SPRINGER BRIEFS IN BUSINESS

Stefan Garding
Andrea Bruns

Complaint
Management and
Channel Choice
An Analysis
of Customer
Perceptions

123
SpringerBriefs in Business
More information about this series at http://www.springer.com/series/8860
Stefan Garding Andrea Bruns
•

Complaint Management
and Channel Choice
An Analysis of Customer Perceptions

123
Stefan Garding Andrea Bruns
Düsseldorf Bad Wünnenberg
Germany Germany

ISSN 2191-5482 ISSN 2191-5490 (electronic)

SpringerBriefs in Business
ISBN 978-3-319-18178-3 ISBN 978-3-319-18179-0 (eBook)
DOI 10.1007/978-3-319-18179-0

Library of Congress Control Number: 2015938431

Springer Cham Heidelberg New York Dordrecht London

© The Author(s) 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.

Printed on acid-free paper

Springer International Publishing AG Switzerland is part of Springer Science+Business Media

(www.springer.com)
Contents

1 Organisational Complaint Management. . . . . . . . . . . . . . . . . . . . . 1

1.1 Relevance of Organisational Complaint Management . . . . . . . . . 1
1.2 Advantages of Complaining Customers. . . . . . . . . . . . . . . . . . . 3
1.3 Importance of Customer Actions . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 Adequacy of Complaint Channels . . . . . . . . . . . . . . . . . . . . . . 5
1.5 Scientific Contribution to Existing Research . . . . . . . . . . . . . . . 7
1.6 Structure of the Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2 Moving Towards Successful Complaint Management . . . . . . . . . . . 13

2.1 Impact of Customer Complaint Behaviour . . . . . . . . . . . . . . . . . 13
2.2 Typology of Communication Channels . . . . . . . . . . . . . . . . . . . 16
2.3 Emergence of Online Social Networks . . . . . . . . . . . . . . . . . . . 18
2.4 Fundamentals of Customer Satisfaction
with Complaint Handling . . . . . . . . . . . . . . . . . . .......... 20
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......... 23

3 Conceptualising Customer Expectations

with Complaint Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.1 Factor 1: Procedural Justice . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.2 Factor 2: Interactional Justice . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3 Factor 3: Distributive Justice . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4 Factor 4: Personal Attitudes. . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.5 Factor 5: Social Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

v
vi Contents

4 Analysis of Customers’ Complaint Channel Choice

and Complaint Behaviour. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.2 Questionnaire Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.2.1 Structure of Questionnaire. . . . . . . . . . . . . . . . . . . . . . . 36
4.2.2 Pre-Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.3 Operationalisation of the Variables . . . . . . . . . . . . . . . . . . . . . . 39
4.4 Research Agenda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.5 Study 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.5.1 Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.5.2 Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.5.3 Manipulation and Realism Check. . . . . . . . . . . . . . . . . . 44
4.5.4 Construct Measurement . . . . . . . . . . . . . . . . . . . . . . . . 45
4.5.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.6 Study 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.6.1 Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.6.2 Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.6.3 Manipulation and Realism Checks . . . . . . . . . . . . . . . . . 56
4.6.4 Construct Measurement . . . . . . . . . . . . . . . . . . . . . . . . 57
4.6.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.7 Channel Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.8 Identification of Customer Subgroups . . . . . . . . . . . . . . . . . . . . 68
4.8.1 Homogenous Subsets of Complainers . . . . . . . . . . . . . . . 68
4.8.2 Analysis of Age Clusters . . . . . . . . . . . . . . . . . . . . . . . 69
4.8.3 Relevance of Social Interactions
for Network Members . . . . . . . . . . . . . ............ 71
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ 73

5 Conclusions for Organisational Complaint Management

and Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.1 Managerial Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.2 Contribution to Existing Research . . . . . . . . . . . . . . . . . . . . . . 80
5.3 Limitations and Implications for Future Research . . . . . . . . . . . . 80
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Appendix A: Literature Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

Appendix B: Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

Appendix C: Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

Appendix D: Multivariate Tests for Gender and Age Differences

(Study 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Contents vii

Appendix E: Multivariate Tests for Gender and Age Differences

(Study 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Abbreviations

ANOVA Analysis of Variance

B2B Business to Business
B2C Business to Consumer
CCB Customer Complaint Behaviour
CRP Customer Referral Program
DJ Distributive Justice
EFA Exploratory Factor Analysis
IJ Interactional Justice
KMO Kaiser-Meyer-Olkin Measure
M Mean Value
PA Personal Attitudes
PJ Procedural Justice
SD Standard Deviation
SI Social Interaction
SN Social Networks

ix
Figures

Figure 1.1 Complaint stages from a customer perspective. . . . . . . . . . . 5

Figure 2.1 Classification of customer complaining behaviour . . . . . . . . 16
Figure 2.2 Users of online social networks globally in 2014 . . . . . . . . . 19
Figure 3.1 Theoretical framework . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Figure 4.1 Data analysis flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Figure 4.2 Overview mean values for PJ—study 1 . . . . . . . . . . . . . . . 49
Figure 4.3 Overview mean values for IJ—study 1 . . . . . . . . . . . . . . . . 50
Figure 4.4 Overview mean values for DJ—study 1 . . . . . . . . . . . . . . . 51
Figure 4.5 Overview mean values for PA—study 1 . . . . . . . . . . . . . . . 52
Figure 4.6 Crosstab appreciation for advice and complaint
channel choice—study 1. . . . . . . . . . . . . . . . . . . . ...... 54
Figure 4.7 Crosstab appreciation for experience
and complaint channel choice—study 1 . . . . . . . . . . . . . . . 54
Figure 4.8 Overview mean values for PJ—study 2 . . . . . . . . . . . . . . . 61
Figure 4.9 Overview mean values for IJ—study 2 . . . . . . . . . . . . . . . . 62
Figure 4.10 Overview mean values for DJ—study 2 . . . . . . . . . . . . . . . 63
Figure 4.11 Overview mean values for PA—study 2 . . . . . . . . . . . . . . . 64
Figure 4.12 Crosstab appreciation for advice and complaint
channel choice—study 2. . . . . . . . . . . . . . . . . . . . ...... 65
Figure 4.13 Crosstab appreciation for experience and complaint
channel choice—study 2. . . . . . . . . . . . . . . . . . . . . . . . . . 66
Figure 4.14 CCB clusters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Figure 4.15 Overview mean values for complaint likelihood. . . . . . . . . . 71
Figure 4.16 CCB decision tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

xi
Tables

Table 1.1 Positioning of this study within research on complaint

management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Table 2.1 Complaint behaviour outcomes . . . . . . . . . . . . . . . . . . . . . . 15
Table 4.1 Categories of questions . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Table 4.2 Descriptive statistics—study 1 . . . . . . . . . . . . . . . . . . . . . . 43
Table 4.3 Reliability and construct measurements—study 1 . . . . . . . . . 45
Table 4.4 Robust tests of equality of means—study 1 . . . . . . . . . . . . . 47
Table 4.5 ANOVA results for procedural justice—study 1 . . . . . . . . . . 47
Table 4.6 Results post hoc tests—study 1 . . . . . . . . . . . . . . . . . . . . . 48
Table 4.7 Descriptive statistics—study 2 . . . . . . . . . . . . . . . . . . . . . . 56
Table 4.8 Reliability and construct measurements—study 2 . . . . . . . . . 57
Table 4.9 Robust tests of equality of means—study 2 . . . . . . . . . . . . . 59
Table 4.10 ANOVA results for procedural justice—study 2 . . . . . . . . . . 59
Table 4.11 Results post hoc tests—study 2 . . . . . . . . . . . . . . . . . . . . . 60
Table 4.12 Robust tests of equality of customers’
complaint likelihood . . . . . . . . . . . . . . . . . . . . . . . . ..... 70
Table 4.13 Results post hoc tests complaint likelihood . . . . . . . . ..... 70

xiii
Abstract

Most customers do not complain when dissatisfied after making a purchase. Being
unaware of the great number of unvoiced complaints by unhappy customers,
companies often experience loss in revenue, market share and eventually hitherto
loyal customers. Companies are therefore well advised to motivate customers to
express their complaints and to incorporate an effective complaint management as
part of their corporate strategy. By following this advice, dissatisfied customers can
be turned into satisfied ones to gain customer loyalty and prevent them from
churning. However, a precondition for customers to address complaints to com-
panies and subsequently be satisfied with the complaint handling is the availability
of adequate complaint channels.
This research provides a detailed comparison of communication channels
regarding their adequacy to voice complaints to companies from a customer per-
spective. In particular, a comparison of traditional complaint channels (mail, e-mail,
phone, face-to-face) is provided. Moreover, it is analysed whether social networks
might be utilised as a new complaint channel. Two empirical studies are conducted
examining the adequacy of complaint channels. Based on the concepts of justice
theory and personal attitudes, participants’ perceptions on each complaint channel
are measured as well as their need for social interaction in the context of
complaining.
The results of both studies show that traditional complaint channels are perceived
differently regarding their adequacy to voice a complaint. However, traditional
complaint channels are seen as preferable when compared to social networks.
Nevertheless, social networks work well as an additional complaint channel, espe-
cially with a view to younger customers.

Keywords Complaint channel Complaint management
Customer complaint



behaviour Customer satisfaction Social networks

xv
Chapter 1
Organisational Complaint Management

Abstract This chapter highlights the relevance of organisational complaint

management. Advantages of customer complaints are outlined from a company
perspective as well as from a customer perspective in the context of business-
to-customer relationships. It is emphasized that complaining customers are bene-
ficial for companies as they can identify weaknesses of service encounters or
products. After a successful service recovery, companies might have the opportu-
nity of regaining previously dissatisfied customers. However, a larger part of cus-
tomers stay nevertheless silent after a dissatisfying incident. Therefore, this chapter
focuses on the importance of complaint channels as precondition for customers to
voice their complaints to companies. Several complaint channels are identified,
including traditional ones (such as mail, e-mail, phone, face-to-face) and a potential
emerging channel: online social networks. Finally, this chapter elaborates the
scientific contribution of this study and outlines the structure of the remaining
chapters.

Keywords Complaint channels Complaint handling Customer behaviour



Organisational complaint management Post-purchase satisfaction

1.1 Relevance of Organisational Complaint Management

Among the most favoured corporate strategies in the 21st century, service excel-
lence is recognised as one of the most desirable (Johnston 2004). Up to 70 % of
today’s global economy is based on services (Mayer et al. 2009) in fiercely com-
petitive sectors (Bernstein and Federgruen 2004), making the intention behind such
initiatives obvious. The delivery of high-quality standards is a key element within
the customer-company relationship as many companies’ attempt to exceed expec-
tations to create satisfied and loyal customers (Khalifa 2004).
However, in the context of supply and demand, not all transactions lead to
customer satisfaction. Dissatisfaction can occur for various reasons, there might, for
instance be a discrepancy between customers’ expectations regarding a product or

© The Author(s) 2015 1

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0_1
2 1 Organisational Complaint Management

service and their actual experience (Lapré and Tsikriktsis 2006). Some products and
services may quite simply be flawed. Given the remarkable range of goods and
services available, failures are likely to happen and cannot always be prevented
(Heskett et al. 1990).
Whenever this does occur, customers must have the choice to complain to the
responsible company. To avoid customer churn, these complaints can provide the
company with a second chance to meet expectations (Gruber et al. 2009).
Companies are therefore well advised to incorporate post-purchase services and
recovery procedures into the life-cycle management of their merchandise (Blodgett
et al. 1995).
Although an effective and satisfactory complaint management can lead to major
advantages for companies (to a loyal customer base being one such advantage)
(Homburg and Fürst 2005), companies often manage complaints ineffectively, and
do not encourage customers to proactively submit their complaints (Gruber et al.
2009; Hansen et al. 2010). Some companies do not recognise the value of an
effective complaint management for their benefit. Consequently, given that an
active complaint management is not operated, these companies even do not provide
communication channels that can be utilised by customers to voice complaints.
Research and practice show that the existence of one or more complaint channels
does not necessarily imply that customers really complain. On the one hand,
regardless of the availability of complaint channels, only a minority of customers
actually do so in case of dissatisfaction (Chebat et al. 2005). Often, these prefer to
churn silently after a dissatisfying shopping experience or after a service failure has
occurred. Negative complaint experience can also lead to customers’ silence about
their dissatisfaction (Chebat et al. 2005). On the other hand, some complaint
management policies tend to discourage customers from stating a problem
(Voorhees and Brady 2005). Therefore, companies should actively motivate cus-
tomers to complain. An efficiently handling of complaints and eventually finding
satisfying solutions is beneficial for all parties involved.
Given the advantages of encouraging customers to complain, companies should
provide adequate complaint channels to implement a successful complaint man-
agement. Since the availability of complaint channels is the first step towards
successful complaint management, this work aims at answering the following
question:
Challenging Question in Complaint Management
Which communication channels are adequate options to voice a complaint from a customer
perspective?

To answer this question, established and conventional communication channels

like phone, e-mail, face-to-face and mail are examined. However, modern com-
munication channels gain increasing importance, such as social networks (Kaplan
and Haenlein 2010). Therefore, this channel is paid particular attention regarding its
possible function as an adequate complaint channel. Throughout this book, the
reader will be given opportunity to assess how customers rate various modes of
communication.
1.2 Advantages of Complaining Customers 3

1.2 Advantages of Complaining Customers

Complaints from customers have two major advantages for companies which can
lead to benefits and enhancements on both an individual and an accumulated level.
First, through complaints, companies can gain satisfy customers. The imple-
mentation of effective customer-relationship management is positively linked to a
company’s revenues (Zeithaml 2000) and to leveraging costs (Cugini et al. 2007).
A successfully solved complaint moderates customer satisfaction and prevents
negative customer behaviour towards companies, “such as bad-mouthing, com-
plaints to third-parties, boycott[s], and exit[s]” (Mattila and Wirtz 2004). Moreover,
it is observed that customers’ satisfaction can even be higher after a product or
service failure is followed by a successful complaint management than satisfaction
would be without anything having occurred (Krishna et al. 2014). For example,
customers receive an overcharged bill from their mobile service provider and are
therefore dissatisfied. Once they complain to the company by phone they might
experience a friendly service agent, get immediate resolution (i.e. a new bill), and
are for instance granted a monetary compensation to make up for the inconve-
nience. This successful complaint handling might turn these previously dissatisfied
customers to very satisfied customers due to the exemplary behaviour on the
company’s part. This post-complaint satisfaction level might even be higher before
the service incident, because the customers experience individual problem resolu-
tion and therefore regain trust and positive emotions in the relationship with the
company (Schoefer 2008). This phenomenon is called the recovery paradox (e.g.
Schoefer 2008; Smith 1998). Additional research shows that even future advertising
and promotions might become more effective because of post-complaint satisfac-
tion (Luo and Homburg 2007). Moreover, customers even tend to consider
choosing their vendors because of the anticipated quality of post-purchase services
in the event of product or service failure (Blodgett et al. 1995). Consequently,
active complaint management has the potential to increase customer loyalty, as it
might be able to convert dissatisfied customers into satisfied ones and can be
implemented as a defensive marketing strategy (Blodgett et al. 1995; Cho et al.
2002a). Thus, companies should view customer complaints as opportunities rather
than as nuisances (e.g. Harari 1992; Michel et al. 2009).
Second of all, recurring complaints about identical or comparable failures enable
companies to review and analyse processes on an accumulative level, leading to
insights concerning potential process improvements (Westbrook and Fornell 1984).
Companies can also take the information collected from complaints into account as
a source of strategic business intelligence. For instance, information about tech-
nological trends, customers’ perception of competitors, and the success of com-
mercial models can be gathered (Larivet and Brouard 2010).
Thus, complainants should not be suppressed; instead, they should be recognised
as a source of valuable data that can contribute to a company’s performance
(Johnston 2004) and to an increased level of customer satisfaction. From a strategic
corporate perspective, customer satisfaction is a precondition for “repeat sales,
4 1 Organisational Complaint Management

positive word-of-mouth and customer loyalty” (Bearden and Teel 1983). Despite
research attempts to understand customer complaint behaviour and a company’s
increasing efforts to listen to their customers, it is widely known that non-voicing is
still the most likely outcome in cases of a dissatisfied customer (e.g. Chebat et al.
2005; Warland et al. 1975).
Through online social networks customers are more than ever capable of
exchanging experiences and opinions (Algesheimer et al. 2010). Consequently, the
quality of company’s products and services is to a high degree transparent to
customers nowadays. Thus, shortcomings in products or services can be spread
quickly, e.g. via electronical word-of-mouth, using social networks, e.g. Facebook
(Pfeffer et al. 2013). This might significantly impact future revenues.
As a conclusion: complaint management represents an important element within
a defensive marketing strategy and it supports a company’s ability to react to
market needs and changes fast and effectively (Kasabov and Warlow 2010).
Successful complaint management supports a company in protecting its market
share by keeping its customer base satisfied, even if dissatisfaction could occur.
This ability is important, given that service markets are increasingly saturated and
that competitors will try to entice customers away. According to this synthesis, the
relevance of investigating the potential adequacy of traditional channels and new
trends, such as the online social networks wave (Kaplan and Haenlein 2010),
becomes apparent. For a successful complaint management and therefore satisfied
customers, it is really important to understand costumers’ opinions of the complaint
channels open to them.

1.3 Importance of Customer Actions

Customers who choose to remain silent and do not complain can affect the company
in a negative way. Since companies are typically unaware of individual sales of
defective, low-quality goods and services or instances of unfair treatment, it is up to
the customers themselves to voice their complaints and claim their rights (Chebat
et al. 2005). However, research reveals that only a minority of customers do indeed
complain (Singh and Wilkes 1996). On average, 66–95 % of all customers remain
silent rather than protest (e.g. Davidow 2003; Voorhees and Brady 2005). Individual
customer complaints can be seen as the “tip of the iceberg” (Johnston and Mehra
2002), representing a great number of unvoiced complaints and consequently dis-
satisfied customers. Dissatisfied customers who remain silent are harmful for com-
panies given that companies are not given the opportunity to make amends (McCole
2004). These customers are expected to be less loyal compared to customers who
experienced successful service recovery (Blodgett et al. 1995). Silent customers
therefore cause harm to a company since its relationship to its customers determines
its long-term viability (Conlon and Murray 1996) by generating sustainable revenue.
Assuming that a failure has occurred, the complaint process consists of three
steps as illustrated in Fig. 1.1: customers’ (i) perception of a problem, followed by
1.3 Importance of Customer Actions 5

(i)
Perception of a
problem

(iii) (ii)
Problem solving Voicing a
efforts complaint

Fig. 1.1 Complaint stages from a customer perspective

(ii) voicing a complaint to the company, and eventually the company’s (iii) problem
solving efforts (Conlon and Murray 1996). Given that new failures might occur in
the course of the complaint management process, a circular reference is anticipated.
From a customer’s perspective, the perception of a problem represents the initial
cause for upcoming dissatisfaction. If the customer remains silent, the company
usually has no opportunity to find a solution. In contrast, if the customer decides to
take action, a second step becomes necessary, i.e. the customer needs to voice a
complaint. Therefore, encouraging customers to do this can be seen as a major
significant challenge for companies. Complaints are a precondition for a company
taking steps to resolve the issues involved (Halstead 1991) and to perform their
problem solving efforts.
In summary, the important role of the availability of complaint channels and
their management becomes apparent. As already outlined early in 1986 by
Technical Assistance Research Programs (TARP), customers often experience a
lack of sufficient information to contact a company. With this in mind, this study
examines the portfolio of complaint channels companies should offer to motivate
their customers to provide critical feedback. In particular, the focus of this work is
an investigation into the adequacy of complaint channels as a means both for
customers to voice their complaints and for companies to work out solutions.

1.4 Adequacy of Complaint Channels

The improvement of organisational complaint management should be seen as a

necessity from both a managerial and a theoretical perspective. To guide subsequent
analysis, four management questions are listed below.
6 1 Organisational Complaint Management

First, the main purpose of this work is to determine the adequacy of traditional
communication channels (such as mail, phone, face-to-face and e-mail) in the
context of complaint management. It is investigated whether these channels are
adequate instruments for customers to state their complaints and to approach
companies. Thus, the following management question is formulated:
Management Question 1
To what extent do customers perceive each traditional communication channel as an
adequate one for complaints?

Second, the adequacy of online social networks as a potential new complaint

channel is analysed. Given the extreme growth rates and acceptance of online social
networks in everydays life (Statista 2014), it is to be assumed that this communi-
cation channel represents a convenient way for customers to approach companies.
However, it might be questioned whether customers actually would accept handling
their relationship with the company by the use of those online networks. Social
networks are not suggested to substitute existing complaint channels, but rather to
extent a company’s complaint channel portfolio. Consequently, the following
question must be addressed:
Management Question 2
To what extent do customers perceive online social networks as an additional and adequate
communication channel to voice complaints to companies?

Third, a synthesis of all complaint channels involved is required. In order to

extract valuable insights for companies, this work compares the adequacy of dif-
ferent complaint channels detailing their advantages and pitfalls. This leads to the
following question:
Management Question 3
To what extent do complaint channels differ in their adequacy of customers’ perceptions?

Finally, online social networks are fundamentally different from other complaint
channels due to their unique characteristics and their ability to stimulate social
interaction among customers. Thus, social networks might have the potential to
combine elements of private and public actions, thereby leading to higher levels of
customer satisfaction. Private actions are defined as the informal exchange of
information and experience with a customer’s personal environment (Day and
Bodur 1978). Public actions represent all circumstances where a customer seeks
direct redress from the company or by means of official third parties such as legal
authorities (Mattila and Wirtz 2004). Considering that private actions have an even
greater destructive effect for a company (von der Heyde Fernandes and Pizzuti dos
Santos 2008), social networks might represent an integrative approach for com-
panies to monitor and to intervene in a scenario of word-of-mouth activities.
Besides, in the context of customer acquisition the facilitation of customer inter-
action by means of customer referral programs (CRP) has been found to be beneficial
for companies (Garnefeld et al. 2013). Garnefeld et al. (2013) even recommend
companies to consider “building CRPs into social media campaigns to take advantage
1.4 Adequacy of Complaint Channels 7

of the increasing social commerce opportunities”. It is therefore expected that social

interaction is not only relevant in customer acquisition, but also profitable for com-
panies within complaint management. According to Malafi et al. (1993), customers
often seek advice and learn from the experiences made by other customers. Given that
conventional communication channels do not have the interactive capability of online
social networks the role of both social interaction and social influence needs to be
analysed in detail. This circumstance is illustrated by the following question:
Management Question 4
To what extent do customers appreciate the opportunity of social interaction with other
customers in the context of voicing a complaint?

1.5 Scientific Contribution to Existing Research

The managerial part of complaint management as a field of research often lies in the
focus of scientific interest. Complaint management studies and theories are as highly
diverse as the recommendations that follow from this. Therefore, this analysis
centres on a particular subset in order to conduct a precise and expedient analysis.
For example, a variety of research has shown that customers evaluate their
decision to complain by considering several factors: the probability of the com-
plaint’s success, the accessibility of the company, the anticipated response by the
company, and customers’ personal characteristics and attitudes (e.g. Robertson and
Shaw 2009). The combination of these factors enables the calculation of an effort/
benefit ratio regarding the customer (Crié 2003). Obviously, the probability of
making a complaint is higher when customers expect to benefit (Robertson and
Shaw 2009) by this. However, not all customers expecting a successful complaint
will indeed communicate their dissatisfaction to the company. For example, inse-
cure customers are less likely to complain (Bodey and Grace 2006). These factors
also influence customers’ satisfaction with the complaint handling process. When it
takes a long time for customers to identify the available communication channels
for making a complaint, the satisfaction with the handling process will be lower
than when procedures are easy to access. This can affect future behaviour
(Voorhees and Brady 2005).
However, scientists also identify research gaps: Especially, the lack of research
regarding (1) the determinants of channel choice (Wirtz and Mattila 2004) and (2)
the influence of social interaction (Malafi et al. 1993) has not been addressed to
date. Both issues are subjects of this study and explained in more detail.
First, and most importantly, Mattila and Wirtz (2004) state that the determinants
of customers’ channel choice have not been analysed sufficiently, although a
detailed understanding of channel choice might lead to an increase in the capture of
customer complaints (Mattila and Wirtz 2004). Moreover, the authors already
hinted that the “anonymity found in internet chat-rooms and bulletin boards” (Wirtz
and Mattila 2004) might lead to new corporate challenges with regard to complaint
8 1 Organisational Complaint Management

management. Therefore, it seems advisable to follow this up with more research on

complaint channels.
Second, Malafi et al. (1993) mentioned that there is need for more research on
the importance of social influence, i.e. advice and experience, on customer com-
plaint behaviour. In their pioneering research, the authors revealed that customers,
who have been advised by friends to complain, are more likely to do so than
customers who have not received such encouragement in advance. Customers who
have been advised not to complain are even less likely to do so. In light of the vast
impact of social circumstances more attention ought to be paid to this.
With regard to recent technical developments, the gaps in the research described
above become even more significant. In the light of the fact that Malafi et al.
outlined the importance of social influence in 1993, during a decade when online
social networks did not yet exist, their research in the present is more crucial than
ever. The current work attempts to address suggestions offered by Malafi et al.
(1993) and Mattila and Wirtz (2004) by investigating (i) the suitability of traditional
complaint channels, (ii) the adequacy of online social networks as a complaint
channel, (iii) the differences of conventional as well as emerging communication
channels, from a customer perspective and finally (iii) the importance of social
interaction in the course of complaining.
In reference to existing theories in the research area of complaint management,
this study is positioned as follows: Within the field of post-purchase satisfaction,
three dimensions of satisfaction are identified (Table 1.1). The first dimension is
immediate post-purchase satisfaction, which is based on the ratio of pre-purchase
expectation and post-purchase experience, whereby experience is equal to or higher
than customer expectation (Lapré and Tsikriktsis 2006). At this stage, in the case of
dissatisfaction (i.e. experienced outcome is worse than expected outcome),
customers decide whether to complain about shortcomings to the company. Once
the complaint has been addressed customers evaluate their satisfaction level with
the handling of the complaint according to the perceived fairness (which can be
regarded as being synonymous with justice theory) as the second dimension.
During the third stage, post-complaint intention, customer satisfaction is determined

Table 1.1 Positioning of this study within research on complaint management

Post-purchase dimension
Stage Complaining decision Complaint handling Post-complaint intention
Satisfaction Post-purchase Satisfaction with Post-complaint
dimension satisfaction complaint handling satisfaction
Variable Experience Fairness Remedy
Research E.g. Huppertz (2007), E.g. Smith and Bolton E.g. Blodgett et al.
studies Kim et al. (2010) (2002), Voorhees and (1995), Davidow (2003)
Brady (2005)
Positioning of ✗ ✓ ✗
this study
1.5 Scientific Contribution to Existing Research 9

by the solution provided and the outcome of the voiced complaint (remedy),
leading to post-complaint satisfaction as the third dimension. Depending on the
post-complaint satisfaction level, customers decide whether to repurchase at the
specific company or to switch when shopping in the future.
As illustrated, this investigation contributes to existing theories on customer
satisfaction with complaint handling processes. In particular, the study sheds light
on the adequacy of conventional complaint channels and the feasibility of incor-
porating social networks as a new channel in order to increase satisfaction by taking
social circumstances into account.
However, with regard to online social networks, it needs to be emphasised that
there is only little research on online complaint management and corresponding
customer behaviour so far (Harrison-Walker 2001); therefore, the influence of this
complaint channel has to be taken into question. Contrary evidence does not yet
exist (Breitsohl et al. 2010) and new studies in this field are expected to be pub-
lished (Cho et al. 2002b). Thus, this study closes the research gap while relying on a
theoretical background.

1.6 Structure of the Study

The structure of this study is as follows. Chapter 2 outlines potential customer

reactions to service failures, comprising direct complaints and further alternatives to
vent frustration (Sect. 2.1). Building upon this discussion, Sect. 2.2 provides a
typology of communication channels by differentiating among interactive, semi-
interactive and remote complaint channels. The concept of online social networks is
then introduced (Sect. 2.3) aiming at a deeper understanding of this potential new
complaint channel. The second Chapter concludes with an illustration of applicable
theories in the field of complaint management (Sect. 2.4).
In Chap. 3, customer expectations with complaint channels are discussed. This
in depth analysis is framed by scientific theories. Five factors are revealed and
related to each analysed complaint channel (Sects. 3.1–3.5).
In Chap. 4, the empirical analysis is described. First, the research designs
(Sect. 4.1) and the questionnaire design are introduced (Sect. 4.2), including the
surveyed sample for two empirical studies. This is followed by the questionnaire
design as the operationalization of the variables (Sect. 4.3). Section 4.4 shows the
research agenda. Two empirical studies have been conducted and are analysed in
detail (Sects. 4.5 and 4.6). The results are summarised and compared in Sect. 4.7.
Finally, subgroups of complaining customers are analysed (Sect. 4.8).
Chapter 5 discusses implications for practitioners and scientists (Sect. 5.1).
Further, the contribution of this study to existing research (Sect. 5.2) is outlined.
Implications for future research are given in Sect. 5.3.
10 1 Organisational Complaint Management

Take away
Complaint management represents an important part of corporate strategies,
given that the avoidance of dissatisfied customer churn protects future rev-
enues. Since customers need to voice their complaints in order to experience
compensation, companies are advised to implement adequate communication
channels for customers to make complaints. However, not all complaint
channels are equally suited from a customer’s perspective; these should
therefore be selected carefully by companies.

References

Algesheimer R, Borle S, Dholakia UM, Singh SS (2010) The impact of customer community
participation on customer behaviors: an empirical investigation. Mark Sci 29(4):756–769.
doi:10.1287/mksc.1090.0555
Bearden WO, Teel JE (1983) Selected determinants of consumer satisfaction and complaint
reports. J Mark Res 20(1):21–28
Bernstein F, Federgruen A (2004) A general equilibrium model for industries with price and
service competition. Oper Res 52(6):868–886. doi:10.1287/opre.1040.0149
Blodgett JG, Wakefield KL, Barnes JH (1995) The effects of customer service on consumer
complaining behavior. J Serv Mark 9(4):31–42. doi:10.1108/08876049510094487
Bodey K, Grace D (2006) Segmenting service “complainers” and “non-complainers” on the basis
of consumer characteristics. J Serv Mark 20(3):178–187. doi:10.1108/08876040610665634
Breitsohl J, Khammash M, Griffiths G (2010) E-business complaint management: perceptions and
perspectives of online credibility. J Enterp Inf Manag 23(5):653–660. doi:10.1108/
17410391011083083
Chebat J, Davidow M, Codjovi I (2005) Silent voices: why some dissatisfied consumers fail to
complain. J Serv Res 7(4):328–342. doi:10.1177/1094670504273965
Cho Y, Im I, Hiltz R, Fjermestad J (2002a) An analysis of online customer complaints:
implications for Web complaint management. In: Proceedings of the 35th annual Hawaii
international conference on system sciences vol. 00(c), pp 2308–2317. doi:10.1109/HICSS.
2002.994162
Cho Y, Im I, Hiltz R, Fjermstad J (2002b) The effects of post-purchase evaluation factors on online
vs. Offline customer complaining behavior. Implic Cust Loyalty Adv Consum Res 29:318–326
Conlon DE, Murray NM (1996) Customer perceptions of corporate responses to product
complaints. The role of explanations. Acad Manag J 39(4):1040–1056
(1986) Consumer complaint handling in America: an update study, Washington D.C.
Crié D (2003) Consumers’ complaint behaviour. Taxonomy, typology and determinants: Towards
a unified ontology. J Database Mark Cust Strategy Manag 11(1):60–79. doi:10.1057/palgrave.
dbm.3240206
Cugini A, Carù A, Zerbini F (2007) The cost of customer satisfaction: a framework for strategic
cost management in service industries. Euro Acc Revs 16(3):499–530. doi:10.1080/
09638180701507130
Davidow M (2003) Organizational responses to customer complaints: what works and what
doesn’t. J Serv Res 5(3):225–250. doi:10.1177/1094670502238917
Day RL, Bodur M (1978) Consumer response to dissatisfaction with services and intangibles. Adv
Consum Res 5:263–272
References 11

Garnefeld I, Eggert A, Helm SV, Tax SS (2013) Growing existing customers’ revenue streams
through customer referral programs. J Mark 77(4):17–32. doi:10.1509/jm.11.0423
Gruber T, Szmigin I, Voss R (2009) Developing a deeper understanding of the attributes of
effective customer contact employees in personal complaint-handling encounters. J Serv Mark
23(6):422–435. doi:10.1108/08876040910985889
Halstead D (1991) Consumer attitudes toward complaining and the prediction of multiple
complaint responses. Adv Consum Res 18(1):210–216
Hansen T, Wilke R, Zaichkowsky J (2010) Managing consumer complaints: differences and
similarities among heterogeneous retailers. Int J Retail Distribut Manag 38(1):6–23. doi:10.
1108/09590551011016304
Harari O (1992) Thank heaven for complainers. Manag Rev 81(1):59–61
Harrison-Walker LJ (2001) E-complaining: a content analysis of an Internet complaint forum.
J Serv Mark 15(5):397–412. doi:10.1108/EUM0000000005657
Heskett JL, Sasser WE, Hart CWL (1990) The profitable art of service recovery. Harvard Bus Rev
68(4):148–156
Homburg C, Fürst A (2005) How organizational complaint handling drives customer loyalty: an
analysis of the mechanistic and the organic approach. J Mark 69(3):95–114. doi:10.1509/jmkg.
69.3.95.66367
Huppertz JW (2007) Firms’ complaint handling policies and consumer complaint voicing.
J Consum Mark 24(7):428–437. doi:10.1108/07363760710834843
Johnston R, Mehra S (2002) Best-practice complaint management. Acad Manag Exec 4(4):145–154
Johnston R (2004) Towards a better understanding of service excellence. Manag Serv Quality 14
(2/3):129–133. doi:10.1108/09604520410528554
Kaplan AM, Haenlein M (2010) Users of the world, unite! The challenges and opportunities of
social media. Bus Horiz 53(1):59–68. doi:10.1016/j.bushor.2009.09.003
Kasabov E, Warlow AJ (2010) Towards a new model of “customer compliance” service provision.
Eur J Mark 44(6):700–729. doi:10.1108/03090561011032685
Khalifa AS (2004) Customer value: a review of recent literature and an integrative configuration.
Manag Decis 42(5):645–666. doi:10.1108/00251740410538497
Kim MG, Wang C, Mattila AS (2010) The relationship between consumer complaining behavior
and service recovery: An integrative review. Int J Contemp Hosp Manag 22(7):975–991.
doi:10.1108/09596111011066635
Krishna A, Dangayach GS, Sharma S (2014) Service recovery paradox: the success parameters.
Global Bus Rev 15(2):263–277. doi:10.1177/0972150914523567
Lapré MA, Tsikriktsis N (2006) Organizational learning curves for customer dissatisfaction.
Heterogeneity across airlines. Manag Sci 52(3):352–366. doi:10.1287/mnsc.1050.0462
Larivet S, Brouard F (2010) Complaints are a firm’s best friend. J Strategic Mark 18(7):537–551.
doi:10.1080/0965254X.2010.529155
Luo X, Homburg C (2007) Neglected outcomes of customer satisfaction. J Mark 71(2):133–149.
doi:10.1509/jmkg.71.2.133
Malafi TN, Cini MA, Taub SL, Bertolami J (1993) Social influence and the decision to complain.
Investigations on the role of advice. J Consum Satisf Dissatisfaction Complain Behav 6:81–89
Mattila AS, Wirtz J (2004) Consumer complaining to firms: the determinants of channel choice.
J Serv Mark 18(2):147–155. doi:10.1108/08876040410528746
Mayer DM, Ehrhart MG, Schneider B (2009) Service attribute boundary conditions of the service
climate-customer satisfaction link. Acad Manag J 52(5):1034–1050
McCole P (2004) Dealing with complaints in services. Int J Contemp Hosp Manag 16(6):345–354.
doi:10.1108/09596110410550789
Michel S, Bowen D, Johnston R (2009) Why service recovery fails: Tensions among customer,
employee, and process perspectives. J Serv Manag 20(3):253–273. doi:10.1108/
09564230910964381
Pfeffer J, Zorbach T, Carley KM (2013) Understanding online firestorms: Negative word-of-mouth
dynamics in social media networks. J Mark Commun 20(1–2):117–128. doi:10.1080/
13527266.2013.797778
12 1 Organisational Complaint Management

Robertson N, Shaw RN (2009) Predicting the likelihood of voiced complaints in the self-service
technology context. J Serv Res 12(1):100–116. doi:10.1177/1094670509333789
Schoefer K (2008) The role of cognition and affect in the formation of customer satisfaction
judgements concerning service recovery encounters. J Consum Behav 7(3):210–221. doi:10.
1002/cb.246
Singh J, Wilkes RE (1996) When consumers complain: a path analysis of the key antecedents of
consumer complaint response estimates. J Acad Mark Sci 24(4):350–365. doi:10.1177/
0092070396244006
Smith AK (1998) An experimental investigation of customer reactions to service failure and
recovery encounters: paradox or peril? J Serv Res 1(1):65–81. doi:10.1177/
109467059800100106
Smith AK, Bolton RN (2002) The effect of customers’ emotional responses to service failures on
their recovery effort evaluations and satisfaction judgments. J Acad Mark Sci 30(1):5–23.
doi:10.1177/03079450094298
Statista (2014) Soziale Netzwerke. Dossier 2014
von der Heyde Fernandes D, Pizzuti dos Santos C (2008) The antecedents of the consumer
complaining behavior. Adv Consum Res 35:584–593
Voorhees CM, Brady MK (2005) A Service Perspective on the Drivers of Complaint Intentions.
J Serv Res 8(2):192–204. doi:10.1177/1094670505279702
Warland RH, Herrmann RO, Willits J (1975) Dissatisfied consumers: who gets upset and who
takes action. J Consum Aff 9(2):148–163. doi:10.1111/j.1745-6606.1975.tb00559.x
Westbrook RA, Fornell C (1984) The vicious circle of consumer complaints. J Mark 48(3):68–78
Wirtz J, Mattila AS (2004) Consumer responses to compensation, speed of recovery and apology
after a service failure. Int J Serv Ind Manag 15(2):150–166. doi:10.1108/09564230410532484
Zeithaml VA (2000) Service Quality, Profitability, and the Economic Worth of Customers: What
We Know and What We Need to Learn. J Acad Mark Sci 28(1):67–85. doi:10.1177/
0092070300281007
Chapter 2
Moving Towards Successful Complaint
Management

Abstract This chapter analyses the impact of customer complaint behaviour and
classifies potential types of behaviour after a dissatisfying incident. Besides the
opportunity of contacting a company directly, three alternative options are dis-
cussed which can be utilised by customers to vent their frustration. Thereafter, four
different traditional complaint channels are introduced. This part is followed by a
definition of social networks as a potentially emerging complaint channel. The
remaining part of this chapter emphasizes the importance of complaint channels and
describes three scientific theories which are important for customer satisfaction
analysis: justice theory, behaviour theory and attribution theory. Each theory is
described in detail and linked to the context of the study.

Keywords Attribution theory Behaviour theory Communication channels




Customer complaint behaviour Customer dissatisfaction Customer satisfaction



Justice theory Online social networks

This chapter begins with a system of customer complaint behaviour and highlights
potential customer choices once a product or service failure has occurred
(Sect. 2.1). A typology of existing communication channels for contacting com-
panies is provided (Sect. 2.2) and extended by incorporating social networks into
existing models (Sect. 2.3). After this, fundamental aspects of the underlying sci-
entific theories are discussed (Sect. 2.4).

2.1 Impact of Customer Complaint Behaviour

Customer Complaint Behaviour (CCB) has several antecedents, characteristics, and

implications. To provide a consistent basis for this study, the term customer
complaint behaviour can be defined as “a customer’s protest to a firm with the goal
of obtaining an exchange, a refund or an apology” (Larivet and Brouard 2010).
However, researchers have shown that a certain dissatisfaction threshold needs to
be crossed in order for customers to take action (Rust and Chung 2006). Thus, not

© The Author(s) 2015 13

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0_2
14 2 Moving Towards Successful Complaint Management

all types of a customer’s protest are voiced directly to the company (Sect. 2.2).
The analysis at hand deals with business-to-customer complaints (B2C) and does
not consider corporate complaints in the business-to-business (B2B) segment.
According to Henneberg et al. (2009), this distinction is constitutive since such
relationships are different. In the B2C context, from the company perspective, the
goal of complaint handling is the cognition and remedy of individual and systemic
problems affecting the company’s customers (Huppertz et al. 2003).
The reasons for customer protests, i.e. causes of initial customer dissatisfaction,
are multifaceted. Shortcomings in products, slow service, unreasonable employee
behaviour, product damage, and delivery problems are amongst the most common
issues (Estelami 2000). Several antecedents for CCB have been identified. For
instance, customer response depends on the type of service failure (Hirschman
1970), customers’ attitude to complaints (Richins 1982) and their emotions (Smith
and Bolton 2002). Among others, these factors mediate the process, namely the
decision whether to complain or not. Researchers have established a general clas-
sification of complaint reasons and intentions by distinguishing whether customers
have suffered monetary loss due to the failure (e.g. Gilly and Gelb 1982; Mayer
et al. 1995). It must also be mentioned that not all claims for compensation or
redress are justified, since some customers behave opportunistically and unrea-
sonable (Wirtz and McColl-Kennedy 2010).1 However, such variations are exclu-
ded from further investigation given that their intention is completely different from
justified complainers. The management questions do not cover opportunistic cus-
tomer behaviour.
When customers are dissatisfied with a product or a service, several possible
ways to react are open to them. One option is that they can leave the company or
engage in private complaining, another is choosing a form of public action (von der
Heyde Fernandes and Pizzuti dos Santos 2008) such as voicing the complaint to the
company. Dacin and Davidow (1997) structured the different opportunities and
modelled potential CCB outcomes as seen from the company perspective
(Table 2.1). Within this framework two dimensions are defined: (i) a company’s
involvement with consumer dissatisfaction and (ii) a consumer’s involvement in his
social network.
Each dimension is divided into two attributes. Dimension 1, i.e. consumer’s
involvement in social network: this represents the customer’s decision whether to
take private (internal) or public actions (external). Dimension 2, i.e. company’s
involvement with consumer dissatisfaction: this represents whether the company is
informed about consumer actions (involved) or not (not involved). As shown in
Table 2.1 consumers may choose to behave in the following way:

1
Further readings on unjustified and opportunistic customer complaint behaviour: Harris (2010),
Jacoby and Jaccard (1981), Reynolds and Harris (2005).
2.1 Impact of Customer Complaint Behaviour 15

Table 2.1 Complaint behaviour outcomes

Dimension 2
Company’s involvement with consumer
dissatisfaction
Involved Not involved
Dimension 1 Internal (i) Exit or boycott (ii) Consumer’s social net
Consumer’s involvement (word-of-mouth)
in social network External (iii) Organisation (iv) Third party
(redress/complaint)
Source Adapted from Dacin and Davidow (1997, Table 1, p. 452)

(i) Internal/involved: Consumers do not contact others. They have made a

silent decision, e.g. switching to another company for future purchases.
(ii) Internal/not involved: Consumers choose private complaining. Consumers
decide to talk to friends and family (by word-of-mouth) to complain about
the company.
(iii) External/involved: Consumers choose to voice their complaints directly to
the company. Conventional complaint management procedures and tech-
niques can be used to handle these complaints.
(iv) External/not involved: Consumers involve external parties. However, con-
sumers do not approach the focal company but third parties such as gov-
ernmental and customers’ protection institutions instead.
This study focuses on the (iii) external/involved and (ii) internal/not involved
combination of attributions: First, this research addresses the (iii) external/involved
combination to investigate how companies can further encourage customer com-
plaints. Second, strategies to minimise the disadvantages of the (ii) internal/not
involved combination, e.g. customer complaints to friends and family, are identi-
fied. As social networks are positioned in this field, the integration of such an online
platform as a further communication channel can directly involve companies in
customers’ word-of-mouth activities, where they have not yet been included.
Moreover, customer satisfaction might increase if social networks meet the criteria
for an appropriate communication channel in the context of complaint management.
To conclude, considering social networks as a new complaint channel might not
only be a convenient way to increase customer satisfaction but may also obviate bad
word-of-mouth by motivating customers to contact companies directly who had
previously addressed their complaints to friends and family. Thus, one goal of this
study is to analyse whether social networks work as an appropriate new complaint
channel. To clarify the effects of using social networks as a “new” complaint
channel is of high managerial importance. One the one hand, encouraging non-
voicers and preventing bad word-of-mouth are necessary for successful complaint
management (e.g. Blodgett and Anderson 2000; Choi and Mattila 2008; Rust and
Chung 2006). On the other hand, companies are advised to develop corporate
strategies for online social networks.
16 2 Moving Towards Successful Complaint Management

2.2 Typology of Communication Channels

Typically there are several communication channels available for customers to

address their complaints to companies: writing a letter (mail), sending an e-mail,
calling a hotline (phone), or visiting a shop (face-to-face) (Halstead 1991). All
channels entail specific characteristics and are basically expected to be suitable for
providing customer satisfaction with the complaint management process in most
cases (Mattila and Wirtz 2004). According to past and recent research, complaints
are most frequently addressed through a face-to-face dialogue at a shop, followed
by phone, e-mail, and regular mail complaints (e.g. Matos et al. 2009; Tax et al.
1998).
The availability of communication channels is a compulsory component of
corporate complaint management (Halstead 1991). The term complaint channel can
be defined as the medium by which a customer submits a complaint to a company
(Mattila and Wirtz 2004). These channels are also used by companies to respond to
customer complaints (Gilly and Gelb 1982). Building upon this research and in line
with the suggestions of Mattila and Wirtz (2004) this study examines four con-
ventional channels: face-to-face, phone, e-mail, and mail (5th level in Fig. 2.1).
As shown in Fig. 2.1, the process of CCB consists of five steps: (1) Starting with
an initial dissatisfaction incident, customers decide whether to take action or not
(1st level). In case they remain silent (i.e. take no action) these customers stay

Dissatisfaction
Incident

1st level: Take no

Take Action
Behavioural vs Action
Non Behavioural Actions

2nd level:
Private vs Public Private Action Public Action
Action

3rd level: Negative Seek Redress Complain to

Specific Actions Boycott Legal Action
WOM Directly 3rd Parties

4th level:
Tendency Toward Semi-
Interactive Remote
Type of Channel Interactive

5th level:
Channel of
Social
Communication Face-to-Face Phone Letter Email
Networks

Fig. 2.1 Classification of customer complaining behaviour. Source Adapted from Mattila and
Wirtz (2004, Fig. 1, p. 148)
2.2 Typology of Communication Channels 17

unhappy due to an unresolved complaint and companies may experience future loss
in revenues due to customer churn caused by customer dissatisfaction. However,
once customers decide to act, (2) they have to choose between private and public
actions (2nd level). Whereas private actions comprise all types of word-of-mouth
activities, public actions describe customers to either approach the company
directly, or, alternatively, venting their frustration by engaging lawyers or 3rd
parties to solve the problem. When the customer has decided to (3) seek redress
directly (3rd level) at the company, the customer (4) has to decide for an interaction
level with the company. Mattila and Wirtz (2004) assess and specify this decision
by noting customers’ tendency to prefer either interactive or remote channels (4th
level). Finally, the customer can (5) choose a communication channel to complain
(5th level) as illustrated in Fig. 2.1.
As it can be seen in Fig. 2.1, Wirtz and Mattila (2004) distinguish the tendency
towards a type of communication channel only between interactive and remote
communication channels.
The first typology, interactive communication channels, represents all channels
in which customers have direct contact with a company’s employees and can
discuss and interact immediately, as for example by means of face-to-face and
complaints on the phone. On the one hand, not all customers favour direct inter-
action. According to research, most customers are afraid of confronting companies
with an interactive complaint (Lerman 2006). Especially face-to-face confrontations
are likely to escalate because of impoliteness between a customer and the employee
(van Jaarsveld et al. 2010). On the other hand, some customers prefer phone calls,
one reason being to experience individualised, personal treatment (Johnston and
Mehra 2002).
The second typology, remote communication channels, comprises written
communication such as mail and e-mail complaints. Upon notifying the company
customers have to wait for a company response (Mattila and Wirtz 2004). For
instance, some male complainers are found to be comfortable complaining in
writing a mail or e-mail, which allows them to structure their complaints more
accurately (Grougiou and Pettigrew 2009). Female customers sometimes wish to
obviate the potential embarrassment inherent in interactive channels and are thus
also likely to choose remote channels (Grougiou and Pettigrew 2009). Although the
majority of companies support most of the mentioned communication channels and
customers have general channel preferences, the implementation, execution, and
supervision of each channel affects customers’ satisfaction levels (Blodgett et al.
1995). These attributes can be consolidated as a major part of a company’s com-
plaint management policy (Huppertz 2007). Section 2.4 highlights the major
determinants of customer satisfaction in complaint management policies, which are
essential for the conceptualisation in Chap. 3.
Apart from these conventional typologies comprising communication channels
are already offered by the majority of complaint management systems, a third
typology represents a new opportunity of communication that has emerged in recent
years—social networks. In the context of complaint management the social net-
works channel is embedded in existing research by an extension of Mattila and
18 2 Moving Towards Successful Complaint Management

Wirtz’s (2004) approach. As Fig. 2.1 shows, customer complaining through social
networks can be considered a public action (2nd level), given that the target
company is involved. By contacting the company customers seek direct redress (3rd
level). So far, only two comprehensive channel classifications, interactive and
remote, are represented in the 4th level of the model. By applying social networks
to this typology, a third classification, semi-interactive communication channels, is
included. This element illustrates the specific characteristics of social networks and
represents a consolidation of the interactive and remote dimensions. On the one
hand, social networks facilitate written complaints, a key characteristic of the
remote category. On the other hand, social networks are far more interactive than
conventional remote approaches (mail and e-mail) though not as much as a bidi-
rectional face-to-face discussion or phone call. Thus, the combination of the two
approaches is a unique feature of social networks. Consequently, as a conjecture,
this communication channel is seen as having an impact on overall customer
satisfaction in the context of complaint handling processes. However, this manner
of communication and interaction is so far applied only seldom in these situations
(Lee and Lee 2006) and therefore tested in the course of this research.

2.3 Emergence of Online Social Networks

From a traditional perspective social networks can be delineated as a combination

of relations among individuals where the characteristics of the linkages influence
the social behaviour of the persons involved (Tichy et al. 1979). In principle, social
network theory has been widely researched for decades (Parkhe et al. 2006).
Several studies have investigated and enhanced this theory interdisciplinary, for
instance from an interpersonal (Brass et al. 2004) and interorganisational per-
spective (Provan et al. 2007). More precisely, the determinants and antecedents of
social networks have been investigated along with the role of strong and weak ties
and knowledge transfer (e.g. Inkpen and Tsang 2005; Nelson 1989).
Apart from this traditional academic background a further aspect of social net-
work theory has become apparent in the course of the 21st century: individuals have
started participating in and interacting through online communities, such as
Facebook, Google+, and LinkedIn. According to the literature various terms to
describe this new social networking mechanism are widely used, namely descrip-
tions such as ‘online communities’ (Dellarocas 2006), ‘virtual communities’ (Porter
and Donthu 2008), ‘social media’ (Kaplan and Haenlein 2010), and ‘online social
networks’ (Ellison et al. 2007). This study always uses the term social networks and
adopts Miller et al.’s (2009) definition of social networks:
“Online communities, consisting of people who engage in computer-supported social
interaction,…[allowing] members to continuously express and access others’ opinions,…
[providing] a highly accessible and efficient source for evaluating and adjusting one’s own
thoughts and actions in light of input from socially relevant peers within a community.”
(Miller et al. 2009)
2.3 Emergence of Online Social Networks 19

In the context of this study, the scope of social networks is further narrowed
down. Within complaint management online social networks are used as an
instrument to contact companies directly. Thus, social interactions take place
between the customer and the company by means of social networks.
The most famous instances of such online communities are Facebook, Google+,
and LinkedIn (Busemann and Gscheidle 2011) whereas qzone is strong in the Asian
market. Figure 2.2 provides an overview of users of online social networks
worldwide in 2014. The huge importance of this channel becomes even more
apparent when considering that the most widely used online social network
(Facebook) has more than one billion users.
The relevance of this communication channel is incontestable; almost 36 % of
the entire German population is already participating in at least one social network
(van Eimeren and Frees 2011). This ratio is steadily increasing, particularly given
that older people are becoming more receptive to social networks (Spahr and Arns
2012). The importance of this medium is growing, as 55 % of all members use their
accounts daily (Busemann and Gscheidle 2011). Among the six major social net-
works activities, ‘searching for information’ and ‘getting updates from online
friends’ is stated by users as being the two major sources of interests (Busemann
and Gscheidle 2011).
However, customers might utilise online social networks not only for contacting
companies directly but also for venting their frustration about this towards their
friends and families. The latter behaviour constitutes private action and is therefore
not part of this study. This research focuses on an opportunity for companies to
utilise social networks as a new complaint channel. In this regard, customers
directly contact companies by sending a message through online social networks.
To sum up, social networks might represent a new communication channel for
handling customer complaints, because they are already widely used by customers.
By incorporating this communication channel as a complaint channel, companies
might be able to increase customer post-complaint satisfaction with complaint
management systems and to improve the customer-company relationship.

Users of online social networks 2014 (globally in million)

1276

644

300 277 255 230 220 200 136

Fig. 2.2 Users of online social networks globally in 2014. Source Adapted from Statista (2014)
20 2 Moving Towards Successful Complaint Management

2.4 Fundamentals of Customer Satisfaction with Complaint

Handling

The majority of studies reveal that customers’ selection of companies, and therefore
their purchase decisions regarding certain products or services are typically influ-
enced by several factors (Blodgett et al. 1995). Among these the quality of customer
service is a key determinant (Anand et al. 2011; Venkatesh and Agarwal 2006).
Customer services can be subdivided into pre- and post-purchase services (Mitchell
and Boustani 1994). The first part, pre-purchase services, comprises all company
activities which focus on raising customer awareness (e.g. advertising) and sup-
porting customer purchase decisions (e.g. sales agents). When a customer buys a
product or service, responsibility shifts to the company’s post-purchase services.
These services are, for instance, up-selling or cross-selling activities as well as
customer care policies. One major activity centres on the company’s capability to
handle dissatisfied customers. This study focuses on customers’ post-purchase
behaviour revealing insights into customer complaint management.
Dissatisfaction motivates customers to complain the reasons for their disap-
pointment. This was a main research topic in the 1960s. In particular, research
concentrated on the multidimensional topics of customer behaviour and complaint
management. Cardozo (1965), for instance, was interested in the interdependence of
customer effort, expectation, and satisfaction. Moreover, according to Gilly and
Gelb (1982), customer behaviour in this context is predominantly affected by
customers’ perceptions of a product or service. Customers perceptions might lead to
satisfaction or dissatisfaction because customers tend to measure their perceived
post-purchase satisfaction as a ratio of ex-ante expectation and ex-post experience
(Lapré and Tsikriktsis 2006). In other words, taking Oliver (1980) into account,
customer dissatisfaction is characterised by their expectation level and corre-
sponding expectancy disconfirmation. More precisely, any disappointments of
expectation decrease customers’ satisfaction level more than any excess increases it
(Anderson and Sullivan 1993). In the wake of dissatisfaction, post-purchase
behaviour might result in customer complaints (Dacin and Davidow 1997).
Therefore, complaints can be observed as unique “recovery opportunities” (Kim
et al. 2010) for companies (Gilly 1987).
The traditional scope examined satisfaction by investigating the impact of dis-
satisfaction on complaint behaviour (e.g. Churchill and Suprenant 1982; Day 1977;
Landon 1980) and by establishing a model to frame customers’ tendencies to
complain directly to companies or to friends and family (e.g. Day 1984; Oliver
1980). From a more general perspective, three areas of research are identified and
inferred as appropriate descriptive models in the context of this study: (1) justice
theory, (2) behaviour theory, and (3) attribution theory. These theories are expected
to be an eligible foundation for testing the adequacy of complaint channels. The
theories referred to may not influence customer satisfaction directly; however, they
have different effects on CCB. Diverse CCB, in turn, eventually leads to different
levels of customer satisfaction.
2.4 Fundamentals of Customer Satisfaction with Complaint Handling 21

First, customers tend to be satisfied with a company’s complaint management

when they perceive to be treated fairly. These results are derived from justice
theory. Customers’ satisfaction with a complaint handling process becomes rele-
vant after customers have decided to address their complaints to the respective
company. Much research indicates that the level of satisfaction with complaint
management systems is primarily influenced by the extent of a company’s fairness.
This can be described as the manner in which a company communicates with the
complaining customer (Homburg and Fürst 2005). Moreover, under consideration
of Morrisson and Huppertz (2010), customers expect companies to treat com-
plaining customers in the same way, regardless of the individual customer’s value.
This idea is supported by further studies, which demonstrate that monetary com-
pensation is often not as important as fair treatment and an apology (Wirtz and
Mattila 2004).
This topic is investigated in justice theory, a first major category in this field
of research. Three elements are embedded in the construct of justice theory:
(i) procedural justice, (ii) interactional justice, and (iii) distributive justice (Larivet
and Brouard 2010). Procedural justice refers to customers’ perceived fairness of the
policies and procedures that companies have in place to handle complaints (McCole
2004). Interactional justice describes the interpersonal treatment of the complaining
customer, i.e. politeness and helpfulness of companies (McCole 2004). Distributive
justice is defined as the fairness of the outcome and the provided remedy (McCole
2004).
Second, Singh (1990) groups different behavioural clusters into four main
response styles in CCB on the basis of behaviour theory. The underlying
behaviour theory represents the next important area of relevance to this field of
research. By applying the taxonomy-approach, four clusters have been identified as
a proper guideline to classify complainers (Singh 1990). The first group is named
passives. These customers usually do not voice complaints, i.e. they behave pas-
sively. The second cluster is characterized by voicers which are customers, who
usually complain actively to the company when dissatisfied. In contrast, irates are
mainly engaged in private complaining by addressing their complaints to friends
and family instead of approaching the company. The fourth cluster is called
activists. These customers voice their complaints in a formal way through the use of
third parties. Addressees are e.g. customer protection agencies or lawyers. In
support of this categorisation, Siddiqui and Tripathi’s (2010) contemporary
research in the banking sector reveals similar findings.
A fundamental precondition of this theory is that companies probably have no
means of influencing customers’ basic attitudes, though they are able to motivate
them to overcome inertia, to name but one issue, (Kim et al. 2010) through the use
of complaint systems (Bodey and Grace 2006). The contributions of Matos et al.
(2009) support previous studies by demonstrating the substantial moderating effect
of customers’ attitudes towards complaining on complaint intentions. In particular,
the authors investigate the relationship between service recovery after a complaint
has been made and subsequent customer satisfaction. The most researched
personality traits are self-confidence and conservatism (Bodey and Grace 2006),
22 2 Moving Towards Successful Complaint Management

word-of-mouth behaviour (Halstead 1991), and being afraid of a confrontation

(Dacin and Davidow 1997).
However, considering Mattila and Wirtz (2004), the validity of those classifi-
cations can be questioned, because such typologies fail to integrate the antecedents
of customers’ channel choice. Further research shows that a customer's perception
of the company’s response to customer complaints also influences the probability of
a customer complaining at all (Halstead 1991). Thus, not only does customer
behaviour and attitude impact CCB, but also the anticipated company reaction.
Third, attribution theory establishes the link between product failure and
customer response (Folkes 1984). In other words, this theory predicts that the
perceived cause of a product failure influences customers’ responses and therefore
influences the mode of the complaint (Peterson and Kerin 1979).
According to attribution theory, both stability and controllability influence
customers’ satisfaction levels (Blodgett et al. 1995). Stability refers to the perceived
likelihood of a similar failure occurring again, whereas controllability describes
customers’ perception as to whether a failure could easily have been prevented by
the company (Choi and Mattila 2008). Scholars and practitioners recognise a set of
obvious preconditions for customer satisfaction. For example, if a product’s
or service’s shortcoming is completely unreasonable from the customer’s point of
view and could easily been prevented by the company, the company will likely be
unable to remedy customer dissatisfaction (Choi and Mattila 2008). On the basis of
these perceptions customers form their opinion about a company and subsequently
decide whether it makes any sense to complain. The construct of attribution theory
is obviously linked to and dependent on such variations of customer behaviour:
A customer’s choice of action is directly related to specific reasons for a product
failure (Folkes 1984).
In addition to the three branches of research mentioned above, other researchers
have investigated individual elements of complaint management. First, Mattila and
Wirtz (2004) analysed the likelihood of channel choice by customers in order to
voice complaints. They showed that the choice depends on customers’ expected
outcome. For example, customers who want to “vent their frustration” choose mail
or e-mail to complain (Wirtz and Mattila 2004). Second, Gilly (1987) focuses on
post-complaint processes and analyses repurchase behaviour after a complaint has
been solved by a company. The results of the study show that the relationship
between customers’ complaints and their repurchase behaviour is mediated by their
“cognitive processes regarding the complaint response” (Gilly 1987). Third, Wirtz
and McColl-Kennedy (2010) investigate the impact of opportunistic customer
behaviour on service recovery. When customers experience procedural, distributive
or interactional justice to be low, they tend to act opportunistically. Wirtz and
McColl-Kennedy (2010) show that customers are more likely to act like this when
dealing with larger companies and do not want to build a long-term
relationship. Fourth, scholars have analysed the relationship between demographic
characteristics and complaining behaviour. It is widely accepted as true that
demographic variables exert only a weak influence (von der Heyde Fernandes and
Pizzuti dos Santos 2008). Variations in customer behaviour are primarily caused by
2.4 Fundamentals of Customer Satisfaction with Complaint Handling 23

different attitudes rather than by demographic characteristics. Fifth, researchers

have also observed that emotions mediate customer perceptions and have therefore
to be considered in a company’s handling of complaints (e.g. Chebat and
Slusarczyk 2005; Schoefer 2008). In particular, it is shown that positive as well as
negative emotions have an influence (positive/negative) on service recovery satis-
faction (Schoefer 2008). However, only a few literature reviews summarise the
status quo of research on complaint management (e.g. Gelbrich and Roschk 2010;
Orsingher et al. 2010).
It can be concluded that the field of complaint management has been widely
researched and several perspectives on the antecedents and determinants of cus-
tomer satisfaction with complaint management systems have been revealed.
However, not all findings are complementary; in fact, some are, as has been
described, contradictory. Thus, the current analysis extends existing research with
regard to customer satisfaction by analysing the adequacy of complaint channels.

Take away
Only a minority of customers decide to complain directly to the responsible
company by utilising one of the available complaint channels. Apart from the
traditional complaint channels, i.e. mail, e-mail, phone, and face-to-face,
social networks are introduced as a potential new complaint channel.
Customer satisfaction with complaint handling is disassembled into several
dimensions, detailed in the research literature on the topic: procedural,
interactional, distributive justice, as well as behavioural and attribution
theory.

References

Anand KS, Pac MF, Veeraraghavan S (2011) Quality-speed conundrum: trade-offs in customer-
intensive services. Manag Sci 57(1):40–56. doi:10.1287/mnsc.1100.1250
Anderson EW, Sullivan MW (1993) The antecedents and consequences of customer satisfaction
for firms. Mark Sci 12(2):125–143. doi:10.1287/mksc.12.2.125
Blodgett JG, Anderson RD (2000) A Bayesian network model of the consumer complaint process.
J Serv Res 2(4):321–338. doi:10.1177/109467050024002
Blodgett JG, Wakefield KL, Barnes JH (1995) The effects of customer service on consumer
complaining behavior. J Serv Mark 9(4):31–42. doi:10.1108/08876049510094487
Bodey K, Grace D (2006) Segmenting service “complainers” and “non-complainers” on the basis
of consumer characteristics. J Serv Mark 20(3):178–187. doi:10.1108/08876040610665634
Brass DJ, Galaskiewicz J, Greve HR, Tsai W (2004) Taking stock of networks and organizations.
A multilevel perspective. Acad Manag J 47(6):795–817
Busemann K, Gscheidle C (2011) Ergebnisse der ARD/ZDF-Onlinestudie 2011. Web 2.0:Aktive
Mitwirkung verbleibt auf niedrigem Niveau. Media Perspektiven, Frankfurt am Main
Cardozo RN (1965) An experimental study of customer effort, expectation, and satisfaction.
J Mark Res 2(3):244–249
24 2 Moving Towards Successful Complaint Management

Chebat J, Slusarczyk W (2005) How emotions mediate the effects of perceived justice on loyalty in
service recovery situations: an empirical study. J Bus Res 58(5):664–673. doi:10.1016/j.
jbusres.2003.09.005
Choi S, Mattila A (2008) Perceived controllability and service expectations: Influences on
customer reactions following service failure. J Bus Res 61(1):24–30. doi:10.1016/j.jbusres.
2006.05.006
Churchill GA, Suprenant C (1982) An investigation into the determinants of customer satisfaction.
J Mark Res 19(4):491–504
Dacin PA, Davidow M (1997) Understanding and influencing consumer complaint behavior.
improving organizational complaint management. Adv Consum Res 24(1):450–456
Day RL (1977) Extending the concept of consumer satisfaction. Adv Consum Res 4:149–154
Day RL (1984) Modeling choices among alternative responses to dissatisfaction. Adv Consum Res
11:496–499
de Matos CA, Rossi CAV, Veiga RT, Vieira VA (2009) Consumer reaction to service failure and
recovery: the moderating role of attitude toward complaining. J Serv Mark 23(7):462–475.
doi:10.1108/08876040910995257
Dellarocas C (2006) Strategic manipulation of internet opinion forums: implications for consumers
and firms. Manag Sci 52(10):1577–1593. doi:10.1287/mnsc.1060.0567
Ellison NB, Steinfield C, Lampe C (2007) The benefits of facebook “friends:” social capital and
college students’ use of online social network sites. J Comput Med Commun 12(4):1143–1168.
doi:10.1111/j.1083-6101.2007.00367.x
Estelami H (2000) Competitive and procedural determinants of delight and disappointment in
consumer complaint outcomes. J Serv Res 2(3):285–300. doi:10.1177/109467050023006
Folkes VS (1984) Consumer reactions to product failure: an attributional approach. J Consum Res
10(4):398–409
Gelbrich K, Roschk H (2010) A meta-analysis of organizational complaint handling and customer
responses. J Serv Res 14(1):24–43. doi:10.1177/1094670510387914
Gilly MC (1987) Postcomplaint processes: from organizational response to repurchase behavior.
J Consum Aff 21(2):213–293. doi:10.1111/j.1745-6606.1987.tb00204.x
Gilly MC, Gelb BD (1982) Post-Purchase consumer processes and the complaining consumer.
J Consum Res 9(3):323–328
Grougiou V, Pettigrew S (2009) Seniors’ attitudes to voicing complaints: a qualitative study.
J Mark Man 25(9):987–1001. doi:10.1362/026725709X479336
Halstead D (1991) Consumer attitudes toward complaining and the prediction of multiple
complaint responses. Adv Consum Res 18(1):210–216
Harris LC (2010) Fraudulent consumer returns: exploiting retailers’ return policies. Eur J Mark 44
(6):730–747. doi:10.1108/03090561011032694
Henneberg SC, Gruber T, Reppel A, Ashnai B, Naudé P (2009) Complaint management
expectations: an online laddering analysis of small versus large firms. Ind Mark Manag 38
(6):584–598. doi:10.1016/j.indmarman.2009.05.008
Hirschman AO (1970) Exit, voice, and loyalty. Responses to decline in firms, organizations, and
states. Harvard University Press, Cambridge, Mass
Homburg C, Fürst A (2005) How organizational complaint handling drives customer loyalty: an
analysis of the mechanistic and the organic approach. J Mark 69(3):95–114. doi:10.1509/jmkg.
69.3.95.66367
Huppertz JW (2007) Firms’ complaint handling policies and consumer complaint voicing.
J Consum Mark 24(7):428–437. doi:10.1108/07363760710834843
Huppertz JW, Mower E, Associates (2003) An effort model of first-stage complaining behavior.
J Consum Satisf Dissatisf Complain Behav 16:132–144
Inkpen AC, Tsang EWK (2005) Social capital, networks, and knowledge transfer. Acad Manag
Rev 30(1):146–165. doi:10.5465/AMR.2005.15281445
Jacoby J, Jaccard JJ (1981) The sources, meaning and validity of consumer complaint behavior:
a psychological analysis. J Retail 57(3):4–24
Johnston R, Mehra S (2002) Best-practice complaint management. Acad Manag Exec 4(4):145–154
References 25

Kaplan AM, Haenlein M (2010) Users of the world, unite! The challenges and opportunities of
social media. Bus Horiz 53(1):59–68. doi:10.1016/j.bushor.2009.09.003
Kim MG, Wang C, Mattila AS (2010) The relationship between consumer complaining behavior
and service recovery: An integrative review. Int J Contemp Hosp Manag 22(7):975–991.
doi:10.1108/09596111011066635
Landon EL (1980) The direction of consumer complaint research. Adv Consum Res 7:335–338
Lapré MA, Tsikriktsis N (2006) Organizational learning curves for customer dissatisfaction.
Heterog Across Airl Manag Sci 52(3):352–366. doi:10.1287/mnsc.1050.0462
Larivet S, Brouard F (2010) Complaints are a firm’s best friend. J Strategic Mark 18(7):537–551.
doi:10.1080/0965254X.2010.529155
Lee SJ, Lee Z (2006) An experimental study of online complaint management in the online
feedback forum. J Organ Comp Elec Commer 16(1):65–85. doi:10.1080/10919390609540291
Lerman D (2006) Consumer politeness and complaining behavior. J Serv Mark 20(2):92–100.
doi:10.1108/08876040610657020
Mattila AS, Wirtz J (2004) Consumer complaining to firms: the determinants of channel choice.
J Serv Mark 18(2):147–155. doi:10.1108/08876040410528746
Mayer RC, Davis JH, Schoorman FD (1995) An integrative model of organizational trust. Acad
Manag Rev 20(3):709. doi:10.2307/258792
McCole P (2004) Dealing with complaints in services. Int J Contemp Hosp Manag 16(6):345–354.
doi:10.1108/09596110410550789
Miller KD, Fabian F, Lin S (2009) Strategies for online communities. Strat Mgmt J 30(3):305–
322. doi:10.1002/smj.735
Mitchell V, Boustani P (1994) A preliminary investigation into pre- and post-purchase risk
perception and reduction. Eur J Mark 28(1):56–71. doi:10.1108/03090569410049181
Morrisson O, Huppertz JW (2010) External equity, loyalty program membership, and service
recovery. J Serv Mark 24(3):244–254. doi:10.1108/08876041011040640
Nelson RE (1989) The strength of strong ties: social networks and intergroup conflict in
organizations. Acad Manag J 32(2):377–401
Oliver RL (1980) A cognitive model of the antecedents and consequences of satisfaction decisions.
J Mark Res 17(4):460–469. doi:10.2307/3150499
Orsingher C, Valentini S, Angelis M (2010) A meta-analysis of satisfaction with complaint
handling in services. J Acad Mark Sci 38(2):169–186. doi:10.1007/s11747-009-0155-z
Parkhe A, Wasserman S, Ralston DA (2006) New frontiers in network theory development. Acad
Manag Rev 31(3):560–568
Peterson RA, Kerin R (1979) An information processing theory of consumer choice by James R.
Bettman. Book Review. J Mark 43(3):124–126
Porter CE, Donthu N (2008) Cultivating trust and harvesting value in virtual communities. Manag
Sci 54(1):113–128. doi:10.1287/mnsc.1070.0765
Provan KG, Fish A, Sydow J (2007) Interorganizational networks at the network level: a review of
the empirical literature on whole networks. J Manag 33(3):479–516. doi:10.1177/
0149206307302554
Reynolds KL, Harris LC (2005) When service failure is not service failure: an exploration of the
forms and motives of “illegitimate” customer complaining. J Serv Mark 19(5):321–335.
doi:10.1108/08876040510609934
Richins ML (1982) An investigation of consumers’ attitudes toward complaining. Adv Consum
Res 9(1):502–506
Rust RT, Chung TS (2006) Marketing models of service and relationships. Mark Sci 25(6):560–
580. doi:10.1287/mksc.1050.0139
Schoefer K (2008) The role of cognition and affect in the formation of customer satisfaction
judgements concerning service recovery encounters. J Consum Behav 7(3):210–221. doi:10.
1002/cb.246
Siddiqui MH, Tripathi SN (2010) An analytical study of complaining attitudes: with reference to
the banking sector. J Target Meas Anal Mark 18(2):119–137. doi:10.1057/jt.2010.2
Singh J (1990) A typology of consumer dissatisfaction response styles. J Retail 66(1):57–99
26 2 Moving Towards Successful Complaint Management

Smith AK, Bolton RN (2002) The effect of customers’ emotional responses to service failures on
their recovery effort evaluations and satisfaction judgments. J Acad Mark Sci 30(1):5–23.
doi:10.1177/03079450094298
Spahr C, Arns T (2012) Rasanter Zuwachs: Ältere entdecken soziale Netzwerke, Berlin
Statista (2014) Soziale Netzwerke. Dossier 2014
Tax SS, Brown SW, Chandrashekaran M (1998) Customer evaluations of service complaint
experiences. implications for relationship marketing. J Mark 62(2):60–76
Tichy NM, Tushman ML, Fombrun C (1979) Social network analysis for organizations. Acad
Manag Rev 4(4):507–519
van Eimeren B, Frees B (2011) Ergebnisse der ARD/ZDF-Onlinestudie 2011. Drei von vier
Deutschen im Netz – ein Ende des digitalen Grabens in Sicht? Media Perspektiven, Frankfurt
am Main
van Jaarsveld DD, Walker DD, Skarlicki DP (2010) The role of job demands and emotional
exhaustion in the relationship between customer and employee incivility. J Manag 36(6):1486–
1504. doi:10.1177/0149206310368998
Venkatesh V, Agarwal R (2006) Turning visitors into customers: a usability-centric perspective on
purchase behavior in electronic channels. Manag Sci 52(3):367–382. doi:10.1287/mnsc.1050.
0442
von der Heyde Fernandes D, Pizzuti dos Santos C (2008) The antecedents of the consumer
complaining behavior. Adv Consum Res 35:584–593
Wirtz J, Mattila AS (2004) Consumer responses to compensation, speed of recovery and apology
after a service failure. Int J Serv Ind Manag 15(2):150–166. doi:10.1108/09564230410532484
Wirtz J, McColl-Kennedy JR (2010) Opportunistic customer claiming during service recovery.
J Acad Mark Sci 38(5):654–675. doi:10.1007/s11747-009-0177-6
Chapter 3
Conceptualising Customer Expectations
with Complaint Channels

Abstract This chapter outlines the theoretical framework of the empirical studies.
Five main factors are identified which guide the conceptualisation of customer
satisfaction with complaint channels: procedural justice, interactional justice, dis-
tributive justice, personal attitudes and social interactions. Each factor is described
in detail and related to one of the underlying scientific theories: justice theory,
attribution theory and behaviour theory. As a next step, these factors are linked to
the five complaint channels (mail, e-mail, phone, face-to-face, social networks) in
order to allow for subsequent empirical testing of customers’ perceptions. Based on
this methodology, the adequacy of complaint channels can be tested from a cus-
tomer perspective.

Keywords Distributive justice

Interactional justice
Personal attitudes



Procedural justice Social interactions

Based on the analysis of different theories (justice theory, behaviour theory, and
attribution theory) which are used to illustrate customer satisfaction with complaint
handling (Sect. 2.4), this chapter establishes a detailed connection between the
fundamentals of these theories and complaint channels. In this context, concrete
factors are abstracted from these three theories in order to establish a framework for
the questionnaire and the subsequent statistical analysis, thus revealing the under-
lying factors.
The following investigation builds upon (i) procedural justice, (ii) interactional
justice, (iii) distributive justice, (iv) personal attitudes, and (v) social interaction.
These factors operationalise customers’ satisfaction level with complaint channels
and are therefore utilised in this study.
These five factors are found to influence customer satisfaction levels with
complaint management; a relationship that has been extensively researched. More
than twenty previous studies have been identified as assessing these theories in the
context mentioned. Twelve authors emphasize the importance of procedural justice,
while eleven studies concentrate on interactional justice. The distributive justice
factor has been highlighted as being important by another fifteen studies. Personal
attitudes are stressed as an important element in seven studies, social interaction in

© The Author(s) 2015 27

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0_3
28 3 Conceptualising Customer Expectations …

Complaint channels Dependent variables

Mail Procedural

E-mail Justice dimensions Interactional

Phone Distributive

Social networks Personal attitudes

Face-to-face Social interactions

Customer characteristics

Fig. 3.1 Theoretical framework

ten. An overview of these studies is provided in Appendix A. Their publication

dates range from 1977 to 2010, indicating a continuous relevance throughout the
research process. Most of the previous studies show empirical characteristics, i.e.
surveys and field research; only six studies contain meta-analysis or pure
conceptualisation.
Figure 3.1 illustrates the theoretical framework for this research. Each of the five
complaint channels is tested in its relationship to customer perceptions of the three
justice dimensions (procedural, interactional, distributive) as well as with regard to
personal attitudes and social interaction. Based on customer characteristics, i.e.
demographics, customers will be clustered later on in order to determine further
findings for certain sub-groups. In the remaining part of this chapter, each of the
five factors is defined in detail.

3.1 Factor 1: Procedural Justice

Procedural justice refers to the extent of fairness a company shows in its complaint
handling processes and policies (e.g. McCollough et al. 2000; Voorhees and Brady
2005). According to Smith et al. (1999), procedural justice describes the way
decisions are made and disputes are resolved. More specifically, this includes
factors such as ease of accessibility (Blodgett et al. 1995). It is widely found that
customers prefer a certain method for contacting companies (Sect. 2.2). Robertson
and Shaw (2009) illustrate how customers take into consideration whether it will be
easy for them to contact the focal company (i.e. user-friendliness). Expenditure of
time refers to the effort a customer must make to complete a complaint procedure
(i.e. time-consumption), as well as the amount of time (i.e. response time) it takes
for a company to respond (Davidow 2003). Moreover, procedural justice involves
3.1 Factor 1: Procedural Justice 29

flexibility as a synonym for a company’s ability to adapt its procedures to fit

individual requirements (Tax et al. 1998).
Therefore, it is assumed that complaint channels differ in the extent of their
adequacy from a procedural justice perspective. For instance, as it usually takes
longer to answer mailings, the adequacy of this channel might be considered as
lower compared to the faster response time, e-mails or immediate feedback in face-
to-face situations offer. Given that the majority of social network members log in on
a daily basis (van Eimeren and Frees 2011), it can be inferred that these individuals
are familiar with the opportunity of interaction offered by these platforms.
Moreover, social networks incorporate search engines of their own enabling the
searches for unknown members, or, in this case, companies. Thus, complainants do
not need to take the pre-emptive step of searching for a company’s contact details
somewhere else and are, instead, directed immediately to the company’s site within
the respective online social network. Customers can utilize the well-known pro-
cedures of social networks to address their concerns. In addition, the receipt of
messages through social networks is faster than by conventional channels, given
that messages are delivered immediately and customers usually check for new
messages daily.
All in all procedural justice is based on the three dimensions (1) user-friendliness,
(2) time-consumption, and (3) response time. Based on the selective illustrations,
it is questioned whether communication channels differ regarding the aspect of
procedural justice as perceived by the customer and might therefore be more or less
equally adequate in the context of complaint management.

3.2 Factor 2: Interactional Justice

Interactional justice is established as a description of interpersonal fairness (Michel

et al. 2009). Schoefer (2008) describes it as the treatment a customer receives
during a recovery process. More precisely, interactional justice embodies the degree
of a company’s honesty and politeness in the eyes of the customer (Tax et al. 1998).
Interactional justice is delineated as the employees’ behaviour and their efforts. Tax
et al. (1998) describe this dimension as the amount of energy a company puts into
resolving a problem (Tax et al. 1998). Researchers have also noticed the influence
of respect and displays of helpfulness on complaining customers (Seiders and Berry
1998). In this study honesty and helpfulness constitute interactional justice.
Customers expect companies to vary in their behaviour across diverse com-
munication channels. For instance, a formal mail response by the company might be
impersonal, generic, and lacking a personal note. In contrast, a face-to-face dis-
cussion reflects an employee’s positive or negative behaviour directly. Moreover,
since social networks are characterised as a medium through which customers
usually talk to friends in a private environment, users might expect companies to
adapt a style of communication appropriate to social networks when using them to
respond to complaints, namely in a friendly and honest manner. From a company
30 3 Conceptualising Customer Expectations …

perspective, impolite behaviour might lead to negative word-of-mouth in social

networks.
Therefore, it is assumed that complaint channels differ in their adequacy from a
customer perspective. It is posited that the type of interaction between customer and
company depend on the utilised complaint channel.

3.3 Factor 3: Distributive Justice

Focusing on the final outcome of a complaint, distributive justice is defined as a

customer’s perception of the degree of fairness involved in the company’s efforts to
find a remedy (Maxham and Netemeyer 2002). Compensatory elements might be,
for example, refunds, or rebates (Chebat et al. 2005). Seiders and Berry (1998)
mention three pertinent characteristics: equity, equality, and need. First, equity
refers to the customer perception that the remedy reflects the customer’s input
within the exchange. Second, equality implies customers’ expectation of equality in
the way they are treated. Third, need is defined as the extent to which an outcome is
based on the requirements and needs of customers.
In the context of this study, distributive justice refers to customers’ anticipation
of the outcome of a complaint. It is assumed that degrees of fairness vary depending
on the utilised complaint channel. For instance, when sending a formal letter, the
company is expected to be more engaged than when answering informal complaints
within social networks. It has to be questioned whether customers anticipate that a
complaint by means of social networks leads to a justified outcome. Besides,
customers expectation of the outcome level with regard to e-mail and phone
complaints are not clear given that e-mail answers are expected to be more reliable
than social network communication on the one hand, and formal mail complaints,
which hold less promise, on the other hand.
Overall, an investigation of the distributive dimension is expected to lead to
interesting insights given that different levels of customer perception are posited for
each channel that is taken into consideration.

3.4 Factor 4: Personal Attitudes

Individuals have specific behaviours, attitudes, and preferences (Oliver and Linda
1981). With regard to CCB, customers also have diverse preferences. In principle, if
a company lives up to all expectations, customer satisfaction will likely be high
(Anderson and Sullivan 1993). These personal attitudes are multi-faceted, e.g. Luria
et al. (2009) argue that customers prefer to complain informally (i.e. unofficially)
due to the fear of raising a protest against a superior opponent—the company.
Apart from the extent of this intimidation, perception influences both the
likelihood of complaining and the satisfaction about the complaining process
3.4 Factor 4: Personal Attitudes 31

(Dacin and Davidow 1997). Halstead (1991) emphasises that some customers do
not voice formal complaints since they are afraid of being seen as troublemakers.
Thus, following these findings, companies are advised to offer convenient (i.e.
informal) opportunities for customers to complain in order to adapt to customer
attitudes and increase their satisfaction. Therefore, companies should strive to
match customer preferences and attitudes and offer, for example, convenient, i.e.
informal, ways for contacting the company in order to voice complaints. It is
assumed that customers welcome the incorporation of social networks into com-
plaint management systems. Among the most important causations, it is assumed
that customers perceive social networks as an important informal medium for
interaction with companies, compared to conventional channels such as mail
complaints.
Depending on personal attitudes of the customers, it is not yet clear whether all
communication channels differ in the extent of customer evaluation of complaint
channels. This study sheds some light on the influence of personal attitudes on
customer satisfaction with the complaint channels.

3.5 Factor 5: Social Interaction

Originally developed in the context of social support theory, Gelbrich (2010)

stresses that individuals rely on resources and information from their social envi-
ronment. Such information, expressed as advice, is expected to help people cope
with problems (Gelbrich 2010). By allowing customers to interact with each other
as well as encourage the exchange of experiences and the sharing of advices,
companies are able to influence CCB (Dacin and Davidow 1997), guiding them
towards making a direct complaint to the company. By emphasising social inter-
actions among customers, companies encourage complaining within an area their
customers control, potentially leading to higher customer satisfaction and an
enhanced company image (Dacin and Davidow 1997). As pointed out by Gelbrich
(2010), empathy and understanding from other customers can provide an outlet for
frustration and in the end lead to an increase in satisfaction. Malafi (1991) explicitly
investigated the role of social influence on both CCB and customer satisfaction. The
authors showed that most customers appreciate the opportunity for social interac-
tion in order to obtain advice, share feelings, and exchange experiences.
Therefore this construct represents a major characteristic of the investigation. As
pointed out by past research (Malafi et al. 1993), social interaction shows that
customers appreciate getting advice from and sharing their experiences with others.
By means of social networks, customers have the opportunity to interact within
their social environment and to simultaneously contact the company. None of the
conventional communication channels is capable of allowing customers to perform
both actions at the same time. Thus, customer satisfaction is expected to be higher
when this channel is used. However, it is anticipated that customers, who intend to
complain anyway are not equally receptive to and reliant on social interaction
32 3 Conceptualising Customer Expectations …

compared to those who mostly do not complain at all or do so seldom. It is therefore

important to examine the relationship between the customers’ appreciation of social
interaction and their general preference for making a complaint.
Furthermore, it can be assumed that the appreciation of social interaction relies
neither on demographic factors nor on the preferred communication channel, but is
instead generally present among all demographic clusters of customers. Therefore,
by extending the investigation of Malafi (1991), it is assumed that the appreciation
of social interaction does not rely on demographic determinants, such as gender,
age, or the preferred communication channel.
Finally, it is posited that social interaction is not linked to customers’ evaluation
of complaint management systems in general. In other words, the appreciation of
social interaction does not vary among the assessments of the other antecedents of
satisfaction. Therefore, it is postulated that the appreciation of social interaction
does not vary on the basis of customers’ evaluations of the justice dimensions and
personal attitudes.

Take away
An examination on the adequacy of complaint channels is important. Apart
from the well-known justice dimensions (interactional, procedural, distribu-
tive), further factors have been found to be relevant: personal attitudes of
costumers and their desire for social interaction with others. Customers are
therefore expected to vary in their evaluation of complaint channels on basis
of their individual characteristics. When offering complaint channels, com-
panies should have their target group in mind.

References

Anderson EW, Sullivan MW (1993) The antecedents and consequences of customer satisfaction
for firms. Mark Sci 12(2):125–143. doi:10.1287/mksc.12.2.125
Blodgett JG, Wakefield KL, Barnes JH (1995) The effects of customer service on consumer
complaining behavior. J Serv Mark 9(4):31–42. doi:10.1108/08876049510094487
Chebat J, Davidow M, Codjovi I (2005) Silent voices: why some dissatisfied consumers fail to
complain. J Serv Res 7(4):328–342. doi:10.1177/1094670504273965
Dacin PA, Davidow M (1997) Understanding and influencing consumer complaint behavior:
improving organizational complaint management. Adv Consum Res 24(1):450–456
Davidow M (2003) Organizational responses to customer complaints: what works and what
doesn’t. J Serv Res 5(3):225–250. doi:10.1177/1094670502238917
Gelbrich K (2010) Anger, frustration, and helplessness after service failure: coping strategies and
effective informational support. J Acad Mark Sci 38(5):567–585. doi:10.1007/s11747-009-
0169-6
Halstead D (1991) Consumer attitudes toward complaining and the prediction of multiple
complaint responses. Adv Consum Res 18(1):210–216
Luria G, Gal I, Yagil D (2009) Employees’ willingness to report service complaints. J Serv Res 12
(2):156–174. doi:10.1177/1094670509344214
References 33

Malafi TN (1991) The impact of social influence on consumer complaint behavior. J Consum
Satisfaction Dissatisfaction Complaining Behav 4:144–150
Malafi TN, Cini MA, Taub SL, Bertolami J (1993) Social influence and the decision to complain:
investigations on the role of advice. J Consum Satisfaction Dissatisfaction Complaining Behav
6:81–89
Maxham JG, Netemeyer RG (2002) Modeling customer perceptions of complaint handling over
time: the effects of perceived justice on satisfaction and intent. J Retail 78(4):239–252. doi:10.
1016/S0022-4359(02)00100-8
McCollough MA, Berry LL, Yadav MS (2000) An empirical investigation of customer satisfaction
after service failure and recovery. J Serv Res 3(2):121–137. doi:10.1177/109467050032002
Michel S, Bowen D, Johnston R (2009) Why service recovery fails: tensions among customer,
employee, and process perspectives. J Serv Manag 20(3):253–273. doi:10.1108/
09564230910964381
Oliver RL, Linda G (1981) Effect of satisfaction and its antecedents on consumer preference and
intention. Adv Consum Res 8:88–93
Robertson N, Shaw RN (2009) Predicting the likelihood of voiced complaints in the self-service
technology context. J Serv Res 12(1):100–116. doi:10.1177/1094670509333789
Schoefer K (2008) The role of cognition and affect in the formation of customer satisfaction judgements
concerning service recovery encounters. J Consum Behav 7(3):210–221. doi:10.1002/cb.246
Seiders K, Berry LL (1998) Service fairness: what it is and why it matters. Acad Manag Exec 12
(2):8–20. doi:10.5465/AME.1998.650513
Smith AK, Bolton RN, Wagner J (1999) A model of customer satisfaction with service encounters
involving failure and recovery. J Mark Res 36(3):356. doi:10.2307/3152082
Tax SS, Brown SW, Chandrashekaran M (1998) Customer evaluations of service complaint
experiences: implications for relationship marketing. J Mark 62(2):60–76
van Eimeren B, Frees B (2011) Ergebnisse der ARD/ZDF-Onlinestudie 2011. Drei von vier
Deutschen im Netz – ein Ende des digitalen Grabens in Sicht? Media Perspektiven, Frankfurt
am Main
Voorhees CM, Brady MK (2005) A service perspective on the drivers of complaint intentions.
J Serv Res 8(2):192–204. doi:10.1177/1094670505279702
Chapter 4
Analysis of Customers’ Complaint
Channel Choice and Complaint Behaviour

Abstract This chapter introduces two empirical studies to analyse the adequacy of
complaint channels. In detail, five communication channels (mail, e-mail, phone,
face-to-face and social networks) as complaint channels are examined from a cus-
tomer perspective with regard to five dimensions: procedural justice, interactional
justice, distributive justice, personal attitudes and social interaction. First, to test the
adequacy of the complaint channels, a one-way analysis of variance is conducted for
each dimension identified. Differences among the adequacy of complaint channels
towards customer perceptions are revealed. Second, a cluster analysis provides
detailed insights into customers’ preferences regarding complaint channel choice.

Keywords Cluster analysis Communication channel Complaint channel





Customer complaint behaviour Empirical studies One-way analysis of variance

In order to investigate the suitability of complaint channels two questionnaire-based

studies are conducted. This chapter outlines the underlying research mechanisms of
the studies.
First, Sect. 4.1 describes the research design. Second, Sect. 4.2 introduces the
questionnaire design. Third, the operationalisation of the independent and the
dependent variables is explained (Sect. 4.3). Fourth, Sect. 4.4 presents the research
agenda. Fifth, the following two Sections describe the studies and their results in
detail (Sects. 4.5 and 4.6). Sixth, a comparison of the involved complaint channels
is drawn up (Sect. 4.7). Finally, Sect. 4.8 shows further results for different sub-
groups of complainers.

4.1 Research Design

The analysis of the research question ‘which communication channels are adequate
options to voice a complaint from a customer perspective?’ is based on primary data
using a questionnaire and does not rely on explicit secondary sources (Blumberg et al.
2008). Furthermore, the data is gathered using a computer-delivered and self-
administered questionnaire; therefore the study reflects an ex-post facto design under

© The Author(s) 2015 35

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0_4
36 4 Analysis of Customers’ Complaint Channel Choice …

field conditions (Albaum et al. 2011) performed as cross-sectional field research

(Blumberg et al. 2008). The formulated management questions are inferred from an
extensive literature review. Hence, the analysis follows a deductive approach in a pure
research context (Babbie 2007). This study is exploratory in that it is the first to
analyse social networks as a complaint channel comparing it to more established ones.
In both quantitative studies, the respondents are instructed to answer the ques-
tions with reference to a fictitious case description (Sect. 4.2.1). Thus, subsequent
data analysis does not build on actual observations but instead on a hypothetical
situation, asking for customer preferences and attitudes. This methodology is also
known as scenario method and is widely accepted being often used in the field of
customer research and complaint management (e.g. Churchill and Suprenant 1982;
von der Heyde and Pizzuti dos Santos 2008; Morel et al. 1997; Robertson and Shaw
2009; Smith et al. 1999; Wirtz and McColl-Kennedy 2010). A sophisticated pre-test
preceded the distribution of the questionnaire.
All ethical standards are adhered to. The study ensures anonymity, i.e. it is not
possible to draw conclusions from the results of the questionnaire about the indi-
vidual respondent, whose participation is voluntary. Moreover, respondents are
informed about the usage of the collected data in a short briefing before answering
the questionnaire survey (Baker 2003). In general, although the research topic is not
likely to be in any way touchy, the entire investigation is approached deontologi-
cally (Macdonald and Beck-Dudley 1994).

4.2 Questionnaire Design

In this section the questionnaire is presented. For the convenience of the partici-
pants the questionnaire for both studies is designed and offered in the English and
German languages. Due to the internationality of this publication the German
version of the questionnaire is not attached. The questionnaire was hosted by
NetQuestionnaires Nederland B.V. and distributed via hyperlink to the respondents.

4.2.1 Structure of Questionnaire

One questionnaire is used in both studies. A cover page welcomes the participants
and familiarises them with the topic in order to motivate them to complete the
questionnaire (Raab-Steiner and Benesch 2010). With regard to Diamantopoulos
and Schlegelmilch (1996), the purpose of the survey is explained. Participants are
then told that their participation is voluntary, anonymous and that it is not possible
to draw personal conclusions from their answers. As the respond-rate for long
questionnaires is typically low (Linsky 1975), the required average duration of
8 minutes (average result of pre-testing) is mentioned in order to avoid early
termination.
4.2 Questionnaire Design 37

The questionnaire itself is divided into two parts (Appendix B). Part one, the
introductory part, asks for demographic data such as gender, and age. They are
asked whether they are a member of a social network. If the answer is in the
affirmative they are asked about the frequency of their log-ins. Another question is
whether they have used a social network as a complaint channel in the past. These
two questions are skipped for participants who are not members of social networks,
in which case an alternative text is provided before answering question 10.
Participants are next asked to evaluate their tendency to complain in general and to
name their preferred communication channel. Here, participants are able to enter
complaint channels not listed. The data gathered in this part is primarily used to
classify sub-segments for the subsequent detailed empirical analysis.
The second part of the questionnaire is the main part: It is introduced with a short
scenario description. Participants are asked to imagine the described situation which
illustrates a service failure. The scenario description contains available communi-
cation channels for addressing a complaint to the fictitious company. Respondents
are asked to put themselves in the described situation while answering the questions.
The scenario described ensures that respondents answer with regard to an equal
information basis. In this way biases due to personal experience and memory lapses
are reduced, which is otherwise “common in results based on retrospective self-
reports” (Smith et al. 1999). The questionnaire consists of 21 questions (part one:
7 questions; part two: 14 questions). If any respondent indicates that he or she has no
membership in a social network, two follow-up questions are skipped. In this case,
part one consists of only 5 questions, resulting in a total number of 19. To sum up,
the classes of questions are modelled in the questionnaire are shown in Table 4.1.
After the scenario description, a manipulation check is conducted to test for the
success of the described service failure. Participants are further asked to choose a
complaint channel they would use in the given situation. Participants are next asked
to indicate their opinion about the five examined communication channels (mail,
e-mail, phone, social networks, and face-to-face) with regard to the four factors
(1) procedural justice, (2) interactional justice, (3) distributive justice, and
(4) personal attitude. Afterwards, they are asked to evaluate their preferences about
getting advice from other customers and the exchange of experience with other
customers (social interaction). Participants are also asked whether they would
appreciate social networks as a complaint channel and whether they are annoyed if
this option of contacting the company is not offered. Participants also evaluate the
user-friendliness of the complaint channels. Finally, participants have the chance to
comment on the questionnaire itself. To receive the results of the study, participants
could enter their e-mail address (Blumberg et al. 2008). These participants have
been sent a summary of the results.
According to the requirements suggested by Baker (2003), the items are con-
structed in a clear manner without the use of specific terminology. Furthermore, the
use of double-barrelled and double-negative questions is avoided, as suggested in
the literature (e.g. Lietz 2010). In order to avoid ambiguity and lack of compre-
hensibility all questions are pre-tested extensively, as described in Sect. 4.2.2. The
questionnaire is shown in Appendix B.
38 4 Analysis of Customers’ Complaint Channel Choice …

Table 4.1 Categories of questions

Simple category scale According to this scale, respondents either have to agree or
disagree, confirm or reject statements, or simply indicate their
gender. For the respondents’ convenience, the option ‘no
preference’ can also be selected. The inclusion of this alternative
prevents dropouts (Albaum et al. 2011).
Multiple choice single- This scale is used when respondents are asked to indicate their age
response scale in terms of selecting the appropriate age cluster. Albaum et al.
(2011) suggest to include the option ‘not specified’ to avoid
dropouts. These are even more likely when personal data is
requested.
Multiple choice multiple- This category allows for multiple answers, as for instance,
response scale whenever respondents are asked for their preferred
communication channel.
Single- and multiple-rating This scale is the most used in the conducted survey. Typically,
Likert scale respondents are asked to rate the given communication channels
with regard to a certain question. 7-point Likert scales are
incorporated as response scales with a middle point are expected
to increase validity and reliability (Lietz 2010). Furthermore,
taking Blumberg et al. (2008) into account, seven scale options
are most appropriate to reduce central tendency at the same time.
Finally, 7-point Likert scales are widely used in the research field
of complaint management and customer research (e.g. Achrol
1991; Brady and Cronin 2001; Choi and Mattila 2008; de Matos
et al. 2009). Generally, the most negative anchor point is
positioned on the left (e.g. ‘in no case suitable’), and the most
positive anchor point is positioned on the right (e.g. ‘very
suitable’). A detailed description of intervals and captions of all
Likert scales questions is illustrated in Appendix C.
Open question Although almost all of the survey questions are closed questions,
which provide predetermined answers, one multiple-choice answer
item and one more question are designed in an open format. First,
with reference to question six, the respondent is allowed to suggest
one further channel that might not have been given. Second, at the
end of the questionnaire, all respondents are offered the chance to
comment on the questionnaire. By means of this mechanism, the
overall integrity of the questionnaire is assured and methodical
errors, if any, can be revealed (Babbie 2007).

4.2.2 Pre-Test

For the pre-testing, four respondents were provided with the German version of the
questionnaire. They were asked to fill-out the entire questionnaire and to measure
the time they needed to complete. They were also asked to take notes about aspects
of the questionnaire with which they were uncomfortable and to explain why this
might be the case. Then the respondents were interviewed and asked to clarify their
subjective perception concerning the intention of each question. This pre-testing
procedure is utilised in order to ensure unambiguousness of each question
4.2 Questionnaire Design 39

(Hunt et al. 1982). Furthermore, as suggested by Baker (2003), the respondents

were asked with regard to their opinion about the following: (i) the appropriate
order of questions, (ii) task difficulty, (iii) skip patterns, and (iv) their overall
interest and attention. An English version of the questionnaire was given to a
bilingual pilot-respondent and who was interviewed in the same manner as
described above. To avoid language differences (Raab-Steiner and Benesch 2010),
this person was asked to comment on the questions’ intention in German. In this
manner, consistency between the German and the English versions was verified.
The results of the pre-test show that it takes approximately 8 minutes to complete
the questionnaire, which is considered to be an acceptable duration (Galesic and
Bosnjak 2009). Both the questions and the answer options are generally compre-
hensible and well defined. One question was deleted owing to redundancy. Also,
six questions were deleted due to the suggestions of two pilot-respondents to reduce
complexity. Re-testing the adjusted questions with the two respondents indicated an
increased level of understanding. Three respondents had also recommended
restructuring the order of the questions by repositioning three questions. Repeated
testing led to approval of these suggestions.

4.3 Operationalisation of the Variables

In this section the used independent variable and dependent variables are
operationalised.
Independent Variables
The independent variable is given by respondents’ complaint channel choice. This
questionnaire follows the study from Mattila and Wirtz (2004). They analysed the
communication channel mail, e-mail, phone and face-to-face. The channel ‘social
networks’ is added as this study focuses on this communication possibility to voice
a complaint via this channel, which is growing in importance. As the study is
adopted from the previous research of Mattila and Wirtz (2004) validity can be
assumed. Moreover, the answers for multiple-rating Likert scales regarding the
mentioned communication channels have also been pre-designed by Mattila and
Wirtz (2004). As the questionnaire is based on a hypothetical situation, respondents
are told which complaint channels are available “[t]o prevent … [respondents] from
being constrained by preconceived notions that certain channels are unavailable”
(Wirtz and Mattila 2004). In this study five complaint channels are given:
(a) Mail
(b) E-mail
(c) Phone
(d) Social network
(e) Face-to-face
40 4 Analysis of Customers’ Complaint Channel Choice …

Dependent Variables
To answer the stated management questions five dependent variables are tested.
They are revealed by former researchers. All variables were adapted to match the
underlying management questions. When not stated otherwise, the dependent
variables are measured on a 7-point Likert scale. An overview of all variables is
given in Appendix C.

First, procedural justice (PJ) is measured by three items. The items are taken
from McCollough et al. (2000). The first item (PJ_1) reflects time-consumption. In
particular, participants are asked to evaluate the different complaint channels with
regard to the time needed to communicate a complaint. The second component of
procedural justice is speed of response (PJ_2). Participants are asked to evaluate the
five complaint channels with regard to their expectations of receiving an immediate
response from the company in reply to their complaint. The third component is
user-friendliness (PJ_3). Participants are asked to evaluate the five complaint
channels with regard to user-friendliness. The items are measured on a 7-point
Likert scale given 1 = “Not time-consuming” and 7 = “Very time-consuming” for
PJ_1; 1 = “Not suitable” and 7 = “Very suitable” for PJ_2; 1 = “Not user-friendly”
and 7 = “Very user-friendly” for PJ_3.
The second dependent variable is interactional justice (IJ). Participants are first
of all asked to evaluate the complaint channels with regard to the honesty of the
company in question (IJ_1). The second item asked for the evaluation of the
complaint channels with regard to helpfulness (IJ_2). The items are taken from
Seiders and Berry (1998) and are expected to represent the latent variable inter-
actional justice. Again, both items are measured on a 7-point Likert scale with
1 = “Not honest” and 7 = “Very honest” for IJ_1; 1 = “Not helpful” and 7 = “Very
helpful” for IJ_2.
Distributive justice (DJ) constitutes the third dependent variable. In particular,
it is asked which communication channel is the most reliable one, i.e. which
complaint communicated through a certain channel will lead to the desired out-
come. Due to the unambiguous character of the distributive justice dimension, no
latent variable is used. The single-items scale is taken from Maxham and
Netemeyer (2002) and measured with 1 = “Not likely” and 7 = “Very likely”.
Fourth, personal attitudes (PA) are measured as a dependent variable.
Complaint channels are evaluated with regard to formalities and as to how official
the complaint process is perceived by participants. Again, a single-item measure-
ment is used, which is adopted from Luria et al. (2009) with the scale anchored on
1 = “Not formal” and 7 = “Very formal”.
The fifth dependent variable is social interaction (SI) and measured by two
items. Item 1 (SI_1) asked participants whether they would appreciate advice from
other customers of the fictitious company via an informal forum. The degree of
appreciation derived from talking to other customers with similar experiences via an
informal forum is measured using the second item for social interaction (IS_2).
These items are taken from Malafi et al. (1991, 1993). Given that the communi-
cation channels, except social networks, do not allow for social interaction, Likert
4.3 Operationalisation of the Variables 41

scales cannot be employed to assess the levels of satisfaction. Both items are
categorical distinct variables, measured with 1 = ‘yes’, 2 = ‘no’, and 3 = ‘no
preference’ for SI_1 and SI_2.

4.4 Research Agenda

In this section the analytical framework is outlined. First, to introduce the data-basis
for both studies, the samples are described. Second, descriptive statistics are given.
Third, to check for the manipulation of the described situation and the familiarity of
the sample with the research topic, manipulation and realism checks are conducted.
Fourth, the underlying constructs are measured. (1) Cronbach alpha is calculated for
multi-item scales to ensure reliability of the construct. (2) An Exploratory Factor
Analysis (EFA) is conducted to verify the underlying structure of the dependent
variables.
For the detailed empirical analysis, only participants who indicated that they are
a member of at least one social network are considered. It is assumed that those who
are not registered members of social networks are not able to evaluate the specific
characteristics of social networks and would therefore bias the results.
For the empirical analysis, the data of both studies has to be restructured in order
to allow for the intended analyses. For instance, the respondents’ evaluations of
each communication channel have been classified as variables by the SPSS import
process. However, it is necessary to transform these variables into cases. Hence all
respondents are represented by five rows, one for the evaluation of each commu-
nication channel. Whenever participants answered questions where no reference to
communication channels is given, one row to represent these values would theo-
retically be sufficient. However, these values are copied and are also represented
five times (same values considered in each row), given that each respondent is
registered by five rows. Consequently, the degrees of freedom for the following
analysis are usually five times higher than the actual collected data. However, this is
not expected to bias the results given that the ratios remain equal.
A one-way ANalysis Of VAriance (ANOVA) with post hoc tests is conducted
for both studies. Therefore, differences between the five examined complaint
channels with regard to the dependent variables’ mean values can be analysed. To
check for the assumption of variance homogeneity between the groups, a Leven’s
test is conducted. Further, a chi-square test as well as an independent t-test are
conducted to analyse the influence of the complaint channels of customers’ need for
social interaction.
To gain detailed insights in customers’ choice of complaint channel and their
complaint behaviour subgroups are analysed. First, a cluster analysis is conducted.
After that a one-way ANOVA is performed to analyse age clusters in more detail.
Finally, a decision tree is presented. Figure 4.1 gives an overview of the entire data
analysis procedure.
42 4 Analysis of Customers’ Complaint Channel Choice …

Data Analysis

Manipulation and Construct

Descriptive Statistics
Realism Checks Measurement

Cronbach Alpha

Exploratory
Factor Analysis

Analysis of Variance
Independent t-Test Chi-Square Test
(ANOVA)

Levene Test

Post-hoc Tests

Analysis of Variance
Cluster Analysis Classification Tree
(ANOVA)
Analysis of Subgroups

Fig. 4.1 Data analysis flow

4.5 Study 1

In the following sections study 1 is described. Section 4.5.1 gives a short overview
of the sample. Section 4.5.2 follows with a detailed description of the samples
descriptive statistics. Next, the success of the manipulation and participants’
familiarity with the research topic are checked (Sect. 4.5.3). In Sect. 4.5.4 the
construct measurement is conducted. Finally, the results are documented in
Sect. 4.5.5.

4.5.1 Sample

This research study makes a contribution to existing theories in the field of com-
plaint management, which are predominantly used to investigate preferences, atti-
tudes, and the behaviour of customers (e.g. Henneberg et al. 2009; Voorhees and
Brady 2005). Therefore, this research analyses people, i.e. customers. As it is
impossible to conduct a complete survey, a sampling procedure is used (Kromrey
2006). The data-gathering procedure is designed as follows:
4.5 Study 1 43

1,260 employees of a local German savings bank were asked to participate in the
study. Thanks to the bank sponsoring the survey, the questionnaire was distributed
to all employees. An announcement to the employees requesting voluntary par-
ticipation was posted on the bank’s intranet on 13 January 2012. The participants
were explicitly asked to respond as private customers and not within their banking
context. To comply with ethical standards, the questionnaire as well as the par-
ticipation of the employees was approved prior to this by the bank’s executive
board. Participation was 28 % within the first four days following the official
statement—an adequate number in view of the intended aim, no reminders were
sent.

4.5.2 Descriptive Statistics

In this Section the data screening and cleaning procedure is described. After that
follows the description of the descriptive statistics for the participants.
Data screening and cleaning
414 responses were drawn from the sample, which amounted to a response rate of
33 %. All questionnaires are checked for data anomalies. Questionnaires which
were either incomplete or showed implausible answers were deleted from the
sample. Also, respondents who indicated not being comfortable with the structure
of the questionnaire were also deleted. Participants could express their opinion
about the questionnaire itself at the end of the survey. 13 respondents had reser-
vations about the structure of the questionnaire; these answers were therefore
deleted to avoid biased results. In total the main sample contains 367
questionnaires.
Participants
In this study the average age of participants ranges between 30 and 39 years. Of
these 367 participants 52.6 % are female. An overview of age and gender distri-
bution is given in Table 4.2. With an increase in age, social network membership
declines. This is in line with the study of van Eimeren and Frees (2011) and
therefore strengthening the appropriateness of the entire sample.

Table 4.2 Descriptive statistics—study 1

Description Age cluster of respondents
<20 20–29 30–39 40–49 50–59 60–69
Sample size 1,260
Responses 414
Completed questionnaires 367 1 79 95 120 66 6
Female (%) 53 100 54 59 55 46 17
SN member 193 1 73 61 45 13 0
44 4 Analysis of Customers’ Complaint Channel Choice …

4.5.3 Manipulation and Realism Check

To check whether the manipulation of the described scenario is successful, par-

ticipants’ dissatisfaction with the fictitious company is calculated. Participants
indicated that they are highly dissatisfied with the company after the described
service failure (CASE_DISSA: M = 5.23, SD = 1.98; measure with 1 = “Not
annoyed at all”, 7 = “Very annoyed”). In particular, 86.6 % of the participants
indicate that they are dissatisfied with the company. Therefore, the manipulation of
the service failure is successful.
Participants show a high familiarity with the research topic. To check the par-
ticipants’ familiarity with the purpose of this research they were asked to describe
their (1) complaint behaviour (COMP.Pref), (2) social network membership
(NET_MEMBER), and (3) social network activities (NET_FREQ). First, more than
60 % of the participants formulated a complaint towards the company after having
gone through a dissatisfying shopping experience (product or service; M = 4.99,
SD = 1.57). This represents a high ratio in comparison to past research (Voorhees
and Brady 2005). Second, 52.6 % of the study participants are a member of a social
network. Third, from this subgroup, 80.3 % of the participants indicated that they
visit their social network several times a week (i.e. 35.2 %), or even on a daily basis
(i.e. 45.1 %). Therefore, it is to be assumed that the participants are familiar with the
described situation and the topic of the questionnaire. However, only 3.1 % of the
active social network users indicated having used social networks as a complaint
channel in the past. This result is quite unexpected and emphasises the importance
of the underlying research topic.
Participants’ channel preferences (CASE_CHAN) with regard to the described
situation were also looked into. The majority of the participants would choose
classical complaint channels, as for instance calling the company (phone = 42.9 %)
or sending an e-mail (e-mail: 28.4 %). 19.9 % of the participants went to the
company in person to communicate their complaint via a face-to-face conversation.
8.2 % would resorted to writing a letter. Only 5 % of the participants opted for
social networks.
For a detailed look into social networks as a complaint channel only participants
with a membership of at least one social network were considered. It turned out that
56.8 % of the participants with social network activities valued advice from other
customers (IS_1) and that 59.6 % appreciated sharing their experiences with a
company with these (IS_2). About 10.1 % of all participants welcomed social
networks as a complaint channel (KANO_P), 1.9 % would be annoyed if this option
is deleted (KANO_N).
The incorporated complaint channels are sufficient. The open questions
(PREF_CHANE1; OPEN_QUESTION) show that only a minority of 6 participants
requested further channels. These specific participants suggested facsimile (5 par-
ticipants) and homepage contact forms (1 participants) as additional complaint
channels. Facsimile as complaint channel is not considered in the analysis due to
two reasons. First, in comparison with other technologies, this type of
4.5 Study 1 45

communication is not often used nowadays (Statistisches Bundesamt 2014).

Second, the effort for customer to send a facsimile is comparable to sending a mail
by regular post. It can therefore be concluded, that no important communication
channel is missing which would otherwise distort the results.
All in all, more than 70 participants submitted their e-mail addresses and showed
interest in the survey results. It can be inferred that the topic was seen as interesting
and thereby containing the motivation to give honest and usable answers.

4.5.4 Construct Measurement

To evaluate the reliability of the dependent variables procedural and interactional

justice Cronbach alpha is calculated (Cronbach 1951). As these two dependent
variables are latent variables, the measurement via multi-item scales is necessary. In
the literature a limit of 0.7 is suggested to assure that the measurement is free of
random error (Nunnally 1978).
In general, the items anchor to 1 with the most negative value, and to 7 with the
most positive value. As PJ_1 (time-consumption) is designed in a reversed manner
(1 = most positive value and 7 = most negative value), a new item (PJ_1r) is
calculated by revising the initial values. Table 4.3 shows a Cronbach alpha of 0.577
for all three items taken into account. As this value cannot be significantly
improved by eliminating one item, all three items are considered to represent the
factor procedural justice (Table 4.3). Furthermore, as column “Corrected item total
correlation” in Table 4.3 shows, all values exceed the required threshold of 0.3.
Therefore, it can be applied that the three items measure the same underlying
construct: procedural justice, although the Cronbach alpha of 0.577 is not really
satisfying. However, as the study is more exploratory, it seems reasonable to
conduct further analyses with this factor (Schmitt 1996; Singh 1990). The proce-
dure is repeated for interactional justice (Table 4.3). The two items for interactional
justice show Cronbach alpha of 0.824 which exceeds the common threshold.
As research on social networks as a complaint channel is still in an early stage, an
exploratory factor analysis is conducted to test for the underlying structure of pro-
cedural and interactional justice. First, the results of the Kaiser-Meyer-Olkin measure

Table 4.3 Reliability and construct measurements—study 1

Procedural M SD Rotated KMO Bartlett’s Corrected Cronbach Cronbach
justice factor test test of item total alpha if item alpha
loadings sphericity correlation deleted
PJ_1r 3.96 1.88 0.881 0.663 0.000 0.31 0.594 0.577
PJ_2 4.84 1.82 0.541 0.41 0.436
PJ_3 4.98 1.72 0.611 0.45 0.391
Interactional justice
IJ_1 4.51 1.65 0.876 0.7 – 0.824
IJ_2 4.63 1.61 0.904 0.7 –
46 4 Analysis of Customers’ Complaint Channel Choice …

(KMO-test) show a value above 0.6 and the Bartlett’s test of sphericity is significant
(p = 0.000) which indicate that the data is suitable for factor analysis (Pallant 2007).
The results are shown in Table 4.3. Furthermore, the results show that two factors are
extracted as the Kaiser’s criterion reveals two factors with an Eigenvalue above 1
(factor 1 = 2.39; factor 2 = 1.24). These two factors explain 70.63 % of the variance.
The rotated factor loadings reveal that procedural justice contains out of the three
items PJ_1r (i.e. time-consumption), PJ_2 (i.e. speed of response) and PJ_3 (i.e. user-
friendliness). For interactional justice, the factor loadings of items IJ_1 (i.e. honesty)
and IJ_2 (i.e. helpfulness) exceed the common thresholds (Backhaus 2011). For
further data analysis, the factors’ mean values are calculated.
As distributive justice (DJ) and personal attitudes (PA) are measured on a single-
item scale, and as the items of social interaction (SI) are categorical distinct vari-
ables, no further calculation is necessary at this point. But because personal attitude
is negative formulated in the questionnaire, it has to be reversed. Hence, a new
variable (PAr) is calculated.
Next, a one-way ANOVA with post hoc test is performed to test for differences
in customer perceptions of the four factors (interactional, procedural, distributive
justice and personal attitudes) towards complaint channels. Finally, the desire for
social interaction is tested by a chi-square test and an independent t-test.

4.5.5 Results

To test for the effects of the different complaint channels on the dependent variables
a one-way ANOVA with post hoc test is conducted. ANOVAs are performed in
order to determine potential differences among customers’ evaluations of the
channels.
Levene’s test1 is conducted which shows that the variances are, however,
significantly different within the four groups (PJ: F(4, 955) = 1.576, p = 0.178; IJ: F
(4, 955) = 004, p = 0.004; DJ: F(4, 955) = 8.841, p = 0.000; PAr: F(4, 955)
= 17.726, p = 0.000). Subsequent efforts to reach homogeneity by means of data
transformation do not lead to improvements. Further analysis, therefore, for IJ, DJ
and PAr refers to the Brown-Forsythe robust test instead of the original ANOVA.
The violation of homogeneity is not expected to result in bias, given that the
purpose of this study is rather exploratory and relies on quite a large sample. The
results of the Brown-Forsythe and Welch tests show that there are significant
differences (all p < 0.001) with regard to each variable PJ, IJ, DJ and PAr
(Table 4.4). It must be noted that ANOVA shows the same significant differences.
Further, as variance homogeneity is given for PJ, ANOVA results are shown in
Table 4.5.

1
It has again to be emphasised that the degrees of freedom increased because of data restructuring.
Thus, each respondent is characterised by 5 df, one for each channel.
4.5 Study 1 47

Table 4.4 Robust tests of equality of means—study 1

Variable Statisticsa df1 df2 Sig.
Procedural justice (PJ) Welch 101.036 4 477.099 0.000
Brown-Forsythe 102.0950 4 943.128 0.000
Interactional justice (IJ) Welch 87.759 4 476.761 0.000
Brown-Forsythe 85.858 4 937.800 0.000
Distributive justice (DJ) Welch 96.368 4 475.673 0.000
Brown-Forsythe 105.083 4 909.670 0.000
Personal attitudes (PAr) Welch 194.080 4 472.997 0.000
Brown-Forsythe 201.728 4 872.616 0.000
a
Asymptotically F distributed

Table 4.5 ANOVA results for procedural justice—study 1

Procedural justice (PJ) Square sum df1 Fa Sig.
Between groups 507.079 4 102.095 0.000
Within groups 1185.804 955
Sum 1692.883 959
a
Asymptotically F distributed

The results show that there are significant differences between the complaint
channels with regard to the different dependent variables. A post hoc test is per-
formed to analyse the complaint channel differences for the four dependent vari-
ables in more detail. As most of the dependent variables show heterogeneity of
variance, the Games-Howell test is conducted instead of the intended Tukey-Test.
An overview of the results of the Games-Howell test is given in Table 4.6.
In the following the results for the different dependent variables are shown and
the results of the post hoc tests for the four factors, procedural justice, interactional
justice, distributive justice as well as personal attitudes are introduced. Then, social
interaction is examined.
Procedural justice
Complaining can be very time-consuming, as companies need time to respond to a
complaint. It must also be considered that different complaint channels offer varying
levels of user-friendliness. These aspects are measured with procedural justice. An
overview of the mean values of the different complaint channels for the dependent
variable procedural justice is given in Fig. 4.2.

The conducted post hoc tests show significant differences between the complaint
channel mail and e-mail (MMail = 3.43, SDMail = 1.15; ME-mail = 5.25, SDE-
mail = 1.01; p = 0.000). Hence, to communicate a complaint via e-mail is perceived
as more convenient (i.e. less time-consuming, faster response, and user-friendly)
than communicating complaints via normal mail. Furthermore, complaints which
are filed by phone are also perceived as more convenient than those by letter
48 4 Analysis of Customers’ Complaint Channel Choice …

Table 4.6 Results post hoc tests—study 1

Complaint channel Mean SD
Procedural justice
Mail 5.41 1.48
E-mail 5.05a 1.38
Phone 5.37b 1.32
Social network 3.34 1.59
Face-to-face 6.07 1.12
Interactional justice
Mail 4.72c 1.37
d
E-mail 4.52 1.32
Phone 4.88e 1.27
Social network 3.18 1.36
Face-to-face 5.54 1.12
Distributive justice
Mail 5.41f 1.48
E-mail 5.05g 1.38
Phone 5.37h 1.32
Social network 3.34 1.59
Face-to-face 6.07 1.12
Personal attitude
Mail 1.57 0.95
E-mail 2.90i 1.21
Phone 3.07j 1.40
Social network 5.16 1.58
Face-to-face 2.21 1.40
Notes
If not stated otherwise, complaint channels differ significantly from each other at p < 0.05
a
Does not differ significantly from b
c,d,e
Do not differ from each other significantly
f,g,h
Do not differ from each other significantly
i
Does not differ significantly from j

(MMail = 3.43, SDMail = 1.15; MPhone = 5.40, SDPhone = 1.12; p = 0.000). Compared
to mail, complaints via social networks or face-to-face are both perceived as more
convenient, too (MMail = 3.43, SDMail = 1.15; MSocialnetwork = 4.17,
SDSocialnetwork = 1.20; p = 0.000; MFacetoface = 4.71, SDFacetoface = 1.09; p = 0.000).
The results show that there is no significant difference between the perceived
procedural justice of the complaint channels phone and e-mail (ME-mail = 5.25, SDE-
mail = 1.01; MPhone = 5.40, SDPhone = 1.12; p = 0.625). Thus, customers perceive
both complaint channels as being convenient. Social network is also compared to
e-mail. The results show that e-mail is perceived as more convenient than social
networks (ME-mail = 5.25, SDE-mail = 1.01; MSocialnetwork = 4.17,
SDSocialnetwork = 1.20; p = 0.000). Customers perceive e-mail complaints to be more
4.5 Study 1 49

Procedural justice
7

1
Mail E-mail Phone Social network Face-to-face

Fig. 4.2 Overview mean values for PJ—study 1

convenient than those made face-to-face regarding the dimensions of procedural

justice (ME-mail = 5.25, SDE-mail = 1.01; MFacetoface = 4.71, SDFacetoface = 1.09;
p = 0.000). The results reveal a significant difference between the complaint
channels phone and social network (MPhone = 5.40, SDPhone = 1.12;
MSocialnetwork = 4.17, SDSocialnetwork = 1.20; p = 0.000). Hence, calling a company
on the phone to communicate a complaint is perceived as being more convenient
than complaining via a social network, e.g. Facebook; complaining via face-to-face
conversation is also perceived as more convenient than seeking company contact
via social networks (MSocialnetwork = 4.17, SDSocialnetwork = 1.17; MFacetoface = 4.71,
SDFacetoface = 1.09; p = 0.000).
Summary
As a summary, the classical complaint channels e-mail and phone are perceived as being
the most convenient (i.e. time-consumption low, fast response from the company, and a
user-friendly design of the complaint channel) channels by customers. They are followed
by complaining via a face-to-face conversation, social networks come next in this order,
mail as a complaint channel comes in last.

Interactional justice
Interactional justice summarises the complaint channel characteristics honesty and
helpfulness. Figure 4.3 offers an overview of the mean values of interactional
justice with regard to the different complaint channels.
50 4 Analysis of Customers’ Complaint Channel Choice …

Interactional justice
7

1
Mail E-mail Phone Social network Face-to-face

Fig. 4.3 Overview mean values for IJ—study 1

Results of the post hoc tests reveal that the complaint channels mail, e-mail, and
phone are perceived as being equally helpful/honest. Therefore no significant dif-
ferences exist (MMail = 4.72, SDMail = 1.37; ME-mail = 4.52, SDE-mail = 1.32; p = 0.602;
MMail = 4.72, SDMail = 1.37; MPhone = 4.88, SDPhone = 1.27; p = 0.765; ME-mail = 4.52,
SDE-mail = 1.32; MPhone = 4.88, SDPhone = 1.27; p = 0.056). Mail is perceived as being
more interactional than social networks (MMail = 4.72, SDMail = 1.37;
MSocialnetwork = 3.18, SDSocialnetwork = 1.36; p = 0.000). However, a face-to-face
complaint is perceived as giving more helpful and honest results than a complaint via
mail (MMail = 4.72, SDMail = 1.37; MFacetoface = 5.54, SDFacetoface = 1.12; p = 0.000).
When comparing the complaint channels e-mail and social network, again, e-mail is
perceived as being more helpful than a social network (ME-mail = 4.52, SDE-mail = 1.32;
MSocialnetwork = 3.18, SDSocialnetwork = 1.36; p = 0.000). However, when comparing a
face-to-face conversation about, for instance, a service failure to a complaint via
e-mail, the latter is perceived as providing less interaction (ME-mail = 4.52, SDE-mail
= 1.32; MFacetoface = 5.54, SDFacetoface = 1.12; p = 0.000). The same applies for face-to-
face versus phone complaints (MPhone = 4.88, SDPhone = 1.27; MFacetoface = 5.54,
SDFacetoface = 1.12; p = 0.0000). Comparing making a phone call with using social
networks in this regard the results show that phone calls offer more interactional
possibilities than the new complaint channel social network (MPhone = 4.88,
SDPhone = 1.27; MSocialnetwork = 3.18, SDSocialnetwork = 1.36; p = 0.000). Face-to-face
handling of a complaint, when compared to using a social network is perceived as less
interactional (MFacetoface = 5.54, SDFacetoface = 1.12; MSocialnetwork = 3.18,
SDSocialnetwork = 1.36; p = 0.000).
4.5 Study 1 51

Summary
Participants see a face-to-face complaint as the most honest and helpful way to commu-
nicate a complaint. The complaint channels mail, e-mail and phone, i.e. calling the com-
pany, come next. Social networks show the lowest values, which indicate a lower degree of
trustworthiness for social networks.

Distributive justice
Next, distributive justice is analysed as a dependent variable. Here, complaint
channels are evaluated with regard to the perceived likelihood of complaints being
handled successfully, therefore leading to the desired outcome (e.g. an appropriate
compensation). Figure 4.4 shows the mean values.

The complaint channels mail, e-mail and phone are perceived as being equal in
their potential to generate a desirable outcome after the complaint has been made.
Post hoc tests results, in particular, show no significant differences (MMail = 5.41,
SDMail = 1.48; ME-mail = 5.05, SDE-mail = 1.38; p = 0.092; MMail = 5.41,
SDMail = 1.48; MPhone = 5.37, SDPhone = 1.32; p = 0.998; ME-mail = 5.05, SDE-mail
= 1.38; MPhone = 5.37, SDPhone = 1.32; p = 0.132). These results are similar as the
complaint channel adequacy with regard to interactional justice. However, cus-
tomers evaluate the likelihood of a companies’ willingness to consider a customer
complaint and targeting the customers’ desired outcome higher for mail than for a
social network (MMail = 5.41, SDMail = 1.48; MSocialnetwork = 3.34,
SDSocialnetwork = 1.59; p = 0.000). The same applies for the complaint channels
e-mail and social networks (ME-mail = 5.05, SDE-mail = 1.38; MSocialnetwork = 3.34,

Distributive justice
7

1
Mail E-mail Phone Social network Face-to-face

Fig. 4.4 Overview mean values for DJ—study 1

52 4 Analysis of Customers’ Complaint Channel Choice …

SDSocialnetwork = 1.59; p = 0.000). When contrasting mail and e-mail with making a
face-to-face complaint customers see a higher chance for reaching a successful
outcome in a face-to-face encounter than via mail or e-mail (MFacetoface = 6.07,
SDFacetoface = 1.12; MMail = 5.41, SDMail = 1.48; p = 000; MFacetoface = 6.07,
SDFacetoface = 1.12; ME-mail = 5.05, SDE-mail = 1.38; p = 0.000). Face-to-face also has
the upper hand in this regard when comparing it to using the phone for making a
complaint. (MFacetoface = 6.07, SDFacetoface = 1.12; MPhone = 5.37, SDPhone = 1.32;
p = 0.000). The phone, though, is perceived as being more practical with regard to
distributive justice compared to social networks (MPhone = 5.37, SDPhone = 1.32;
MSocialnetwork = 3.34, SDSocialnetwork = 1.59; p = 0.000). Also, the comparison of
social networks and face-to-face communicated complaints shows that face-to-face
complaints are more beneficial than complaints via social networks (MFacetoface
= 6.07, SDFacetoface = 1.12; MSocialnetwork = 3.34, SDSocialnetwork = 1.59; p = 0.000).
Summary
Similar to interactional justice, face-to-face complaints are attributed the highest probability
of getting to a desired outcome when compared to the other complaint channels. Mail, e-
mail and phone come next. Again, participants assume that a complaint via social networks
will probably not lead to the outcome desired.

Personal attitude
Whether or not participants perceive a complaint channel as official or formal is
measured by personal attitude. Figure 4.5 shows the mean values.

Personal attitude
7

1
Mail E-mail Phone Social network Face-to-face

Fig. 4.5 Overview mean values for PA—study 1

4.5 Study 1 53

The post hoc tests reveal awaited results. Mail in comparison to the other four
complaint channel is perceived as the most official and formal complaint channel
(MMail = 1.57, SDMail = 0.95; ME-mail = 2.90, SDE-mail = 1.21; p = 0.000; MPhone
= 3.07, SDPhone = 1.40; p = 0.000; MSocialnetwork = 5.16, SDSocialnetwork = 1.58;
p = 0.000; MFacetoface = 2.21, SDFacetoface = 1.39; p = 0.000). E-mail and phone are
not perceived differing in degrees of formality (ME-mail = 2.90, SDE-mail = 1.21;
MPhone = 3.07, SDPhone = 1.40; p = 0.699). As expected, complaining via e-mail is
perceived being much more official in nature than complaining via a social network
(ME-mail = 2.90, SDE-mail = 1.21; MSocialnetwork = 5.16, SDSocialnetwork = 1.58;
p = 0.000). The same is true for e-mail and face-to-face encounters (ME-mail = 2.90,
SDE-mail = 1.21; MFacetoface = 2.21, SDFacetoface = 1.39; p = 0.000). Complaining on
the phone is perceived as more official than a communicated complaint though a
social network, but as less official than complaining face-to-face (MPhone = 3.07,
SDPhone = 1.40; MSocialnetwork = 5.16, SDSocialnetwork = 1.58; p = 0.000;
MFacetoface = 2.21, SDFacetoface = 1.39; p = 0.000). A face-to-face complaint is
perceived as far more official and formal than one via social networks
(MFacetoface = 2.21, SDFacetoface = 1.39; MSocialnetwork = 5.16, SDSocialnetwork = 1.58;
p = 0.000), which constitutes the most unofficial complaint channel.
Summary
Social networks are perceived as the most informal and unofficial complaint channel. Next
in this order are e-mail and phone, then face-to-face conversation; the most formal com-
plaint channel being mail.

Social interaction
To analyse whether social interaction, measured as the degree to which customers
appreciate advice from other customers and the exchange of experience with these
influences the customers’ complaint channel preference; chi-square tests are con-
ducted. The validity of the chi-square test is ensured, i.e. no more than 20 % of the
expected frequencies are below 5 and none are below 1 (Backhaus 2011). For this
analysis, only customers with a clear preference for or against advice and the
exchange of experiences are examined.

The results of the chi-square test for independence indicate a significant relat-
edness between customers’ appreciation for other customers’ advice and their
channel preference (χ2 = (4, n = 825) = 13.23, p = 0.01, Cramer’s V = 0.13). An
overview is given in Fig. 4.6 (to simplify the Figure, the number of cases is divided
by 5 to illustrate the real number of participants). The results show, that participants
who prefer other customers’ advice usually choose phone or e-mail as a complaint
channel when compared to customers who do not appreciate advice from other
customers.
The same procedure is applied to test the influence of the appreciation of the
exchange of experience with other customers. The result of the chi-square test
indicates no significant interdependence between the appreciation of experience
exchange and channel preference (χ2 = (4, n = 870) = 7.09, p = 0.13, Cramer’s
V = 0.09). An overview is given in Fig. 4.7.
54 4 Analysis of Customers’ Complaint Channel Choice …

70

60

50

40

30

20

10

0
Mail E-mail Phone Social network Face-to-face
Advice Yes Advice No

Fig. 4.6 Crosstab appreciation for advice and complaint channel choice—study 1

70

60

50

40

30

20

10

0
Mail E-mail Phone Social network Face-to-face
Experience Yes Experience No

Fig. 4.7 Crosstab appreciation for experience and complaint channel choice—study 1
4.5 Study 1 55

There was also a test as to whether the need for social interaction has an effect on
the probability, in general, of a customer making a complaint. Two independent
t-tests are conducted. COMP.Pref, i.e. the likelihood to complain in general in case
of dissatisfaction, represents the dependent variable (1 = “In no case”; 7 = “Certainly
yes”), whereas the independent variables are given by (1) appreciation for advice
(“yes”/“no”), and (2) appreciation for experience exchange (“yes”/“no”).
The result of the independent t-test for the appreciation of advice shows no
significant differences (Myes = 4.84, SDyes = 1.30; Mno = 4.89, SDno = 1.47;
t = 0.390; p = 0.717), whereas the appreciation for experience exchange shows
significant differences (Myes = 4.97, SDyes = 1.27; Mno = 4.72, SDno = 1.53;
t = 2.15; p = 0.033). Thus, customers with a high need of social interaction tend to
complain more.
Summary
The need for social interaction has an influence on customers’ complaint channel prefer-
ences and their probability to complain in general.

4.6 Study 2

Study 1 is based on a quite homogeneous sample, as all participants work in the

same company, live in the same area, and presumably belong to the same social
class. This can lead to biased results. To minimise sample-based biases and to
ensure a higher generalisability of the results, a second study is conducted. To
ensure comparability the second study is based on the same questionnaire as the one
used in study 1.
Section 4.6.1 describes the sample. Then, the descriptive statistics are outlined
(Sect. 4.6.2). In Sect. 4.6.3, the manipulation and realism check for the study is
conducted and the construct measurement examined (Sect. 4.6.4). Section 4.6.5
shows the results of the study.

4.6.1 Sample

In addition to study 1, convenience sampling is employed in study 2. Potential

participants received an online link via e-mail, which directed them to the ques-
tionnaire. Participation was voluntary and anonymity ensured. Participants were
also asked to forward the link to acquaintances, a method also known as snowball
sampling (Schnell et al. 2011). In hopes of increasing the response rate, the e-mail
includes a personalised message. According to recent research, these sampling
techniques are appropriate, since it is the relationship among variables, rather than
the precise extent of the effect, on which the focus of the investigation lies
(Blumberg et al. 2008).
56 4 Analysis of Customers’ Complaint Channel Choice …

4.6.2 Descriptive Statistics

First, the data screening and cleaning procedure is described. After this the par-
ticipants’ descriptive statistics are introduced.
Data screening and cleaning
Again, all questionnaires are checked for completeness and outliers. From the 189
received questionnaires, 34 are deleted from the sample because of incompleteness.
Consequently, 155 questionnaires remain in the sample for study 2.
Participants
The average participant in this study is between 20 and 29. About 41.9 % of the
participants are female. An overview of age and gender distribution is given in
Table 4.7. Moreover, similar to study 1, Table 4.7 shows that with increasing age
social network membership decreases (van Eimeren and Frees 2011).

4.6.3 Manipulation and Realism Checks

Similar to study 1, the success of the manipulation is ensured. Again the partici-
pants’ dissatisfaction with the fictitious company after the described service failure
is calculated. Participants are highly dissatisfied with the company (M = 5.76,
SD = 1.03; measure with 1 = “Not annoyed at all”, 7 = “Very annoyed”).
Specifically, 90.3 % show a high dissatisfaction with the company. Hence, the
manipulation of the service failure therefore is regarded as successful.
A high degree of familiarity with the research topic is ensured by asking par-
ticipants about their (1) complaint behaviour, (2) social network membership, and
(3) social network activities. First, 63.3 % of the participants complain to a com-
pany when they are dissatisfied with a service or a product (M = 4.910, SD = 1.45).
Similarly to study 1, participants in past studies showed a comparatively lesser
willingness to complain (Voorhees and Brady 2005). Second, 91 % of the partic-
ipants indicated being members in at least one social network, of these more than
80 % log in on a daily basis (several times a week = 12.8 %; daily = 80.1 %). It can
therefore be assumed that the participants are familiar with the described situation

Table 4.7 Descriptive statistics—study 2

Description Age cluster of respondents
<20 20–29 30–39 40–49 50–59
Sample size –
Responses 189
Completed questionnaires 155 13 112 19 8 3
Female (%) 41.9 84.6 52.7 63.2 75.0 66.7
SN member 141 12 109 13 6 1
4.6 Study 2 57

and the research topic. However, similar to study 1, the number of participants, who
have used social networks as a complaint channel in the past is quite low: 5.7 %.
Participants were also asked about their complaint channel preferences
(CASE_CHAN) with regard to the described situation. Similar to study 1, classic
complaint channels are preferred (e-mail: 21.9 %; phone: 54.2 %). 18.7 % of the
participants would favour communicating their complaint via a face-to-face con-
versation and 3.9 % would send a letter. Only 1.3 % would complain via a social
network.
In order to analyse social networks as a complaint channel in more detail, the
sample is restricted to social networks members as a subgroup. 75.9 % of the
participants active in a social network appreciate advice from other customers in an
informal forum (SI_1), 73 % like to hear about and relate to similar experience
(SI_2). 29.1 % of the social network members would appreciate social networks as
a complaint channel (KANO_P), and 10.6 % (KANO_N) are annoyed if this
complaint channel is not offered.
The incorporation of the complaint channels is reconsidered and rated as being
sufficient. Only 4 participants requested further channels (Facsimile = 1;
Twitter = 2, and homepage contact form = 1). Twitter, as an online social network,
necessitates no separate mention, while homepage contact forms, likewise, are not
dissimilar to e-mail messages and shall therefore not be treated as a different cat-
egory here. No important communication channel is thus missing—and therefore
no distortion of results.

4.6.4 Construct Measurement

Cronbach alpha is calculated for the two dependent variables procedural and
interactional justice (Cronbach 1951). Similar to study 1, for procedural justice,
item 1 (PJ_1; time-consumption) is designed in a reversed manner (1 = most
positive value and 7 = most negative value), and a new item (PJ_1r) is calculated by
revising the initial values. The results show a Cronbach alpha of 0.554, when all
three items are taken into account (Table 4.8). As the elimination of one item would

Table 4.8 Reliability and construct measurements—study 2

Procedural M SD Rotated KMO Bartlett’s Corrected Cronbach Cronbach
justice factor test test of item total alpha if item alpha
loadings sphericity correlation deleted
PJ_1r 3.76 1.82 0.856 0.667 0.000 0.299 0.558 0.554
PJ_2 4.60 1.93 0.719 0.347 0.486
PJ_3 4.92 1.67 0.728 0.465 0.308
Interactional justice
IJ_1 4.53 1.48 0.859 0.685 – 0.813
IJ_2 4.58 1.54 0.909 0.685 –
58 4 Analysis of Customers’ Complaint Channel Choice …

not increase the reliability of the construct, all three items are considered to rep-
resent the factor procedural justice (Table 4.8). Column “Corrected item total
correlation” shows that the values of item 2 and 3 exceed the required threshold of
0.3. However, item 1 shows only a value of 0.299. The literature suggests that the
value is still appropriate, as the inter-item correlation shows values in the range of
0.2–0.4 (Briggs and Cheek 1986). The three items therefore measure the same
underlying construct: namely procedural justice; although the results are, once
again, not satisfactory (Singh 1990). However, due to the exploratory interest of
this study, procedural justice is calculated using these three items (Schmitt 1996).
The results for interactional justice are also documented in Table 4.8. All values are
above the common threshold.
Similar to study 1, an exploratory factor analysis is conducted. The data is suitable
for a factor analysis, which is shown by the results of the Kaiser-Meyer-Olkin
measure (KMO-test) and the Bartlett’s test of sphericity (Table 4.8). Again,
according to the Kaiser’s criterion, two factors can be extracted (factor 1 = 2.39;
factor 2 = 1.24). These two factors explain 72.74 % of the variance. Factor analysis
reveals that procedural justice contains out of the three items PJ_1r (i.e. time-con-
sumption), PJ_2 (i.e. speed of response) and PJ_3 (i.e. user-friendliness), and
interactional justice out of two items IJ_1 (i.e. honesty) and IJ_2 (i.e. helpfulness). All
factor loadings exceed the common thresholds (Table 4.8). These results are con-
sistent with study 1. For further data analysis, the factors’ mean values are calculated.
As already mentioned in study 1, distributive justice (DJ) and personal attitudes
(PA) are measured on a single-item scale and social interaction (SI) is measured by
categorically distinct variables. Therefore, no further calculation is necessary at this
point. Further, in line with study 1, the item for personal attitude is reversed and a
new variable (PAr) is calculated.

4.6.5 Results

To analyse the data with regard to the management questions, a one-way ANOVA is
conducted with the four factors (interactional, procedural, distributive justice and
personal attitudes) as dependent variables and the complaint channels as an inde-
pendent variable. This approach is consistent with the one chosen in study 1. Finally,
to analyse social interaction, a chi-square test and an independent t-test are conducted.
Again, a one-way ANOVA is conducted. First, the Levene’s test is conducted
(PJ: F(4, 705) = 0.820, p = 0.513; IJ: F(4, 705) = 7.156, p = 0.000; DJ: F(4, 705)
= 16.721, p = 0.000; PAr: F(4, 705) = 7.943, p = 0.000), which tests for variance
homogeneity between the groups. Again, subsequent efforts to reach homogeneity
by means of data transformation do not lead to improvements. Thus, further
analysis for IJ, DJ and PAr refers to the Brown-Forsythe robust test. The results
of the Brown-Forsythe and Welch tests show that there are significant differences
(all p < 0.001) with regard to each variable PJ, IJ, DJ, and PAr (Table 4.9). It must
be noted that ANOVA shows the same significant differences.
4.6 Study 2 59

Further, as variance equality is given for PJ, ANOVA results are shown in
Table 4.10. The results show significant differences between the complaint channels
with regard to the different dependent variables. Post hoc tests are performed to
analyse the complaint channel differences for the four dependent variables in more
detail. Again, as most of the dependent variables show heterogeneity of variance,
the Games-Howell test is conducted instead of the intended Tukey-Test. The results
of the Games-Howell test are summarised in Table 4.11.
Similarly to study 1, the results of the post hoc tests for procedural justice,
interactional justice, distributive justice and personal attitude are introduced.
Further, the results of the analysis of social interaction are revealed.
Procedural Justice
Procedural justice represents time-consumption, i.e. the time a company needs to
respond to a complaint and also the channel’s user-friendliness. An overview of the
mean values is given in Fig. 4.8.

The results from the post hoc tests show that the complaint channel mail sig-
nificantly differs from all other complaint channel options (MMail = 2.87,
SDMail = 1.01; ME-mail = 4.91, SDE-mail = 0.97; p = 0.000; MPhone = 5.27,
SDPhone = 1.03; p = 0.000; MSocialnetwork = 4.37, SDSocialnetwork = 1.11; p = 0.000;
MFacetoface = 4.70, SDFacetoface = 0.98; p = 0.000). Hence, mail is perceived as being
the most time-consuming and least user-friendly complaint channel. Apart from this
the time needed for companies to react to such a complaint is also perceived as
being the longest compared to the other complaint channels. Moreover, e-mail is
perceived as being less convenient than phone (ME-mail = 4.91, SDE-mail = 0.97;

Table 4.9 Robust tests of equality of means—study 2

Variable Statisticsa df1 df2 Sig.
Procedural justice (PJ) Welch 119.885 4 349.822 0.000
Brown-Forsythe 117.365 4 693.282 0.000
Interactional justice (IJ) Welch 69.020 4 348.565 0.000
Brown-Forsythe 65.580 4 649.409 0.000
Distributive justice (DJ) Welch 89.443 4 347.997 0.000
Brown-Forsythe 96.696 4 629.233 0.000
Personal attitudes (PAr) Welch 176.403 4 348.223 0.000
Brown-Forsythe 186.131 4 652.856 0.000
a
Asymptotically F distributed

Table 4.10 ANOVA results for procedural justice—study 2

Procedural justice (PJ) Square sum df1 F Sig.
Between groups 489.052 4 117.365 0.000
Within groups 729.214 700
Sum 1218.266 704
60 4 Analysis of Customers’ Complaint Channel Choice …

Table 4.11 Results post hoc tests—study 2

Complaint channel Mean SD
Procedural justice
Mail 2.87 1.01
E-mail 4.91a 0.97
Phone 5.29 1.03
Social network 4.37c 1.11
Face-to-face 4.70b 0.98
Interactional justice
Mail 4.42d 1.44
e
E-mail 4.48 1.16
Phone 4.97 1.00
Social network 3.34 1.26
Face-to-face 5.54 1.12
Distributive justice
Mail 4.79f 1.48
E-mail 4.93g 1.09
Phone 5.53 1.08
Social network 3.27 1.47
Face-to-face 5.99 1.01
Personal attitude
Mail 1.56 1.00
E-mail 2.75h 1.12
Phone 3.06i 1.23
Social network 5.46 1.44
Face-to-face 2.66j 1.42
Notes
If not stated otherwise, complaint channels differ significantly from each other at p < 0.05
a,c
Does not differ significantly from b
d
Does not differ significantly from e
f
Does not differ significantly from g
h,i,j
Do not differ from each other significantly

MPhone = 5.27, SDPhone = 1.03; p = 0.014), but more convenient than social net-
works (ME-mail = 4.91, SDE-mail = 0.97; MSocialnetwork = 4.37, SDSocialnetwork = 1.11;
p = 0.000). However, there is no difference between the complaint channels e-mail
and a face-to-face conversation (ME-mail = 4.91, SDE-mail = 0.97; MFacetoface = 4.70,
SDFacetoface = 0.98; p = 0.395). With regard to procedural justice the complaint
channel phone is also preferred, when compared to social networks (MPhone = 5.27,
SDPhone = 1.03; MSocialnetwork = 4.37, SDSocialnetwork = 1.11; p = 0.000) and face-to-
face conversation (MPhone = 5.27, SDPhone = 1.03; MFacetoface = 4.70,
SDFacetoface = 0.98; p = 0.000). The results show no significant difference between
social networks complaint and face-to-face complaint (MSocialnetwork = 4.37,
SDSocialnetwork = 1.11; MFacetoface = 4.70, SDFacetoface = 0.98; p = 0.070).
4.6 Study 2 61

Procedural justice
7

1
Mail E-mail Phone Social network Face-to-face

Fig. 4.8 Overview mean values for PJ—study 2

Summary
With regard to procedural justice (i.e. amount of time-consumption; response time of the
company and a user-friendly design of the complaint channel) a phone call is perceived as
being the most convenient. Next in sequence are e-mail, face-to-face complaints and social
networks, whereby the convenience of face-to-face conversations and social network
complaints is perceived as being equal; a letter sent by mail is seen as the most inconvenient
complaint channel.

Interactional Justice
Interactional justice measures the helpfulness and honesty of the complaint channel.
The mean values are displayed in Fig. 4.9.

As Fig. 4.9 shows, the complaint channels mail and e-mail are not significantly
different from each other (MMail = 4.43, SDMail = 1.44; ME-mail = 4.48, SDE-
mail = 1.16; p = 0.997). Hence, the helpfulness and honesty of these two complaint
channel is evaluated as equal by customers. However, complaining via mail is
perceived as less helpful than a complaint via phone (MMail = 4.43, SDMail = 1.44;
MPhone = 4.97, SDPhone = 1.00; p = 0.003) or through a face-to-face conversation
(MMail = 4.43, SDMail = 1.44; MFacetoface = 5.54, SDFacetoface = 1.02; p = 0.000),
though more helpful than complaints communicated using social networks
(MMail = 4.43, SDMail = 1.44; MSocialnetwork = 3.34, SDSocialnetwork = 1.26;
p = 0.000). Calling a company (ME-mail = 4.48, SDE-mail = 1.16; MPhone = 4.97,
SDPhone = 1.00; p = 0.002) or visiting a store (ME-mail = 4.48, SDE-mail = 1.16;
MFacetoface = 5.54, SDFacetoface = 1.02; p = 0.000) to communicate a complaint is
62 4 Analysis of Customers’ Complaint Channel Choice …

Interactional justice
7

1
Mail E-mail Phone Social network Face-to-face

Fig. 4.9 Overview mean values for IJ—study 2

perceived as more helpful and honest than sending an e-mail, although complaining
via e-mail leads to a greater degree of interactional justice than a complaint via
social networks (ME-mail = 4.48, SDE-mail = 1.16; MSocialnetwork = 3.34,
SDSocialnetwork = 1.26; p = 0.000). The same is true for calling a company
(MPhone = 4.97, SDPhone = 1.00; MSocialnetwork = 3.34, SDSocialnetwork = 1.26;
p = 0.000) or visiting it (MPhone = 4.97, SDPhone = 1.00; MFacetoface = 5.54,
SDFacetoface = 1.02; p = 0.000) instead of contacting it by way of social networks.
With regard to interactional justice, a face-to-face conversation to communicate a
complaint is preferred to social networks (MFacetoface = 5.54, SDFacetoface = 1.02;
MSocialnetwork = 3.34, SDSocialnetwork = 1.26; p = 0.000).
Summary
Face-to-face complaints are chosen when honesty and helpfulness are asked for. Next in
sequence are phone, mail and e-mail—in that order, with social networks coming in last.

Distributive justice
Whether or not a complaint leads to the customers’ desired outcome is measured
with distributive justice. An overview of the mean values is given in Fig. 4.10.

Again, an evaluation of the complaint channels mail and e-mail show these to be
equal with regard to distributive justice. Hence, no significant results are found
(MMail = 4.79, SDMail = 1.48; ME-mail = 4.93, SDE-mail = 1.09; p = 0.908).
Compared to making a complaint via mail, distributive justice is higher for the
complaint channels phone (MMail = 4.79, SDMail = 1.48; MPhone = 5.53,
4.6 Study 2 63

Distributive justice
7

1
Mail E-mail Phone Social network Face-to-face

Fig. 4.10 Overview mean values for DJ—study 2

SDPhone = 1.08; p = 0.000) and face-to-face conversation (MMail = 4.79,

SDMail = 1.48; MFacetoface = 5.99, SDFacetoface = 1.01; p = 0.000), but lower for
social networks (MMail = 4.79, SDMail = 1.48; MSocialnetwork = 3.27,
SDSocialnetwork = 1.45; p = 0.000). Complaining via e-mail is perceived as less useful
than communicating a complaint with a phone call (ME-mail = 4.93, SDE-mail = 1.09;
MPhone = 5.53, SDPhone = 1.08; p = 0.000) or face-to-face (ME-mail = 4.93, SDE-
mail = 1.09; MFacetoface = 5.99, SDFacetoface = 1.01; p = 0.000). Compared to social
networks e-mail is perceived as being more practical (ME-mail = 4.93, SDE-
mail = 1.09; MSocialnetwork = 3.27, SDSocialnetwork = 1.45; p = 0.000). Customers
evaluate the distributive justice of the complaint channel phone higher than that of
social networks (MPhone = 5.53, SDPhone = 1.08; MSocialnetwork = 3.27,
SDSocialnetwork = 1.45; p = 0.000), yet lower than when making a face-to-face
complaint (MPhone = 5.53, SDPhone = 1.08; MFacetoface = 5.99, SDFacetoface = 1.01;
p = 0.003).The distributive justice for social networks is also evaluated as lower
than a face-to-face complaint (MSocialnetwork = 3.27, SDSocialnetwork = 1.45;
MFacetoface = 5.99, SDFacetoface = 1.01; p = 0.000).
Summary
When customers desire a certain outcome, face-to-face complaints are the mode of choice.
Next comes calling the company, after that writing a letter and sending an email. Social
networks would be chosen as a last option, as customers assess this channel as being less
expedient compared to the others.
64 4 Analysis of Customers’ Complaint Channel Choice …

Personal attitude
The factor personal attitudes measure customers’ evaluation of the complaint
channels with regard to their formality and as how formal they are perceived to be.
Figure 4.11 shows the mean values. As personal attitude is measured in a reversed
manner, a higher mean value shows a less formal and official complaint channel.

Complaining in a letter sent by normal mail is perceived as the most formal way
to communicate a complaint compared to the other four examined complaint
channels (MMail = 1.56, SDMail = 0.995; ME-mail = 2.75, SDE-mail = 1.12; p = 0.000;
MPhone = 3.06, SDPhone = 1.23; p = 0.000; MSocialnetwork = 5.46,
SDSocialnetwork = 1.44; p = 0.000; MFacetoface = 2.66, SDFacetoface = 1.42; p = 0.000).
Comparing e-mail and phone, e-mail and face-to-face conversation, as well as
phone and face-to-face conversation, these channels are perceived as equally formal
and official—no significant differences are found (ME-mail = 2.75, SDE-mail = 1.12;
MPhone = 3.06, SDPhone = 1.23; p = 0.190; MFacetoface = 2.66, SDFacetoface = 1.42;
p = 0.974; MPhone = 3.06, SDPhone = 1.23; MFacetoface = 2.66, SDFacetoface = 1.42;
p = 0.091). Moreover, e-mail is seen as more formal and official than complaining
via social networks (ME-mail = 2.75, SDE-mail = 1.12; MSocialnetwork = 5.46,
SDSocialnetwork = 1.44; p = 0.000). The same goes for phone and social networks
(MPhone = 3.06, SDPhone = 1.23; MSocialnetwork = 5.46, SDSocialnetwork = 1.44;
p = 0.000). Social networks are also regarded as less formal and official than a face-
to-face conversation (MSocialnetwork = 5.46, SDSocialnetwork = 1.44; MFacetoface = 2.66,
SDFacetoface = 1.42; p = 0.000).

Personal attitude
7

1
Mail E-mail Phone Social network Face-to-face

Fig. 4.11 Overview mean values for PA—study 2

4.6 Study 2 65

Summary
Mail is perceived as the most formal way to complain, e-mail, phone and face-to-face
complaint, in the order listed, are next. Social networks are the most informal and unofficial
way to communicate a complaint.

Social Interaction
Similar to study 1, chi-square tests are conducted to analyse whether social inter-
action influences the customers’ complaint channel preferences. The validity of the
chi-square test is ensured, i.e. no more than 20 % of the expected frequencies are
below 5 and none is below 1 (Backhaus 2011). Again, only customers with a clear
preference for or against getting advice and the exchange of experiences are
examined.

In line with the results from study 1, the chi-square test for independence
indicates a significant link between customers’ appreciation for other customers’
advice and their channels preferences (χ2 = (4, n = 595) = 30.783, p = 0.000,
Cramer’s V = 0.21). An overview is given in Fig. 4.12. Again, customers with a
high appreciation for other customers advice usually complain via phone, e-mail or
face-to-face.
The result of the chi-square test indicates a significant degree of relatedness
between the appreciation of experience, exchange with other customers, and
channel preference (χ2 = (4, n = 600) = 15.03, p = 0.005, Cramer’s V = 0.13). An
overview is given in Fig. 4.13. Although customers appreciate the exchange of

60

50

40

30

20

10

0
Mail E-mail Phone Social network Face-to-face
Advice Yes Advice No

Fig. 4.12 Crosstab appreciation for advice and complaint channel choice—study 2
66 4 Analysis of Customers’ Complaint Channel Choice …

60

50

40

30

20

10

0
Mail E-mail Phone Social network Face-to-face
Experience Yes Experience No

Fig. 4.13 Crosstab appreciation for experience and complaint channel choice—study 2

experience with other customers, participants choose conventional complaint

channels such as e-mail, phone and face-to-face.
To test whether the need for social interaction affects customers’ complaint
probability in general, two independent t-tests are conducted. Again, Comp.Pref
represents the dependent variable, whereas the independent variables are given by
(1) appreciation for advice (“yes”/“no”), and (2) appreciation for experience
exchange (“yes”/“no”).
The results of the independent t-test for the appreciation of advice show sig-
nificant differences (Myes = 3.98, SDyes = 145; Mno = 4.33, SDno = 1.12; t = 2.24;
p = 0.03). Thus, customers with a less need for social interaction, i.e. advice from
other customers, tend to complain more. However, the appreciation of experience
(SI_2) exchange with other customers shows no significant differences
(Myes = 3.98, SDyes = 1.42; Mno = 4.06, SDno = 1.48; t = 0.469; p = 0.639).
Summary
The need for social interaction has an influence on customers’ complaint channel prefer-
ences. Customers appreciating other customers’ advice tend to complain in a less direct
way.
4.7 Channel Comparison 67

4.7 Channel Comparison

Both studies reveal that classic complaint channels, such as e-mail, phone and face-
to-face, are perceived as more appropriate compared to the potentially new and
emerging complaint channel option provided by social networks. In particular,
customers who desire a fast and user-friendly complaint handling process (proce-
dural justice) would choose phone or e-mail to communicate the complaint.
Another option is complaining face-to-face. As this implies going to the store in
question, this involves a far greater investment in time, than simply calling the
company. Social networks are rather inadequate as a complaint channel, as they are
seen as a time-consuming and user-unfriendly way (study 1). Surprisingly, study 2
shows participants to evaluate face-to-face complaints as being equal as a mode of
complaint to social networks owing to the response time. In particular, as face-to-
face complaints involve a visit at the store, posting a complaint in a social network
has a slower response rate. However, complaints via mail are judged as being the
most time-consuming option available as correspondence via mail is very slow.
When regarding honesty and helpfulness (interactional justice) as the focus of a
company’s complaint handling, the face-to-face complaint channel is perceived as
the most adequate one. Here, customers are able to speak directly with an
employee. Customers can see the employee’s facial cues and interpret gestures
thereby evaluating the honesty of the employee regarding the complaint handling.
Phone, mail and e-mail are also perceived as adequate ways to communicate a
complaint. As these complaint channels are more integrated and known by cus-
tomers, they are felt to be more trustworthy. Once again social networks come in
last.
The results for distributive justice are similar to interactional justice. When
customers want a company to do something, a face-to-face complaint is rated as the
most appropriate complaint channel. Here, customers have a good bargaining
position and can react to employees’ arguments and explanations. Then, the other
complaint channels (phone, mail, and e-mail) come next in sequence. When calling
a company customers can also react quickly and directly to the arguments brought
forth by the employee, an immediacy which is limited when resorting to mail and e-
mail complaint channels. Customers do not rely on social networks as a complaint
channel, when they have fixed expectations for reaching a desired outcome, e.g.
after a severe service failure.
Customers perceive mail as the most formal and official complaint channel. The
other conventional complaint channels, i.e. e-mail, phone, and face-to-face are also
perceived as formal, but less so than mail. Social networks represent the most
informal of all complaint channels.
68 4 Analysis of Customers’ Complaint Channel Choice …

4.8 Identification of Customer Subgroups

As the research topic is quite exploratory a cluster analysis is conducted to gain a

detailed knowledge about the adequateness of the different complaint channels,
especially social networks (Sect. 4.8.1). Age clusters are then analysed with regard
to the customers’ complaint likelihood (Sect. 4.8.2). Finally, to gain a more detailed
look into customers’ choice of complaint channels, a decision tree is compiled
(Sect. 4.8.3).

4.8.1 Homogenous Subsets of Complainers

Figure 4.14 shows the result of the cluster analysis. The variables age (Age), gender
(Gender), general complaint behaviour (COMP.Pref), preferred complaint channel
[with regard to the described situation (Preferred Channel)], log in frequency (LOG
IN SN) in social networks, the appreciation of advice (IS_1) and exchange for
experience (IS_2) are used to cluster participants. The four main clusters of

Fig. 4.14 CCB clusters

4.8 Identification of Customer Subgroups 69

respondents are calculated by means of nonlinear canonical correlation (Field

2009). Both samples are matched for this analysis. Each cluster is represented by
one quadrant and points to common characteristic of each subset of respondents.
The following clusters are identified by the cluster analysis:
“Young non-complainers”: As shown in the first quadrant (upper right), younger
respondents (typically below 20) are active users of social networks on a daily basis
and are thankful for advice from and an exchange of experience with their peer-
group. These users are characterized by a low probability of complaining in case of
dissatisfaction (the values for the variable COMP.Pref in this quadrant is relatively
low, which indicates a low complaint probability). However, if they do decide to
complain, they appreciate the possibility of doing this within social networks. This
cluster shows similarities to Singh’s (1990) cluster of “passives”, as both show a
low tendency to take actions.
“Middle-aged man”: The second quadrant (upper left) represents mostly male
respondents between 30 and 49 years of age. These respondents generally prefer to
complain by mail or e-mail. This result is in line with the results from Grougiou and
Pettigrew (2009). No conclusion about their perception of social networks or their
need for receiving advice from others can be drawn.
“Grumpy oldies”: The third quadrant (lower left) shows participants indicating a
very high probability of complaining (high values for variable COMP.Pref from 5
to 7). These respondents are of 50–59 years of age and do not rely on other people’s
experience or advices. With regard to social networks usage, these customers can be
characterized as either “mostly inactive” or logging in on a “weekly” basis. This
cluster shows parallels to Singh’s (1990) “voicers”.
“Neglectful female”: The fourth quadrant (lower right) depicts the subgroup with
the lowest probability for complaining. These are typically young female partici-
pants who are generally disinterested regarding advice others or the exchange of
experiences. They preferably complain by phone or face-to-face. This cluster is
assumed to be the most likely “non-voicing” subgroup, and is similar to the
“passives” from Singh’s (1990) cluster analysis.

4.8.2 Analysis of Age Clusters

The cluster analysis hints at age having an influence on customers’ complaint

likelihood. In particular, there are evidences that young customers complain less
(young non-complainers), whereas older customers show a high likelihood to
complain (grumpy oldies). To shed light on this relationship, a one-way ANOVA is
conducted. Age clusters form the independent variable and, as a dependent variable,
the complaint likelihood (COMP.Pref) is chosen. Complaint likelihood is measured
on a 7 point Likert scale with 1 = “In no case” and 7 = “Certainly yes”.
70 4 Analysis of Customers’ Complaint Channel Choice …

First, Levene’s test is conducted (F(5, 516) = 5.46, p = 0.000), which analyses
the variance in homogeneity between the groups. Subsequent efforts to reach
homogeneity by means of data transformation again did not lead to any improve-
ments. Further analysis for COMP.Pref refers to the Brown-Forsythe robust test.
The results of the Brown-Forsythe and Welch tests show that there are significant
differences (all p < 0.001) with regard to the dependent variable in Table 4.12;
ANOVA shows the same significant differences.
The results show that there are significant differences between the age clusters
and the complaint likelihood. Therefore a post hoc test is conducted. Again, as the
dependent variable shows heterogeneity of variance, the Games-Howell test is
applicable instead of the intended Tukey-Test. The results of the Games-Howell test
are summarised in Table 4.13.
An overview of the mean values of the complaint likelihood for the different age
clusters is given in Fig. 4.15. The results of the post hoc test show a significant
difference for customers younger than 20, customers between 30 and 39
(M<20 = 3.71, SD<20 = 1.14; M30–39 = 4.79, SD30–39 = 1.22; p = 0.041), between 40
and 49 (M<20 = 3.71, SD<20 = 1.14; M40–49 = 4.99, SD40–49 = 1.20; p = 0.011), as
well as for the age cluster of the 50–59 years old customers (M<20 = 3.71,
SD<20 = 1.14; M50–59 = 5.30, SD50–59 = 1.15; p = 0.002). Further, significant
differences are found between the 20 to 29 year olds, the 40 to 49 year olds (M20–
29 = 4.36, SD20–29 = 1.54; M40–49 = 4.99, SD40–49 = 1.20; p = 0.001) and the 50 to
59 year olds (M20–29 = 4.36, SD20–29 = 1.54; M50–59 = 5.30, SD50–59 = 1.15;
p = 0.000). To sum up, people within the younger age clusters exhibit a lower
likelihood to complain than older customers.

Table 4.12 Robust tests of equality of customers’ complaint likelihood

Variable Statisticsa df1 df2 Sig.
Complaint likelihood Welch 8.533 5 40.759 0.000
Brown-Forsythe 10.318 5 151.849 0.000
a
Asymptotically F distributed

Table 4.13 Results post hoc tests complaint likelihood

Age clusters Mean SD
Complaint likelihood
<20 3.71a 1.14
20–29 4.36b 1.54
30–39 4.79c 1.22
40–49 4.99d 1.20
50–59 5.30e 1.15
60–69 5.17 0.98
Notes
Significant at p < 0.05
a
Differs significantly from c, d, e
b
Differs significantly from d, e
4.8 Identification of Customer Subgroups 71

Complaint likelihood
7

1
< 20 20 – 29 30 – 39 40 – 49 50 – 59 60 – 69
Age clusters

Fig. 4.15 Overview mean values for complaint likelihood

4.8.3 Relevance of Social Interactions for Network Members

To further describe the importance of the determinants of satisfaction in the context

of social interaction, a decision tree is compiled (Fig. 4.16). By means of this
statistical procedure, hierarchical classifications and critical paths for decisions can
be analysed. The first component of social interaction, advice (SI_1n), is chosen as
the dependent variable since previous investigations within this study have shown
this to be slightly more important than experience (SI_2n).
The most robust predictor of SI_1n is log in frequency (NET_FREQ).2 More
than 75 % of the participants who log in daily assess advice from others to be
important in the course of complaining compared to 63 % of the mostly inactive
users. If we continue to the next most important classification, the second predictor
can be considered to be complaining preference (COMP.Pref). Those customers
who frequently use social networks and usually do not complain are even more
receptive to advice than customers who frequently use social networks and show a
high likelihood of complaining (78.5 and 69.2 %). For customers who show an
average probability of complaining, the need for advice is stronger in females than
in males (81.1 to 65.9 %). These findings again strengthen the proposition that
social networks might be adequate for a subset of customers.

2
A detailed description of the values for each variable is provided in Appendix C.
72 4 Analysis of Customers’ Complaint Channel Choice …

Fig. 4.16 CCB decision tree

Take away
Even the traditional complaint channels, i.e. mail, e-mail, phone, and face-to-
face, differ widely in customers’ perceptions. However, these channels are
mostly judged as being better suited than social networks for making a
complaint. Especially in light of interactional and distributive justice, tradi-
tional complaint channels are more accepted by customers for voicing a
complaint. However, for younger customers, social networks can be imple-
mented as additional complaint channel to motivate the usually silent young,
as a target group, to come forward with a complaint.
References 73

References

Achrol RS (1991) Evolution of the marketing organization: new forms for turbulent environments.
J Mark 55(4):77. doi:10.2307/1251958
Albaum G, Wiley J, Roster C, Smith SM (2011) Visiting item non-responses in internet survey
data collection. Int J Market Res 53(5):687. doi:10.2501/IJMR-53-5-687-703
Babbie ER (2007) The practice of social research, 11th edn. Thomson Wadsworth, Belmont
Backhaus K (2011) Multivariate Analysemethoden. Eine anwendungsorientierte Einführung, 13.,
überarb. Aufl. Springer-Lehrbuch. Springer, Berlin [u.a.]
Baker MJ (2003) Data collection—questionnaire design. Mark Rev 3(3):343–370. doi:10.1362/
146934703322383507
Blumberg B, Cooper DR, Schindler PS (2008) Business research methods, 2nd edn. McGraw-Hill
Higher Education, London
Brady MK, Cronin JJ (2001) Some new thoughts on conceptualizing perceived service quality:
a hierarchical approach. J Mark 65(3):34–49. doi:10.1509/jmkg.65.3.34.18334
Briggs SR, Cheek JM (1986) The role of factor analysis in the development and evaluation of
personality scales. J Pers 54:106–148
Choi S, Mattila A (2008) Perceived controllability and service expectations: influences on
customer reactions following service failure. J Bus Res 61(1):24–30. doi:10.1016/j.jbusres.
2006.05.006
Churchill GA, Suprenant C (1982) An investigation into the determinants of customer satisfaction.
J Mark Res 19(4):491–504
Cronbach LJ (1951) Coefficient alpha and the internal structure of tests. Psychometrika 16(3):297–
334
de Matos CA, Rossi CAV, Veiga RT, Vieira VA (2009) Consumer reaction to service failure and
recovery: the moderating role of attitude toward complaining. J Serv Mark 23(7):462–475.
doi:10.1108/08876040910995257
Diamantopoulos A, Schlegelmilch BB (1996) Determinants of industrial mail survey response:
a survey-on-surveys analysis of researchers’ and managers’ views. J Mark Manag 12(6):505–531.
doi:10.1080/0267257X.1996.9964432
Field A (2009) Discovering statistics using SPSS, 3rd edn. SAGE, London
Galesic M, Bosnjak M (2009) Effects of questionnaire length on participation and indicators of
response quality in a web survey. Pub Opin Q 73(2):349–360. doi:10.1093/poq/nfp031
Grougiou V, Pettigrew S (2009) Seniors’ attitudes to voicing complaints: a qualitative study.
J Mark Man 25(9):987–1001. doi:10.1362/026725709X479336
Henneberg SC, Gruber T, Reppel A, Ashnai B, Naudé P (2009) Complaint management
expectations: an online laddering analysis of small versus large firms. Ind Mark Manage 38
(6):584–598. doi:10.1016/j.indmarman.2009.05.008
Hunt SD, Sparkman RD Jr, Wilcox JB (1982) The pretest in survey research: issues and
preliminary findings. J Mark Res 19(2):269–273
Kromrey H (2006) Empirische Sozialforschung. Modelle und Methoden der standardisierten
Datenerhebung und Datenauswertung, 11., überarb. Aufl. UTB, vol 1040. Lucius & Lucius,
Stuttgart
Lietz P (2010) Research into questionnaire design: a summary of the literature. Int J Adv 52
(2):249. doi:10.2501/S147078530920120X
Linsky AS (1975) Stimulating responses to mailed questionnaires: a review. Pub Opin Q 39(1):82.
doi:10.1086/268201
Luria G, Gal I, Yagil D (2009) Employees’ willingness to report service complaints. J Serv Res 12
(2):156–174. doi:10.1177/1094670509344214
Macdonald JE, Beck-Dudley CL (1994) Are deontology and teleology mutually exclusive? J Bus
Ethics 13(8):615–623. doi:10.1007/BF00871809
Malafi TN (1991) The impact of social influence on consumer complaint behavior. J Consum
Satisfaction, Dissatisfaction Complaining Behav 4:144–150
74 4 Analysis of Customers’ Complaint Channel Choice …

Malafi TN, Cini MA, Taub SL, Bertolami J (1993) Social influence and the decision to complain.
Investigations on the role of advice. J Consum Satisfaction, Dissatisfaction Complaining Behav
6:81–89
Mattila AS, Wirtz J (2004) Consumer complaining to firms: the determinants of channel choice.
J Serv Mark 18(2):147–155. doi:10.1108/08876040410528746
Maxham JG, Netemeyer RG (2002) Modeling customer perceptions of complaint handling over
time: the effects of perceived justice on satisfaction and intent. J Retail 78(4):239–252. doi:10.
1016/S0022-4359(02)00100-8
McCollough MA, Berry LL, Yadav MS (2000) An Empirical Investigation of Customer
Satisfaction after Service Failure and Recovery. J Serv Res 3(2):121–137. doi:10.1177/
109467050032002
Morel KP, Poiesz TB, Wilke HA (1997) Motivation, capacity and opportunity to complain.
Towards a comprehensive model of consumer complaint behavior. Adv Consum Res 24:464–
469
Nunnally J (1978) Psychometric theory, 2nd edn. McGraw-Hill Higher Education, New York
Pallant J (2007) SPSS survival manual. A step by step guide to data analysis using SPSS for
Windows, 3rd edn. Open University Press, Maidenhead
Raab-Steiner E, Benesch M (2010) Der Fragebogen. Von der Forschungsidee zur SPSS/PASW-
Auswertung, 2nd edn. Facultas.wuv, Wien
Robertson N, Shaw RN (2009) Predicting the likelihood of voiced complaints in the self-service
technology context. J Serv Res 12(1):100–116. doi:10.1177/1094670509333789
Schmitt N (1996) Uses and abuses of coefficient alpha. Psychol Assess 8(4):350–353. doi:10.1037/
1040-3590.8.4.350
Schnell R, Hill PB, Esser E (2011) Methoden der empirischen Sozialforschung, 9th edn.
Oldenbourg Verlag, München
Seiders K, Berry LL (1998) Service fairness: what it is and why it matters. Acad Manag Executive
12(2):8–20. doi:10.5465/AME.1998.650513
Singh J (1990) A typology of consumer dissatisfaction response styles. J Retail 66(1):57–99
Smith AK, Bolton RN, Wagner J (1999) A model of customer satisfaction with service encounters
involving failure and recovery. J Mark Res 36(3):356. doi:10.2307/3152082
Statistisches Bundesamt (2014) Statistisches Jahrbuch. Deutschland und internationales.
Statistisches Bundesamt, Wiesbaden
van Eimeren B, Frees B (2011) Ergebnisse der ARD/ZDF-Onlinestudie 2011. Drei von vier
Deutschen im Netz – ein Ende des digitalen Grabens in Sicht? Media Perspektiven, Frankfurt
am Main
von der Heyde Fernandes D, Pizzuti dos Santos C (2008) The antecedents of the consumer
complaining behavior. Adv Consum Res 35:584–593
Voorhees CM, Brady MK (2005) A service perspective on the drivers of complaint intentions.
J Serv Res 8(2):192–204. doi:10.1177/1094670505279702
Wirtz J, Mattila AS (2004) Consumer responses to compensation, speed of recovery and apology
after a service failure. Int J Serv Ind Manag 15(2):150–166. doi:10.1108/09564230410532484
Wirtz J, McColl-Kennedy JR (2010) Opportunistic customer claiming during service recovery.
J Acad Mark Sci 38(5):654–675. doi:10.1007/s11747-009-0177-6
Chapter 5
Conclusions for Organisational Complaint
Management and Future Research

Abstract In this chapter the main question “which communication channels are
adequate options to voice a complaint from a customer perspective” is answered.
Four main managerial implications are drawn from the results of the empirical
studies. First, traditional communication channels like mail, e-mail, phone and face-
to-face are evaluated regarding their adequacy in the context of customer com-
plaints. Second, social networks as a potential new complaint channel are exam-
ined. Third, traditional complaint channels and social networks are compared
regarding their adequacy to voice a complaint. Fourth, social interaction is identi-
fied as a mediator for customers’ perception of complaint channel adequacy. To
conclude this chapter, managerial implications are suggested. Besides, the contri-
butions to existing literature as well as the limitations of the studies are revealed.

Keywords Complaint channel comparison Managerial implications
Customer



complaint channel choice Limitation Future research

This chapter summarises the main findings of this study and derives managerial
recommendations from these findings (Sect. 5.1). The contribution to existing lit-
erature is highlighted (Sect. 5.2) and, finally, the limitations are demonstrated and
suggestions for topics for further research are made (Sect. 5.3).

5.1 Managerial Implications

Guided by the most important issue four management questions have been for-
mulated at the beginning of this study. Detailed answers for each of these are given
in the following.
The first management question focuses on the customers’ perception of each of
the traditional communication channels’ adequacy for voicing complaints to
companies. It can be concluded that basically all traditional communication chan-
nels can be used for complaints. The respondents of both surveys indicate that they
perceive mail, e-mail, phone, and face-to-face as being adequate, though customers

© The Author(s) 2015 75

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0_5
76 5 Conclusions for Organisational Complaint Management …

do not rate all of these channels in the same way. Major differences, for instance,
were shown to exist in customers’ perception of the complaint channel mail for the
procedural justice dimension. As procedural justice implies factor time as an
important one, it is reasonable that mail is perceived as inferior to the other com-
plaint channels, all of which allow for an almost immediate response. The major
reason for this evaluation is obvious; writing a letter, sending a letter, and getting a
response letter takes more time and effort for customers and companies than the
other channels require. Mail also fares badly with regard to personal attitudes.
Complaining via mail is perceived as the most formal and official way. This can
lead to a decrease in customers’ satisfaction with the overall complaint handling
process (Luria et al. 2009), given that customers often fear formality as a mode of
interaction having the potential for a confrontation with a company (Sect. 3.4).
Generally speaking, all conventional communication channels are perceived as
adequate complaint channels. However, due to the afore mentioned negative
aspects of mail complaints, companies should consider eliminating this complaint
option depending on their customer segment. Further grounds for this are that only
12 % of participants (of the total in both studies) would choose mail to complain.
As long as companies do not serve specific customers, namely those which would
prefer mail complaints (e.g. “middle-aged men”; Sect. 4.8.1), mail complaints
generally do not represent any sort of significance among the complaints received
by companies. However, depending on the type of product or service, customers
might prefer a more formal complaint channel. For instance, customers of com-
panies offering expensive products or services might show a greater preference for
mail complaints. In this case, companies should offer mail as a complaint channel to
offer a formal way to complain and therefore to increase customer satisfaction with
the complaint handling process.
Furthermore, complaints by mail entail extensive costs for companies.
Envelopes must be opened, read and digitalised to save the customer history in most
of these cases; answers have to be printed out and postal services paid for. By
means of minimising mail complaints, companies might therefore increase cus-
tomer satisfaction (due to the more immediate responses the other complaint
channels offer) and decrease costs (due to more efficient complaint handling).
Nevertheless, before eliminating mail as a complaint channel, companies should
analyse the use of this channel. If many customers use this channel to complain,
elimination can lead to customer dissatisfaction.
Management Implication 1
Conventional mail correspondence as a complaint channel is disadvantageous with regard
to the factors procedural justice and personal attitude. Therefore, companies should analyse
customers’ usage of this complaint channel. If an analysis reveals a low usage rate, com-
panies could consider the elimination of this complaint channel.

The second management question asks for the extent of customers’ perception of
online social networks as an additional and adequate communication channel to
voice complaints to companies. As the results of the surveys show, social networks
5.1 Managerial Implications 77

do not seem to be an adequate option for the communication of complaints in

general. When the stakes are a potential financial loss for customers, because of the
dissatisfaction incident, customers’ trust in social networks is low. This is shown by
low values for interactional and distributional justice. The reasons for this evalu-
ation could be reasoned as follows:
Many customers use social networks on a daily basis. In particular, about 60 % of
the participants indicate that they log in on a daily basis (cumulative for both
studies). This indicates a high trust level for this communication channel. However,
it might be the case that customers distinguish between their private social life, i.e.
communicating with friends and family via social networks, and their more business
related life, i.e. shopping and communicating with companies. Therefore, they prefer
more formal and official ways of making a complaint and do not accept the private
environments of social networks for this. Where financial loss is involved, customers
might have less confidence in this informal way of voicing complaints, but would
prefer using one of the conventional communication channels instead. As social
networks are a fairly new communication channel, they still have to prove them-
selves as reliable complaint channels despite their rapid growth rates. This is sup-
ported by the results, that only 1.2 % of all participants would choose social
networks as a complaint channel if other conventional channels are available.
Results of the cluster analysis further support the assumption that social networks are
relatively new compared to conventional channels and that customers who use social
networks on a daily basis and are therefore already familiar with them, would choose
social networks as complaint channels (Sect. 4.8.1, “young non-complainers”).
Complaining via social networks is visible to a wider public. Other members of
the social networks are thus able to read the complaint after it has been posted on
the companies’ social network site instead of being communicated privately
between the sender (i.e. the customer) and the recipient (i.e. the company).
Customers can be uncomfortable with this and might therefore choose complaint
channels offering a more private mode of communication.
Nevertheless, companies should be present in social networks. In the event that
their customer base prefers an informal way of complaining, these customers do
have the ability to choose social networks as a complaint channel. However, for
other customer groups the company should provide other complaint channels, like
phone number and e-mail-addresses in order to receive complaints.
Given that complaint management is only one of many customer services pro-
vided, social networks might enable companies to gain further insights into their
customers’ behaviour. Price trends, product opportunities and changes in customer
perception, to name just a few examples, may well be detected in time using social
networks as a kind of early-warning system. One downside for companies might be
the increased transparency for the company’s competitors, since all information,
including customer complaints, are to some extent publicly visible. At this stage
only little can be said about the future of social networks in the long run, since they
might just turn out to be a fad. Companies wanting to launch social networks as a
complaint channel and ensure the acceptance of these have to improve customers’
78 5 Conclusions for Organisational Complaint Management …

trust through marketing campaigns. Especially, complaints which are voiced via
social networks in an early stage should be handled flawlessly in order to reduce
prejudices. Thereby, the question of sustainability is raised: the implied uncertainty
about payback of investments.
Management Implication 2
Social networks as a complaint channel are only suitable for subgroups. It is recommended
that companies analyse their customer base in order to determine the suitability of social
networks as a complaint channel. If social networks are found to be eligible as a new
complaint channel, companies should start marketing campaigns to reduce prejudices and
manage customer complaints perfectly.

The third management question addresses the comparison between the con-
ventional complaint channels and the new opportunity offered by social networks.
The results of the survey show that conventional complaint channels are perceived
by customers as being more suitable to their needs when compared to social net-
works. This supports results by Matos et al. (2009) and Tax et al. (1998). The
implementation of social networks as a complaint channel therefore does not have
the potential to improve overall customer satisfaction with handling complaints.
However, social networks can serve as an additional complaint channel for cus-
tomer groups like “young non-complainers”. This customer-cluster usually does not
complain at all, but when they decide to do so, they would choose social networks.
Hence, companies can motivate this specific sub-cluster of customers to voice their
complaints through these. Given that companies are able, for instance, to com-
pensate customers for a service failure, this can lead to more satisfied customers and
profitable long-term relationships (Cho et al. 2002; Wirtz and Mattila 2004).
Conventional complaint channels generally do not differ significantly in the way
customers perceive them. Two communication channels can partly be seen as
exceptions. First, as already discussed, mail complaints are perceived as inferior
compared to other channels for the procedural justice and personal attitudes.
Second, face-to-face complaints seem to represent the most adequate complaint
channel with regard to the interactional justice dimension. By choosing this
channel, customers directly talk to an employee and can immediately experience
honesty and helpfulness in a face-to-face situation to an extent, which cannot be
realised by the other communication channels.
To sum up, the following management implication can be drawn by comparing
the complaint channels:
Management Implication 3
Traditional complaint channels cannot be ranked with regard to their suitability for voicing
complaints. However, face-to-face situations do have the best chance to remedy the dis-
satisfaction incident by providing a solution in a honest and helpful manner. Furthermore,
companies should implement social networks as an additional complaint channel to
motivate a sub-group of customers, i.e. young non-complainers, to voice their complaints.
5.1 Managerial Implications 79

Management question 4 asks for customers’ social interaction preferences. The

results show that the need for social interaction has a partial influence on customers’
complaint channel choice. Usually, customers with a high need for social interac-
tion choose to phone to complain. Furthermore, study 1 reveals that customers with
a high need for social interaction also show a higher complaint likelihood in
general. Customers, who appreciate advice from and exchange of experience with
other customers, are more likely to complain directly at the company. However, the
results from study 2 show the opposite result, i.e. customers with a high need for
social interaction are less likely to complain in this way. This is reasonable as
customers with a high need for social interaction appreciate the opinions of other
customers or advice that is given to them. Hence, these customers (i.e. young non-
complainers) are more likely to communicate their complaints to friends, their
family or to use social media as a network for contacting the company, thus getting
into contact with other customers. These different results can be caused by sig-
nificant differences in age-distribution between the two survey samples (t = 22.45,
p = 0.000). However, neither age nor gender affects the results regarding the other
dependent variables (Appendices D and E).
These results support implementing social networks as an additional complaint
channel. In this way, customers, who usually stay silent, can voice their complaints
and satisfy their need to interact with other customers.
Results from a more detailed analysis of the age cluster also show that older
customers tend to complain more than younger ones. This result is contrary to that
by von der Heyde Fernandes and Pizzuti dos Santos (2008). Companies should try
to change younger customers’ complaint behaviour. Research has shown that
customer complaints can benefit internal processes and increase the customers’
satisfaction with a company (Bearden and Teel 1983; Blodgett et al. 1995; Cho
et al. 2002). Moreover, motivating younger customers to complain can also lead to
long-term relationships. As complaining can lead to increased satisfaction after a
successful complaint management, loyalty can increase (Bearden and Teel 1983).
As younger customers (i.e. <29 years old) might be influenced by a satisfying
transaction with a company, a long-term relationship can be developed. As the
results revealed, social networks are the favourite complaint channel for this
younger customer group; companies should implement social networks as an
additional complaint channel.
Management Implication 4
Companies should recognise the need of some of their customers for social interaction, i.e.
getting advice and sharing experiences with other customers. In the context of complaint
management, companies should again consider to utilise social networks as a complaint
channel to satisfy the need for social interaction for a certain segment of their customer
base. To motivate younger customers to complain, companies should implement social
networks as a complaint channel.
80 5 Conclusions for Organisational Complaint Management …

5.2 Contribution to Existing Research

This study provides valuable contributions to existing literature in the field of

complaint management. Basically, this research extends the research field by an
investigation of the adequacy of complaint channels. So far, only a few papers have
taken this perspective as outlined in Sect. 2.2. Apart from an investigation of four
conventional complaint channels, this study is the first to include social networks as
a potential new complaint channel. The adequacy of all complaint channels is
measured by operationalising widely established factors and theories.
Existing research has been adopted and extended. Foremost, the major study of
Wirtz and Mattila (2004) has been updated by including the ‘semi-interactive’
dimension (Sect. 2.2) classification of the customers’ complaining behaviour.
Moreover, as already described in detail, the studies contribute to the research of
Malafi (1991) and Malafi et al. (1993). Since the studies have been published in
1991 and 1993, this research shows that the concept of social interaction within
complaint management is still relevant. In contrast to the studies published more
than twenty years ago, customers nowadays have more opportunities for social
interaction through the use of online social networks.

5.3 Limitations and Implications for Future Research

Despite pre-testing and using two different samples to avoid bias, limitations are
inevitable. First, given the cross-sectional character of the studies, they only show a
moment in time. With regard to the fast developing and changing environment of
social networks, the validity of the findings might come to be called into question in
the future. Second, given that the questionnaire is computer-administered, a certain
extent of internet affinity by the respondents is presupposed. The results might
therefore be different when compared to data that might have been collected by
distributing paper-based questionnaires to participants having no closeness to new
technologies. This kind of an approach might have yielded an entirely different
outcome. Third, participants were given a concrete case study, and although this
approach has its advantages, a different scenario might have been evaluated dif-
ferently, too. Participant’s behaviour in and response to a real life situation could
well have been quite different compared to this theoretical setup. In this context,
social desirability might also have biased the answers. For instance, customers may
perhaps have been unwilling to admit that they usually do not complain to com-
panies and might therefore have given untrue answers (e.g. Question 5). Fourth, the
information value of the findings might to some extent be characterized by con-
straints, as the study does not distinguish between customers’ complaints about
monetary and non-monetary issues—just to name one aspect. Future studies might
perhaps address this research topic to gain a more detailed insight into customer
complaint behaviour.
5.3 Limitations and Implications for Future Research 81

The authors would like to encourage others to conduct more future research
related to communication channels in the context of complaint voicing, as customer
attitudes and preferences toward the different channels should be analysed in more
detail. This investigation indicates that the preferred complaint channel depends, for
instance, on the complaining preferences in general. With reference to social net-
works, the specific role of social desirability and social opinion leaders in online
communities should be considered.
Additionally, more research on demographic determinants and on internet
complaining is required. Besides, the antecedents of channel usage across certain
types of failures (e.g. monetary loss, non-monetary loss, different branches) need to
be researched in more detail.

Take away
Customers prefer well-known traditional communication channels like phone,
e-mail or face-to-face conversations to voice a complaint. Social networks
can serve as an additional complaint channel for younger customers.
However, the implementation of social networks as a complaint channel will
take time. Therefore, companies have to reduce prejudices against the usage
of social networks in a B2C context. A successful implementation of social
networks can motivate younger customers to complain. Thus, companies can
gain valuable insights into younger customers’ behaviour, satisfy them, for
instance, after a service failure, and build a stronger customer-company
relationship.

References

Bearden WO, Teel JE (1983) Selected determinants of consumer satisfaction and complaint
reports. J Mark Res 20(1):21–28
Blodgett JG, Wakefield KL, Barnes JH (1995) The effects of customer service on consumer
complaining behavior. J Serv Mark 9(4):31–42. doi:10.1108/08876049510094487
Cho Y, Im I, Hiltz R, Fjermestad J (2002) An analysis of online customer complaints: implications
for Web complaint management. In: Proceedings of the 35th annual Hawaii international
conference on system sciences, vol 00(c), pp 2308–2317. doi:10.1109/HICSS.2002.994162
de Matos CA, Rossi CAV, Veiga RT, Vieira VA (2009) Consumer reaction to service failure and
recovery: the moderating role of attitude toward complaining. J Serv Mark 23(7):462–475.
doi:10.1108/08876040910995257
Luria G, Gal I, Yagil D (2009) Employees’ willingness to report service complaints. J Serv Res 12
(2):156–174. doi:10.1177/1094670509344214
Malafi TN (1991) The impact of social influence on consumer complaint behavior. J Consum
Satisfaction, Dissatisfaction Complaining Behav 4:144–150
Malafi TN, Cini MA, Taub SL, Bertolami J (1993) Social influence and the decision to complain.
Investigations on the role of advice. J Consum Satisfaction, Dissatisfaction Complaining Behav
6:81–89
82 5 Conclusions for Organisational Complaint Management …

Tax SS, Brown SW, Chandrashekaran M (1998) Customer evaluations of service complaint
experiences. Implications for relationship marketing. J Mark 62(2):60–76
von der Fernandes Heyde D, Pizzuti dos Santos C (2008) The antecedents of the consumer
complaining behavior. Adv Consum Res 35:584–593
Wirtz J, Mattila AS (2004) Consumer responses to compensation, speed of recovery and apology
after a service failure. Int J Serv Ind Manag 15(2):150–166. doi:10.1108/09564230410532484
Appendix A
Literature Overview

The following table illustrates both the research topic and the research design of
previous studies in the context of complaint management. To highlight the
importance of the chosen factors as outlined in Chap. 3, it is stated (“Yes”)
whenever past studies have attributed importance to the following factors: proce-
dural justice (PJ), interactional justice (IJ), distributive justice (DJ), personal atti-
tudes (PA), and social interaction (SI). None of the considered previous studies
explicitly negated the importance of one or more of the suggested factors.
It can be concluded that the circumstances leading to a satisfactory outcome have
often been researched and verified by applying field studies using questionnaires;
counter-evidence has not been found. Therefore, the underlying research design of
this investigation seems to be appropriate (Chap. 4). See Table A.1.

Table A.1 Literature review on complaint management

Reference paper Authors suggest factors to be
relevant*,**
Authors Research topic Research design PJ IJ DJ PA SI
1 Andreassen Antecedents to Field study, – – Yes – –
(2000) satisfaction with interviews
service recovery
2 Boshoff Service recovery Field study, – – Yes – –
(1997) options, satisfaction interviews
3 Chebat et al. Mediating effects of Field study, Yes Yes Yes – –
(2005) emotions on interviews
perceived justice in
service recovery
4 Dacin and Understanding and Literature review – – – Yes Yes
Davidow influencing consumer
(1997) complaint behaviour
(continued)

© The Author(s) 2015 83

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0
84 Appendix A: Literature Overview

Table A.1 (continued)

Reference paper Authors suggest factors to be
relevant*,**
Authors Research topic Research design PJ IJ DJ PA SI
5 Day (1977) Consumer Conceptualisation – – – – Yes
satisfaction
6 Day (1984) Responses to Conceptualisation Yes – Yes Yes –
dissatisfaction
7 Gelbrich Emotions after Experiment, – – – – Yes
(2010) service failure, questionnaire
informational support
8 Gelbrich Organisational Meta-analysis Yes Yes Yes – –
and Roschk complaint handling
(2010) and consumer
responses
9 Grougiou Attitudes to voicing Field study, – – – – Yes
and complaints interviews
Pettigrew
(2009)
10 Halstead Consumer attitudes Field study, – – – Yes –
(1991) towards complaining questionnaire
11 Homburg Organisation’s Field study, Yes Yes Yes – –
and Fürst complaint questionnaire
(2005) management,
customer justice
evaluation,
satisfaction
12 Kim et al. Relationship between Conceptualisation Yes Yes Yes Yes Yes
(2010) CCB and service
recovery
13 Luria et al. Employees’ Field study, – – – Yes –
(2009) willingness to report questionnaire
service failures
14 Malafi Impact of social Analysis of – – – – Yes
(1991) influence on previous study
consumer complaint
behaviour
15 Malafi et al. Investigations on the Experiment, – – – – Yes
(1993) role of advice questionnaire
16 Maxham Customer perceptions Field study, Yes Yes Yes – –
and of complaint handling questionnaire
Netemeyer over time
(2002)
17 Robertson Voiced complaints in Experiment, – – Yes Yes –
and Shaw the self-service questionnaire
(2009) technology context
(continued)
Appendix A: Literature Overview 85

Table A.1 (continued)

Reference paper Authors suggest factors to be
relevant*,**
Authors Research topic Research design PJ IJ DJ PA SI
18 Seiders and Antecedents of Field study, Yes Yes Yes – –
Berry service fairness interviews
(1998)
19 Smith et al. Satisfaction, service Experiment; Yes Yes Yes – –
(1999) encounter, recovery questionnaire
20 Smith and Customers’ emotional Experiment, Yes Yes Yes – –
Bolton responses to service questionnaire
(2002) failure, satisfaction
judgements
21 Stephens Conceptual Experiment, – – – – Yes
and framework of interviews
Gwinner customer complaint
(1998) behaviour
22 Strahle et al. Consumer Field study, – – – – Yes
(1992) complaining questionnaire
behaviour
23 Sundaram Word-of-mouth Field study, – – – – Yes
et al. (1998) communication interviews
24 Tax et al. Customer evaluations Field study, Yes Yes Yes – –
(1998) of service complaint questionnaire
experiences
25 Voorhees Drivers of complaint Field study, Yes Yes Yes Yes –
and Brady intentions and interviews
(2005) complaint handling
26 Wirtz and Justice, attribution Experiment, Yes Yes Yes – –
Mattila processes, questionnaire
(2004) satisfaction,
behavioural
responses
Σ 12 11 15 7 10
*Based on authors’ findings, whereas parameters can be: relevant (‘Yes’), not relevant (‘No’), or
not considered (‘–’)
**Illustrated constructs are not exclusive to further findings of the respective authors, i.e. only
constructs with reference to the five factors are considered
Appendix B
Questionnaire

Dear Participant

Thank you for your willingness to participate in this survey.

The questionnaire asks for your preference and attitude regarding complaints to
companies.

• Your participation is voluntary. The survey results will be treated anonymously.

• Please select the answers which best describe your situation.
• The survey takes approx. 8 minutes to complete.
• The results will be published as part of my master’s thesis and potentially as a
scientific essay.

Kind regards
Stefan Garding

Maastricht University
1. Please select your gender.
Male
Female
2. Please select your age.
Below 20
20 - 29
30 - 39
40 - 49
50 - 59
60 - 69
Above 69
Not specified
3. Are you a member of at least one online social network (e.g. Facebook, Google+,
StudiVz, MeinVz, SchülerVz, Xing, or LinkedIn)?
Yes
No
4. How often do you log in on an average to the social network you use most?
Mostly inactive
Fewer than seven times per week
Every day

© The Author(s) 2015 87

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0
88 Appendix B: Questionnaire

5. Do you complain to companies in case of dissatisfaction with one of their products

or services?
In no Probably More Maybe More Probably Certainly
case no likely no likely yes yes yes

6. How do you usually address (communication channel) your complaints to

companies?
You may select up to two answers.
By phone
E-Mail
Mail
Face-to-face (at a shop)
Other: ____________
I do not complain to companies
7. Have you ever used social networks in order to address a complaint to a company?
Yes
No
Not specified
Please think of the following example when answering subsequent questions:

“You subscribed for a mobile phone account with a mobile network operator a
month ago. The call rates are fair and the service standards are reasonable. After
the first month of use, you receive the first invoice. Although the monthly service
fee was supposed to be 25 Euro, you were charged 30 Euro.”

You are now considering complaining to the company in order to receive a revised
invoice and to be invoiced in the future according to the agreed upon amount.

The mobile network operator provides the following communication channels to its
customers. All information required for using these channels is known to you. Thus,
you can freely choose the channel you prefer in order to complain.
• Mail: Send a letter to the company
• E−Mail: Write an e-mail to the company
• By phone: Call the company's hotline
• Social networks: Contact the company by visiting its page on a social
network (private message, post on wall, guestbook entry, or similar
action)
• Face-to-face: Go to one of the company's local shops
8. How annoyed would you be about this shortcoming?
Not Very
annoyed annoyed
at all
1 2 3 4 5 6 7
Appendix B: Questionnaire 89

9. Please select the communication channel you would choose in this situation to
address your complaint to the company.
Mail
E−Mail
By phone
Social networks
Face-to-face
I would not complain
Please evaluate all communication channels in all of the following questions.

[Note: If respondents stated previously that they are not participating in social
networks, the following message appeared instead.]
Please evaluate all communication channels in all of the following questions. When
evaluating the communication channel “social networks” please consider your
general perceptions of this channel (from media reports, or discussions with your
colleagues).
10. How time-consuming is the use of each communication channel to address a
complain?
Not Very
time- time-
consuming consuming
at all
1 7
2 3 4 5 6
Mail

E-Mail
By
phone
Social
networks
Face-to-
face
11. Please evaluate the suitability of each communication channel to receive an
immediate response from the company to your complaint?

In no Quite More More

case unsuitable likely Moderate likely Quite Very
suitable unsuitable suitable suitable suitable
Mail

E-Mail
By
phone
Social
networks
Face-to-
face
90 Appendix B: Questionnaire

12. Please evaluate the official/formal character of a complaint, addressed to the

company by using the respective communication channel.

In no Quite More More

case informal likely Moderate likely Quite Very
formal informal formal formal formal
Mail

E-Mail
By
phone
Social
networks
Face-to-
face
13. Please evaluate the extent to which the company will be honest with you according
to your general experience with each communication channel.
Not
honest Very
at all honest
1 2 3 4 5 6 7
Mail

E-Mail
By
phone
Social
networks
Face-to-
face
14. Please evaluate the extent to which the company will be helpful according to your
general experience with each communication channel.
Not
helpful Very
at all helpful
1 2 3 4 5 6 7
Mail

E-Mail
By
phone
Social
networks
Face-to-
face
Appendix B: Questionnaire 91

15. Please evaluate the likelihood that the complaint, addressed by using the respective
communication channel, will lead to your desired outcome/success.
In no
case Very
likely likely
1 2 3 4 5 6 7
Mail

E-Mail
By
phone
Social
networks
Face-to-
face
16. Would you appreciate it if you could get advices from other customers of the
company with regard to the shortcoming via an informal forum?
Yes
No
No preference
17. Would you appreciate talking to customers with similar experiences via an informal
forum?
Yes
No
No preference
18. Would you appreciate it if you could address a complaint to a company via social
networks?
Yes
No
No preference
19. Would you be annoyed if you could not address a complaint to a company via social
networks?
Yes
No
No preference
92 Appendix B: Questionnaire

20. Please evaluate the user-friendliness of each communication channel in principle,

regardless of the likelihood of success of your complaint.
Not
user- Very
friendly user-
at all friendly
1 2 3 4 5 6 7
Mail

E-Mail
By
phone
Social
networks
Face-to-
face
21. Do you have any further comments you would like to make? If so, please type in
the field below.
If you are interested in the results, please also type your e−mail address in the field
below.
When you have finished, proceed by clicking “submit”.
______________________________
______________________________
______________________________

The questionnaire has been successfully completed and transmitted.

Thank you for your participation.
Appendix C
Variables

See Table C.1.

Table C.1 Overview of variables

Q# Value Variable Min Max
1 1 = Male GENDER 1 2
2 = Female
2 1 = <20 AGE_CLUSTER 1 5
2 = 20–29
3 = 30–39
4 = 40–49
5 = 50–59
6 = 60–69
7 = >69
8 = Not spec.
3 1 = Yes NET_MEMBER 1 1
2 = No
4 1 = Mostly inactive NET_FREQ 1 3
2 = Fewer 7 times
3 = Every day
5 7-point Likert scale COMP.Pref 1 7
1 = In no case
7 = Certainly yes
6 1 = By phone PREF_CHAN 1 5
2 = E-Mail
3 = Mail
4 = Face-to-face
E1 = Other: ______
5 = Don’t complain
7 1 = Yes PAST_COM 1 3
2 = No
3 = Not specified
8 7-point Likert scale CASE_DISSA 2 7
1 = Not annoyed
7 = Very annoyed
(continued)

© The Author(s) 2015 93

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0
94 Appendix C: Variables

Table C.1 (continued)

Q# Value Variable Min Max
9 1 = Mail CASE_CHAN 1 5
2 = E-Mail
3 = By phone
4 = Social networks
5 = Face-to-face
6 = Don’t complain
10 7-point Likert scales PJ_1A [Mail] 1 7
1 = Not time-consuming PJ_1B [E-Mail] 1 7
7 = Very time-consuming PJ_1C [Phone] 1 7
PJ_1D [Social Network] 1 7
PJ_1E [Face-to-face] 1 7
11 7-point Likert scales PJ_2A [Mail] 1 7
1 = Not suitable PJ_2B [E-Mail] 1 7
7 = Very suitable PJ_2C [Phone] 1 7
PJ_2D [Social Network] 1 7
PJ_2E [Face-to-face] 1 7
12 7-point Likert scales PA_A [Mail] 1 7
1 = Not formal PA_B [E-Mail] 1 7
7 = Very formal PA_C [Phone] 1 7
PA_D [Social Network] 1 7
PA_E [Face-to-face] 1 7
13 7-point Likert scales IJ_1A [Mail] 1 7
1 = Not honest IJ_1B [E-Mail] 1 7
7 = Very honest IJ_1C [Phone] 1 7
IJ_1D [Social Network] 1 7
IJ_1E [Face-to-face] 1 7
14 7-point Likert scales IJ_2A [Mail] 1 7
1 = Not helpful IJ_2B [E-Mail] 1 7
7 = Very helpful IJ_2C [Phone] 1 7
IJ_2D [Social Network] 1 7
IJ_2E [Face-to-face] 1 7
15 7-point Likert scales DJ_A [Mail] 1 7
1 = Not likely DJ_B [E-Mail] 1 7
7 = Very likely DJ_C [Phone] 1 7
DJ_D [Social Network] 1 7
DJ_E [Face-to-face] 1 7
16 1 = Yes SI_1 1 3
2 = No
3 = No preference
17 1 = Yes SI_2 1 3
2 = No
3 = No preference
18 1 = Yes KANO_P 1 3
2 = No
3 = No preference
(continued)
Appendix C: Variables 95

Table C.1 (continued)

Q# Value Variable Min Max
19 1 = Yes KANO_N 1 3
2 = No
3 = No preference
20 7-point Likert scales PJ_3A [Mail] 1 7
1 = Not user-friendly PJ_3B [E-Mail] 1 7
7 = Very user-friendly PJ_3C [Phone] 1 7
PJ_3D [Social Network] 1 7
PJ_3E [Face-to-face] 1 7
21 Free-text field OPEN_QUESTION / /
Appendix D
Multivariate Tests for Gender and Age
Differences (Study 1)

See Tables D.1 and D.2.

Table D.1 T-test: to test for gender differences

Gender M SD t p
PJ Male 4.59 1.32 0.08 0.939
Female 4.60 1.33
IJ Male 4.52 1.51 1.01 0.312
Female 4.62 1.50
DJ Male 5.03 1.62 0.25 0.8
Female 5.06 1.70
PA Male 2.94 1.82 0.76 0.446
Female 3.03 1.77

© The Author(s) 2015 97

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0
98 Appendix D: Multivariate Tests for Gender and Age Differences (Study 1)

Table D.2 Manova: to test for age differences

Age clusters M SD Multivariate tests
F p
PJ <20 3.93 1.55 1.308 0.143
20–29 4.63 1.37
30–39 4.63 1.33
40–49 4.54 1.29
50–59 4.43 1.27
Sum 4.59 1.33
IJ <20 3.40 1.14
20–29 4.45 1.60
30–39 4.65 1.48
40–49 4.69 1.37
50–59 4.53 1.54
Sum 4.57 1.50
DJ <20 2.60 2.19
20–29 4.95 1.73
30–39 5.17 1.64
40–49 5.10 1.48
50–59 5.02 1.78
Sum 5.05 1.66
PA <20 4.40 2.19
20–29 3.06 1.87
30–39 2.91 1.84
40–49 2.90 1.61
50–59 3.08 1.73
Sum 2.98 1.79
Appendix E
Multivariate Tests for Gender and Age
Differences (Study 2)

See Tables E.1 and E.2.

Table E.1 T-test: to test for gender differences

Gender M SD t p
PJ Male 4.39 1.34 0.739 0.460
Female 4.47 1.29
IJ Male 4.61 1.37 1.312 0.190
Female 4.48 1.41
DJ Male 4.90 1.51 0.09 0.929
Female 4.91 1.59
PA Male 3.16 1.81 0.965 0.335
Female 3.03 1.77

© The Author(s) 2015 99

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0
100 Appendix E: Multivariate Tests for Gender and Age Differences (Study 2)

Table E.2 Manova: to test for age differences

Age clusters M SD Multivariate tests
F p
PJ <20 4.47 1.34 1.29 0.196
20–29 4.41 1.32
30–39 4.40 1.36
40–49 4.63 1.17
50–59 5.27 0.95
Sum 4.43 1.32
IJ <20 4.67 1.36
20–29 4.54 1.40
30–39 4.45 1.29
40–49 4.70 1.42
50–59 5.60 1.03
Sum 4.55 1.39
DJ <20 4.92 1.52
20–29 4.87 1.56
30–39 5.09 1.56
40–49 4.93 1.46
50–59 6.00 0.00
Sum 4.90 1.55
PA <20 3.33 1.87
20–29 3.11 1.80
30–39 2.97 1.82
40–49 3.00 1.49
50–59 1.00 0.00
Sum 3.10 1.79
Glossary

Attribution theory is a psychological theory. In the context of this study it

comprises two key elements. Stability refers to the perceived likelihood that a
similar failure is likely to occur again, whereas controllability describes cus-
tomers’ perception as to whether a failure could easily have been prevented by
the company (Choi and Mattila 2008)
Behaviour theory refers to Singh’s (1990) classification of four groups of com-
plainers based on their individual complaining behaviour: passives, voicers,
irates, and activists
Complaint channel is a certain communication channel (e.g. phone, e-mail) which
is utilised by customers to voice a complaint to a company
Customer complaint behaviour (CCB) describes the portfolio of potential cus-
tomer behaviour after a product or service failure. Customers can remain silent
or voice their complaints in a public or private manner
Dependent variable shows the estimated effect (output) of a statistical model. In
contrast, the independent variable represents the causes (input). The value of a
dependent variable depends to a certain extent on the values of one or more
independent variables
Distributive justice is part of justice theory. Is defined as the fairness of the
outcome and the provided remedy (McCole 2004)
Fairness theory is a synonym to Justice theory
Independent variable represents the input variable which influences the value of
the dependent variable
Interactional justice is part of justice theory. Describes the interpersonal treatment
of the complaining customer, i.e. politeness and helpfulness of companies
(McCole 2004)
Justice theory covers three dimensions of customer evaluation. In the context of
this study, the dimensions of justice theory describe customer evaluations of
companies’ complaint handling. These dimensions are interactional, procedural,
and distributive justice

© The Author(s) 2015 101

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0
102 Glossary

Non-voicers are customers who do not take public or private action after a dis-
satisfaction incident. In the context of this study, customers do not voice their
complaints to the company but remain silent instead
Personal attitudes can also be described as individual characteristics and influence
customer complaint behaviour. Customers differ for example in their general
preferences of choosing a certain complaint channel, their likelihood of com-
plaining, and their fear of confrontation with a company
Private actions represent all actions taken by customers to voice a complaint in a
customers’ private environment, as for instance to friends and families. In
contrast, customers might take official public actions
Procedural justice is part of justice theory. Refers to customers’ perceived fair-
ness of the policies and procedures that companies have in place to handle
complaints (McCole 2004)
Product failure specifies the circumstance where a purchased product does not
fulfil customer expectations due to malfunction or shortcomings in design.
Product, failure often leads to customer dissatisfaction
Public actions describe all actions taken by customers to voice a complaint to
official parties. This involves either directly the responsible company or 3rd
parties like e.g. lawyers or consumer protection agencies
Recovery paradox characterises a phenomenon where customers show higher
satisfaction levels after successful recovery of a dissatisfaction incident after
complaining, compared to the satisfaction level before the incident occurs
Service failure depicts a situation where customers experience shortcomings and
resulting dissatisfaction in services a company has provided
Voicers are customers who take public or private action after a dissatisfaction
incident. In the context of this study, customers voice their complaints to the
company
Index

A Customer satisfaction, 1, 3, 6, 9, 15, 16, 18, 20,

Activists, 21 21, 23, 31, 76, 78
Adequacy of complaint channels, 5, 20, 23, 32,
80 D
Advice, 7, 31, 40, 53, 64, 67, 68, 70 Decision tree, 42, 67, 70, 71
Analysis of variance (ANOVA), 41, 42, 46, 47, Defensive marketing strategy, 3, 4
58, 59, 69 Degree of freedom, 41
Attribution theory, 22, 27 Demographic, 22, 23, 28, 32, 37, 81
Dependent variable, 40, 41, 45–47, 51, 54, 57,
B 59, 65, 69
Bartlett’s test of sphericity, 45, 46, 57 Descriptive statistics, 41–43, 55, 56
Behaviour theory, 21, 27 Distributive justice, 21, 27, 30, 37, 40, 46, 51,
Brown-Forsythe, 46, 47, 58, 69 58, 83, 101
Business to business, 14
Business to customer, 1 E
E-mail, 2, 16, 17, 22, 30, 44, 45, 47–52, 55, 57,
C 59–63, 66
Channel choice, 7, 22, 53, 54, 65, 79 Emotions, 3, 14, 23, 84
Chi-square test, 41, 46, 53, 54, 58, 64 Empirical analysis, 37, 41
Cluster analysis, 41, 67, 68, 77 Equality, 30, 47, 58, 69
Communication channel, 2, 6, 7, 13, 15–19, 31, Equity, 30
37, 39, 41, 57, 76, 81, 101 Expectation, 1, 8, 20, 30, 67
Compensation, 3, 14, 21 Experience, 2, 3, 5, 7, 8, 17, 20, 31, 37, 41, 53,
Complaint behaviour, 13, 15, 20, 41, 56, 67, 54, 66, 79
79, 83–85 Exploratory, 36, 45, 46, 57, 67
Complaint channel, 1, 2, 5–9, 13, 15, 16, 19, Exploratory factor analysis(EFA), 41, 45, 57
20, 23, 27–31, 35–37, 39–41, 44–48,
63, 65–71, 75–81 F
Complaint voicing, 81 Face-to-face, 2, 16, 17, 29, 44, 48–51, 56, 60,
Construct measurement, 42, 45, 55, 57 66, 75, 78
Controllability, 22 Facsimile, 44, 45
Cronbach alpha, 41, 45, 57 Fairness theory, 8
Customer characteristics, 28 Female, 17, 43, 56, 68
Customer complaint behaviour (CCB), 4, 8, 13,
14, 16, 20–22, 30, 31, 68, 71, 84 G
Customer dissatisfaction, 14, 17, 20, 22, 76 Games-Howell test, 47, 59, 69
Customer expectation, 8, 27 Gender, 43, 55, 79
Customer loyalty, 3, 4

© The Author(s) 2015 103

S. Garding and A. Bruns, Complaint Management and Channel Choice,
SpringerBriefs in Business, DOI 10.1007/978-3-319-18179-0
104 Index

H Private actions, 6, 17, 102

Helpfulness, 21, 29, 40, 46, 58, 61, 66 Procedural justice, 21, 27, 28, 40, 45, 47, 59,
Homogeneity, 41, 46, 47, 58, 69 76, 78, 102
Honesty, 29, 40, 46, 58, 61, 66, 78 Product, 1, 3, 4, 14, 20, 22, 76, 102
Product failure, 22, 102
I Public actions, 6, 14, 17, 102
Incident, 3, 77
Independent variable, 39, 54, 65, 69 Q
Informal, 6, 30, 31, 41, 77 Questionnaire, 9, 27, 35, 37, 39, 43, 46, 83
Interactional justice, 21, 22, 27, 29, 40, 45, 46,
48, 49, 57, 58, 60–62 R
Interactive, 7, 9, 17, 18 Realism check, 55, 56
Irates, 21 Recovery paradox, 3, 102
Remedy, 14, 21, 22, 30, 78, 101
J Remote, 9, 17, 18
Justice theory, 20, 21 Research agenda, 9, 35
Research design, 35, 83
K Respondent, 36–39, 41, 67, 80
Kaiser-Meyer-Olkin (KMO), 45, 57 Results, 9, 21, 22, 35, 37, 39, 41, 43, 45–47,
55, 58, 59, 66, 77, 79, 80
L Rotated factor loadings, 46
Levene’s test, 46, 58, 69
Likert scale, 38–41 S
Limitations, 75, 80 Self-confidence, 21
Semi-interactive, 9, 18, 80
M Service, 1, 3, 4, 20, 23, 67
Mail, 2, 5, 16–18, 22, 29, 37, 39, 45, 48, 49, Service failure, 2, 3, 9, 13, 37, 44, 78, 101, 102
51, 52, 60 Social influence, 7, 8, 31
Male, 17, 67, 97, 99 Social interaction, 6–8, 27, 31, 40, 46, 53, 54,
Managerial Implication, 75 58, 64, 65, 79, 80, 83
Manipulation, 37, 41, 42, 44, 55, 56 Social networks, 4, 6, 8, 9, 15, 18, 19, 29, 31,
Multiple choice, 38 37, 41, 44, 56, 66, 67, 70, 77, 79
Speed of response, 40, 46, 58
N Stability, 22, 101
Need, 4, 5, 8, 9, 13, 29, 30, 47, 54, 59, 65, 79, Subgroups, 35, 41
81
Non-voicers, 15 T
T-test, 41, 54
O Time-consumption, 29, 40, 60
Operationalisation, 35, 39 Twitter, 57
Opportunistic, 14, 22
U
P User-friendliness, 29, 37, 47
Passives, 21, 67, 68, 101
Personal attitudes, 27, 30, 31, 46, 47, 58, 63, V
76, 78, 102 Voicers, 21, 68, 102
Phone, 3, 5, 17, 18, 30, 37, 39, 48, 101
Post-hoc test, 46, 47, 59, 69 W
Post-purchase satisfaction, 8, 20 Welch, 46, 58
Primary data, 35 Word-of-mouth, 4, 6, 15, 17, 30