answers provided by an airline to VPOs
Airline type Budget airline: Also called low-cost carriers which are airlines that offer lower fares and fewer amenities than
traditional full-service airlines
Flag airline is an airline that is owned or subsidized by the country in which they are registered. Normally
associated with a traditional full-service carrier
Source: Own work

more price sensitive therefore are willing to put an extra effort algorithm emphasizes the use of composite reliability because it
to follow up on their complaints. is generally regarded as a more appropriate criterion of internal
consistency reliability because it considers the weight of each
manifest variable, whereas Cronbach’s alpha assumes all
Multivariate analysis manifest variables have equal weights (Hair Jr et al., 2014). All
Previously to addressing the hypotheses of the study, evidence reliability scores seem to be acceptable, i.e. above 0.70. Note
that the partial structural model satisfies certain minimum that some variables are only reflected by one observable
evaluation criteria must be provided, especially because two of variable therefore reliability equals 1.
the variables are latent, namely, VPOs’ engagement and airline Table 2 shows that the two latent variables have inter-item
response. Internal consistency reliability (Cronbach’s alpha reliability on two measures, namely, composite reliability and
and composite reliability) should be above 0.70. PLS-SEM Cronbach’s alpha, and discriminant validity using the Fornell–
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

Figure 2 Airline response to complaints

Table 2 Evaluation criteria of latent variables

AVE Composite reliability R-square Cronbach’s alpha
Airline type 1 1 0 1
Firm reply 0.825 0.9339 0.245 0.8937
VPOs engagement 0.7936 0.9498 0.1942 0.9314
Author’s complaint/follow-up 1 1 0 1
Discriminant validity: heterotrait-monotrait ratio of correlations (HTMT)
Construct Firm reply VPOs engagement Author’s complaint/follow-up Airline type
Firm reply
VPOs engagement 0.1127
Author’s complaint/follow-up 0.5409 0.4019
Airline type 0.1214 0.0646 0.0417
Source: Own work

Larcker criterion and the HTMT ratio of correlations response from an airline. Because the value of the coefficient is
(Henseler et al., 2015). positive and the code of the airline was 1 = budget and 2 for flag
Convergent validity can be demonstrated by observing airlines, we can conclude flag airlines will be more likely to
average variance extracted (AVE) greater than 0.50. The two respond to complaints and follow-ups. As expected, the
latent variables, namely, firm reply and other passengers’ authors of the complaints and their follow-ups also predict the
engagement have AVE equal to 0.89 and 0.79, respectively. airline’s response. The higher the number of follow-ups
Discriminant validity was tested using the HTMT ratio of performed by the claimant, the more likely an airline will
correlations which is based on the average correlations of the respond (H1 is supported). VPOs’ engagement as endorsers
indicators across latent constructs. The acceptable level of influences the author of the complaints to follow-up his/her
discriminant validity is <0.85 (Henseler et al., 2015). As seen in complaint (H3 is supported). However, online bystanders that
Table 2, the correlations range from 0.0417 to 0.5409 engaged with complaining customers and that interpellated the
confirming discrimination between the latent constructs. firm do not receive a response from the airline (H2 is not
Table 3 reproduces the path coefficients of the model and the supported). Figure 3 shows the main findings of the posited
T-statistics. Complainants’ follow-ups R2 (adj) = 0.14 and firm relations.
response R2 (adj) = 0.26 are both statistically significant. Cohen’s
f2 for VPOs engagement ! complainants’ follow-ups = 0.16 and
Discussion
complainants’ follow-ups ! airline response = 0.36 are both
statistically significant. Customer complaints on social media represent a major
From Table 3 we can conclude that airline type is statistically concern for companies due to the easiness with which the
significant indicating this variable predicts beyond chance a complaints can be posted and the number of people that can be
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

Table 3 Path coefficients

Original sample Sample mean SD SE T-statistics
Airline type fi airline response 0.1476 0.147 0.045 0.045 3.2777
VPOs’ engagement fi airline response 0.0713 0.0181 0.2591 0.2591 0.2751
Author’s complaint/follow-up fi airline response 0.5045 0.5289 0.1827 0.1827 2.7612
VPOs engagement fi Author’s follow-up 0.4406 0.5854 0.1591 0.1591 2.7695
Source: Own work

Figure 3 Indicators’ loadings and path coefficients

Number of “thumb Number of “Like”

Number of + up” to complaint of the complaint
posted
comments
0.784
0.846 0.937

0.585 VPOs Control Variable:

Claimant’s engagements Type of airline
comments
-0.018
R2 = 0.144

0.147

0.528
Airline
Response
2
R = 0.261

0.897 0.924

Number of Interacons between

answers of the claimant and the
airline airline

Source: Own work

reached. The impact of complaints on online bystanders, the response rate in our study may also reflect the assumption of
interaction of bystanders with complainants and their effect on the airlines that there was no need for a response since the flight
a company response have hardly been tested again in situ cancellations (and non-reimbursement tickets as a result of a
empirical data or real consumer experience rather than based crisis) were supposed to be understood by consumers. And they
on consumer perceptions based on scenarios. Our study could have felt uncomfortable with their responses being visible
captured data from an actual event of dissatisfaction with to others on the platform. Alternatively, it could reflect a lack of
airlines and the effect of online bystanders on airline response human resources given the pandemic. More explanations
to complaints posted on Facebook in a dynamic content among follow under the section related to specific response to VPOs.
hundreds of active and nonactive online bystanders.
Our results show that, on average, only a quarter of the Airline responsiveness to customers’ complaints and
companies replied to complaining customers. In a normal their follow-ups
situation, a firm’s first response to a complaint, obtained from The positive relationship between complainants’ follow-ups
surveys, has varied between 33% and 48%, but customers that and airlines responses suggests that airlines are more reactive
actually engage in a conversation with the firm drops to 19.9% and are pressed to respond as a result of the increasing number
(Istanbulluoglu, 2017). Note that our study not only considers of follow-up comments. The results are consistent with the
the first response by the firm (that may be automatic) but also organizational accountability view that posits as the volume and
engagement with the firm. visibility of social media customers complaints and their follow-
Perhaps, airlines felt that replying to complaints on social ups increase, the accountability pressures on organizations
media would bring more complaints therefore they abstained. expand (Karunakaran et al., 2021). It is possible that
There is evidence that complaints, when not addressed companies perceive that, as the number of complaining
properly, bring more complaints (Ma et al., 2015). The lower comments increases, the more other consumers can view the
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

posts, and get on the wagon creating further negative impact on the fact they target price-sensitive customers, they assume
the firm. The results of this study are consistent with the customers will not complain if not attended when they
reaction of managers to customers’ complaints that fear the complain and that they will continue using the service due to
complaints on social media will damage the brand publicly. A the low price.
caveat is important; our study is based mainly on the “voices”
of customers. Other studies (Ma et al., 2015; Sun et al., 2021) Virtually present others’ engagement with an author’s
that have looked at the quality of the response (e.g. promptness, complaints and follow-ups
effectiveness) or the relationship with the brand, indicate that Our research contributed to the investigation of the complexity
managers are over reacting to customers’ voices of complaint. and dynamics of intertwined complaints and responses not only
When the firm service intervention provides complaining between firm and complainant but also between complainants
customers with a prompt and effective response, the company and active online bystanders in social media. An increase of
action leads to fewer future complaints (Sun et al., 2021). It has comments by VPOs as endorsers motivate the authors of a
also been noted that some firm responses may have been complaint to follow up on their complaints, and gives them
automatic therefore it is questionable if the claimants actually power through WOM (Minazzi, 2015). Our results concur
influenced the firm. We believe this is not the case because firm with social learning theory that an individual can learn specific
response consists of two observable variables. We concede that behaviors by simply observing others but contradicts the social
one may be biased as it reflects dichotomous behavior (the firm influence theory that suggests that virtual exposure to unknown
replied or not) but the second reflects the number of others should have little social influence. Our study provides
interactions between the claimants and the firm. evidence of the effect of complaints on third-party online
bystanders and the potential effect on firm response. We should
Airline responsiveness to online bystander’s note here that our sample only considered the interaction of
engagements VPOs with the complainants in a positive manner (endorser
We indicated at the beginning of this article that this research role). There is evidence that VPOs engage negatively with
responds to recent calls on the effect of organizational complainants, sometimes in an uncivil way (Bacile et al., 2020).
complaint responses on active observers rather than simply the
complainant (Armstrong et al., 2021a). We find that airlines do Conclusions and managerial implications
not respond to VPOs despite the fact they endorse the Several practical implications have derived from the finding of
complainants and address the airlines to respond. The results this study. Our study reveals that Facebook represents a viable
were unexpected since given the social media affordances, the channel when passengers seek urgent solutions from airlines in
impact of negative brand evaluations, reputational damage and critical events and airlines. Most companies did not reply to
spread of platforms, organizations are more pressured to the complaints, perhaps because of the crisis they faced. Many
account to various constituents (Karunakaran et al., 2021; companies diverted the attention of the customer to other
Wang et al., 2021; Van Noort and Willemsen, 2012). Different channels corroborating previous study (Van Herck et al., 2020;
reasons can explain this unexpected result. First, it is possible Einwiller and Steilen, 2015). For instance, an airline stated that
that VPOs, by engaging with complainants, made the “Hi, unfortunately we do not have access to customer systems
companies less propense to reply. Gunarathne et al. (2017) on here as Facebook is run by the marketing team. Please, can
found that complainants that mentioned another individual in you get in touch with our customer services team via Twitter
the complaint reduced the chance of receiving a response by (@____) and they should be able to help you.” Airlines need to
42.17%, and if a competing airline was mentioned in the rethink the role of social media and invest in dedicated
complaint, the odds of receiving a response is reduced by marketing staff to attend customer enquiries. Diverting
16.55%. In our study, it is the VPO that mentions another attention to other channels, normally private, may irritate
party, namely, the complainant. customers even more, especially if they have already tried more
Perhaps VPOs did not receive a response from the airlines personal ones such as the phone.
because they are afraid of being judged in public (Gunarathne Second, we noticed that some airlines were using standard
et al., 2017) and for fear the replies could trigger more instead of personalized answers to customers such as “Hi, we’re
complaints and intensify their emotions (Ma et al., 2015). doing our best to assist every one of our customers. . . .” or “We
Another possible reason is that airlines did not feel obliged to are currently experiencing high volumes and are doing our
respond to VPOs’ engagement because the airlines did not utmost to assist as many passengers as possible. . .” which might
consider this segment as relevant, and more so, in a critical not increase the communication effectiveness between the
situation of massive failures that was out of their hands due to parties (Kerkhof and Beukeboom, 2011) or meeting the
the pandemic. Finally, airlines may have purposefully used a expectation companies are expending an effort in addressing
reactive reticence strategy to limit online bystanders’ emotional their problem (Van Noort and Willemsen, 2012). These types
engagement (Wang et al., 2021). of texts were visible to all users who would have noticed the
repetition of identical or similar answers, exacerbating the
Airline responsiveness by type of airline situation of the customers. Answers can be easily automated by
Our data show that there is a significant difference between flag a machine but can have their own drawbacks. With the advent
and low-cost airlines in terms of number or responses; hence, of more affordable artificial intelligence applications,
the control variable is statistically significant. Flag airlines companies will recognize the source and nature of the message
received more complaints, and on average, they replied more and prepare answers totally customizable (Rosenmayer et al.,
times than low-cost airlines. Low-cost airliners, we suppose, by 2018).
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

At first sight, the results from our study show that interactive of customers are unhappy with airlines because they did not
virtual social presence does not make a difference when it respond to their complaints (Gunarathne et al., 2017).
comes to firms’ responses to online bystanders. Firms should Although the data collected in our study was for a shorter
be cautious about their non-response to active online period and during the COVID-19 pandemic, the results may
bystanders. Our study shows that the VPOs influence the focal not be unrepresentative of a firm response. Some studies
customer’s behavior by inciting the latter to persist and follow- conducted during periods of crises have found that many
up on their complaints. Firms failing to address active online organizations do not respond to the messages of consumers or,
bystanders can generate further criticism from focal customers’ in the best case, only 15% of the companies responded to
complaints and the VPOs. It has been tested that company messages (Crijns et al., 2017).
responsiveness and redress can positively influence online Moreover, we have only considered the interactive virtual
bystanders in reducing their likelihood to complain and presence whose valence is positive. There were very few
forming positive attitudes toward the brand (Armstrong et al., negative interactions with the claimants therefore we left them
2021b). out. We do not believe this bias could have affected the results.
The airlines may perceive they already have done their job by Because our interest was to determine the influence of active
replying to focal complainants; however, we believe companies VPOs in social media on complaining customers and firm
should acknowledge the virtual presence of others in the service response, we studied the general response of the company to
recovery. At the time that companies have implemented the complaint and the number the firm interacted with the
webcare and focused on addressing the focal complainant, complainants. In this regard, we surmise that some of the firm
firms should also pay attention to third-party complainants on responses were automatic; therefore, we have to be cautious in
social media. It is understandable that firms may not want to interpreting the influence of VPO and claimants. Having said
answer third-party online bystanders’ reactions not to open that, the firm response also included the number of interactions
another front and antagonize consumers that view a complaint between the claimant and the firm response.
as legitimate. It has been noted that when an employee This research did not receive any specific grant from funding
intervenes to a nonsupportive comment made to a complainant agencies in the public, commercial or not-for profit sectors.
by a VPO, the likelihood to complain increases (Armstrong
et al., 2021a). It is also known that VPOs can negatively interact
with complainants in an uncivil way leaving the company more References
exposed and as perceived to fail its duty of care toward Abboud, L., As’ad, N., Bilstein, N., Costers, A., Henkens, B.
complainants (Bacile et al., 2020). Thus, a nonresponse from a and Verleye, K. (2021), “From third party to significant
firm increases the negative experience of the claimant and the other for service encounters: a systematic review on third-
perception of fairness of VPOs (Xu et al., 2016). Compared to party roles and their implications”, Journal of Service
no-response, Weitzl and Hutzinger (2017) find that Management, Vol. 32 No. 4, pp. 533-559.
accommodative marketer-initiated webcare responses (an Abney, A.K., Pelletier, M.J., Ford, T.R. and Horky, A.B.
apology) yield significantly more favorable brand-related (2017), “#IHateYourBrand: adaptive service recovery
outcomes within bystanders while defensive responses do not. strategies on Twitter”, Journal of Services Marketing, Vol. 31
Finally, this study provides support to the crowd-based No. 3, pp. 281-294.
accountability framework (Karunakaran et al., 2021) positing Armstrong, C., Kulcynski, A. and Brennan, S. (2021a), “For
the view that organizations use the “diffractive reactivity” all to see: social risk and observable complaining on
response by occasionally engaging with complainants, offering Facebook”, European Journal of Marketing, Vol. 56 No. 1,
rote replies or simply refraining from posting online to pp. 31-71.
complainants or online bystanders. Armstrong, C., Kulczynski, A. and Brennan, S. (2021b), “Cue
the complaint: the visual cues of Facebook business pages
Limitation and further studies and their influence on consumer complaint behaviour”,
Our study is certainly not flawless. We did not analyze the Journal of Marketing Management, Vol. 37 Nos 11/12,
messages or the firm response in detail, for example, to pp. 1027-1057.
determine linguistically different components in the message. A Austin, M. (2012), “70% of customer complaints on social
more detailed study could analyze the responses to determine if media ignored”, available at: www.techradar.com/news/
there is a relationship between the response of a company and world-of-tech/roundup/70-of-customer-complaints-on-social-
how strong the message is perceived as a threat to the company. media-ignored-1081434 (accessed 27 Jan 2023).
Another potential limitation is the overwhelming number of Bacile, T.J., Krallman, A., Wolter, J.S. and Beachum, N.D.
complaints in a brief period that may affect the airline’s (2020), “The value disruption of uncivil other-customers
responsiveness. In a normal situation, a firm first response to a during online service recovery”, Journal of Services Marketing,
complaint, obtained from surveys, has varied between 33% and Vol. 34 No. 4, pp. 483-498.
48%, but customers that actually engage in a conversation with Barcelos, R.H., Dantas, D.C. and Senecal, S. (2018), “Watch
the firm drops to 19.9% (Istanbulluoglu, 2017). Note that our your tone: how a brand’s tone of voice on social media
study not only considers the first response by the firm (that may influences consumer responses”, Journal of Interactive
be automatic) but also engagement with the firm. It is possible Marketing, Vol. 41 No. 1, pp. 60-80.
that response rate is inflated in survey studies because of recall. Becker, L.C. and Pizzuti, C. (2017), “C2C value creation:
A study that actually considers a firm response as an exchange social anxiety and retail environment”, Journal of Research in
between a complainant and the firm indicates a high percentage Interactive Marketing, Vol. 11 No. 4, pp. 398-415.
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

Bever, L. (2019), “Doctor who was dragged, screaming from implications for practitioners”, Journal of Services Marketing,
the united airlines flight finally breaks silence”, The Vol. 24 No. 6, pp. 417-429.
Washington Post. Gelbrich, K. and Roschk, H. (2011), “A meta-analysis of
Champoux, V., Durgee, J. and McGlynn, L. (2012), organizational complaint handling and customer responses”,
“Corporate Facebook pages: when ‘fans’ attack”, Journal of Journal of Service Research, Vol. 14 No. 1, pp. 24-43.
Business Strategy, Vol. 33 No. 2, pp. 22-30. Gregoire, Y., Salle, A. and Tripp, T.M. (2015), “Managing
Chang, L.-Y. and Hung, S.-C. (2013), “Adoption and loyalty social media crises with your customers: the good, the bad,
toward low cost carriers: the case of Taipei–Singapore and the ugly”, Business Horizons, Vol. 58 No. 2, pp. 173-182.
passengers”, Transportation Research Part E: Logistics and Gunarathne, P., Rui, H. and Seidmann, A. (2017), “Whose
Transportation Review, Vol. 50, pp. 29-36. and what social media complaints have happier resolutions?
Chen, K., Chen, J., Zhan, W. and Sharma, P. (2020), “When Evidence from twitter”, Journal of Management Information
in Rome! Complaint contagion effect in multi-actor service Systems, Vol. 34 No. 2, pp. 314-340.
ecosystems”, Journal of Business Research, Vol. 121, Hair, J.F., Sarstedt, M., Pieper, T.M. and Ringle, C.M.
pp. 628-641. (2012), “The use of partial least squares structural equation
Chen, Q., He, Y. and Alden, D.L. (2014), “Social presence in modeling in strategic management research: a review of past
service failure: why it might not be a bad thing”, Customer practices and recommendations for future applications”,
Needs and Solutions, Vol. 1 No. 4, pp. 288-297. Long Range Planning, Vol. 45 Nos 5/6, pp. 320-240.
Chin, W.W. (2010), “How to write up and report PLS He, Y., Chen, Q. and Alden, D.L. (2012), “Consumption in
analyses”, in Esposito Vinzi, V., Chin, W.W., Henseler, J. the public eye: the influence of social presence on service
and Wang, H. (Eds), Handbook of Partial Least Squares: experience”, Journal of Business Research, Vol. 65 No. 3,
Concepts, Methods and Applications, Springer Berlin pp. 302-10.
Heidelberg, Berlin, Heidelberg, pp. 655-690. Hennig-Thurau, T., Gwinner, K.P., Walsh, G. and Gremler,
Chu, S.-C. (2011), “Viral advertising in social media”, Journal D.D. (2004), “Electronic word-of-mouth via consumer-
of Interactive Advertising, Vol. 12 No. 1, pp. 30-43. opinion platforms: what motivates consumers to articulate
Colm, L., Ordanini, A. and Parasuraman, A. (2017), “When themselves on the internet?”, Journal of Interactive Marketing,
Vol. 18 No. 1, pp. 38-52.
service customers do not consume in isolation: a typology of
Henseler, J., Ringle, C.M. and Sarstedt, M. (2015), “A new
customer copresence influence modes (CCIMs)”, Journal of
criterion for assessing discriminant validity in variance-based
Service Research, Vol. 20 No. 3, pp. 223-239.
structural equation modeling”, Journal of the Academy of
Crijns, H., Cauberghe, V., Hudders, L. and Claeys, A.-S.
Marketing Science, Vol. 43 No. 1, pp. 115-135.
(2017), “How to deal with online consumer comments
Hermida, A., Fletcher, F., Korell, D. and Logan, D. (2012),
during a crisis? The impact of personalized organizational
“Share, like, recommend”, Journalism Studies, Vol. 13 Nos 5/
responses on organizational reputation”, Computers in
6, pp. 815-824.
Human Behavior, Vol. 75, pp. 619-631.
Hogreve, J., Bilstein, N. and Hoerner, K. (2019), “Service
Davidow, M. (2003), “Organizational responses to customer
recovery on stage: effects of social media recovery on virtually
complaints: what works and what doesn’t”, Journal of Service
present others”, Journal of Service Research, Vol. 22 No. 4,
Research, Vol. 5 No. 3, pp. 225-250. pp. 421-439.
DiStaso, M.W., Vafeiadis, M. and Amaral, C. (2015), Hogreve, J., Bilstein, N. and Mandl, L. (2017), “Unveiling the
“Managing a health crisis on Facebook: how the response recovery time zone of tolerance: when time matters in service
strategies of apology, sympathy, and information influence recovery”, Journal of the Academy of Marketing Science,
public relations”,. Public Relations Review, Vol. 41 No. 2, Vol. 45 No. 6, pp. 866-83.
pp. 222-231. Huang, R. and Ha, S. (2020), “The effects of warmth-oriented
Du, J., Fan, X. and Feng, T. (2014), “Group emotional and competence-oriented service recovery messages on
contagion and complaint intentions in group service failure: observers on online platforms”, Journal of Business Research,
the role of group size and group familiarity”, Journal of Vol. 121, pp. 616-627.
Service Research, Vol. 17 No. 3, pp. 326-338. Huang, R. and Ha, S. (2022), “How do consumer-to-
Einwiller, S.A. and Steilen, S. (2015), “Handling complaints consumer interactions affect bystanders on corporate social
on social network sites: an analysis of complaints and media (CSM)? (In)civility in advocates’ responses and
complaint responses on Facebook and Twitter pages of large complainant-bystander psychological distance”, The Service
US companies”, Public Relations Review, Vol. 41 No. 2, Industries Journal, pp. 1-24.
pp. 195-204. Hutzinger, C. and Weitzl, W. (2021), “Co-creation of online
F. Hair, J., Jr, Sarstedt, M., Hopkins, L. and G. Kuppelwieser, service recoveries and its effects on complaint bystanders”,
V. (2014), “Partial least squares structural equation Journal of Business Research, Vol. 130, pp. 525-538.
modeling (PLS-SEM)”, European Business Review, Vol. 26 Istanbulluoglu, D. (2017), “Complaint handling on social
No. 2, pp. 106-21. media: the impact of multiple response times on consumer
Ferrara, E. and Yang, Z. (2015), “Measuring emotional satisfaction”, Computers in Human Behavior, Vol. 74 No. 1,
contagion in social media”, Plos One, Vol. 10 No. 11. pp. 72-82.
Fisk, R., Grove, S., Harris, L.C., Keeffe, D.A., Daunt, K.L., Javornik, A., Filieri, R. and Gumann, R. (2020), “‘Don’t forget
Russell-Bennett, R. and Wirtz, J. (2010), “Customers others are watching, too!’ The effect of conversational
behaving badly: a state of the art review, research agenda and human voice and reply length on observers’ perceptions of
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

complaint handling in social media”, Journal of Interactive Nicholls, R. (2010), “New directions for customer-to-
Marketing, Vol. 50 No. 1, pp. 100-119. customer interaction research”, Journal of Services Marketing,
Karunakaran, A., Orlikowski, W.J. and Scott, S.V. (2021), Vol. 24 No. 1, pp. 87-97.
“Crowd-based accountability: examining how social media Purnawirawan, N., Dens, N. and De Pelsmacker, P. (2015),
commentary reconfigures organizational accountability”, “‘We (b) care’: how review set balance moderates the
Organization Science, Vol. 33 No. 1, pp. 170-193. appropriate response strategy to negative online reviews”,
Kelman, H. (1958), “Compliance, identification, and Journal of Service Management, Vol. 26 No. 3, pp. 486-515.
internalization: three processes of attitude change”, Journal Rosenmayer, A., McQuilken, L., Robertson, N. and Ogden, S.
of Conflict Resolution, Vol. 2 No. 1, pp. 51-60. (2018), “Omni-channel service failures and recoveries:
Kerkhof, P.D.U.S. and Beukeboom, C.J. (2011), “Crisis PR in refined typologies using Facebook complaints”, Journal of
social media. An experimental study of the effects of Services Marketing, Vol. 32 No. 3, pp. 269-285.
organizational crisis responses on Facebook”, paper Sarstedt, M., Hair, J.F. and Ringle, C.M. (2022), “PLS-
presented at the 61st Annual ICA Conference, 26-30 May SEM: indeed a silver bullet – retrospective observations
2011, Boston (USA). and recent advances”, Journal of Marketing Theory and
Khan, M.L. (2017), “Social media engagement: what Practice, pp. 1-15.
motivates user participation and consumption on Schaefers, T. and Schamari, J. (2016), “Service recovery via
YouTube?”, Computers in Human Behavior, Vol. 66, social media: the social influence effects of virtual presence”,
pp. 236-247. Journal of Service Research, Vol. 19 No. 2, pp. 192-208.
Lariviere, B., Bowen, D., Andreasse, T., Kunz, W., Sirianni, Sen, S. and Lerman, D. (2007), “Why are you telling me this?”,
N., Voss, C., Wunderlich, V. and De Keyser, A. (2017), Journal of Interactive Marketing, Vol. 21 No. 4, pp. 76-94.
“Service encounter 2.0: an investigation into the roles of Sharma, I., Jain, K. and Behl, A. (2020), “Effect of service
technology, employees and customers”, Journal of Business transgressions on distant third-party customers: the role of
moral identity and moral judgment”, Journal of Business
Research, Vol. 79 No. 10, pp. 238-246.
Lee, D., Hosanagar, K. and Nair, H.S. (2018), “Advertising Research, Vol. 121, pp. 696-712.
Shawky, S., Kubacki, K., Dietrich, T. and Weaven, S. (2020),
content and consumer engagement on social media:
“A dynamic framework for managing customer engagement
evidence from Facebook”, Management Science, Vol. 64
on social media”, Journal of Business Research, Vol. 121,
No. 11, pp. 5105-5131.
pp. 567-577.
Lopez-Lopez, I., Palazon, M. and Sanchez-Martinez, J.A.
Sreejesh, S., M.r, A. and Ponnam, A. (2019), “Can online
(2021), “Why should you respond to customer complaints
service recovery interventions benignly alter customers’
on a personal level? The silent observer’s perspective”,
negative review evaluations? Evidence from the hotel
Journal of Research in Interactive Marketing, Vol. 15 No. 4,
industry”, Journal of Hospitality Marketing & Management,
pp. 661-684.
Vol. 28 No. 6, pp. 711-742.
Ma, L., Sun, B. and Kekre, S. (2015), “The squeaky wheel gets
St
ríteský, V. and Stranska, A. (2015), “Complaining behavior
the grease - an empirical analysis of customer voice and firm
in social media”, Socialin_es Technologijos, Vol. 5 No. 1,
intervention on twitter”, Marketing Science, Vol. 34 No. 5, pp. 22-31.
pp. 627-645. Sun, S., Gao, Y. and Rui, H. (2021), “Does active service
Maher, A. and Sobh, R. (2014), “The role of collective angst intervention drive more complaints on social media? The
during and after a service failure”, Journal of Services roles of service quality and awareness”, Journal of
Marketing, Vol. 28 No. 3, pp. 223-232. Management Information Systems, Vol. 38 No. 3,
Makarem, S. and Jae, H. (2016), “Consumer boycott behavior: pp. 579-611.
an exploratory analysis of twitter feeds”, Vol. 50 No. 1, Tellis, G.J., MacInnis, D.J., Tirunillai, S. and Zhang, Y.
pp. 193-223. (2019), “What drives virality (sharing) of online digital
Mason, K. (2001), “Marketing low cost airline services to content? The critical role of information, emotion, and brand
business travellers”, Journal of Air Transport Management, prominence”, Journal of Marketing, Vol. 83 No. 4, pp. 1-20.
Vol. 7 No. 2, pp. 103-109. Van Herck, R., Decock, S. and De Clerck, B. (2020),
Miller, B. and Morris, R.G. (2014), “Virtual peer effects in “‘Can you send us a PM please?’ Service recovery
social learning theory”, Crime & Delinquency, Vol. 62 No. 12, interactions on social media from the perspective of
pp. 1543-1569. organizational legitimacy”, Discourse, Context & Media,
Minazzi, R. (2015), “The digitization of word-of-mouth”, in Vol. 38, p. 100445.
Minazzi, R. (Ed.) Social Media Marketing in Tourism and Van Noort, G. and Willemsen, M. (2012), “Online damage
Hospitality, Springer International Publishing, Cham, control: the effects of proactive versus reactive webcare
pp. 21-45. interventions in consumer-generated and brand generated
Naylor, R.W., Lamberton, C.P. and West, P.M. (2012), platforms”, Journal of Interactive Marketing, Vol. 26 No. 3,
“Beyond the ‘like’ button: the impact of mere virtual pp. 131-140.
presence on brand evaluations and purchase intentions in van Vaerenbergh, I., Vermeir, I. and Lariviere, B. (2013),
social media settings”, Journal of Marketing, Vol. 76 No. 6, “Service recovery’s impact on customers next-in-line”,
pp. 105-120. Managing Service Quality, Vol. 23 No. 6, pp. 495-512.
Negroni, C. (2009), “With video, a traveler fights back”, The Wan, L.C., Chan, E.K.Y. and Su, L. (2011), “When will
New York Times, 28 October. customers care about service failures that happened to
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

strangers? The role of personal similarity and regulatory silent bystanders”, Journal of Business Research, Vol. 80,
focus and its implication on service evaluation”, pp. 164-175.
International Journal of Hospitality Management, Vol. 30 Xu, Y., Yap, S.F.C. and Hyde, K.F. (2016), “Who is talking,
No. 1, pp. 213-220. who is listening? Service recovery through online customer-
Wang, X., Reger, R. and Pfarrer, M. (2021), “Faster, hotter, to-customer interactions”, Marketing Intelligence & Planning,
and more linked in: managing social disapproval in the social Vol. 34 No. 3.
media era”, Academy of Management Review, Vol. 46 No. 2,
pp. 275-298. Corresponding author
Weitzl, W. and Hutzinger, C. (2017), “The effects of marketer- Hernan E. Riquelme can be contacted at: hernan.riquelme@
and advocate-initiated online service recovery responses on hotmail.com
 Virtually present others Journal of Services Marketing
Rosa E. Rios, Hernan E. Riquelme and Alessandro Comai

Appendix

Table A1 Service recovery in social media and third-party interactive presence

Authors Method/focus Findings
Schaefers and Experimental Mere (passive) VPO enhances the effects of recovery success but not the negative
Schamari (2016) Focus: Interactive virtually present results. Interactive VPO leads to polarization. When experts try to mitigate the
others (VPO) affect complainants’ unsuccessful service recovery, they are perceived as less credible. However, when these
reaction to service recovery experts disconfirm a successful recovery, they are perceived as credible
Weitzl and Hutzinger Experimental. Focus: online Compared to a no-response, accommodative marketer-initiated webcare responses (an
(2017) bystanders apology) and advocate-initiated webcare yield significant more favorable brand-related
outcomes within bystanders while defensive responses do not
Hogreve et al. (2019) Experimental. Food delivery, ISP When service recovery is successful, its transparency signals quality, elicit trust and
setting improves virtually present others WOM and purchase intentions. When unsuccessful, it
Virtually present others results in negative implications
Bacile et al. (2020) Mixed method (netnographic and C2C engage sometimes in uncivil communications that impair service recovery and
experimental). Focus: C2C third-party observers and complainants hold the service provider accountable to
(complainants and observers) manage the service environment. Companies that address incivility are perceived
positively
Javornik et al. (2020) Experimental Use of conversational human voice leads to more positive observer perceptions of
Focus: online observers complaint handling compared to corporate voice. Satisfaction with complaint handling
indirectly influences observers’ WOM intentions. Interactional justice fully mediates the
process
Sharma et al. (2020) Experimental Individuals high on moral identity engage in negative eWOM on social media when
Focus: online bystanders observing an ethical transgression
Chen et al. (2020) Experimental The complaint behavior of others increases focal customers’ complaint intention. The
Focus: others’ complaint behavior contagion effect is moderated by focal customers’ social identification with other
on customers’ complaint intentions customers, credibility and prior relationship between focal customers and the service
provider
Hutzinger and Weitzl Experimental Bystander advocates interaction with complainants affect the perception of virtually
(2021) Focus: online bystanders present others’ recovery response and prepurchase brand attitude. This is true for
individuals low in SNI. Both, marketers’ and advocates’ comments affect high SNI
individuals
Armstrong et al. Experimental, US Facebook pages Virtually present others influence participants’ decision to complain in the Facebook
(2021a) Focus: complainant page. Employee intervention directed at nonsupportive comment made to a
complainant by a virtually present other increases the likelihood to complain
Lopez-Lopez et al. Experimental. Twitter accounts Personalized response to a complaint, compared to an automatic one, leads to more
(2021) Focus: silent observers favorable brand image, purchase intention and eWOM. Response attribution and
emotions elicited during the firm–complainant interaction mediates previous effects.
Personalized response given to an influencer is negatively perceived by silent observers
Huang and Ha (2022) Experimental Online advocate’s defensive response affects the bystander’s service evaluation.
Focus: online bystanders Bystanders who feel closer to a service failure generate stronger reactions to the service
recovery provided by advocates
This study Empirical Impact of interactive VPO on customer complaints’ follow-ups and firm response
Focus: interactive online
bystanders
Source: Own work

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]